Reconsideration of the National Ambient Air Quality Standards for Particulate Matter, 16202-16406 [2024-02637]
Download as PDF
16202
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
Table of Contents
The following topics are discussed in
this preamble:
ENVIRONMENTAL PROTECTION
AGENCY
40 CFR Parts 50, 53, and 58
[EPA–HQ–OAR–2015–0072; FRL–8635–02–
OAR]
RIN 2060–AV52
Reconsideration of the National
Ambient Air Quality Standards for
Particulate Matter
Environmental Protection
Agency (EPA).
ACTION: Final rule.
AGENCY:
Based on the Environmental
Protection Agency’s (EPA’s)
reconsideration of the air quality criteria
and the national ambient air quality
standards (NAAQS) for particulate
matter (PM), the EPA is revising the
primary annual PM2.5 standard by
lowering the level from 12.0 mg/m3 to
9.0 mg/m3. The Agency is retaining the
current primary 24-hour PM2.5 standard
and the primary 24-hour PM10 standard.
The Agency also is not changing the
secondary 24-hour PM2.5 standard,
secondary annual PM2.5 standard, and
secondary 24-hour PM10 standard at this
time. The EPA is also finalizing
revisions to other key aspects related to
the PM NAAQS, including revisions to
the Air Quality Index (AQI) and
monitoring requirements for the PM
NAAQS.
SUMMARY:
DATES:
This final rule is effective May 6,
2024.
The EPA has established a
docket for this action under Docket ID
No. EPA–HQ–OAR–2015–0072. All
documents in the docket are listed on
the https://www.regulations.gov
website. Although listed in the index,
some information is not publicly
available, e.g., CBI or other information
whose disclosure is restricted by statute.
Certain other material, such as
copyrighted material, is not placed on
the internet and will be publicly
available only in hard copy form.
Publicly available docket materials are
available electronically through https://
www.regulations.gov.
FOR FURTHER INFORMATION CONTACT: Dr.
Lars Perlmutt, Health and
Environmental Impacts Division, Office
of Air Quality Planning and Standards,
U.S. Environmental Protection Agency,
Mail Code C539–04, Research Triangle
Park, NC 27711; telephone: (919) 541–
3037; fax: (919) 541–5315; email:
perlmutt.lars@epa.gov.
SUPPLEMENTARY INFORMATION:
ddrumheller on DSK120RN23PROD with RULES3
ADDRESSES:
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
Executive Summary
I. Background
A. Legislative Requirements
B. Related PM Control Programs
C. Review of the Air Quality Criteria and
Standards for Particulate Matter
1. Reviews Completed in 1971 and 1987
2. Review Completed in 1997
3. Review Completed in 2006
4. Review Completed in 2012
5. Review Initiated in 2014
a. 2020 Proposed and Final Decisions
b. Reconsideration of the 2020 PM NAAQS
Final Action
D. Air Quality Information
1. Distribution of Particle Size in Ambient
Air
2. Sources and Emissions Contributing to
PM in the Ambient Air
3. Monitoring of Ambient PM
4. Ambient Concentrations and Trends
a. PM2.5 Mass
b. PM2.5 Components
c. PM10
d. PM10–2.5
e. UFP
5. Characterizing Ambient PM2.5
Concentrations for Exposure
a. Predicted Ambient PM2.5 and Exposure
Based on Monitored Data
b. Comparison of PM2.5 Fields in
Estimating Exposure and Relative to
Design Values
6. Background PM
II. Rationale for Decisions on the Primary
PM2.5 Standards
A. Introduction
1. Background on the Current Standards
2. Overview of the Health Effects Evidence
a. Nature of Effects
i. Mortality
ii. Cardiovascular Effects
iii. Respiratory Effects
iv. Cancer
v. Nervous System Effects
vi. Other Effects
b. Public Health Implications and At-Risk
Populations
c. PM2.5 Concentrations in Key Studies
Reporting Health Effects
i. PM2.5 Exposure Concentrations
Evaluated in Experimental Studies
ii. Ambient PM2.5 Concentrations in
Locations of Epidemiologic Studies
d. Uncertainties in the Health Effects
Evidence
3. Summary of Exposure and Risk
Estimates
a. Key Design Aspects
b. Key Limitations and Uncertainties
c. Summary of Risk Estimates
B. Conclusions on the Primary PM2.5
Standards
1. CASAC Advice
2. Basis for the Proposed Decision
3. Comments on the Proposed Decision
4. Administrator’s Conclusions
C. Decisions on the Primary PM2.5
Standards
III. Rationale for Decisions on the Primary
PM10 Standard
A. Introduction
PO 00000
Frm 00002
Fmt 4701
Sfmt 4700
1. Background on the Current Standard
2. Overview of Health Effects Evidence
a. Nature of Effects
i. Mortality
ii. Cardiovascular Effects
iii. Respiratory Effects
iv. Cancer
v. Metabolic Effects
vi. Nervous System Effects
B. Conclusions on the Primary PM10
Standard
1. CASAC Advice
2. Basis for the Proposed Decision
3. Comments on the Proposed Decision
4. Administrator’s Conclusions
C. Decisions on the Primary PM10 Standard
IV. Communication of Public Health
A. Air Quality Index Overview
B. Air Quality Index Category Breakpoints
for PM2.5
1. Summary of Proposed Revisions
a. Air Quality Index Values of 50, 100, and
150
b. Air Quality Index Values of 200 and
Above
2. Summary of Significant Comments on
Proposed Revisions
a. Air Quality Index Values of 50, 100, and
150
b. Air Quality Index Values of 200 and
Above
c. Other Comments
3. Summary of Final Revisions
C. Air Quality Index Category Breakpoints
for PM10
D. Air Quality Index Reporting
1. Summary of Proposed Revisions
2. Summary of Significant Comments on
Proposed Revisions
3. Summary of Final Revisions
V. Rationale for Decisions on the Secondary
PM Standards
A. Introduction
1. Background on the Current Standards
a. Non-Visibility Effects
b. Visibility Effects
2. Overview of Welfare Effects Evidence
a. Nature of Effects
i. Visibility
ii. Climate
iii. Materials
3. Summary of Air Quality and
Quantitative Information
a. Visibility Effects
i. Target Level of Protection in Terms of a
PM2.5 Visibility Index
ii. Relationship Between the PM2.5
Visibility Index and the Current
Secondary 24-Hour PM2.5 Standard
b. Non-Visibility Effects
B. Conclusions on the Secondary PM
Standards
1. CASAC Advice
2. Basis for the Proposed Decision
3. Comments on the Proposed Decision
4. Administrator’s Conclusions
C. Decisions on the Secondary PM
Standards
VI. Interpretation of the NAAQS for PM
A. Amendments to Appendix K:
Interpretation of the NAAQS for
Particulate Matter
B. Amendments to Appendix N:
Interpretation of the NAAQS for PM2.5
VII. Amendments to Ambient Monitoring and
Quality Assurance Requirements
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
A. Amendment to 40 CFR Part 50
(Appendix L): Reference Method for the
Determination of Fine Particulate Matter
as PM2.5 in the Atmosphere—Addition of
the Tisch Cyclone as an Approved
Second Stage Separator
B. Issues Related to 40 CFR Part 53
(Reference and Equivalent Methods)
C. Changes to 40 CFR Part 58 (Ambient Air
Quality Surveillance)
D. Incorporating Data From NextGeneration Technologies
VIII. Clean Air Act Implementation
Requirements for the Revised Primary
Annual PM2.5 NAAQS
A. Designation of Areas
B. Section 110(a)(1) and (2) Infrastructure
SIP Requirements
C. Implementing the Revised Primary
Annual PM2.5 NAAQS in Nonattainment
Areas
D. Implementing the Primary and
Secondary PM10 NAAQS
E. Prevention of Significant Deterioration
and Nonattainment New Source Review
Programs for the Revised Primary
Annual PM2.5 NAAQS
F. Transportation Conformity Program
G. General Conformity Program
IX. Statutory and Executive Order Reviews
A. Executive Order 12866: Regulatory
Planning and Review and Executive
Order 14094: Modernizing Regulatory
Review
B. Paperwork Reduction Act (PRA)
C. Regulatory Flexibility Act (RFA)
D. Unfunded Mandates Reform Act
(UMRA)
E. Executive Order 13132: Federalism
F. Executive Order 13175: Consultation
and Coordination With Indian Tribal
Governments
G. Executive Order 13045: Protection of
Children From Environmental Health
and Safety Risks
H. Executive Order 13211: Actions
Concerning Regulations That
Significantly Affect Energy Supply,
Distribution or Use
I. National Technology Transfer and
Advancement Act (NTTAA)
J. Executive Order 12898: Federal Actions
To Address Environmental Justice in
Minority Populations and Low-Income
Populations and Executive Order 14096:
Revitalizing Our Nation’s Commitment
to Environmental Justice for All
K. Congressional Review Act (CRA)
References
Executive Summary
This document presents the
Administrator’s final decisions for the
reconsideration of the 2020 final
decision on the primary (health-based)
and secondary (welfare-based) National
Ambient Air Quality Standards
(NAAQS) for Particulate Matter (PM).
More specifically, this document
summarizes the background and
rationale for the Administrator’s final
decisions to revise the primary annual
PM2.5 standard by lowering the level
from 12.0 mg/m3 to 9.0 mg/m3; to retain
the current primary 24-hour PM2.5
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
standard (at a level of 35 mg/m3); to
retain the primary 24-hour PM10
standard; and, not to change the
secondary PM standards at this time. In
reaching his final decisions, the
Administrator considered the currently
available scientific evidence in the 2019
Integrated Science Assessment (2019
ISA) and the Supplement to the 2019
ISA (ISA Supplement), quantitative and
policy analyses presented in the 2022
Policy Assessment (2022 PA), advice
from the Clean Air Scientific Advisory
Committee (CASAC), and public
comments on the proposal. The EPA has
established primary and secondary
standards for PM2.5, which includes
particles with diameters generally less
than or equal to 2.5 mm, and PM10,
which includes particles with diameters
generally less than or equal to 10 mm.
The standards include two primary
PM2.5 standards: an annual average
standard, averaged over three years,
with a level of 12.0 mg/m3, and a 24hour standard with a 98th percentile
form, averaged over three years, and a
level of 35 mg/m3. It also includes a
primary PM10 standard with a 24-hour
averaging time, and a level of 150 mg/
m3, not to be exceeded more than once
per year on average over three years.
Secondary PM standards are set equal to
the primary standards, except that the
level of the secondary annual PM2.5
standard is 15.0 mg/m3.
The most recent of the PM NAAQS
was completed in December 2020. In
that review, the EPA retained the
primary and secondary NAAQS,
without revision (85 FR 82684,
December 18, 2020). Following
publication of the 2020 final action,
several parties filed petitions for review
and petitions for reconsideration of the
EPA’s final decision.
In June 2021, the Agency announced
its decision to reconsider the 2020 PM
NAAQS final action.1 The EPA decided
to reconsider the December 2020
decision because the available scientific
evidence and technical information
indicated that the current standards may
not be adequate to protect public health
and welfare, as required by the Clean
Air Act. The EPA noted that the 2020
PA concluded that the scientific
evidence and information called into
question the adequacy of the primary
PM2.5 standards and supported
consideration of revising the level of the
primary annual PM2.5 standard to below
the current level of 12.0 mg/m3 while
retaining the primary 24-hour PM2.5
1 The press release for this announcement is
available at: https://www.epa.gov/newsreleases/epareexamine-health-standards-harmful-soot-previousadministration-left-unchanged.
PO 00000
Frm 00003
Fmt 4701
Sfmt 4700
16203
standard (U.S. EPA, 2020b). The EPA
also noted that the 2020 PA concluded
that the available scientific evidence
and information did not call into
question the adequacy of the primary
PM10 or secondary PM standards and
supported consideration of retaining the
primary PM10 standard and secondary
PM standards without revision (U.S.
EPA, 2020b).
The final decisions presented in this
document on the primary PM2.5
standards have been informed by key
aspects of the available health effects
evidence and conclusions contained in
the 2019 ISA and ISA Supplement,
quantitative exposure/risk analyses and
policy evaluations presented in the 2022
PA, advice from the CASAC 2 and
public comment received as part of this
reconsideration.3 The health effects
evidence newly available in this
reconsideration, in conjunction with the
full body of evidence critically
evaluated in the 2019 ISA, supports a
causal relationship between long- and
short-term exposures and mortality and
cardiovascular effects, and the evidence
supports a likely to be a causal
relationship between long-term
exposures and respiratory effects,
nervous system effects, and cancer. The
longstanding evidence base, including
animal toxicological studies, controlled
human exposure studies, and
epidemiologic studies, reaffirms, and in
some cases strengthens, the conclusions
from past reviews regarding the health
effects of PM2.5 exposures.
Epidemiologic studies available in this
reconsideration demonstrate generally
positive, and often statistically
significant, PM2.5 health effect
associations. Such studies report
associations between estimated PM2.5
exposures and non-accidental,
cardiovascular, or respiratory mortality;
cardiovascular or respiratory
hospitalizations or emergency room
visits; and other mortality/morbidity
outcomes (e.g., lung cancer mortality or
incidence, asthma development). The
scientific evidence available in this
reconsideration, as evaluated in the
2019 ISA and ISA Supplement, includes
2 In 2021, the Administrator announced his
decision to reestablish the membership of the
CASAC. The Administrator selected seven members
to serve on the chartered CASAC, and appointed a
PM CASAC panel to support the chartered CASAC’s
review of the draft ISA Supplement and the draft
PA as a part of this reconsideration (see section
I.C.6.b below for more information).
3 More information regarding the CASAC review
of the draft ISA Supplement and the draft PA,
including opportunities for public comment, can be
found in the following Federal Register notices: 86
FR 54186, September 30, 2021; 86 FR 52673,
September 22, 2021; 86 FR 56263, October 8, 2021;
87 FR 958, January 7, 2022.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16204
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
a number of epidemiologic studies that
use various methods to characterize
exposure to PM2.5 (e.g., ground-based
monitors and hybrid modeling
approaches) and to evaluate associations
between health effects and lower
ambient PM2.5 concentrations. There are
a number of recent epidemiologic
studies that use varying study designs
that reduce uncertainties related to
confounding and exposure
measurement error. The results of these
analyses provide further support for the
robustness of associations between
PM2.5 exposures and mortality and
morbidity. Moreover, the Administrator
notes that recent epidemiologic studies
strengthen support for health effect
associations at lower PM2.5
concentrations, with these new studies
finding positive and significant
associations when assessing exposure in
locations and time periods with lower
annual mean and 25th percentile
concentrations than those evaluated in
epidemiologic studies available at the
time of previous reviews. Additionally,
the experimental evidence (i.e., animal
toxicological and controlled human
exposure studies) strengthens the
coherence of effects across scientific
disciplines and provides additional
support for potential biological
pathways through which PM2.5
exposures could lead to the overt
population-level outcomes reported in
epidemiologic studies for the health
effect categories for which a causal
relationship (i.e., short- and long-term
PM2.5 exposure and mortality and
cardiovascular effects) or likely to be
causal relationship (i.e., short- and longterm PM2.5 exposure and respiratory
effects; and long-term PM2.5 exposure
and nervous system effects and cancer)
was concluded.
The available evidence in the 2019
ISA continues to provide support for
factors that may contribute to increased
risk of PM2.5-related health effects
including lifestage (children and older
adults), pre-existing diseases
(cardiovascular disease and respiratory
disease), race/ethnicity, and
socioeconomic status. For example, the
2019 ISA and ISA Supplement conclude
that there is strong evidence that Black
and Hispanic populations, on average,
experience higher PM2.5 exposures and
PM2.5-related health risks than nonHispanic White populations. In
addition, studies evaluated in the 2019
ISA and ISA Supplement also provide
evidence indicating that communities
with lower socioeconomic status (SES),
as assessed in epidemiologic studies
using indicators of SES including
income and educational attainment are,
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
on average, exposed to higher
concentrations of PM2.5 compared to
higher SES communities.
The quantitative risk assessment, as
well as policy considerations in the
2022 PA, also inform the final decisions
on the primary PM2.5 standards. The risk
assessment in this reconsideration
focuses on all-cause or nonaccidental
mortality associated with long- and
short-term PM2.5 exposures. The
primary analyses focus on exposure and
risk associated with air quality that
might occur in an area under air quality
conditions that just meet the current
and potential alternative standards. The
risk assessment estimates that the
current primary PM2.5 standards could
allow a substantial number of PM2.5associated premature deaths in the
United States, and that public health
improvements would be associated with
just meeting all of the alternative (more
stringent) annual and 24-hour standard
levels modeled. Additionally, the
results of the risk assessment suggest
that for most of the U.S., the annual
standard is the controlling standard and
that revision to that standard has the
most potential to reduce PM2.5
exposure-related risk. The analyses are
summarized in this document and in
the proposal and are described in detail
in the 2022 PA.
In its advice to the Administrator, in
its review of the 2021 draft PA, the
CASAC concurred that the currently
available health effects evidence calls
into question the adequacy of the
primary annual PM2.5 standard. With
regard to the primary annual PM2.5
standard, the majority of the CASAC
concluded that the level of the standard
should be revised within the range of
8.0 to 10.0 mg/m3, while the minority of
the CASAC concluded that the primary
annual PM2.5 standard should be revised
to a level of 10.0 to 11.0 mg/m3. With
regard to the primary 24-hour PM2.5
standard, the CASAC did not reach
consensus on the adequacy of the
current standard. The majority of the
CASAC concluded that the primary 24hour PM2.5 was not adequate and that
the level of the standard should be
revised to within the range of 25 to 30
mg/m3, while the minority of the CASAC
concluded that the standard was
adequate and should be retained,
without revision. Additionally, in their
review of the 2019 draft PA, the CASAC
did not reach consensus on the
adequacy of the primary annual PM2.5
standard, with the minority
recommending revision and the
majority recommending the standard be
retained. In their review of the 2019
draft PA, the CASAC reached consensus
regarding the adequacy of the primary
PO 00000
Frm 00004
Fmt 4701
Sfmt 4700
24-hour PM2.5 standard, concluding that
the standard should be retained.
In considering how to revise the suite
of primary PM2.5 standards to provide
the requisite degree of protection, the
Administrator recognizes that the
current annual standard and 24-hour
standard, together, are intended to
provide public health protection against
the full distribution of short- and longterm PM2.5 exposures. Further, he
recognizes that changes in PM2.5 air
quality designed to meet either the
annual or the 24-hour standard would
likely result in changes to both longterm average and short-term peak PM2.5
concentrations.
As in 2012, the Administrator
concludes that the most effective way to
reduce total population risk associated
with both long- and short-term PM2.5
exposures is to set a generally
controlling annual standard, and to
provide supplemental protection against
the occurrence of peak 24-hour PM2.5
concentrations by means of a 24-hour
standard set at the appropriate level.
Based on the current evidence and
quantitative information, as well as
consideration of CASAC advice and
public comments, the Administrator
concludes that the current primary
annual PM2.5 standard is not adequate to
protect public health with an adequate
margin of safety. The Administrator
notes that the CASAC was unanimous
in its advice on the 2021 draft PA
regarding the need to revise the annual
standard. In considering the appropriate
level for a revised annual standard, the
Administrator concludes that a standard
set at a level of 9.0 mg/m3 reflects his
judgment about placing the most weight
on the strongest available evidence
while appropriately weighing the
uncertainties.
With regard to the primary 24-hour
PM2.5 standard, the Administrator finds
the available scientific evidence and
quantitative information to be
insufficient to call into question the
adequacy of the public health protection
afforded by the current 24-hour
standard. He further notes that a more
stringent annual standard set at a level
of 9.0 mg/m3 is expected to reduce both
average (annual) concentrations and
peak (daily) concentrations. The
Administrator also notes that, in their
review of the 2021 draft PA, the CASAC
did not reach consensus on whether
revisions to the primary 24-hour PM2.5
standard are warranted at this time. He
also notes that, in their review of the
2019 draft PA, the CASAC did reach
consensus that the primary 24-hour
PM2.5 standard should be retained. The
Administrator concludes that the 24hour standard should be retained to
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
continue to provide requisite protection
against short-term peak PM2.5
concentrations, particularly when
considered in conjunction with the
protection provided by the suite of
standards and the decision to revise the
annual standard to a level of 9.0 mg/m3.
The primary PM10 standard is
intended to provide public health
protection against health effects related
to exposures to PM10–2.5, which are
particles with a diameter between 10 mm
and 2.5 mm. The final decision to retain
the current 24-hour PM10 standard has
been informed by key aspects of the
available health effects evidence and
conclusions contained in the 2019 ISA,
the policy evaluations presented in the
2022 PA, advice from the CASAC and
public comments. Specifically, the
health effects evidence for PM10–2.5
exposures is somewhat strengthened
since past reviews, although the
strongest evidence still only provides
support for a suggestive of, but not
sufficient to infer, causal relationship
with long- and short-term exposures and
mortality and cardiovascular effects,
short-term exposures and respiratory
effects, and long-term exposures and
cancer, nervous system effects, and
metabolic effects. In reaching his final
decision on the primary PM10 standard,
the Administrator recognizes that, while
the available health effects evidence has
expanded, recent studies are subject to
the same types of uncertainties that
were judged to be important in previous
reviews. He also recognizes that, in their
review of the 2019 draft PA and the
2021 draft PA, the CASAC generally
agreed that it was reasonable to retain
the primary 24-hour PM10 standard
given the available scientific evidence,
including retaining PM10 as the
indicator. He concludes that the newly
available evidence does not call into
question the adequacy of the current
primary PM10 standard, and retains that
standard, without revision.
With respect to the secondary PM
standards, this reconsideration focuses
on visibility, climate, and materials
effects.4 The Administrator’s final
4 Consistent with the 2016 Integrated Review Plan
(U.S. EPA, 2016), other welfare effects of PM, such
as ecological effects, are being considered in the
separate, on-going review of the secondary NAAQS
for oxides of nitrogen, oxides of sulfur and PM.
Accordingly, the public welfare protection provided
by the secondary PM standards against ecological
effects such as those related to deposition of
nitrogen- and sulfur-containing compounds in
vulnerable ecosystems is being considered in that
separate review. Thus, the Administrator’s
conclusion in this reconsideration of the 2020 final
decision is focused only and specifically on the
adequacy of public welfare protection provided by
the secondary PM standards from effects related to
visibility, climate, and materials and hereafter
‘‘welfare effects’’ refers to those welfare effects.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
decision to not change the current
secondary standards at this time has
been informed by key aspects of the
currently available welfare effects
evidence as well as the conclusions
contained in the 2019 ISA and ISA
Supplement; quantitative analyses of
visibility impairment; policy
evaluations presented in the 2022 PA;
advice from the CASAC; and public
comments. Specifically, the welfare
effects evidence available in this
reconsideration is consistent with the
evidence available in previous reviews
and supports a causal relationship
between PM and visibility, climate, and
materials effects. With regard to
visibility effects, the Administrator
notes that he judges that the evidence
supports a target level of protection of
27 dv. He further notes that the results
of quantitative analyses of visibility
impairment suggest that in areas that
meet the current secondary 24-hour
PM2.5 standard that estimated light
extinction in terms of a 3-year visibility
metric would be at or well below the
target level of protection. With regard to
climate and materials effects, while the
evidence has expanded since previous
reviews, significant limitations and
uncertainties remain in the evidence.
While the evidence has expanded since
previous reviews, the available
scientific evidence remains insufficient
to allow the Administrator to make a
reasoned judgment about what specific
standard(s) would be requisite to protect
against known or anticipated adverse
effects to public welfare from PM’s
effects on materials damage or climate.In their review of the 2019 draft PA and
the 2021 draft PA, the CASAC did not
recommend revising the secondary PM
standards. In considering the available
evidence and quantitative information,
with its inherent uncertainties and
limitations, the Administrator judges
that it is appropriate not to change the
secondary PM standards at this time.
The final revisions to the primary
annual PM2.5 NAAQS trigger a process
under which States (and Tribes, if they
choose) make recommendations to the
Administrator regarding designations,
identifying areas of the country that
either meet or do not meet the new or
revised PM NAAQS. Those areas that do
not meet the revised PM NAAQS will
need to develop plans that demonstrate
how they will meet the standards. As
part of these plans, states have the
opportunity to advance environmental
justice, in this case for overburdened
communities in areas with high PM
concentrations above the NAAQS, by
using the tools described in the current
PM NAAQS implementation guidance
PO 00000
Frm 00005
Fmt 4701
Sfmt 4700
16205
(80 FR 58010, 58136, August 25, 2016).
The EPA is not making changes to any
of the current PM NAAQS
implementation programs in this final
rulemaking.
On other topics, the EPA is finalizing
two sets of changes to the PM2.5 subindex of the Air Quality Index (AQI).
First, the EPA is continuing to use the
approach used in the revisions to the
AQI in 2012 (77 FR 38890, June 29,
2012) of setting the lower breakpoints
(50, 100 and 150) based on the levels of
the primary annual and 24-hour PM2.5
standards. In so doing, the EPA is
revising the AQI value of 50 to 9.0 mg/
m3 and is retaining the AQI values of
100 and 150 at 35.4 mg/m3 and 55.4 mg/
m3, respectively. Second, the EPA is
revising the upper AQI breakpoints (200
and above), and replacing the linearrelationship approach used in 1999 (64
FR 42530, August 4, 1999) to set these
breakpoints, with an approach that more
fully considers the PM2.5 health effects
evidence from controlled human
exposure and epidemiologic studies that
has become available in the last 20
years. The EPA is also revising the AQI
values of 200, 300 and 500 to 125.4 mg/
m3, 225.4 mg/m3, and 325.4 mg/m3,
respectively. In addition, this final rule
revises the daily reporting requirement
from 5 days per week to 7 days per
week, while also reformatting appendix
G and providing clarifications.
With regard to monitoring-related
activities, the EPA finalizes revisions to
data calculations and ambient air
monitoring requirements for PM to
improve the usefulness and
appropriateness of data used in
regulatory decision making and to better
characterize air quality in communities
that are at increased risk of PM2.5
exposure and health risk. These changes
are found in 40 CFR part 50 (appendices
K, L, and N), part 53, and part 58 with
associated appendices (A, B, C, D, and
E). These changes include addressing
updates in data calculations, approval of
reference and equivalent methods,
updates in quality assurance statistical
calculations to account for lower
concentration measurements, updates to
support improvements in PM methods,
a revision to the PM2.5 network design
to account for at-risk populations, and
updates to the Probe and Monitoring
Path Siting Criteria for NAAQS
pollutants.
In setting the NAAQS, the EPA may
not consider the costs of implementing
the standards. This was confirmed by
the Supreme Court in Whitman v.
American Trucking Associations, 531
U.S. 457, 465–472, 475–76 (2001), as
discussed in section II.A of this
document. As has traditionally been
E:\FR\FM\06MRR3.SGM
06MRR3
16206
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
done in NAAQS rulemaking, the EPA
prepared a Regulatory Impact Analysis
(RIA) to provide the public with
information on the potential costs and
benefits of attaining several alternative
PM2.5 standard levels. In NAAQS
rulemaking, the RIA is done for
informational purposes only, and the
final decisions on the NAAQS in this
rulemaking are not based on
consideration of the information or
analyses in the RIA. The RIA fulfills the
requirements of Executive Orders
14094, 13563, and 12866. The RIA
estimates the costs and monetized
human health benefits of attaining the
revised and two alternative annual
PM2.5 standard levels and one
alternative 24-hour PM2.5 standard level.
Specifically, the RIA examines the
revised annual standard level of 9.0 mg/
m3 in combination with the current 24hour standard of 35 mg/m3 (i.e., 9.0/35
mg/m3), as well as the following less and
more stringent alternative standard
levels: (1) An alternative annual
standard level of 10.0 mg/m3 in
combination with the current 24-hour
standard (i.e., 10.0/35 mg/m3), (2) an
alternative annual standard level of 8.0
mg/m3 in combination with the current
24-hour standard (i.e., 8.0/35 mg/m3),
and (3) an alternative 24-hour standard
level of 30 mg/m3 in combination with
an alternative annual standard level of
10 mg/m3 (i.e., 10.0/30 mg/m3). The RIA
presents estimates of the costs and
benefits of applying illustrative national
control strategies in 2032 after
implementing existing and expected
regulations and assessing emissions
reductions to meet the current annual
and 24-hour particulate matter NAAQS
(12.0/35 mg/m3).
ddrumheller on DSK120RN23PROD with RULES3
I. Background
A. Legislative Requirements
Two sections of the Clean Air Act
(CAA) govern the establishment and
revision of the NAAQS. Section 108 (42
U.S.C. 7408) directs the Administrator
to identify and list certain air pollutants
and then to issue air quality criteria for
those pollutants. The Administrator is
to list those pollutants ‘‘emissions of
which, in his judgment, cause or
contribute to air pollution which may
reasonably be anticipated to endanger
public health or welfare’’; ‘‘the presence
of which in the ambient air results from
numerous or diverse mobile or
stationary sources’’; and for which he
‘‘plans to issue air quality
criteria. . . .’’ (42 U.S.C. 7408(a)(1)).
Air quality criteria are intended to
‘‘accurately reflect the latest scientific
knowledge useful in indicating the kind
and extent of all identifiable effects on
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
public health or welfare which may be
expected from the presence of [a]
pollutant in the ambient air. . . .’’ (42
U.S.C. 7408(a)(2)).
Section 109 [42 U.S.C. 7409] directs
the Administrator to propose and
promulgate ‘‘primary’’ and ‘‘secondary’’
NAAQS for pollutants for which air
quality criteria are issued [42 U.S.C.
7409(a)]. Section 109(b)(1) defines
primary standards as ones ‘‘the
attainment and maintenance of which in
the judgment of the Administrator,
based on such criteria and allowing an
adequate margin of safety, are requisite
to protect the public health.’’ 5 Under
section 109(b)(2), a secondary standard
must ‘‘specify a level of air quality the
attainment and maintenance of which,
in the judgment of the Administrator,
based on such criteria, is requisite to
protect the public welfare from any
known or anticipated adverse effects
associated with the presence of [the]
pollutant in the ambient air.’’ 6
In setting primary and secondary
standards that are ‘‘requisite’’ to protect
public health and welfare, respectively,
as provided in section 109(b), the EPA’s
task is to establish standards that are
neither more nor less stringent than
necessary. In so doing, the EPA may not
consider the costs of implementing the
standards. See generally Whitman v.
American Trucking Associations, 531
U.S. 457, 465–472, 475–76 (2001).
Likewise, ‘‘[a]ttainability and
technological feasibility are not relevant
considerations in the promulgation of
national ambient air quality standards.’’
American Petroleum Institute v. Costle,
665 F.2d 1176, 1185 (D.C. Cir. 1981);
accord Murray Energy Corporation v.
EPA, 936 F.3d 597, 623–24 (D.C. Cir.
2019).
The requirement that primary
standards provide an adequate margin
of safety was intended to address
uncertainties associated with
inconclusive scientific and technical
information available at the time of
standard setting. It was also intended to
provide a reasonable degree of
protection against hazards that research
5 The legislative history of section 109 indicates
that a primary standard is to be set at ‘‘the
maximum permissible ambient air level . . . which
will protect the health of any [sensitive] group of
the population,’’ and that for this purpose
‘‘reference should be made to a representative
sample of persons comprising the sensitive group
rather than to a single person in such a group.’’ S.
Rep. No. 91–1196, 91st Cong., 2d Sess. 10 (1970).
6 Under CAA section 302(h) (42 U.S.C. 7602(h)),
effects on welfare include, but are not limited to,
‘‘effects on soils, water, crops, vegetation, manmade
materials, animals, wildlife, weather, visibility, and
climate, damage to and deterioration of property,
and hazards to transportation, as well as effects on
economic values and on personal comfort and wellbeing.’’
PO 00000
Frm 00006
Fmt 4701
Sfmt 4700
has not yet identified. See Lead
Industries Association v. EPA, 647 F.2d
1130, 1154 (D.C. Cir. 1980); American
Petroleum Institute v. Costle, 665 F.2d at
1186; Coalition of Battery Recyclers
Ass’n v. EPA, 604 F.3d 613, 617–18
(D.C. Cir. 2010); Mississippi v. EPA, 744
F.3d 1334, 1353 (D.C. Cir. 2013). Both
kinds of uncertainties are components
of the risk associated with pollution at
levels below those at which human
health effects can be said to occur with
reasonable scientific certainty. Thus, in
selecting primary standards that include
an adequate margin of safety, the
Administrator is seeking not only to
prevent pollution levels that have been
demonstrated to be harmful but also to
prevent lower pollutant levels that may
pose an unacceptable risk of harm, even
if the risk is not precisely identified as
to nature or degree. The CAA does not
require the Administrator to establish a
primary NAAQS at a zero-risk level or
at background concentration levels, see
Lead Industries Ass’n v. EPA, 647 F.2d
at 1156 n.51, Mississippi v. EPA, 744
F.3d at 1351, but rather at a level that
reduces risk sufficiently so as to protect
public health with an adequate margin
of safety.
In addressing the requirement for an
adequate margin of safety, the EPA
considers such factors as the nature and
severity of the health effects involved,
the size of the sensitive population(s),
and the kind and degree of
uncertainties. The selection of any
particular approach to providing an
adequate margin of safety is a policy
choice left specifically to the
Administrator’s judgment. See Lead
Industries Ass’n v. EPA, 647 F.2d at
1161–62; Mississippi v. EPA, 744 F.3d at
1353.
Section 109(d)(1) of the Act requires
the review every five years of existing
air quality criteria and, if appropriate,
the revision of those criteria to reflect
advances in scientific knowledge on the
effects of the pollutant on public health
and welfare. Under the same provision,
the EPA is also to review every five
years and, if appropriate, revise the
NAAQS, based on the revised air quality
criteria. Section 109(d)(1) also provides
that the Administrator may review and
revise criteria or promulgate new
standards earlier or more frequently.
Section 109(d)(2) addresses the
appointment and advisory functions of
an independent scientific review
committee. Section 109(d)(2)(A)
requires the Administrator to appoint
this committee, which is to be
composed of ‘‘seven members including
at least one member of the National
Academy of Sciences, one physician,
and one person representing State air
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
pollution control agencies.’’ Section
109(d)(2)(B) provides that the
independent scientific review
committee ‘‘shall complete a review of
the criteria . . . and the national
primary and secondary ambient air
quality standards . . . and shall
recommend to the Administrator any
new . . . standards and revisions of
existing criteria and standards as may be
appropriate. . . .’’ Since the early
1980s, this independent review function
has been performed by the Clean Air
Scientific Advisory Committee (CASAC)
of the EPA’s Science Advisory Board.
As previously noted, the Supreme
Court has held that section 109(b)
‘‘unambiguously bars cost
considerations from the NAAQS-setting
process.’’ Whitman v. Am. Trucking
Associations, 531 U.S. 457, 471 (2001).
Accordingly, while some of these issues
regarding which Congress has directed
the CASAC to advise the Administrator
are ones that are relevant to the standard
setting process, others are not. Issues
that are not relevant to standard setting
may be relevant to implementation of
the NAAQS once they are established.
ddrumheller on DSK120RN23PROD with RULES3
B. Related PM Control Programs
States are primarily responsible for
ensuring attainment and maintenance of
ambient air quality standards once the
EPA has established them. Under
section 110, Part C, and Part D, Subparts
1 and 4 of the CAA, and related
provisions and regulations, States are to
submit, for the EPA’s approval, State
implementation plans (SIPs) that
provide for the attainment and
maintenance of the NAAQS for PM
through control programs directed to
sources of the pollutants involved. The
States, in conjunction with the EPA,
also administer the prevention of
significant deterioration of air quality
program that covers these pollutants
(see 42 U.S.C. 7470–7479). In addition,
Federal programs provide for or result
in nationwide reductions in emissions
of PM and its precursors under Title II
of the Act, 42 U.S.C. 7521–7574, which
involves controls for motor vehicles and
nonroad engines and equipment; the
new source performance standards
under section 111 of the Act, 42 U.S.C.
7411; and the national emissions
standards for hazardous pollutants
under section 112 of the Act, 42 U.S.C.
7412.
C. Review of the Air Quality Criteria and
Standards for Particulate Matter
1. Reviews Completed in 1971 and 1987
The EPA first established NAAQS for
PM in 1971 (36 FR 8186, April 30,
1971), based on the original Air Quality
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
Criteria Document (AQCD) (DHEW,
1969).7 The Federal reference method
(FRM) specified for determining
attainment of the original standards was
the high-volume sampler, which
collects PM up to a nominal size of 25
to 45 mm (referred to as total suspended
particulates or TSP). The primary
standards were set at 260 mg/m3, 24hour average, not to be exceeded more
than once per year, and 75 mg/m3,
annual geometric mean. The secondary
standards were set at 150 mg/m3, 24hour average, not to be exceeded more
than once per year, and 60 mg/m3,
annual geometric mean.
In October 1979 (44 FR 56730,
October 2, 1979), the EPA announced
the first periodic review of the air
quality criteria and NAAQS for PM.
Revised primary and secondary
standards were promulgated in 1987 (52
FR 24634, July 1, 1987). In the 1987
decision, the EPA changed the indicator
for particles from TSP to PM10, in order
to focus on the subset of inhalable
particles small enough to penetrate to
the thoracic region of the respiratory
tract (including the tracheobronchial
and alveolar regions), referred to as
thoracic particles.8 The level of the 24hour standards (primary and secondary)
was set at 150 mg/m3, and the form was
one expected exceedance per year, on
average over three years. The level of
the annual standards (primary and
secondary) was set at 50 mg/m3, and the
form was the annual arithmetic mean,
averaged over three years.
2. Review Completed in 1997
In April 1994, the EPA announced its
plans for the second periodic review of
the air quality criteria and NAAQS for
PM, and in 1997 the EPA promulgated
revisions to the NAAQS (62 FR 38652,
July 18, 1997). In the 1997 decision, the
EPA determined that the fine and coarse
fractions of PM10 should be considered
separately. This determination was
based on evidence that serious health
effects were associated with short- and
long-term exposures to fine particles in
areas that met the existing PM10
standards. The EPA added new
standards, using PM2.5 as the indicator
for fine particles (with PM2.5 referring to
particles with a nominal mean
aerodynamic diameter less than or equal
to 2.5 mm). The new primary standards
7 Prior to the review initiated in 2007 (see below),
the AQCD provided the scientific foundation (i.e.,
the air quality criteria) for the NAAQS. Beginning
in that review, the Integrated Science Assessment
(ISA) has replaced the AQCD.
8 PM
10 refers to particles with a nominal mean
aerodynamic diameter less than or equal to 10 mm.
More specifically, 10 mm is the aerodynamic
diameter for which the efficiency of particle
collection is 50 percent.
PO 00000
Frm 00007
Fmt 4701
Sfmt 4700
16207
were as follows: (1) An annual standard
with a level of 15.0 mg/m3, based on the
3-year average of annual arithmetic
mean PM2.5 concentrations from single
or multiple community-oriented
monitors; 9 and (2) a 24-hour standard
with a level of 65 mg/m3, based on the
3-year average of the 98th percentile of
24-hour PM2.5 concentrations at each
monitor within an area. Also, the EPA
established a new reference method for
the measurement of PM2.5 in the
ambient air and adopted rules for
determining attainment of the new
standards. To continue to address the
health effects of the coarse fraction of
PM10 (referred to as thoracic coarse
particles or PM10–2.5, generally including
particles with a nominal mean
aerodynamic diameter greater than 2.5
mm and less than or equal to 10 mm), the
EPA retained the primary annual PM10
standard and revised the form of the
primary 24-hour PM10 standard to be
based on the 99th percentile of 24-hour
PM10 concentrations at each monitor in
an area. The EPA revised the secondary
standards by setting them equal in all
respects to the primary standards.
Following promulgation of the 1997
PM NAAQS, petitions for review were
filed by several parties, addressing a
broad range of issues. In May 1999, the
U.S. Court of Appeals for the District of
Columbia Circuit (D.C. Circuit) upheld
the EPA’s decision to establish fine
particle standards and to regulate coarse
particle pollution, but vacated the 1997
PM10 standards, concluding that the
EPA had not provided a reasonable
explanation justifying use of PM10 as an
indicator for coarse particles. American
Trucking Associations, Inc. v. EPA, 175
F. 3d 1027 (D.C. Cir. 1999). Pursuant to
the D.C. Circuit’s decision, the EPA
removed the vacated 1997 PM10
standards, and the pre-existing 1987
PM10 standards remained in place (65
FR 80776, December 22, 2000). The D.C.
Circuit also upheld the EPA’s
determination not to establish more
stringent secondary standards for fine
particles to address effects on visibility.
American Trucking Associations v.
EPA, 175 F. 3d at 1027.
9 The 1997 annual PM
2.5 standard was compared
with measurements made at the communityoriented monitoring site recording the highest
concentration or, if specific constraints were met,
measurements from multiple community-oriented
monitoring sites could be averaged (i.e., ‘‘spatial
averaging’’). In the last review (completed in 2012)
the EPA replaced the term ‘‘community-oriented’’
monitor with the term ‘‘area-wide’’ monitor. Areawide monitors are those sited at the neighborhood
scale or larger, as well as those monitors sited at
micro- or middle-scales that are representative of
many such locations in the same core-based
statistical area (CBSA) (78 FR 3236, January 15,
2013).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16208
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
The D.C. Circuit also addressed more
general issues related to the NAAQS,
including issues related to the
consideration of costs in setting NAAQS
and the EPA’s approach to establishing
the levels of NAAQS. Regarding the cost
issue, the court reaffirmed prior rulings
holding that in setting NAAQS the EPA
is ‘‘not permitted to consider the cost of
implementing those standards.’’
American Trucking Associations v.
EPA, 175 F. 3d at 1040–41. Regarding
the levels of NAAQS, the court held that
the EPA’s approach to establishing the
level of the standards in 1997 (i.e., both
for PM and for the ozone NAAQS
promulgated on the same day) effected
‘‘an unconstitutional delegation of
legislative authority.’’ American
Trucking Associations v. EPA, 175 F. 3d
at 1034–40. Although the court stated
that ‘‘the factors EPA uses in
determining the degree of public health
concern associated with different levels
of ozone and PM are reasonable,’’ it
remanded the rule to the EPA, stating
that when the EPA considers these
factors for potential non-threshold
pollutants ‘‘what EPA lacks is any
determinate criterion for drawing lines’’
to determine where the standards
should be set.
The D.C. Circuit’s holding on the cost
and constitutional issues were appealed
to the United States Supreme Court. In
February 2001, the Supreme Court
issued a unanimous decision upholding
the EPA’s position on both the cost and
constitutional issues. Whitman v.
American Trucking Associations, 531
U.S. 457, 464, 475–76. On the
constitutional issue, the Court held that
the statutory requirement that NAAQS
be ‘‘requisite’’ to protect public health
with an adequate margin of safety
sufficiently guided the EPA’s discretion,
affirming the EPA’s approach of setting
standards that are neither more nor less
stringent than necessary.
The Supreme Court remanded the
case to the D.C. Circuit for resolution of
any remaining issues that had not been
addressed in that court’s earlier rulings.
Id. at 475–76. In a March 2002 decision,
the D.C. Circuit rejected all remaining
challenges to the standards, holding that
the EPA’s PM2.5 standards were
reasonably supported by the
administrative record and were not
‘‘arbitrary and capricious.’’ American
Trucking Associations v. EPA, 283 F. 3d
355, 369–72 (D.C. Cir. 2002).
several drafts, the EPA’s National Center
for Environmental Assessment (NCEA)
finalized the AQCD in October 2004
(U.S. EPA, 2004a). The EPA’s Office of
Air Quality Planning and Standards
(OAQPS) finalized a Risk Assessment
and Staff Paper in December 2005 (Abt
Associates, 2005; U.S. EPA, 2005).10 On
December 20, 2005, the EPA announced
its proposed decision to revise the
NAAQS for PM and solicited public
comment on a broad range of options
(71 FR 2620, January 17, 2006). On
September 21, 2006, the EPA
announced its final decisions to revise
the primary and secondary NAAQS for
PM to provide increased protection of
public health and welfare, respectively
(71 FR 61144, October 17, 2006). With
regard to the primary and secondary
standards for fine particles, the EPA
revised the level of the 24-hour PM2.5
standards to 35 mg/m3, retained the level
of the annual PM2.5 standards at 15.0 mg/
m3, and revised the form of the annual
PM2.5 standards by narrowing the
constraints on the optional use of spatial
averaging. With regard to the primary
and secondary standards for PM10, the
EPA retained the 24-hour standards,
with levels at 150 mg/m3, and revoked
the annual standards. The thenAdministrator judged that the available
evidence generally did not suggest a
link between long-term exposure to
existing ambient levels of coarse
particles and health or welfare effects.
In addition, a new reference method
was added for the measurement of
PM10–2.5 in the ambient air in order to
provide a basis for approving Federal
Equivalent Methods (FEMs) and to
promote the gathering of scientific data
to support future reviews of the PM
NAAQS.
Several parties filed petitions for
review following promulgation of the
revised PM NAAQS in 2006. On
February 24, 2009, the D.C. Circuit
issued its opinion in the case American
Farm Bureau Federation v. EPA, 559 F.
3d 512 (D.C. Cir. 2009). The court
remanded the primary annual PM2.5
NAAQS to the EPA because the Agency
had failed to adequately explain why
the standards provided the requisite
protection from both short- and longterm exposures to fine particles,
including protection for at-risk
populations. Id. at 520–27. With regard
to the standards for PM10, the court
upheld the EPA’s decisions to retain the
3. Review Completed in 2006
In October 1997, the EPA published
its plans for the third periodic review of
the air quality criteria and NAAQS for
PM (62 FR 55201, October 23, 1997).
After the CASAC and public review of
10 Prior to the review initiated in 2007, the Staff
Paper presented the EPA staff’s considerations and
conclusions regarding the adequacy of existing
NAAQS and, when appropriate, the potential
alternative standards that could be supported by the
evidence and information. More recent reviews
present this information in the Policy Assessment.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
PO 00000
Frm 00008
Fmt 4701
Sfmt 4700
24-hour PM10 standard to provide
protection from thoracic coarse particle
exposures and to revoke the annual
PM10 standard. Id. at 533–38. With
regard to the secondary PM2.5 standards,
the court remanded the standards to the
EPA because the Agency failed to
adequately explain why setting the
secondary PM standards identical to the
primary standards provided the
required protection for public welfare,
including protection from visibility
impairment. Id. at 528–32. The EPA
responded to the court’s remands as part
of the next review of the PM NAAQS,
which was initiated in 2007 (discussed
below).
4. Review Completed in 2012
In June 2007, the EPA initiated the
fourth periodic review of the air quality
criteria and the PM NAAQS by issuing
a call for information (72 FR 35462, June
28, 2007). Based on the NAAQS review
process, as revised in 2008 and again in
2009,11 the EPA held science/policy
issue workshops on the primary and
secondary PM NAAQS (72 FR 34003,
June 20, 2007; 72 FR 34005, June 20,
2007), and prepared and released the
planning and assessment documents
that comprise the review process (i.e.,
Integrated Review Plan, (IRP; U.S. EPA,
2008), Integrated Science Assessment
(ISA; U.S. EPA, 2009a), Risk and
Exposure Assessment (REA) planning
documents for health and welfare (U.S.
EPA, 2009b, U.S. EPA, 2009c), a
quantitative health risk assessment (U.S.
EPA, 2010a) and an urban-focused
visibility assessment (U.S. EPA, 2010b),
and a Policy Assessment (PA; U.S. EPA,
2011). In June 2012, the EPA announced
its proposed decision to revise the
NAAQS for PM (77 FR 38890, June 29,
2012).
In December 2012, the EPA
announced its final decisions to revise
the primary NAAQS for PM to provide
increased protection of public health (78
FR 3086, January 15, 2013). With regard
to primary standards for PM2.5, the EPA
revised the level of the annual PM2.5
standard 12 to 12.0 mg/m3 and retained
the 24-hour PM2.5 standard, with its
level of 35 mg/m3. For the primary PM10
standard, the EPA retained the 24-hour
standard to continue to provide
protection against effects associated
with short-term exposure to thoracic
coarse particles (i.e., PM10–2.5). With
regard to the secondary PM standards,
the EPA generally retained the 24-hour
11 The history of the NAAQS review process,
including revisions to the process, is discussed at
https://www.epa.gov/naaqs/historical-informationnaaqs-review-process.
12 The EPA also eliminated the option for spatial
averaging.
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
5. Review Initiated in 2014
In December 2014, the EPA
announced the initiation of the current
periodic review of the air quality criteria
for PM and of the PM2.5 and PM10
NAAQS and issued a call for
information (79 FR 71764, December 3,
2014). On February 9 to 11, 2015, the
EPA’s NCEA and OAQPS held a public
workshop to inform the planning for the
review of the PM NAAQS (announced
in 79 FR 71764, December 3, 2014).
Workshop participants, including a
wide range of external experts as well as
the EPA staff representing a variety of
areas of expertise (e.g., epidemiology,
human and animal toxicology, risk/
exposure analysis, atmospheric science,
visibility impairment, climate effects),
were asked to highlight significant new
and emerging PM research, and to make
recommendations to the Agency
regarding the design and scope of the
review. This workshop provided for a
public discussion of the key science and
policy-relevant issues around which the
EPA structured the review of the PM
NAAQS and of the most meaningful
new scientific information that would
be available in the review to inform
understanding of these issues.
The input received at the workshop
guided the EPA staff in developing a
draft IRP, which was reviewed by the
CASAC Particulate Matter Panel and
discussed on public teleconferences
held in May 2016 (81 FR 13362, March
14, 2016) and August 2016 (81 FR
39043, June 15, 2016). Advice from the
CASAC, supplemented by the
Particulate Matter Panel, and input from
the public were considered in
developing the final IRP (U.S. EPA,
2016). The final IRP discusses the
approaches to be taken in developing
key scientific, technical, and policy
documents in the review and the key
policy-relevant issues that frame the
EPA’s consideration of whether the
primary and/or secondary NAAQS for
PM should be retained or revised.
In May 2018, the then-Administrator
issued a memorandum announcing the
Agency’s intention to conduct the
review of the PM NAAQS in such a
manner as to ensure that any necessary
revisions were finalized by December
2020 (Pruitt, 2018). Following this
memo, on October 10, 2018, the thenAdministrator additionally announced
that the role of reviewing the key
assessments developed as part of the
ongoing review of the PM NAAQS (i.e.,
drafts of the ISA and PA) would be
performed by the seven-member
chartered CASAC (i.e., rather than the
CASAC Particulate Matter Panel that
reviewed the draft IRP).14
The EPA released the draft ISA in
October 2018 (83 FR 53471, October 23,
2018). The draft ISA was reviewed by
the chartered CASAC at a public
meeting held in Arlington, VA in
December 2018 (83 FR 55529, November
6, 2018) and was discussed on a public
teleconference in March 2019 (84 FR
8523, March 8, 2019). The CASAC
provided its advice on the draft ISA in
a letter to the then-Administrator dated
April 11, 2019 (Cox, 2019a). The EPA
addressed these comments in the final
ISA, which was released in December
2019 (U.S. EPA, 2019a).
The EPA released the draft PA in
September 2019 (84 FR 47944,
September 11, 2019). The draft PA was
reviewed by the chartered CASAC and
discussed in October 2019 at a public
meeting held in Cary, NC. Public
comments were received via a separate
public teleconference (84 FR 51555,
September 30, 2019). A public meeting
to discuss the chartered CASAC letter
and response to charge questions on the
draft PA was held in Cary, NC, in
October 2019 (84 FR 51555, September
30, 2019), and the CASAC provided its
advice on the draft PA, including its
advice on the current primary and
secondary PM standards, in a letter to
the then-Administrator dated December
16, 2019 (Cox, 2019b). With regard to
the primary standards, the CASAC
recommended retaining the current 24hour PM2.5 and PM10 standards but did
not reach consensus on the adequacy of
the current annual PM2.5 standard. Some
CASAC members expressed support for
retaining the current primary annual
PM2.5 standard while other members
expressed support for revising that
standard in order to increase public
health protection (Cox, 2019b, p. 1 of
letter). These views are described in
13 Consistent with the primary standard, the EPA
eliminated the option for spatial averaging with the
annual standard.
14 Announcement available at: https://
www.regulations.gov/document/EPA-HQ-OAR2015-0072-0223.
ddrumheller on DSK120RN23PROD with RULES3
and annual PM2.5 standards 13 and the
24-hour PM10 standard to address
visibility and non-visibility welfare
effects.
As with previous reviews, petitioners
challenged the EPA’s final rule.
Petitioners argued that the EPA acted
unreasonably in revising the level and
form of the annual standard and in
amending the monitoring network
provisions. On judicial review, the
revised standards and monitoring
requirements were upheld in all
respects. NAM v. EPA, 750 F.3d 921
(D.C. Cir. 2014).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
PO 00000
Frm 00009
Fmt 4701
Sfmt 4700
16209
greater detail in the letter to the thenAdministrator (Cox, 2019b) and in the
notice of final rulemaking (85 FR
82706–82707, December 18, 2020), as
well as below. With regard to the
secondary standards, the CASAC
recommended retaining the current
standards. In response to the CASAC’s
comments, the 2020 final PA
incorporated a number of changes (Cox,
2019b, U.S. EPA, 2020b), as described in
detail in section I.C.5 of the 2020
proposal document (85 FR 24100, April
30, 2020).
a. 2020 Proposed and Final Actions
On April 14, 2020, the EPA proposed
to retain all of the primary and
secondary PM standards, without
revision. These proposed decisions were
published in the Federal Register on
April 30, 2020 (85 FR 24094, April 30,
2020). The EPA’s final decision on the
PM NAAQS was published in the
Federal Register on December 18, 2020
(85 FR 82684, December 18, 2020). In
the 2020 rulemaking, the EPA retained
the primary and secondary PM2.5 and
PM10 standards, without revision. The
then-Administrator’s rationale for his
decisions is described in more detail in
section II, III, and V below, and is
briefly summarized here.
In reaching his final decision to retain
the primary annual and 24-hour PM2.5
standards, the then-Administrator
considered the available scientific
evidence, quantitative information,
CASAC advice, and public comments in
his supporting rationale in the 2020
final action (85 FR 82714, December 18,
2020). In so doing, he concluded that
the available controlled human
exposure studies did not provide
support for additional public health
protection against exposures to peak
PM2.5 concentrations, beyond the
protection provided by the combination
of the current primary annual and 24hour PM2.5 standards. He also noted that
the available epidemiologic studies did
not indicate that associations in those
studies are strongly influenced by
exposures to peak concentrations in the
air quality distribution and thus did not
indicate the need for additional
protection against short-term exposures
to peak PM2.5 concentrations.
Accordingly, and taking into account
consensus CASAC advice to retain the
current primary 24-hour PM2.5 standard,
the then-Administrator concluded the
primary 24-hour PM2.5 standard should
be retained.
With respect to the annual PM2.5
standard, the then-Administrator
recognized that important uncertainties
and limitations that were present in
epidemiologic studies in previous
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16210
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
reviews remained in the evidence
assessed in the 2019 ISA. In considering
the epidemiologic evidence, the thenAdministrator noted that: (1) The
reported mean concentration in the
majority of the key U.S. epidemiologic
studies using ground-based monitoring
data are above the level of the current
annual standard; (2) the mean of the
reported study means (or medians) (i.e.,
13.5 mg/m3) is above the level of the
current primary annual PM2.5 standard
of 12 mg/m3; (3) air quality analyses
show the study means to be lower than
their corresponding design by 10–20%;
and (4) that these analyses must be
considered in light of uncertainties
inherent in the epidemiologic evidence.
The then-Administrator further
considered other available information,
including the risk assessment,
accountability studies, and controlled
human exposure studies, and found
that, in considering all of the evidence
together along with advice from the
CASAC, the suite of primary PM2.5
standards were requisite to protect
public health with an adequate margin
of safety, and should be retained,
without revision.
With regard to the primary PM10
standard, the then-Administrator noted
that the expanded body of evidence has
broadened the range of effects that have
been linked with PM10–2.5 exposures. In
light of that information, as well as
continued uncertainties in the evidence
and advice from the CASAC to retain
the standard, the then-Administrator
judged it appropriate to retain the
primary PM10 standard to provide the
requisite degree of public health
protection against PM10–2.5 exposures,
regardless of location, source of origin,
or particle composition (85 FR 82725,
December 18, 2020).
With regard to the secondary PM
standards, the then-Administrator
concluded that there was insufficient
information available to establish any
distinct secondary PM standards to
address climate and materials effects of
PM. For visibility effects, he found that
in the absence of a monitoring network
for direct measurement of light
extinction, a calculated light extinction
indicator that utilizes the IMPROVE
algorithms continued to provide a
reasonable basis for defining a target
level of protection against PM-related
visibility impairment. He further found
that a visibility index with a 24-hour
averaging time was reasonable based on
its stability and suitability for
representing subdaily periods, and a
form based on the 3-year average of
annual 90th percentile values was
reasonable based on its stability and that
it represents the median of the 20
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
percent worst visibility days which are
targeted under the Regional Haze
program. With regard to the level of a
visibility index, the then-Administrator
judged it appropriate to establish a
target level of protection of 30 dv,
reflecting the upper end of the range of
visibility impairment judged to be
acceptable by at least 50% of study
participants in the available public
preference studies, taking into
consideration the variability, limitations
and uncertainties of the public
preference studies. The thenAdministrator judged that the secondary
24-hour PM2.5 standard with its level of
35 mg/m3 would provide at least the
target level of protection for visual air
quality of 30 dv which he judged
appropriate. Accordingly, taking into
consideration the advice of the CASAC
to retain the current secondary PM
standards, the then-Administrator found
the current secondary standards provide
the requisite degree of protection and
that they should be retained (85 FR
82742, December 18, 2020).
Following publication of the 2020
final action, several parties filed
petitions for review and petitions for
reconsideration of the EPA’s final
decision. The petitions for review were
filed in the D.C. Circuit and the Court
consolidated the cases.15 Following
EPA’s decision to reconsider the 2020
final decision, the Court ordered the
consolidated cases to be held in
abeyance.
b. Reconsideration of the 2020 PM
NAAQS Final Action
Executive Order 13990 directed
review of certain agency actions (86 FR
7037, January 25, 2021).16 An
accompanying fact sheet provided a
non-exclusive list of agency actions that
agency heads should review in
accordance with that order, including
the 2020 Particulate Matter NAAQS
Decision.17
On June 10, 2021, the Agency
announced its decision to reconsider the
2020 PM NAAQS final action because
the available scientific evidence and
technical information indicate that the
current standards may not be adequate
to protect public health and welfare, as
required by the Clean Air Act.18 The
15 See California v. EPA, (D.C. Cir., No. 21–2014
consolidated with Nos. 21–1027, 21–1054).
16 See https://www.whitehouse.gov/briefing-room/
presidential-actions/2021/01/20/executive-orderprotecting-public-health-and-environment-andrestoring-science-to-tackle-climate-crisis/.
17 See https://www.whitehouse.gov/briefing-room/
statements-releases/2021/01/20/fact-sheet-list-ofagency-actions-for-review/.
18 The press release for this announcement is
available at: https://www.epa.gov/newsreleases/epa-
PO 00000
Frm 00010
Fmt 4701
Sfmt 4700
Administrator reached this decision in
part based on the fact that the EPA
noted that the 2020 PA concluded that
the scientific evidence and information
called into question the adequacy of the
primary annual PM2.5 standard and
supported revising the level to below
the current level of 12.0 mg/m3 while
retaining the primary 24-hour PM2.5
standard (U.S. EPA, 2020b). The EPA
also noted that the 2020 PA concluded
that the available scientific evidence
and information supported retaining the
primary PM10 standard and secondary
PM standards without revision (U.S.
EPA, 2020b).
The EPA staff conclusions detailed in
the 2020 PA in combination with the
CASAC advice that informed the
Administrator’s decisions regarding the
2020 final action, studies highlighted by
public comments on the 2020 proposal,
and the numerous studies published
since the literature cutoff date of the
2019 ISA all informed the scope of the
reconsideration.
In its review of the 2019 draft PA,
some members of the CASAC had
recommended that greater attention
should be given to accountability
studies and epidemiologic studies that
employ alternative methods for
confounder control (also referred to as
causal inference or causal modeling
studies) in order to ‘‘more fully account
for effects of confounding, measurement
and estimation errors, model
uncertainty, and heterogeneity’’ in
epidemiologic studies (Cox, 2019b, p. 8
of consensus responses). In addition,
public commenters submitted a number
of recent studies published after the
literature cutoff date for the 2019 ISA
that would have been considered within
the scope of the 2019 ISA. While the
EPA provisionally considered these
studies in responding to public
comments,19 it was determined that, at
the time of the 2020 final action, these
studies were generally consistent with
the evidence assessed in the 2019 ISA
(85 FR 82690, December 18, 2020; U.S.
EPA, 2020a). As such, and consistent
with previous NAAQS reviews, the EPA
concluded that the new studies did not
materially change any of the broad
scientific conclusions regarding the
health and welfare effects of PM in
ambient air made in the air quality
criteria, and therefore, reopening of the
air quality criteria was not warranted
(85 FR 82691, December 18, 2020).
However, at that time, the EPA
reexamine-health-standards-harmful-soot-previousadministration-left-unchanged.
19 The list of provisionally considered studies is
included in Appendix A to the 2020 Response to
Comments document (U.S. EPA, 2020a).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
recognized that its ‘‘provisional
consideration of these studies did not
and could not provide the kind of indepth critical review’’ (85 FR 82690,
December 18, 2020) that studies
undergo in the development of an ISA.
In preparing to reconsider the 2020
final decision for the PM NAAQS, the
Agency revisited the need to reopen the
air quality criteria, given the amount of
time that had passed since the literature
cutoff date of the 2019 ISA (i.e.,
approximately January 2018) and the
volume of literature that had become
available, including those studies
provisionally considered in responding
to comments in 2020. In so doing, the
EPA preliminarily concluded that at
least some of these studies were likely
to be relevant to its reconsideration of
the air quality criteria and the PM
NAAQS and that, in considering public
comments on any proposed decisions
for the reconsideration, these studies
were likely to be raised by public
commenters and would potentially
warrant a reopening of the air quality
criteria. For example, on February 16,
2021, the EPA received two petitions to
reconsider the PM NAAQS. One
petition objected to the EPA’s
provisional consideration of studies
submitted in public comments on the
2020 proposal and suggested that the
provisional consideration was
inadequate because the studies could be
important in determining whether the
existing standards are adequately
protective. See, Petition for
Reconsideration of National Ambient
Air Quality Standards for Particulate
Matter, submitted by American Lung
Association, et al, dated Feb. 16, 2020.
The other petition identified a number
of new studies, including one
epidemiologic study that was published
after the provisional consideration was
completed that could further inform the
concern expressed by the CASAC that
associations reported in epidemiologic
studies do not adequately account for
‘‘uncontrolled confounding and other
potential sources of error and bias.’’ See
Petition for Reconsideration of ‘‘Review
of the National Ambient Air Quality
Standards for Particulate Matter,’’
submitted by the State of California,
dated Feb. 16, 2020. This was also an
uncertainty noted by the thenAdministrator in the 2020 decision, who
also recognized ‘‘that methodological
study designs to address confounding,
such as causal inference methods, are an
emerging field of study.’’ Thus, the
Agency concluded it was appropriate to
reconsider not only the standards but
also the air quality criteria, in light of
public comments during the 2020 PM
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
NAAQS proposal and recent studies
published since the cutoff date of the
2019 ISA, as reflected in petitions. In
deciding to reopen the air quality
criteria, the Agency concluded it was
reasonable to focus on studies that were
most likely to inform decisions on the
appropriate standard, but not to reassess
areas which, based on the assessment of
available science published since the
cutoff date of the 2019 ISA and through
2021, were judged unlikely to have new
information that would be useful for the
Administrator’s decision making. The
Agency accordingly announced that, in
support of the reconsideration, it would
develop a supplement to the 2019 ISA
and a revised PA.
The EPA also explained that the draft
ISA Supplement and draft PA would be
reviewed at a public meeting by the
CASAC, and the public would have
opportunities to comment on these
documents during the CASAC review
process, as well as to provide input
during the rulemaking through the
public comment process and public
hearings on the proposed rulemaking.
On March 31, 2021, the Administrator
announced his decision to reestablish
the membership of the CASAC to
‘‘ensure the agency received the best
possible scientific insight to support our
work to protect human health and the
environment.’’ 20 Consistent with this
memorandum, a call for nominations of
candidates to the EPA’s chartered
CASAC was published in the Federal
Register (86 FR 17146, April 1, 2021).
On June 17, 2021, the Administrator
announced his selection of the seven
members to serve on the chartered
CASAC.21 22 Additionally, a call for
nominations of candidates to a PMspecific panel was published in the
Federal Register (86 FR 33703, June 25,
2021). The members of the PM CASAC
panel were announced on August 30,
2021.23
The draft ISA Supplement was
released in September 2021 (U.S. EPA,
20 The press release for this announcement is
available at: https://www.epa.gov/newsreleases/
administrator-regan-directs-epa-reset-criticalscience-focused-federal-advisory.
21 The press release for this announcement is
available at: https://www.epa.gov/newsreleases/epaannounces-selections-charter-members-clean-airscientific-advisory-committee.
22 The list of members of the chartered CASAC
and their biosketches are available at: https://
casac.epa.gov/ords/sab/r/sab_apex/casac/
mems?p14_
committeeon=2021%20CASAC%20PM%20Panel
&session=17433386035954.
23 The list of members of the PM CASAC panel
and their biosketches are available at: https://
casac.epa.gov/ords/sab/
f?p=105:14:9979229564047:::14:P14_
COMMITTEEON:2021%20CASAC
%20PM%20Panel.
PO 00000
Frm 00011
Fmt 4701
Sfmt 4700
16211
2021a; 86 FR 54186, September 30,
2021), and included a discussion of the
rationale and scope of the Supplement.
As explained therein, the ISA
Supplement focuses on a thorough
evaluation of some studies that became
available after the literature cutoff date
of the 2019 ISA that could either further
inform the adequacy of the current PM
NAAQS or address key scientific topics
that have evolved since the literature
cutoff date for the 2019 ISA. In selecting
the health effects to evaluate within the
ISA Supplement, the EPA focused on
health effects for which the evidence
supported a ‘‘causal relationship’’
because those were the health effects
that were most useful in informing
conclusions in the 2020 PA (U.S. EPA,
2022a, section 1.2.1).24 Consistent with
the rationale for the focus on certain
health effects, in selecting the nonecological welfare effects to evaluate
within the ISA Supplement, the EPA
focused on the non-ecological welfare
effects for which the evidence
supported a ‘‘causal relationship’’ and
for which quantitative analyses could be
supported by the evidence because
those were the welfare effects that were
most useful in informing conclusions in
the 2020 PA.25 Specifically, for nonecological welfare effects, the focus
within the ISA Supplement is on
visibility effects. The ISA Supplement
also considers recent health effects
evidence that addresses key scientific
topics where the literature has evolved
since the 2020 review was completed,
24 As described in section 1.2.1 of the ISA
Supplement: ‘‘In considering the public health
protection provided by the current primary PM2.5
standards, and the protection that could be
provided by alternatives, [the U.S. EPA, within the
2020 PM PA] emphasized health outcomes for
which the ISA determined that the evidence
supports either a ‘causal’ or a ‘likely to be causal’
relationship with PM2.5 exposures’’ (U.S. EPA,
2020b). Although the 2020 PA initially focused on
this broader set of evidence, the basis of the
discussion on potential alternative standards
primarily focused on health effect categories where
the 2019 PM ISA concluded a ‘causal relationship’
(i.e., short- and long-term PM2.5 exposure and
cardiovascular effects and mortality) as reflected in
Figures 3–7 and 3–8 of the 2020 PA (U.S. EPA,
2020b).’’
25 As described in section 1.2.1 of the ISA
Supplement: ‘‘The 2019 PM ISA concluded a
‘causal relationship’ for each of the welfare effects
categories evaluated (i.e., visibility, climate effects
and materials effects). While the 2020 PA
considered the broader set of evidence for these
effects, for climate effects and material effects, it
concluded that there remained ‘substantial
uncertainties with regard to the quantitative
relationships with PM concentrations and
concentration patterns that limit[ed] [the] ability to
quantitatively assess the public welfare protection
provided by the standards from these effects’ (U.S.
EPA, 2020b).’’
E:\FR\FM\06MRR3.SGM
06MRR3
16212
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
specifically since the literature cutoff
date for the 2019 ISA.26
Building on the rationale presented in
section 1.2.1, the ISA Supplement
considers peer-reviewed studies
published from approximately January
2018 through March 2021 that meet the
following criteria:
• Health Effects
Æ U.S. and Canadian epidemiologic
studies for health effect categories
where the 2019 ISA concluded a
‘‘causal relationship’’ (i.e., short- and
long-term PM2.5 exposure and
cardiovascular effects and mortality).
D U.S. and Canadian epidemiologic
studies that employed alternative
methods for confounder control or
conducted accountability analyses (i.e.,
examined the effect of a policy on
reducing PM2.5 concentrations).
• Welfare Effects
Æ U.S. and Canadian studies that
provide new information on public
preferences for visibility impairment
and/or developed methodologies or
conducted quantitative analyses of light
extinction.
ddrumheller on DSK120RN23PROD with RULES3
• Key Scientific Topics
Æ Experimental studies (i.e.,
controlled human exposure and animal
toxicological) conducted at nearambient PM2.5 concentrations
experienced in the U.S.
Æ U.S.- and Canadian-based
epidemiologic studies that examined the
relationship between PM2.5 exposures
and severe acute respiratory syndrome
coronavirus 2 (SARS–CoV–2) infection
and coronavirus disease 2019 (COVID–
19) death.
Æ At-Risk Populations.
D U.S.- and Canadian-based
epidemiologic or exposure studies
examining potential disparities in either
PM2.5 exposures or the risk of health
effects by race/ethnicity or
socioeconomic status (SES).
Given the narrow scope of the ISA
Supplement, it is important to recognize
that the evaluation does not encompass
the full multidisciplinary evaluation
presented within the 2019 ISA that
would result in weight-of-evidence
26 These key scientific topics include
experimental studies conducted at near-ambient
concentrations, epidemiologic studies that
employed alternative methods for confounder
control or conducted accountability analyses,
studies that assess the relationship between PM2.5
exposure and severe acute respiratory syndrome
coronavirus 2 (SARS–CoV–2) infection and
coronavirus disease 2019 (COVID–19) death; and in
accordance with recent EPA goals on addressing
environmental justice, studies that examine
disparities in PM2.5 exposure and the risk of health
effects by race/ethnicity or socioeconomic status
(SES) (U.S. EPA, 2022a, section 1.2.1).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
conclusions on causality (i.e., causality
determinations). The ISA Supplement
critically evaluates and provides key
study-specific information for those
recent studies deemed to be of greatest
significance for informing preliminary
conclusions on the PM NAAQS in the
context of the body of evidence and
scientific conclusions presented in the
2019 ISA.
In developing a revised PA to support
the reconsideration, the EPA considered
the available scientific evidence,
including the evidence presented in the
2019 ISA and ISA Supplement. The
2022 PA considered the quantitative
and technical information presented in
the 2020 PA, in addition to new and
updated analyses conducted since the
2020 final decision. For those health
and welfare effects for which the ISA
Supplement evaluated recently
available studies (i.e., PM2.5-related
health effects and visibility effects), new
updated quantitative analyses were
conducted as a part of the development
of the 2022 PA. The newly available
scientific and technical information
presented in the 2022 PA were
considered in reaching conclusions
regarding the adequacy of the current
standards and any potential alternative
standards. For those health and welfare
effects for which newly available
scientific and technical information
were not evaluated (i.e., PM10–2.5-related
health effects and non-visibility welfare
effects), the conclusions presented in
the 2022 PA rely heavily on the
information that supported the
conclusions in the 2020 PA.
The CASAC PM panel met at a virtual
public meeting in November 2021 to
review the draft ISA Supplement (86 FR
52673, September 22, 2021). A virtual
public meeting was then held in
February 2022, and during this meeting
the chartered CASAC considered the
CASAC PM panel’s draft letter to the
Administrator on the draft ISA
Supplement (87 FR 958, January 7,
2022).
The chartered CASAC provided its
advice on the draft ISA Supplement in
a letter to the EPA Administrator dated
March 18, 2022 (Sheppard, 2022b). In
its review of the draft ISA Supplement,
the CASAC noted that they found ‘‘the
Draft ISA Supplement to be a wellwritten, comprehensive evaluation of
the new scientific information
published since the 2019 PM ISA’’
(Sheppard, 2022b, p. 2 of letter).
Furthermore, the CASAC stated that
‘‘the final Integrated Science
Assessment (ISA) Supplement . . .
deserve[s] the Administrator’s full
consideration and [is] adequate for
rulemaking’’ (Sheppard, 2022b, p. 2 of
PO 00000
Frm 00012
Fmt 4701
Sfmt 4700
letter). The CASAC generally endorsed
EPA’s decisions regarding the limited
scope of the draft ISA Supplement,
stating that ‘‘this limitation [on scope] is
appropriate for the targeted purpose of
the Draft ISA Supplement’’ although the
CASAC noted it would not be
appropriate for ISAs generally, and
recommended that the EPA provide
additional acknowledgment and
explanation for the limited scope
(Sheppard, 2022b, p. 2 of letter; see also
pp. 2–3 of consensus responses). The
EPA specifically noted in the final ISA
Supplement, which was released in May
2022 (U.S. EPA, 2022a; hereafter
referred to as the ISA Supplement
throughout this document) that the
‘‘targeted approach to developing the
Supplement to the 2019 PM ISA for the
purpose of reconsidering the 2020 PM
NAAQS decision does not reflect a
change to EPA’s approach for
developing ISAs for NAAQS reviews.’’
Thus, the evidence presented within the
2019 ISA, along with the targeted
identification and evaluation of new
scientific information in the ISA
Supplement, provides the scientific
basis for the reconsideration of the 2020
PM NAAQS final decision.
The draft PA was released in October
2021 (86 FR 56263, October 8, 2021).
The CASAC PM panel met at a virtual
public meeting in December 2021 to
review the draft PA (86 FR 52673,
September 22, 2021). A virtual public
meeting was then held in February 2022
and March 2022, and during this
meeting the chartered CASAC
considered the CASAC PM panel’s draft
letter to the Administrator on the draft
PA (87 FR 958, January 7, 2022). The
chartered CASAC provided its advice on
the draft PA in a letter to the EPA
Administrator dated March 18, 2022
(Sheppard, 2022a). The EPA took steps
to address these comments in revising
and finalizing the PA. The 2022 PA
considers the scientific evidence
presented in the 2019 ISA and ISA
Supplement and considers the
quantitative and technical information
presented in the 2020 PA, along with
updated and newly available analyses
since the completion of the 2020 review.
For those health and welfare effects for
which the ISA Supplement evaluated
recently available evidence and for
which updated quantitative analyses
were supported (i.e., PM2.5-related
health effects and visibility effects), the
2022 PA includes consideration of this
newly available scientific and technical
information in reaching preliminary
conclusions. For those health and
welfare effects for which newly
available scientific and technical
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
information were not evaluated (i.e.,
PM10–2.5-related health effects and nonvisibility effects), the conclusions
presented in the 2022 PA rely heavily
on the information that supported the
conclusions in the 2020 PA. The final
PA was released in May 2022 (U.S. EPA,
2022b; hereafter referred to as the 2022
PA throughout this document).
Drawing from his consideration of the
scientific evidence assessed in the 2019
ISA and ISA Supplement and the
analyses in the 2022 PA, including the
uncertainties in the evidence and
analyses, and from his consideration of
advice from the CASAC, on January 5,
2023, the Administrator proposed to
revise the level of the primary annual
PM2.5 standard and to retain the primary
24-hour PM2.5 standard, the primary 24hour PM10 standard, and the secondary
PM standards. These proposed
decisions were published in the Federal
Register on January 27, 2023 (88 FR
5558, January 27, 2023). The EPA held
a multi-day virtual public hearing on
February 21–23, 2023 (88 FR 6215,
January 31, 2023). In total, the EPA
received nearly 700,000 comments on
the proposal from members of the
public by the close of the public
comment period on March 28, 2023.
Major issues raised in the public
comments are discussed throughout the
preamble of this final action. A more
detailed summary of all significant
comments, along with the EPA’s
responses (henceforth ‘‘Response to
Comments’’ document), can be found in
the docket for this rulemaking (Docket
No. EPA–HQ–OAR–2015–0072).
As in prior reviews, the EPA is basing
its decision in this reconsideration on
studies and related information in the
air quality criteria, which have
undergone CASAC and public review.
These studies assessed in the 2019
ISA 27 and ISA Supplement 28 and the
2022 PA, and the integration of the
scientific evidence presented in them,
have undergone extensive critical
review by the EPA, the CASAC, and the
public. Decisions on the NAAQS should
be based on studies that have been
27 In addition to the 2020 review’s opening ‘‘call
for information’’ (79 FR 71764, December 3, 2014),
the 2019 ISA identified and evaluated studies and
reports that have undergone scientific peer review
and were published or accepted for publication
between January 1, 2009, through approximately
January 2018 (U.S. EPA, 2019a, p. ES–2). References
that are cited in the 2019 ISA, the references that
were considered for inclusion but not cited, and
electronic links to bibliographic information and
abstracts can be found at: https://hero.epa.gov/hero/
particulate-matter.
28 As described above, the ISA Supplement
represents an evaluation of recent studies that are
of greatest policy relevance and utility to the
reconsideration of the 2020 final decision on the
PM NAAQS (U.S. EPA, 2022a).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
rigorously assessed in an integrative
manner not only by the EPA but also by
the statutorily mandated independent
scientific advisory committee, as well as
the public review that accompanies this
process. It is for this reason that the EPA
preliminarily concluded that the
scientific evidence available since the
completion of the 2019 ISA, including
those raised in public comments on the
proposal in 2020, warranted a partial
reopening of the air quality criteria and
prepared an ISA Supplement to enable
the EPA, the CASAC, and the public to
consider them further. Some
commenters have referred to and
discussed additional individual
scientific studies on the health effects of
PM that were not included in the 2019
ISA or ISA Supplement (‘‘new studies’’)
and that have not gone through this
comprehensive review process. In
considering and responding to
comments for which such ‘‘new’’
studies were cited in support, the EPA
has provisionally considered the cited
studies in the context of the findings of
the 2019 ISA and ISA Supplement. The
EPA’s provisional consideration of these
studies did not and could not provide
the kind of in-depth critical review
described above, but rather was focused
on determining whether they warranted
further reopening the review of the air
quality criteria to enable the EPA, the
CASAC, and the public to consider
them further.
This approach, and the decision to
rely on the studies and related
information in the air quality criteria,
which have undergone CASAC and
public review, is consistent with the
EPA’s practice in prior NAAQS reviews
and its interpretation of the
requirements of the CAA. Since the
1970 amendments, the EPA has taken
the view that NAAQS decisions are to
be based on scientific studies and
related information that have been
assessed as a part of the pertinent air
quality criteria, and the EPA has
consistently followed this approach.
This longstanding interpretation was
strengthened by new legislative
requirements enacted in 1977, which
added section 109(d)(2) of the Act
concerning CASAC review of air quality
criteria. See 71 FR 6114, 61148 (October
17, 2006, final decision on review of
NAAQS for particulate matter) for a
detailed discussion of this issue and the
EPA’s past practice.
As discussed in the EPA’s 1993
decision not to review the O3 NAAQS,
‘‘new’’ studies may sometimes be of
such significance that it is appropriate
to delay a decision in a NAAQS review
and to supplement the pertinent air
quality criteria so the studies can be
PO 00000
Frm 00013
Fmt 4701
Sfmt 4700
16213
taken into account (58 FR 13013–13014,
March 9, 1993). In the present case, the
EPA decided to partially reopen the air
quality criteria and prepared an ISA
Supplement as a part of the
reconsideration to facilitate evaluation
of these studies by the EPA, the CASAC,
and the public. The narrow scope of the
ISA Supplement is supported by EPA’s
provisional consideration of ‘‘new’’
studies submitted in response to public
comments on the 2020 proposal which
concluded that, taken in context, the
‘‘new’’ information and findings do not
materially change any of the broad
scientific conclusions regarding the
health and welfare effects of PM in
ambient air made in the air quality
criteria. Therefore, a full reopening of
the air quality criteria was not
warranted to assess the health and
welfare effects of PM for purposes of the
review.
Accordingly, the EPA is basing the
final decisions in this reconsideration
on the studies and related information
included in the PM air quality criteria
(including the 2019 PM ISA and ISA
Supplement) that have undergone
rigorous review by the EPA, the CASAC,
and the public. The EPA will consider
these ‘‘new’’ studies for inclusion in the
air quality criteria for the next PM
NAAQS review, which the EPA expects
to begin soon after the conclusion of this
reconsideration and which will provide
the opportunity to fully assess these
studies through a more rigorous review
process involving the EPA, the CASAC,
and the public.
D. Air Quality Information
This section provides a summary of
basic information related to PM ambient
air quality. It summarizes information
on the distribution of particle size in
ambient air (section I.D.1), sources and
emissions contributing to PM in the
ambient air (section I.D.2), monitoring
ambient PM in the U.S. (section I.D.3),
ambient PM concentrations and trends
in the U.S. (I.D.4), characterizing
ambient PM2.5 concentrations for
exposure (section I.D.5), and
background PM (section I.D.6).
Additional detail on PM air quality can
be found in Chapter 2 of the 2022 PA
(U.S. EPA, 2022b).
1. Distribution of Particle Size in
Ambient Air
In ambient air, PM is a mixture of
substances suspended as small liquid
and/or solid particles (U.S. EPA, 2019a,
section 2.2) and distinct health and
welfare effects have been linked with
exposures to particles of different sizes.
Particles in the atmosphere range in size
from less than 0.01 to more than 10 mm
E:\FR\FM\06MRR3.SGM
06MRR3
16214
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
in diameter (U.S. EPA, 2019a, section
2.2). The EPA defines PM2.5, also
referred to as fine particles, as particles
with aerodynamic diameters generally
less than or equal to 2.5 mm. The size
range for PM10–2.5, also called coarse or
thoracic coarse particles, includes those
particles with aerodynamic diameters
generally greater than 2.5 mm and less
than or equal to 10 mm. PM10, which is
comprised of both fine and coarse
fractions, includes those particles with
aerodynamic diameters generally less
than or equal to 10 mm. In addition,
ultrafine particles (UFP) are often
defined as particles with a diameter of
less than 0.1 mm based on physical size,
thermal diffusivity or electrical mobility
(U.S. EPA, 2019a, section 2.2).
Atmospheric lifetimes are generally
longest for PM2.5, which often remains
in the atmosphere for days to weeks
(U.S. EPA, 2019a, Table 2–1) before
being removed by wet or dry deposition,
while atmospheric lifetimes for UFP and
PM10–2.5 are shorter and are generally
removed from the atmosphere within
hours, through wet or dry deposition
(U.S. EPA, 2019a, Table 2–1; U.S. EPA,
2022b, section 2.1).
2. Sources and Emissions Contributing
to PM in the Ambient Air
PM is composed of both primary
(directly emitted particles) and
secondary particles. Primary PM is
derived from direct particle emissions
from specific PM sources while
secondary PM originates from gas-phase
precursor chemical compounds present
in the atmosphere that have participated
in new particle formation or condensed
onto existing particles (U.S. EPA, 2019a,
section 2.3). As discussed further in the
2019 ISA (U.S. EPA, 2019a, section
2.3.2.1), secondary PM is formed in the
atmosphere by photochemical oxidation
reactions of both inorganic and organic
gas-phase precursors. Precursor gases
include sulfur dioxide (SO2), nitrogen
oxides (NOX), and volatile organic
compounds (VOC) (U.S. EPA, 2019a,
section 2.3.2.1). Ammonia also plays an
important role in the formation of
nitrate PM by neutralizing sulfuric acid
and nitric acid. Sources and emissions
of PM are discussed in more detail the
2022 PA (U.S. EPA, 2022b, section
2.1.1). Briefly, anthropogenic sources of
PM include both stationary (e.g., fuel
combustion for electricity production
and other purposes, industrial
processes, agricultural activities) and
mobile (e.g., diesel- and gasolinepowered highway vehicles and other
engine-driven sources) sources. Natural
sources of PM include dust from the
wind erosion of natural surfaces, sea
salt, wildfires, primary biological
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
aerosol particles (PBAP) such as bacteria
and pollen, oxidation of biogenic
hydrocarbons, such as isoprene and
terpenes to produce secondary organic
aerosol (SOA), and geogenic sources,
such as sulfate formed from volcanic
production of SO2. Wildland fire, which
encompass both wildfire and prescribed
fire, accounts for 44% of emissions of
primary PM2.5 emissions (U.S. EPA,
2021b). Emissions from wildfire
comprises 29% of primary PM2.5
emissions.
In recent years, the frequency and
magnitude of wildfires have increased
(U.S. EPA, 2019a). The magnitude of the
public health impact of wildfires is
substantial both because of the increase
in PM2.5 concentrations as well as the
duration of the wildfire smoke season,
which is considered to range from May
to November. Wildfire can make a large
contribution to air pollution (including
PM2.5), and wildfire events can threaten
public safety and life. The impacts of
wildfire events can be mitigated through
management of wildland vegetation,
including through prescribed fire.
Prescribed fire (and some wildfires) can
mimic the natural processes necessary
to maintain fire-dependent ecosystems,
minimizing catastrophic wildfires and
the risks they pose to safety, property
and air quality (see, e.g., 81 FR 58010,
58038, August 24, 2016). The EPA views
the strategic use of prescribed fire as an
important tool for reducing wildfire risk
and the severity of wildfires and
wildfire smoke (88 FR, 54118, 54126,
August 9, 2023).29 As noted in the PM
NAAQS proposal, agencies have efforts
in place to reduce the frequency and
severity of human-caused wildfires (88
FR 5570, January 27, 2023).
Wildfire events produce high PM
emissions that may impact the PM
concentrations in ambient air to the
extent that the concentrations result in
an exceedance or violation which may
affect the design value in a given area.
The EPA’s Exceptional Events Rule (81
FR 68216, October 3, 2016) describes
the process by which air agencies may
request to exclude ‘event-influenced’
data caused by exceptional events,
which can include wildfires and
prescribed fires on wildland. The EPA
has issued guidance specifically
addressing exceptional events
demonstrations for both wildfires and
prescribed fires on wildland. These
documents are available on EPA’s
Exceptional Events Program website.30
29 See also: https://www.usda.gov/sites/default/
files/documents/usda-epa-doi-cdc-mou.pdf.
30 See: https://www.epa.gov/air-quality-analysis/
final-2016-exceptional-events-rule-supportingguidance-documents-updated-faqs.
PO 00000
Frm 00014
Fmt 4701
Sfmt 4700
The EPA will develop fire-related
exceptional events implementation
tools, including updates as needed to
existing guidance to facilitate more
efficient processing of PM2.5-related
exceptional events demonstrations for
both the 24-hour and annual standards.
3. Monitoring of Ambient PM
To promote uniform application of
the air quality standards set forth under
the CAA and to achieve the degree of
public health and welfare protection
intended for the NAAQS, the EPA
establishes PM Federal Reference
Methods (FRMs) for both PM10 and
PM2.5 in appendices J and L to 40 CFR
part 50, both of which were amended
following the 2006 and 2012 PM
NAAQS reviews. The current PM
monitoring network relies on FRMs and
automated continuous Federal
Equivalent Methods (FEMs) approved
pursuant to 40 CFR part 53, in part to
support changes necessary for
implementation of the revised PM
standards. Additionally, 40 CFR part 58,
appendices A through E, detail the
requirements to measure ambient air
quality and report ambient air quality
data and related information. More
information on PM ambient monitoring
networks is available in section 2.2 of
the 2022 PA (U.S. EPA, 2022b).
The PM2.5 monitoring program is one
of the major ambient air monitoring
programs with a robust, nationally
consistent network of ambient air
monitoring sites providing mass and/or
chemical speciation measurements. 40
CFR part 58, appendix D, section 4.7
provides the applicable PM2.5 network
design criteria. For most urban
locations, PM2.5 monitors are sited at the
neighborhood scale,31 where PM2.5
concentrations are reasonably
homogeneous throughout an entire
urban sub-region. In each CBSA with a
monitoring requirement, at least one
PM2.5 monitoring station representing
31 For PM , neighborhood scale is defined at 40
2.5
CFR part 58, appendix D, 4.7.1(c)(3) as follows:
Measurements in this category would represent
conditions throughout some reasonably
homogeneous urban sub-region with dimensions of
a few kilometers and of generally more regular
shape than the middle scale. Homogeneity refers to
the particulate matter concentrations, as well as the
land use and land surface characteristics. Much of
the PM2.5 exposures are expected to be associated
with this scale of measurement. In some cases, a
location carefully chosen to provide neighborhood
scale data would represent the immediate
neighborhood as well as neighborhoods of the same
type in other parts of the city. PM2.5 sites of this
kind provide good information about trends and
compliance with standards because they often
represent conditions in areas where people
commonly live and work for periods comparable to
those specified in the NAAQS. In general, most
PM2.5 monitoring in urban areas should have this
scale.
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
area-wide air quality is sited in an area
of expected maximum concentration.32
By ensuring the area of expected
maximum concentration in a CBSA has
a site compared to both the annual and
24-hour NAAQS, all other similar
locations are thus protected. Sites that
represent relatively unique microscale,
localized hot-spot, or unique middle
scale impact sites are only eligible for
comparison to the 24-hour PM2.5
NAAQS.
Under 40 CFR part 50, appendix L,
and 40 CFR part 53, and 40 CFR part 58
appendix D there are three main
methods components of the PM2.5
monitoring program: filter-based FRMs
measuring PM2.5 mass, FEMs measuring
PM2.5 mass, and other samplers used to
collect the aerosol used in subsequent
laboratory analysis for measuring PM2.5
chemical speciation. The FRMs are
primarily used for comparison to the
NAAQS, but also serve other important
purposes, such as developing trends and
evaluating the performance of FEMs.
PM2.5 FEMs are typically continuous
methods used to support forecasting and
reporting of the Air Quality Index (AQI)
but are also used for comparison to the
NAAQS. Samplers that are part of the
Chemical Speciation Network (CSN)
and Interagency Monitoring of Protected
Visual Environments (IMPROVE)
network are used to provide chemical
composition of the aerosol and serve a
variety of objectives. More detail on of
each of these components of the PM2.5
monitoring program and of recent
changes to PM2.5 monitoring
requirements are described in detail in
the 2022 PA (U.S. EPA, 2022b, section
2.2.3).
ddrumheller on DSK120RN23PROD with RULES3
4. Ambient Concentrations and Trends
This section summarizes available
information on recent ambient PM
concentrations in the U.S. and on trends
in PM air quality. Sections I.D.4.a and
I.D.4.b summarize information on PM2.5
mass and components, respectively.
Section I.D.4.c summarizes information
on PM10. Sections I.D.4.d and I.D.4.e
summarize the more limited
information on PM10–2.5 and UFP,
respectively. Additional detail on PM
air quality and trends can be found in
the 2022 PA (U.S. EPA, 2022b, section
2.3).
a. PM2.5 mass
At monitoring sites in the U.S.,
annual PM2.5 concentrations from 2017
to 2019 averaged 8.0 mg/m3 (with the
10th and 90th percentiles at 5.9 and
10.0 mg/m3, respectively) and the 98th
percentiles of 24-hour concentrations
32 40
CFR part 58, app. D, 4.7.1(b)(2).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
averaged 21.3 mg/m3 (with the 10th and
90th percentiles at 14.0 and 29.7 mg/m3,
respectively) (U.S. EPA, 2022b, section
2.3.2.1). The highest ambient PM2.5
concentrations occur in the western
U.S., particularly in California and the
Pacific Northwest (U.S. EPA, 2022b,
Figure 2–15). Much of the eastern U.S.
has lower ambient concentrations, with
annual average concentrations generally
at or below 12.0 mg/m3 and 98th
percentiles of 24-hour concentrations
generally at or below 30 mg/m3 (U.S.
EPA, 2022b, section 2.3.2.1).
Recent ambient PM2.5 concentrations
reflect the substantial reductions that
have occurred across much of the U.S.
(U.S. EPA, 2022b, section 2.3.2.1). From
2000 to 2019, national annual average
PM2.5 concentrations declined from 13.5
mg/m3 to 7.6 mg/m3, a 43% decrease
(U.S. EPA, 2022b, section 2.3.2.1).33
These declines have occurred at urban
and rural monitoring sites, although
urban PM2.5 concentrations remain
consistently higher than those in rural
areas (Chan et al., 2018) due to the
impact of local sources in urban areas.
Analyses at individual monitoring sites
indicate that declines in ambient PM2.5
concentrations have been most
consistent across the eastern U.S. and in
parts of coastal California, where both
annual average and 98th percentiles of
24-hour concentrations declined
significantly (U.S. EPA, 2022b, section
2.3.2.1). In contrast, trends in ambient
PM2.5 concentrations have been less
consistent over much of the western
U.S., with no significant changes since
2000 observed at some sites in the
Pacific Northwest, the northern Rockies
and plains, and the Southwest,
particularly for 98th percentiles of 24hour concentrations (U.S. EPA, 2022b,
section 2.3.2.1). As noted below, some
sites in the northwestern U.S. and
California, where wildfire have been
relatively common in recent years, have
experienced high concentrations over
shorter periods (i.e., 2-hour averages).
The recent deployment of PM2.5
monitors near major roads in large
urban areas provides information on
PM2.5 concentrations near an important
emissions source. For 2016–2018, Gantt
et al. (2021) reported that 52% and 24%
of the time near-road sites reported the
highest annual and 24-hour PM2.5
design value 34 in the CBSA,
respectively. Of the CBSAs with the
highest annual design values at nearroad sites reported by Gantt et al. (2021),
33 See https://www.epa.gov/air-trends/particulatematter-pm25-trends for up-to-date PM2.5 trends
information.
34 A design value is considered valid if it meets
the data handling requirements given in appendix
N to 40 CFR part 50.
PO 00000
Frm 00015
Fmt 4701
Sfmt 4700
16215
those design values were, on average,
0.8 mg/m3 higher than at the highest
measuring non-near-road sites (range is
0.1 to 2.1 mg/m3 higher at near-road
sites). Although most near-road
monitoring sites do not have sufficient
data to evaluate long-term trends in
near-road PM2.5 concentrations,
analyses of the data at one near-roadlike site in Elizabeth, NJ, 35 show that
the annual average near-road increment
has generally decreased between 1999
and 2017 from about 2.0 mg/m3 to about
1.3 mg/m3 (U.S. EPA, 2022b, section
2.3.2.1).
Ambient PM2.5 concentrations can
exhibit a diurnal cycle that varies due
to impacts from intermittent emission
sources, meteorology, and atmospheric
chemistry. The PM2.5 monitoring
network in the U.S. has an increasing
number of continuous FEM monitors
reporting hourly PM2.5 mass
concentrations that reflect this diurnal
variation. The 2019 ISA describes a twopeaked diurnal pattern in urban areas,
with morning peaks attributed to rushhour traffic and afternoon peaks
attributed to a combination of rush hour
traffic, decreasing atmospheric dilution,
and nucleation (U.S. EPA, 2019a,
section 2.5.2.3, Figure 2–32). Because a
focus on annual average and 24-hour
average PM2.5 concentrations could
mask subdaily patterns, and because
some health studies examine PM
exposure durations shorter than 24hours, it is useful to understand the
broader distribution of subdaily PM2.5
concentrations across the U.S. The 2022
PA presents information on the
frequency distribution of 2-hour average
PM2.5 mass concentrations from all FEM
PM2.5 monitors in the U.S. for 2017–
2019. At sites meeting the current
primary PM2.5 standards, these 2-hour
concentrations generally remain below
10 mg/m3, and rarely exceed 30 mg/m3.
Two-hour concentrations are higher at
sites violating the current standards,
generally remaining below 16 mg/m3 and
rarely exceeding 80 mg/m3 (U.S. EPA,
2022b, section 2.3.2.2.3). The extreme
upper end of the distribution of 2-hour
PM2.5 concentrations is shifted higher
during the warmer months, generally
corresponding to the period of peak
wildfire frequency (April to September)
in the U.S. At sites meeting the current
primary standards, the highest 2-hour
concentrations measured rarely occur
outside of the period of peak wildfire
frequency. Most of the sites measuring
35 The Elizabeth Lab site in Elizabeth, NJ, is
situated approximately 30 meters from travel lanes
of the Interchange 13 toll plaza of the New Jersey
Turnpike and within 200 meters of travel lanes for
Interstate 278 and the New Jersey Turnpike.
E:\FR\FM\06MRR3.SGM
06MRR3
16216
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
these very high concentrations are in the
northwestern U.S. and California, where
wildfires have been relatively common
in recent years (see U.S. EPA, 2022b,
Appendix A, Figure A–1). When the
period of peak wildfire frequency is
excluded from the analysis, the extreme
upper end of the distribution is reduced
(U.S. EPA, 2022b, section 2.3.2.2.3).
b. PM2.5 Components
ddrumheller on DSK120RN23PROD with RULES3
Based on recent air quality data, the
major chemical components of PM2.5
have distinct spatial distributions.
Sulfate concentrations tend to be
highest in the eastern U.S., while in the
Ohio Valley, Salt Lake Valley, and
California nitrate concentrations are
highest, and relatively high
concentrations of organic carbon are
widespread across most of the
continental U.S. (U.S. EPA, 2022b,
section 2.3.2.3). Elemental carbon,
crustal material, and sea salt are found
to have the highest concentrations in the
northeast U.S., southwest U.S., and
coastal areas, respectively.
An examination of PM2.5 composition
trends can provide insight into the
factors contributing to overall
reductions in ambient PM2.5
concentrations. The biggest change in
PM2.5 composition that has occurred in
recent years is the reduction in sulfate
concentrations due to reductions in SO2
emissions. Between 2000 and 2015, the
nationwide annual average sulfate
concentration decreased by 17% at
urban sites and 20% at rural sites. This
change in sulfate concentrations is most
evident in the eastern U.S. and has
resulted in organic matter or nitrate now
being the greatest contributor to PM2.5
mass in many locations (U.S. EPA,
2019a, Figure 2–19). The overall
reduction in sulfate concentrations has
contributed substantially to the decrease
in national average PM2.5 concentrations
as well as the decline in the fraction of
PM10 mass accounted for by PM2.5 (U.S.
EPA, 2019a, section 2.5.1.1.6; U.S. EPA,
2022b, section 2.3.1).
c. PM10
At long-term monitoring sites in the
U.S., the 2017–2019 average of 2nd
highest 24-hour PM10 concentration was
68 mg/m3 (with 10th and 90th
percentiles at 28 and 124 mg/m3,
respectively) (U.S. EPA, 2022b, section
2.3.2.4).36 The highest PM10
concentrations tend to occur in the
western U.S. Seasonal analyses indicate
that ambient PM10 concentrations are
generally higher in the summer months
36 The form of the current 24-hour PM
10 standard
is one-expected-exceedance, averaged over three
years.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
than at other times of year, though the
most extreme high concentration events
are more likely in the spring (U.S. EPA,
2019a, Table 2–5). This is due to fact
that the major PM10 emission sources,
dust and agriculture, are more active
during the warmer and drier periods of
the year.
Recent ambient PM10 concentrations
reflect reductions that have occurred
across much of the U.S. (U.S. EPA,
2022b, section 2.3.2.4). From 2000 to
2019, 2nd highest 24-hour PM10
concentrations have declined by about
46% (U.S. EPA, 2022b, section
2.3.2.4).37 Analyses at individual
monitoring sites indicate that annual
average PM10 concentrations have
generally declined at most sites across
the U.S., with much of the decrease in
the eastern U.S. associated with
reductions in PM2.5 concentrations (U.S.
EPA, 2022b, section 2.3.2.4). Annual
2nd highest 24-hour PM10
concentrations have generally declined
in the eastern U.S., while concentrations
in much of the midwest and western
U.S. have remained unchanged or
increased since 2000 (U.S. EPA, 2022b,
section 2.3.2.4).
Compared to previous reviews, data
available from the NCore monitoring
network in the current reconsideration
allows a more comprehensive analysis
of the relative contributions of PM2.5
and PM10–2.5 to PM10 mass. PM2.5
generally contributes more to annual
average PM10 mass in the eastern U.S.
than the western U.S. (U.S. EPA, 2022b,
Figure 2–23). At most sites in the
eastern U.S., the majority of PM10 mass
is comprised of PM2.5. As ambient PM2.5
concentrations have declined in the
eastern U.S. (U.S. EPA, 2022b, section
2.3.2.2), the ratios of PM2.5 to PM10 have
also declined. For sites with days
having concurrently very high PM2.5 and
PM10 concentrations (U.S. EPA, 2022b,
Figure 2–24), the PM2.5/PM10 ratios are
typically higher than the annual average
ratios. This is particularly true in the
northwestern U.S. where the high PM10
concentrations can occur during
wildfires with high PM2.5 (U.S. EPA,
2022b, section 2.3.2.4).
d. PM10–2.5
Since the 2012 review, the availability
of PM10–2.5 ambient concentration data
has greatly increased because of
additions to the PM10–2.5 monitoring
capabilities to the national monitoring
network. As illustrated in the 2022 PA
(U.S. EPA, 2022b, section 2.3.2.5),
annual average and 98th percentile
37 For more information, see https://
www.epa.gov/air-trends/particulate-matter-pm10trends#pmnat.
PO 00000
Frm 00016
Fmt 4701
Sfmt 4700
PM10–2.5 concentrations exhibit less
distinct differences between the eastern
and western U.S. than for either PM2.5
or PM10.
Due to the short atmospheric lifetime
of PM10–2.5 relative to PM2.5, many of the
high concentration sites are isolated and
likely near emission sources associated
with wind-blown and fugitive dust. The
spatial distributions of annual average
and 98th percentile concentrations of
PM10–2.5 are more similar than that of
PM2.5, suggesting that the same dustrelated emission sources are affecting
both long-term and episodic
concentrations (U.S. EPA, 2022b, Figure
2–25). The highest concentrations of
PM10–2.5 are in the southwest U.S. where
widespread dry and windy conditions
contribute to wind-blown dust
emissions. Additionally, compared to
PM2.5 and PM10, changes in PM10–2.5
concentrations have been small in
magnitude and inconsistent in direction
(U.S. EPA, 2022b, Figure 2–25). The
majority of PM10–2.5 sites in the U.S. do
not have a concentration trend from
2000–2019, reflecting the relatively
consistent level of dust emissions across
the U.S. during the same time period
(U.S. EPA, 2022b, section 2.3.2.5).38
e. UFP
Compared to PM2.5 mass, there is
relatively little data on U.S. particle
number concentrations, which are
dominated by UFP. In the published
literature, annual average particle
number concentrations reaching about
20,000 to 30,000 cm3 have been
reported in U.S. cities (U.S. EPA,
2019a). In addition, based on UFP
measurements in two urban areas (New
York City, Buffalo) and at a background
site (Steuben County) in New York,
there is a pronounced difference in
particle number concentration between
different types of locations (U.S. EPA,
2022b, Figure 2–26; U.S. EPA, 2019a,
Figure 2–18). Urban particle number
counts were several times higher than at
the background site, and the highest
particle number counts in an urban area
with multiple sites (Buffalo) were
observed at a near-road location (U.S.
EPA, 2022b, section 2.3.2.6).
Long-term trends in UFP are not
routinely available at U.S. monitoring
38 PM from dust emissions in the National
Emissions Inventory (NEI) remain fairly consistent
from year-to-year, except when there are severe
weather incursions or there is a dust event that
transports or causes major local dust storms to
occur (particularly in the western U.S.). These dust
events and weather incursions needed to effect dust
emissions on a national level are not common and
only seldomly occur. In the emissions trends
analysis presented in the 2022 PA (U.S. EPA,
2022b, section 2.1.1), dust is included in the NEI
sector labeled ‘‘miscellaneous.’’
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
sites. At one background site in Illinois
with long-term data available, the
annual average particle number
concentration declined between 2000
and 2019, closely matching the
reductions in annual PM2.5 mass over
that same period (U.S. EPA, 2022b,
section 2.3.2.6). In addition, a small
number of published studies have
examined UFP trends over time. While
limited, these studies also suggest that
UFP number concentrations have
declined over time along with decreases
in PM2.5 (U.S. EPA, 2022b, section
2.3.2.6). However, the relationship
between changes in ambient PM2.5 and
UFPs cannot be comprehensively
characterized due to the high variability
and limited monitoring of UFPs (U.S.
EPA, 2022b, section 2.3.2.6).
ddrumheller on DSK120RN23PROD with RULES3
5. Characterizing Ambient PM2.5
Concentrations for Exposure
Epidemiologic studies use various
methods to characterize exposure to
ambient PM2.5. The methods used to
estimate PM2.5 concentrations can vary
from traditional methods using
monitoring data from ground-based
monitors to newer methods using more
complex hybrid modeling approaches.
Studies using hybrid modeling
approaches aim to broaden the spatial
coverage, as well as estimate more
spatially-resolved ambient PM2.5
concentrations, by expanding beyond
just those areas with monitors and
providing estimates in areas that do not
have ground-based monitors (i.e., areas
that are generally less densely
populated and tend to have lower PM2.5
concentrations) and at finer spatial
resolutions (e.g., 1 km x 1 km grid cells).
Ground-based PM2.5 monitors are
generally sited in areas of expected
maximum concentration. As such, the
hybrid modeling approaches tend to
broaden the areas captured in the
exposure assessment, and in doing so,
the studies that utilize these methods
tend to report lower mean PM2.5
concentrations than monitor-based
approaches. Further, other aspects of the
approaches applied in the various
epidemiologic studies to estimate PM2.5
exposure and/or to calculate the related
study-reported mean concentration (i.e.,
population weighting, trim mean
approaches) can affect those data values.
More detail related to hybrid modeling
methods, performance of the methods,
and how the reported mean
concentrations compare across
approaches is provided in section
2.3.3.2 of the 2022 PA (U.S. EPA,
2022b). The subsections below discuss
the characterization of PM2.5
concentrations based on monitoring
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
data (I.D.5.a) and using hybrid modeling
approaches (I.D.5.b).
a. Predicted Ambient PM2.5 and
Exposure Based on Monitored Data
Ambient concentrations of PM2.5 are
often characterized using measurements
from national monitoring networks due
to the accuracy and precision of the
measurements and the public
availability of data. For applications
requiring PM2.5 characterizations across
large areas or provide complete coverage
from the site measurements, data
interpolation and averaging techniques
(such as Average Nearest Neighbor
tools, and area-wide or populationweighted averaging of monitors) are
sometimes used (U.S. EPA, 2019a,
chapter 3).
For an area to meet the NAAQS, all
valid design values 39 in that area,
including the highest annual and 24hour design values, must be at or below
the levels of the standards. Because the
monitoring network siting requirements
are specified to capture the high PM2.5
concentrations (U.S. EPA, 2022b,
section 2.2.3), areas meeting an annual
PM2.5 standard with a particular level
would be expected to have long-term
average monitored PM2.5 concentrations
(i.e., averaged across space and over
time in the area) somewhat below that
standard level. This means that the
PM2.5 design value in an area is
associated with a distribution of PM2.5
concentrations in that area, and, based
on monitoring siting requirements,
should represent the highest
concentration location applicable to be
monitored under the PM2.5 NAAQS.
Analyses in the 2022 PA indicate that,
based on recent air quality in U.S.
CBSAs, maximum annual PM2.5 design
values are often 10% to 20% higher
than annual average concentrations (i.e.,
averaged across multiple monitors in
the same CBSA) (U.S. EPA, 2022b,
section 2.3.3.1, Figures 2–28 and 2–29).
This difference between the maximum
annual design value and the average
concentration in an area can vary,
depending on factors such as the
number of monitors, monitor siting
characteristics, and the distribution of
ambient PM2.5 concentrations. Given
that higher PM2.5 concentrations have
been reported at some near-road
monitoring sites relative to the
surrounding area (U.S. EPA, 2022b,
section 2.3.2.2.2), recent requirements
39 For the annual PM
2.5 standard, design values
are calculated as the annual arithmetic mean PM2.5
concentration, averaged over 3 years. For the 24hour standard, design values are calculated as the
98th percentile of the annual distribution of 24hour PM2.5 concentrations, averaged over three
years (appendix N of 40 CFR part 50).
PO 00000
Frm 00017
Fmt 4701
Sfmt 4700
16217
for PM2.5 monitoring at near-road
locations in large urban areas (U.S. EPA,
2022b, section 2.2.3.3) may increase the
ratios of maximum design values to
average annual design values in some
areas. Such ratios may also depend on
how the averages are calculated (i.e.,
averaged across monitors versus across
modeled grid cells, as described below
in section I.5.b). Compared to annual
design values, the analysis in the 2022
PA indicates a more variable
relationship between maximum 24-hour
PM2.5 design values and annual average
concentrations (U.S. EPA, 2022b,
section 2.3.3.1, Figure 2–29).
b. Comparison of PM2.5 Hybrid
Modeling Approaches in Estimating
Exposure and Relative to Design Values
Two types of hybrid approaches that
have been utilized in several key PM2.5
epidemiologic studies in the 2019 ISA
and ISA Supplement include neural
network approaches and a satellitebased method with regression of
residual PM2.5 with land-use and other
variables to improve estimates of PM2.5
concentration in the U.S. As such, the
2022 PA further compares these two
types of approaches across various
scales (e.g., CBSA versus nationwide),
taking into account population
weighting approaches utilized in
epidemiologic studies when estimating
PM2.5 exposure (U.S. EPA, 2022b,
section 2.3.3.2.4). Additionally, the
2022 PA assesses how average PM2.5
concentrations computed in
epidemiologic studies using these
hybrid surfaces compare to the
maximum design values measured at
ground-based monitors. For this
assessment, the 2022 PA evaluates the
DI2019 40 and HA2020 41 hybrid
surfaces, surfaces that are used in
several of the key epidemiologic studies
in the 2022 PA. This analysis is
intended to help inform how the
magnitude of the overall study-reported
mean PM2.5 concentrations in
epidemiologic studies may be
40 This analysis includes an updated version of
the surface used in Di et al. (2016). Predictions in
Di et al. (2016) were for 2000 to 2012 using a neural
network model. The Di et al. (2019) study improved
on that effort in several ways. First, a generalized
additive model was used that accounted for
geographic variations in performance to combine
predictions from three models (neural network,
random forest, and gradient boosting) to make the
final optimal PM2.5 predictions. Second, the
datasets were updated that were used in model
training and included additional variables such as
12-km CMAQ modeling as predictors. Finally, more
recent years were included in the Di et al. (2019)
study.
41 The HA2020 field is based on the V4.NA.03
product available at: https://sites.wustl.edu/acag/
datasets/surface-pm2-5/. The name ‘‘HA2020’’
comes from the references for this product (Hammer
et al., 2020; van Donkelaar et al., 2019).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16218
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
influenced by the approach used to
compute that mean and how that value
might compare to monitor reported
concentrations. The PM2.5 standards are
expected to achieve a pattern of air
quality through the attainment of a
specific design value at each monitor in
the monitoring network. As a result, it
is important to be able to assess the
relationship between monitor
concentrations and patterns of air
quality evaluated in the epidemiologic
studies.
In estimating exposure, some studies
focus on estimating concentrations in
urban areas, while others examine the
entire U.S. or large portions of the
country. In general, the areas that are
not included in the CBSA-only analysis
tend to be more rural or less densely
populated areas, tend to have lower
PM2.5 concentrations, and likely
correspond to those locations where
monitoring data availability is limited or
nonexistent (U.S. EPA, 2022b, section
2.3.3.2.4, Figure 2–37). To evaluate the
differences in mean PM2.5
concentrations across different spatial
scales, the 2022 PA analysis compares
the DI2019 and HA2020 surfaces. At the
national scale, the two surfaces
generally produce similar average
annual PM2.5 concentrations, with the
DI2019 surface being slightly higher
compared to the HA2020 surface. The
average annual PM2.5 concentrations are
also slightly higher using the DI2019
surface compared to the HA2020 surface
when the analyses are conducted for
CBSAs. Also, regardless of which
surface is used, the average annual and
3-year average of the average annual
PM2.5 concentrations for the CBSA-only
analyses are somewhat higher than for
the nationwide analyses (4–8% higher)
(U.S. EPA, 2022b, section 2.3.3.2.4,
Table 2–5).42 Overall, these analyses
suggest that there are only slight
differences in the average PM2.5
concentrations depending on the hybrid
modeling method employed, though
including other hybrid modeling
methods in this comparison could result
in larger differences.
The 2022 PA next evaluates how the
averages of the hybrid model surfaces
compare to regulatory design values
using both the DI2019 and HA2020
surfaces and how population weighting
influences the mean PM2.5
concentration.43 As presented in the
42 For
the national scale, 3-year averages of the
average annual PM2.5 concentrations generally range
from about 5.3 mg/m3 to 8.1 mg/m3, compared to the
CBSA scale, which ranges from 5.7 mg/m3 to 8.7 mg/
m3. (U.S. EPA, 2022b, section 2.3.3.2.4, Table 2–6).
43 For this analysis, the 2022 PA includes CBSAs
with three or more valid design values for the 3year period. The regulatory design values for the
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
2022 PA, the results using the DI2019
and HA2020 surfaces are similar for the
average annual PM2.5 concentrations, for
each 3-year period. When population
weighting is not applied, the average
annual PM2.5 concentrations generally
range from 7.0 to 8.6 mg/m3. When
population weighting is applied, the
average annual PM2.5 concentrations are
slightly higher, ranging from 8.2 to 10.2
mg/m3. As with CBSAs versus the
national comparison above, population
weighting results in a higher average
PM2.5 concentration than when
population weighting is not applied
(U.S. EPA, 2022b, section 2.3.3.2.4,
Table 2–7). For the CBSAs included in
the population weighted analyses, the
average maximum annual design values
generally range from 9.5 to 11.7 mg/m3.
The results are similar for both the
DI2019 and HA2020 surfaces and the
maximum annual PM2.5 design values
measured at the monitors are often 40%
to 50% higher than average annual
PM2.5 concentrations predicted by
hybrid modeling methods when
population weighting is not applied.
However, when population weighting is
applied, the ratio of the maximum
annual PM2.5 design values to the
predicted average annual PM2.5
concentrations are lower than when
population weighting is not applied,
with monitored design values generally
15% to 18% higher than populationweighted hybrid modeling average
annual PM2.5 concentrations (U.S. EPA,
2022b, section 2.3.3.2.4, Table 2–7).
6. Background PM
In this reconsideration, background
PM is defined as all particles that are
formed by sources or processes that
cannot be influenced by actions within
the jurisdiction of concern. U.S.
background PM is defined as any PM
formed from emissions other than U.S.
anthropogenic (i.e., manmade)
emissions. Potential sources of U.S.
background PM include both natural
sources (i.e., PM that would exist in the
absence of any anthropogenic emissions
of PM or PM precursors) and
transboundary sources originating
outside U.S. borders. Background PM is
discussed in more detail in the 2022 PA
(U.S. EPA, 2022b, section 2.4). At
annual and national scales, estimated
background PM concentrations in the
CBSAs were calculated for each 3-year period for
the CBSAs with 3 or more design values in each of
the 3-year periods. Using the maximum design
value for each CBSA and by each 3-year period, the
ratio of maximum design values to modeled average
annual PM2.5 concentrations were calculated, for
each 3-year period. More details about the
analytical methods used for this analysis are
described in section A.6 of Appendix A in the 2022
PA (U.S. EPA, 2022b).
PO 00000
Frm 00018
Fmt 4701
Sfmt 4700
U.S. are small compared to
contributions from domestic
anthropogenic sources.44 For example,
based on zero-out modeling in the last
review of the PM NAAQS, annual
background PM2.5 concentrations were
estimated to range from 0.5–3 mg/m3
across the sites examined. In addition,
speciated monitoring data from
IMPROVE sites can provide some
insights into how contributions from
different sources, including sources of
background PM, may have changed over
time. Such data suggests the estimates of
background concentrations using
speciated monitoring data from
IMPROVE monitors are around 1–3 mg/
m3 and have not changed significantly
since the 2012 review. Contributions to
background PM in the U.S. result
mainly from sources within North
America. Contributions from
intercontinental events have also been
documented (e.g., transport from dust
storms occurring in deserts in North
Africa and Asia), but these events are
less frequent and represent a relatively
small fraction of background PM in
most of the U.S. (U.S. EPA, 2022b,
section 2.4).
II. Rationale for Decisions on the
Primary PM2.5 Standards
This section presents the rationale for
the Administrator’s decision to revise
the primary annual PM2.5 standard
down to a level of 9 mg/m3 and retain
the primary 24-hour PM2.5 standard.
This rationale is based on a thorough
review of the scientific evidence
generally published through January
2018,45 as evaluated in the 2019 ISA
(U.S. EPA, 2019a), on the human health
effects of PM2.5 associated with longand short-term exposures 46 to PM2.5 in
44 Sources that contribute to natural background
PM include dust from the wind erosion of natural
surfaces, sea salt, wildland fires, primary biological
aerosol particles such as bacteria and pollen,
oxidation of biogenic hydrocarbons such as
isoprene and terpenes to produce secondary organic
aerosols (SOA), and geogenic sources such as
sulfate formed from volcanic production of SO2 and
oceanic production of dimethyl-sulfide (U.S. EPA,
2022b, section 2.4). While most of these sources
release or contribute predominantly to fine aerosol,
some sources including windblown dust, and sea
salt also produce particles in the coarse size range
(U.S. EPA, 2019a, section 2.3.3).
45 In addition to the 2020 review’s opening ‘‘call
for information’’ (79 FR 71764, December 3, 2014),
the 2019 ISA identified and evaluated studies and
reports that have undergone scientific peer review
and were published or accepted for publication
between January 1, 2009, through approximately
January 2018 (U.S. EPA, 2019a, p. ES–2). References
that are cited in the 2019 ISA, the references that
were considered for inclusion but not cited, and
electronic links to bibliographic information and
abstracts can be found at: https://hero.epa.gov/hero/
particulate-matter.
46 Short-term exposures are defined as those
exposures occurring over hours up to 1 month,
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
the ambient air. Additionally, this
rationale is based on a thorough
evaluation of some studies that became
available after the literature cutoff date
of the 2019 ISA, as evaluated in the ISA
Supplement, that could either further
inform the adequacy of the current PM
NAAQS or address key scientific topics
that have evolved since the literature
cutoff date for the 2019 ISA, generally
through March 2021 (U.S. EPA,
2022a).47 The Administrator’s rationale
also takes into account: (1) The 2022 PA
evaluation of the policy-relevant
information in the 2019 ISA and ISA
Supplement and presentation of
quantitative analyses of air quality and
health risks; (2) CASAC advice and
recommendations; and (3) public
comments received during the
development of these documents.
In presenting the rationale for the
Administrator’s decisions and its
foundations, section II.A provides
background on the general approach for
this reconsideration and the basis for
the existing standard, and also presents
brief summaries of key aspects of the
currently available health effects and
risk information. Section II.B
summarizes the CASAC advice and the
basis for the proposed conclusions,
addresses public comments received on
the proposal and presents the
Administrator’s conclusions on the
adequacy of the current standards,
drawing on consideration of the
scientific evidence and quantitative risk
information, advice from the CASAC,
and comments from the public. Section
II.C summarizes the Administrator’s
decision on the primary PM2.5
standards.
ddrumheller on DSK120RN23PROD with RULES3
A. Introduction
The general approach for this
reconsideration of the 2020 final
decision on the primary PM2.5 standards
is fundamentally based on using the
EPA’s assessment of the current
scientific evidence and associated
quantitative analyses to inform the
Administrator’s judgment regarding
primary PM2.5 standards that protect
public health with an adequate margin
whereas long-term exposures are defined as those
exposures occurring over 1 month to years (U.S.
EPA, 2019a, section P.3.1).
47 The ISA Supplement represents an evaluation
of recent studies that are of greatest policy
relevance to the reconsideration of the 2020 final
decision on the PM NAAQS. Specifically, the ISA
Supplement focuses on studies of health effects for
which the evidence in the 2019 ISA supported a
‘‘causal relationship’’ (i.e., short- and long-term
PM2.5 exposure and mortality and cardiovascular
effects) because those were the health effects that
were most useful in informing conclusions in the
2020 PA. The ISA Supplement does not include an
evaluation of studies for other PM2.5-related health
effects (U.S. EPA, 2022a).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
of safety. The EPA’s assessments are
primarily documented in the 2019 ISA,
ISA Supplement, and 2022 PA, all of
which have received CASAC review and
public comment (83 FR 53471, October
23, 2018; 83 FR 55529, November 6,
2018; 85 FR 4655, January 27, 2020; 86
FR 52673, September 22, 2021; 86 FR
54186, September 30, 2021; 86 FR
56263, October 8, 2021; 87 FR 958,
January 7, 2022; 87 FR 22207, April 14,
2022; 87 FR 31965, May 26, 2022). In
bridging the gap between the scientific
assessments of the 2019 ISA and ISA
Supplement and the judgments required
of the Administrator in determining
whether the current standards provide
the requisite public health protection,
the 2022 PA evaluates policy
implications of the evaluation of the
current evidence in the 2019 ISA and
ISA Supplement, and the risk
information documented in the 2022
PA. In evaluating the public health
protection afforded by the current
standards, the four basic elements of the
NAAQS (i.e., indicator, averaging time,
level, and form) are considered
collectively.
The final decision on the adequacy of
the current primary PM2.5 standards is a
public health policy judgment to be
made by the Administrator. In reaching
conclusions with regard to the
standards, the decision will draw on the
scientific information and analyses
about health effects and population
risks, as well as judgments about how to
consider the range and magnitude of
uncertainties that are inherent in the
scientific evidence and analyses. This
approach is based on the recognition
that the available health effects evidence
generally reflects a continuum,
consisting of levels at which scientists
generally agree that health effects are
likely to occur, through lower levels at
which the likelihood and magnitude of
the response become increasingly
uncertain. This approach is consistent
with the requirements of the NAAQS
provisions of the Clean Air Act and with
how the EPA and the courts have
historically interpreted the Act
(summarized in section I.A above).
These provisions require the
Administrator to establish primary
standards that, in the judgment of the
Administrator, are requisite to protect
public health with an adequate margin
of safety. In so doing, the Administrator
seeks to establish standards that are
neither more nor less stringent than
necessary for this purpose. The Act does
not require that primary standards be set
at a zero-risk level, but rather at a level
that avoids unacceptable risks to public
PO 00000
Frm 00019
Fmt 4701
Sfmt 4700
16219
health, including the health of sensitive
(also referred to as ‘‘at-risk’’) groups.48
1. Background on the Current Standards
The current primary PM2.5 standards
were retained in 2020 based on the
scientific evidence and quantitative risk
information available at that time, as
well as the then-Administrator’s
judgments regarding the available health
effects evidence and the appropriate
degree of public health protection
afforded by the existing standards (85
FR 82718, December 18, 2020). With the
2020 decision, the then-Administrator
retained the primary annual PM2.5
standard with its level of 12.0 mg/m3
and retained the primary 24-hour PM2.5
standard with its level of 35 mg/m3. The
key considerations and the thenAdministrator’s conclusions regarding
the primary PM2.5 standards in the 2020
review are summarized below.
The health effects evidence base
available in the 2020 review included
extensive evidence from previous
reviews as well as the evidence that had
emerged since the prior review had been
completed in 2012. This evidence base,
spanning several decades, documents
the relationship between short- and
long-term PM2.5 exposure and mortality
or serious morbidity effects. The
evidence available in the 2019 ISA
reaffirmed, and in some cases
strengthened, the conclusions from the
2009 ISA regarding the health effects of
PM2.5 exposures (U.S. EPA, 2019a).
Much of the evidence came from
epidemiologic studies conducted in
North America, Europe, or Asia
examining short-term and long-term
exposures that demonstrated generally
positive, and often statistically
significant, PM2.5 health effect
associations with a range of outcomes
including non- accidental,
cardiovascular, or respiratory mortality;
cardiovascular- or respiratory-related
hospitalizations or emergency
department visits; and other mortality/
morbidity outcomes (e.g., lung cancer
mortality or incidence, asthma
development). Experimental evidence,
as well as evidence from panel studies,
strengthened support for potential
biological pathways through which
PM2.5 exposures could lead to health
effects reported in many populationbased epidemiologic studies, including
support for pathways that could lead to
cardiovascular, respiratory, nervous
system, and cancer-related effects.
48 As noted in section I.A above, the legislative
history describes such protection for the sensitive
group of individuals and not for a single person in
the sensitive group (see S. Rep. No. 91–1196, 91st
Cong, 2d Sess. 10 [1970]); see also Am. Lung Ass’n
v. EPA, 134 F.3d 388, 389 (D.C. Cir. 1998).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16220
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
Based on this evidence, the 2019 ISA
concluded there to be a causal
relationship between long- and shortterm PM2.5 exposure and mortality and
cardiovascular effects, as well as likely
to be causal relationships between longand short-term PM2.5 exposure and
respiratory effects, and between longterm PM2.5 exposure and cancer and
nervous system effects (U.S. EPA,
2019a, section 1.7).
Epidemiologic studies reported PM2.5
health effect associations with mortality
and/or morbidity across multiple U.S.
cities and in diverse populations,
including in studies examining
populations and lifestages that may be
at increased risk of experiencing a
PM2.5-related health effect (e.g., older
adults, children). The 2019 ISA cited
extensive evidence indicating that ‘‘both
the general population as well as
specific populations and lifestages are at
risk for PM2.5-related health effects’’
(U.S. EPA, 2019a, p. 12–1), including
children and older adults, people with
pre-existing respiratory or
cardiovascular disease, minority
populations, and low socioeconomic
status (SES) populations.
The risk information available in the
2020 review included risk estimates for
air quality conditions just meeting the
existing primary PM2.5 standards, and
also for air quality conditions just
meeting potential alternative standards.
The general approach to estimating
PM2.5-associated health risks combined
concentration-response (C–R) functions
from epidemiologic studies with modelbased PM2.5 air quality surfaces,
baseline health incidence data, and
population demographics for 47 urban
areas (U.S. EPA, 2020b, section 3.3,
Figure 3–10, Appendix C). The risk
assessment estimated that the existing
primary PM2.5 standards could allow a
substantial number of PM2.5-associated
deaths in the U.S. Uncertainty in risk
estimates (e.g., in the size of risk
estimates) can result from a number of
factors, including assumptions about the
shape of the C–R relationship with
mortality at low ambient PM2.5
concentrations, the potential for
confounding and/or exposure
measurement error, and the methods
used to adjust PM2.5 air quality.
Consistent with the general approach
routinely employed in NAAQS reviews,
the initial consideration in the 2020
review of the primary PM2.5 standards
was with regard to the adequacy of the
protection provided by the existing
standards.
As an initial matter, the thenAdministrator considered the range of
scientific evidence evaluating these
effects, including studies of at-risk
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
populations, to inform his review of the
primary PM2.5 standards, placing the
greatest weight on evidence of effects for
which the 2019 ISA determined there to
be a causal or likely to be causal
relationship with long- and short-term
PM2.5 exposures (85 FR 82714–82715,
December 18, 2020).
With regard to indicator, the thenAdministrator recognized that,
consistent with the evidence available
in prior reviews, the scientific evidence
continued to provide strong support for
health effects following short- and longterm PM2.5 exposures. He noted the
2020 PA conclusions that the
information continued to support the
PM2.5 mass-based indicator and
remained too limited to support a
distinct standard for any specific PM2.5
component or group of components, and
too limited to support a distinct
standard for the ultrafine fraction. Thus,
the then-Administrator concluded that
it was appropriate to retain PM2.5 as the
indicator for the primary standards for
fine particles (85 FR 82715, December
18, 2020).
With respect to averaging time and
form, the then-Administrator noted that
the scientific evidence continued to
provide strong support for health effects
associations with both long-term (e.g.,
annual or multi-year) and short-term
(e.g., mostly 24-hour) exposures to
PM2.5, consistent with the conclusions
in the 2020 PA. In the 2019 ISA,
epidemiologic and controlled human
exposure studies examined a variety of
PM2.5 exposure durations.
Epidemiologic studies continued to
provide strong support for health effects
associated with short-term PM2.5
exposures based on 24-hour PM2.5
averaging periods, and the EPA noted
that associations with subdaily
estimates are less consistent and, in
some cases, smaller in magnitude (U.S.
EPA, 2019a, section 1.5.2.1; U.S. EPA,
2020b, section 3.5.2.2). In addition,
controlled human exposure and panelbased studies of subdaily exposures
typically examined subclinical effects,
rather than the more serious populationlevel effects that have been reported to
be associated with 24-hour exposures
(e.g., mortality, hospitalizations). Taken
together, the 2019 ISA concluded that
epidemiologic studies did not indicate
that subdaily averaging periods were
more closely associated with health
effects than the 24-hour average
exposure metric (U.S. EPA, 2019a,
section 1.5.2.1). Additionally, while
controlled human exposure studies
provided consistent evidence for
cardiovascular effects following PM2.5
exposures for less than 24 hours (i.e.,
<30 minutes to 5 hours), exposure
PO 00000
Frm 00020
Fmt 4701
Sfmt 4700
concentrations in the studies were wellabove the ambient concentrations
typically measured in locations meeting
the existing standards (U.S. EPA, 2020b,
section 3.2.3.1). Thus, these studies also
did not suggest the need for additional
protection against subdaily PM2.5
exposures (U.S. EPA, 2020b, section
3.5.2.2). Therefore, the thenAdministrator judged that the 24-hour
averaging time remained appropriate (85
FR 82715, December 18, 2020).
With regard to the form of the 24-hour
standard (98th percentile, averaged over
three years), the then-Administrator
noted that epidemiologic studies
continued to provide strong support for
health effect associations with shortterm (e.g., mostly 24-hour) PM2.5
exposures (U.S. EPA, 2020b, section
3.5.2.3) and that controlled human
exposure studies provided evidence for
health effects following single shortterm ‘‘peak’’ PM2.5 exposures. Thus, the
evidence supported retaining a standard
focused on providing supplemental
protection against short-term peak
exposures and supported a 98th
percentile form for a 24-hour standard.
The then-Administrator further noted
that this form also provided an
appropriate balance between limiting
the occurrence of peak 24-hour PM2.5
concentrations and identifying a stable
target for risk management programs
(U.S. EPA, 2020b, section 3.5.2.3). As
such, the then-Administrator concluded
that the available information supported
retaining the form and averaging time of
the current 24-hour standard (98th
percentile, averaged over three years)
and annual standard (annual average,
averaged over three years) (85 FR 82715,
December 18, 2020).
With regard to the level of the
standards, in reaching his final decision,
the then-Administrator considered the
large body of evidence presented and
assessed in the 2019 ISA (U.S. EPA,
2019a), the policy-relevant and riskbased conclusions and rationales as
presented in the 2020 PA (U.S. EPA,
2020b), advice from the CASAC, and
public comments. In particular, in
considering the 2019 ISA and 2020 PA,
he considered key epidemiologic
studies that evaluated associations
between PM2.5 air quality distributions
and mortality and morbidity, including
key accountability studies; the
availability of experimental studies to
support biological plausibility;
controlled human exposure studies
examining effects following short-term
PM2.5 exposures; air quality analyses;
and the important uncertainties and
limitations associated with the
information (85 FR 82715, December 18,
2020).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
As an initial matter, the thenAdministrator considered the protection
afforded by both the annual and 24-hour
standards together against long- and
short-term PM2.5 exposures and health
effects. The Administrator recognized
that the annual standard was most
effective in controlling ‘‘typical’’ PM2.5
concentrations near the middle of the
air quality distribution (i.e., around the
mean of the distribution), but also
provided some control over short-term
peak PM2.5 concentrations. On the other
hand, the 24-hour standard, with its
98th percentile form, was most effective
at limiting peak 24-hour PM2.5
concentrations, but in doing so also had
an effect on annual average PM2.5
concentrations. Thus, while either
standard could be viewed as providing
some measure of protection against both
average exposures and peak exposures,
the 24-hour and annual standards were
not expected to be equally effective at
limiting both types of exposures. Thus,
consistent with previous reviews, the
then-Administrator’s consideration of
the public health protection provided by
the existing primary PM2.5 standards
was based on his consideration of the
combination of the annual and 24-hour
standards. Specifically, he recognized
that the annual standard was more
likely to appropriately limit the
‘‘typical’’ daily and annual exposures
that are most strongly associated with
the health effects observed in
epidemiologic studies. The thenAdministrator concluded that an annual
standard (as the arithmetic mean,
averaged over three years) remained
appropriate for targeting protection
against the annual and daily PM2.5
exposures around the middle portion of
the PM2.5 air quality distribution.
Further, recognizing that the 24-hour
standard (with its 98th percentile form)
was more directly tied to short-term
peak PM2.5 concentrations, and more
likely to appropriately limit exposures
to such concentrations, the thenAdministrator concluded that the
current 24-hour standard (with its 98th
percentile form, averaged over three
years) remained appropriate to provide
a balance between limiting the
occurrence of peak 24-hour PM2.5
concentrations and identifying a stable
target for risk management programs.
However, the then-Administrator
recognized that changes in PM2.5 air
quality to meet an annual standard
would likely result not only in lower
short- and long-term PM2.5
concentrations near the middle of the
air quality distribution, but also in fewer
and lower short-term peak PM2.5
concentrations. The then-Administrator
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
further recognized that changes in air
quality to meet a 24-hour standard, with
a 98th percentile form, would result not
only in fewer and lower peak 24-hour
PM2.5 concentrations, but also in lower
annual average PM2.5 concentrations (85
FR 82715–82716, December 18, 2020).
Thus, in considering the adequacy of
the 24-hour standard, the thenAdministrator noted the importance of
considering whether additional
protection was needed against shortterm exposures to peak PM2.5
concentrations. In examining the
scientific evidence, he noted the limited
utility of the animal toxicological
studies in directly informing
conclusions on the appropriate level of
the standard given the uncertainty in
extrapolating from effects in animals to
those in human populations. The thenAdministrator noted that controlled
human exposure studies provided
evidence for health effects following
single, short-term PM2.5 exposures that
corresponded best to exposures that
might be experienced in the upper end
of the PM2.5 air quality distribution in
the U.S. (i.e., ‘‘peak’’ concentrations).
However, most of these studies
examined exposure concentrations
considerably higher than are typically
measured in areas meeting the standards
(U.S. EPA, 2020b, section 3.2.3.1). In
particular, controlled human exposure
studies often reported statistically
significant effects on one or more
indicators of cardiovascular function
following 2-hour exposures to PM2.5
concentrations at and above 120 mg/m3
(at and above 149 mg/m3 for vascular
impairment, the effect shown to be most
consistent across studies). To provide
insight into what these studies may
indicate regarding the primary PM2.5
standards, the 2020 PA (U.S. EPA,
2020b, p. 3–49) noted that 2-hour
ambient concentrations of PM2.5 at
monitoring sites meeting the current
standards almost never exceeded 32 mg/
m3. In fact, even the extreme upper end
of the distribution of 2-hour PM2.5
concentrations at sites meeting the
primary PM2.5 standards remained wellbelow the PM2.5 exposure
concentrations consistently shown in
controlled human exposure studies to
elicit effects (i.e., 99.9th percentile of 2hour concentrations at these sites is 68
mg/m3 during the warm season). Thus,
the experimental evidence did not
indicate the need for additional
protection against exposures to peak
PM2.5 concentrations, beyond the
protection provided by the combination
of the 24-hour and the annual standards
(U.S. EPA, 2020b, section 3.2.3.1; 85 FR
82716, December 18, 2020).
PO 00000
Frm 00021
Fmt 4701
Sfmt 4700
16221
With respect to the epidemiologic
evidence, the then-Administrator noted
that the studies did not indicate that
associations in those studies were
strongly influenced by exposures to
peak concentrations in the air quality
distribution and thus did not indicate
the need for additional protection
against short-term exposures to peak
PM2.5 concentrations (U.S. EPA, 2020b,
section 3.5.1). The then-Administrator
noted that this was consistent with
CASAC consensus support for retaining
the current 24-hour standard. Thus, the
then-Administrator concluded that the
24-hour standard with its level of 35 mg/
m3 was adequate to provide
supplemental protection (i.e., beyond
that provided by the annual standard
alone) against short-term exposures to
peak PM2.5 concentrations (85 FR 82716,
December 18, 2020).
With regard to the level of the annual
standard, the then-Administrator
recognized that the annual standard,
with its form based on the arithmetic
mean concentration, was most
appropriately meant to limit the
‘‘typical’’ daily and annual exposures
that were most strongly associated with
the health effects observed in
epidemiologic studies. However, the
then-Administrator also noted that
while epidemiologic studies examined
associations between distributions of
PM2.5 air quality and health outcomes,
they did not identify particular PM2.5
exposures that cause effects and thus,
they could not alone identify a specific
level at which the standard should be
set, as such a determination necessarily
required the then-Administrator’s
judgment. Thus, consistent with the
approaches in previous NAAQS
reviews, the then-Administrator
recognized that any approach that used
epidemiologic information in reaching
decisions on what standards are
appropriate necessarily required
judgments about how to translate the
information from the epidemiologic
studies into a basis for appropriate
standards. This approach included
consideration of the uncertainties in the
reported associations between daily or
annual average PM2.5 exposures and
mortality or morbidity in the
epidemiologic studies. Such an
approach is consistent with setting
standards that are neither more nor less
stringent than necessary, recognizing
that a zero-risk standard is not required
by the Clean Air Act (CAA) (85 FR
82716, December 18, 2020).
The then-Administrator emphasized
uncertainties and limitations that were
present in epidemiologic studies in
previous reviews and persisted in the
2020 review. These uncertainties
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16222
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
included exposure measurement error,
potential confounding by copollutants,
increasing uncertainty of associations at
lower PM2.5 concentrations, and
heterogeneity of effects across different
cities or regions (85 FR 82716,
December 18, 2020). The thenAdministrator also noted the advice
given by the CASAC on this matter. As
described in section I.C.5 above, the
CASAC did not reach consensus on the
adequacy of the primary annual PM2.5
standard. ‘‘Some CASAC members’’
expressed support for retaining the
primary annual PM2.5 standard while
‘‘other members’’ expressed support for
revising that standard in order to
increase public health protection (Cox,
2019b, p. 1 of consensus letter). The
CASAC members who supported
retaining the annual standard expressed
their concerns with the epidemiologic
studies, asserting that these studies did
not provide a sufficient basis for
revising the existing standards. They
also identified several key concerns
regarding the associations reported in
epidemiologic studies and concluded
that ‘‘while the data on associations
should certainly be carefully
considered, this data should not be
interpreted more strongly than
warranted based on its methodological
limitations’’ (Cox, 2019b, p. 8 consensus
responses).
Taking into consideration the views
expressed by the CASAC members who
supported retaining the annual
standard, the then-Administrator
recognized that epidemiologic studies
examined associations between
distributions of PM2.5 air quality and
health outcomes, and they did not
identify particular PM2.5 exposures that
cause effects (U.S. EPA, 2020b, section
3.1.2). While the Administrator
remained concerned about placing too
much weight on epidemiologic studies
to inform conclusions on the adequacy
of the primary standards, he noted the
approach to considering such studies in
the 2012 review. In the 2012 review, it
was noted that the evidence of an
association in any epidemiologic study
was ‘‘strongest at and around the longterm average where the data in the study
are most concentrated’’ (78 FR 3140,
January 15, 2013). In considering the
characterization of epidemiologic
studies, the then-Administrator viewed
that when assessing the mean
concentrations of the key short-term and
long-term epidemiologic studies in the
U.S. that used ground-based monitoring
(i.e., those studies where the mean is
most directly comparable to the current
annual standard), the majority of studies
had mean concentrations at or above the
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
level of the existing annual standard,
with the mean of the study-reported
means or medians equal to 13.5 mg/m3,
a concentration level above the existing
level of the primary annual standard of
12 mg/m3. The then-Administrator
further noted his caution in directly
comparing the reported study mean
values to the standard level given that
study-reported mean concentrations, by
design, are generally lower than the
design value of the highest monitor in
an area, which determines compliance.
In the 2020 PA, analyses of recent air
quality in U.S. CBSAs indicated that
maximum annual PM2.5 design values
for a given three-year period were often
10% to 20% higher than average
monitored concentrations (i.e., averaged
across multiple monitors in the same
CBSA) (U.S. EPA, 2020b, Appendix B,
section B.7). He further noted his
concern in placing too much weight on
any one epidemiologic study but instead
judged that it was more appropriate to
focus on the body of studies together
and therefore noted the calculation of
the mean of study-reported means (or
medians). Thus, while the thenAdministrator was cautious in placing
too much weight on the epidemiologic
evidence alone, he noted that: (1) The
reported mean concentration in the
majority of the key U.S. epidemiologic
studies using ground-based monitoring
data were above the level of the existing
annual standard; (2) the mean of the
reported study means (or medians) (i.e.,
13.5 mg/m3) was above the level of the
current standard; 49 (3) air quality
analyses showed the study means to be
lower than their corresponding design
values by 10–20%; and (4) these
analyses must be considered in light of
uncertainties inherent in the
epidemiologic evidence. When taken
together, the then-Administrator judged
that, even if it were appropriate to place
more weight on the epidemiologic
evidence, this information did not call
into question the adequacy of the
current standards (85 FR 82716–17,
December 18, 2020).
In addition to the evidence, the thenAdministrator also considered the
potential implications of the risk
assessment. He noted that all risk
assessments have limitations and that
he remained concerned about the
uncertainties in the underlying
epidemiologic data used in the risk
assessment. The then-Administrator
also noted that in previous reviews,
these uncertainties and limitations have
often resulted in less weight being
49 The median of the study-reported mean (or
median) PM2.5 concentrations is 13.3 mg/m3, which
was also above the level of the existing standard.
PO 00000
Frm 00022
Fmt 4701
Sfmt 4700
placed on quantitative estimates of risk
than on the underlying scientific
evidence itself (e.g., 78 FR 3086, 3098–
99, January 15, 2013). These
uncertainties and limitations included
uncertainty in the shapes of C–R
functions, particularly at low
concentrations; uncertainties in the
methods used to adjust air quality; and
uncertainty in estimating risks for
populations, locations and air quality
distributions different from those
examined in the underlying
epidemiologic study (U.S. EPA, 2020b,
section 3.3.2.4). Additionally, the thenAdministrator noted similar concern
expressed by some members of the
CASAC who support retaining the
existing standards; they highlighted
similar uncertainties and limitations in
the risk assessment (Cox, 2019b). In
light of all of this, the thenAdministrator judged it appropriate to
place little weight on quantitative
estimates of PM2.5-associated mortality
risk in reaching conclusions about the
level of the primary PM2.5 standards (85
FR 82717, December 18, 2020).
The then-Administrator additionally
considered an emerging body of
evidence from accountability studies
that examined past reductions in
ambient PM2.5 and the degree to which
those reductions resulted in public
health improvements. While the thenAdministrator agreed with public
commenters that well-designed and
conducted accountability studies can be
informative, he viewed the
interpretation of such studies in the
context of the primary PM2.5 standards
as complicated by the fact that some of
the available studies had not evaluated
PM2.5 specifically (e.g., as opposed to
PM10 or total suspended particulates),
did not show changes in PM2.5 air
quality, or had not been able to
disentangle health impacts of the
interventions from background trends in
health (U.S. EPA, 2020b, section 3.5.1).
He further recognized that the small
number of available studies that did
report public health improvements
following past declines in ambient PM2.5
had not examined air quality meeting
the existing standards (U.S. EPA, 2020b,
Table 3–3). This included U.S. studies
that reported increased life expectancy,
decreased mortality, and decreased
respiratory effects following past
declines in ambient PM2.5
concentrations. Such studies examined
‘‘starting’’ annual average PM2.5
concentrations (i.e., prior to the
reductions being evaluated) ranging
from about 13.2 to >20mg/m3 (i.e., U.S.
EPA, 2020b, Table 3–3). Given the lack
of available accountability studies
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
reporting public health improvements
attributable to reductions in ambient
PM2.5 in locations meeting the existing
standards, together with his broader
concerns regarding the lack of
experimental studies examining PM2.5
exposures typical of areas meeting the
existing standards, the thenAdministrator judged that there was
considerable uncertainty in the
potential for increased public health
protection from further reductions in
ambient PM2.5 concentrations beyond
those achieved under the existing
primary PM2.5 standards (85 FR 82717,
December 18, 2020).
When the above considerations were
taken together, the then-Administrator
concluded that the scientific evidence
assessed in the 2019 ISA, together with
the analyses in the 2020 PA based on
that evidence and consideration of
CASAC advice and public comments,
did not call into question the adequacy
of the public health protection provided
by the existing annual and 24-hour
PM2.5 standards. In particular, the thenAdministrator judged that there was
considerable uncertainty in the
potential for additional public health
improvements from reducing ambient
PM2.5 concentrations below the
concentrations achieved under the
existing primary standards and that,
therefore, standards more stringent than
the existing standards (e.g., with lower
levels) were not supported. That is, he
judged that more stringent standards
would be more than requisite to protect
the public health with an adequate
margin of safety. This judgment
reflected the Administrator’s
consideration of the uncertainties in the
potential implications of the lower end
of the air quality distributions from the
epidemiologic studies due in part to the
lack of supporting evidence from
experimental studies and retrospective
accountability studies conducted at
PM2.5 concentrations meeting the
existing standards (85 FR 82717,
December 18, 2020).
In reaching this conclusion in the
2020 review, the then-Administrator
judged that the existing standards
provided an adequate margin of safety.
With respect to the annual standard, the
level of 12 mg/m3 was below the lowest
‘‘starting’’ concentration (i.e., 13.2 mg/
m3) in the available accountability
studies that showed public health
improvements attributable to reductions
in ambient PM2.5. In addition, while the
then-Administrator placed less weight
on the epidemiologic evidence for
selecting a standard, he noted that the
level of the annual standard was below
the reported mean (and median)
concentrations in the majority of the key
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
U.S. epidemiologic studies using
ground-based monitoring data (noting
that these means tend to be 10–20%
lower than their corresponding area
design values which is the more
relevant metric when considering the
level of the standard) and below the
mean of the reported means (or
medians) of these studies (i.e., 13.5 mg/
m3). In addition, the then-Administrator
recognized that concentrations in areas
meeting the existing 24-hour and annual
standards remained well-below the
PM2.5 exposure concentrations
consistently shown to elicit effects in
human exposure studies (85 FR 82717–
82718, December 18, 2020).
In addition, based on the thenAdministrator’s review of the science in
the 2020 review, including controlled
human exposure studies examining
effects following short-term PM2.5
exposures, the epidemiologic studies,
and accountability studies conducted at
levels just above the existing annual
standard, he judged that the degree of
public health protection provided by the
existing annual standard is not greater
than warranted. This judgment, together
with the fact that no CASAC member
expressed support for a less stringent
standard, led the then- Administrator to
conclude that standards less stringent
than the existing standards (e.g., with
higher levels) were also not supported
(85 FR 82718, December 18, 2020).
In reaching his final decision in the
2020 review, the then-Administrator
concluded that the scientific evidence
and technical information continued to
support the existing annual and 24-hour
PM2.5 standards. This conclusion
reflected the then-Administrator’s view
that there were important limitations
and uncertainties that remained in the
evidence. The then-Administrator
concluded that these limitations
contributed to considerable uncertainty
regarding the potential public health
implications of revising the existing
primary PM2.5 standards. Given this
uncertainty, and noting the advice from
some CASAC members, he concluded
that the primary PM2.5 standards,
including the indicators (PM2.5),
averaging times (annual and 24-hour),
forms (arithmetic mean and 98th
percentile, averaged over three years)
and levels (12.0 mg/m3, 35 mg/m3), when
taken together, remained requisite to
protect the public health. Therefore, in
the 2020 review, the Administrator
reached the conclusion that the primary
24-hour and annual PM2.5 standards,
together, were requisite to protect public
health from fine particles with an
adequate margin of safety, including the
health of at-risk populations, and
PO 00000
Frm 00023
Fmt 4701
Sfmt 4700
16223
retained the standards, without revision
(85 FR 82718, December 18, 2020).
2. Overview of the Health Effects
Evidence
The information summarized here
and further detailed in section II.B of
the proposal (88 FR 5580, January 27,
2023), is an overview of the policyrelevant aspects of the health effects
evidence available in this
reconsideration; the assessment of this
evidence is documented in the 2019 ISA
(U.S. EPA, 2019a) and ISA Supplement
(U.S. EPA, 2022a) and its policy
implications are further discussed in the
2022 PA (U.S. EPA, 2022b). While the
2019 ISA provides the broad scientific
foundation for this reconsideration,
additional literature has become
available since the cutoff date of the
2019 ISA that expands the body of
evidence related to mortality and
cardiovascular effects for both shortand long-term PM2.5 exposure, which
can inform the Administrator’s
judgment on the adequacy of the current
primary PM2.5 standards. As such, the
ISA Supplement builds on the
information presented within the 2019
ISA with a targeted identification and
evaluation of new scientific information
(U.S. EPA, 2022a, section 1.2). The ISA
Supplement focuses on PM2.5 health
effects evidence where the 2019 ISA
concludes a ‘‘causal relationship,’’
because such health effects are given the
most weight in an Administrator’s
decisions in a NAAQS review. As such,
in selecting the health effects to evaluate
within the ISA Supplement (i.e., newly
available evidence related to short- and
long-term PM2.5 exposure and mortality
and cardiovascular effects), the primary
rationale is based on the causality
determinations for health effect
categories presented in the 2019 PM
ISA, and the subsequent use of the
health effects evidence in the 2020 PM
PA. Specifically, U.S. and Canadian
epidemiologic studies for mortality and
cardiovascular effects, along with
controlled human exposure studies
associated with cardiovascular effects at
near ambient concentrations, were
considered to be of greatest utility in
informing the Administrator’s
conclusions on the adequacy of the
current primary PM2.5 standards.
Additionally, studies examining
associations outside the U.S. or Canada
reflect air quality and exposure patterns
that may be less typical of the U.S., and
thus less likely to be informative for
purposes of reviewing the NAAQS (U.S.
EPA, 2022b, p.1–3). While the ISA
Supplement does not include
information for health effects other than
mortality and cardiovascular effects, the
E:\FR\FM\06MRR3.SGM
06MRR3
16224
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
scientific evidence for other health
effect categories is evaluated in the 2019
ISA, which in combination with the ISA
Supplement represents the complete
scientific record for the reconsideration
of the 2020 final decision.
The ISA Supplement also assessed
accountability studies because these
types of epidemiologic studies were part
of the body of evidence that was a focus
of the 2020 review. Accountability
studies inform our understanding of the
potential for public health
improvements as ambient PM2.5
concentrations have declined over time.
Further, the ISA Supplement considered
studies that employed statistical
approaches that attempt to more
extensively account for confounders and
are more robust to model
misspecification (i.e., used alternative
methods for confounder control),50
given that such studies were highlighted
by the CASAC and identified in public
comments in the 2020 review. Since the
literature cutoff date for the 2019 ISA,
multiple accountability studies and
studies that employ alternative methods
for confounder control have become
available for consideration in the ISA
Supplement and, subsequently, in this
reconsideration.
The ISA Supplement also considered
recent health effects evidence that
addresses key scientific issues where
the literature has expanded since the
completion of the 2019 ISA.51 The 2019
ISA evaluated a couple of controlled
human exposure studies that
investigated the effect of exposure to
near-ambient concentrations of PM2.5
(U.S. EPA, 2019a, section 6.1.10 and
6.1.13). The ISA Supplement adds to
this limited evidence, including a recent
study conducted in young healthy
individuals exposed to near-ambient
PM2.5 concentrations (U.S. EPA, 2022a,
section 3.3.1). Given the importance of
identifying populations at increased risk
of PM2.5-related effects, the ISA
Supplement also included
50 As noted in the ISA Supplement (U.S. EPA,
2022a, p. 1–3): ‘‘In the peer-reviewed literature,
these epidemiologic studies are often referred to as
causal inference studies or studies that used causal
modeling methods. For the purposes of this
Supplement, this terminology is not used to prevent
confusion with the main scientific conclusions (i.e.,
the causality determinations) presented within an
ISA. In addition, as is consistent with the weightof-evidence framework used within ISAs and
discussed in the Preamble to the Integrated Science
Assessments, an individual study on its own cannot
inform causality, but instead represents a piece of
the overall body of evidence.’’
51 As with the epidemiologic studies for long- and
short-term PM2.5 exposure and mortality and
cardiovascular effects, epidemiologic studies of
exposure or risk disparities and SARS–CoV–2
infection and/or COVID–19 death were limited to
those conducted in the U.S. and Canada.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
epidemiologic or exposure studies that
examined whether there is evidence of
exposure or risk disparities by race/
ethnicity or SES. These types of studies
provide additional information related
to factors that may increase risk of
PM2.5-related health effects and provide
additional evidence for consideration by
the Administrator in reaching
conclusions regarding the adequacy of
the current standards. In addition, the
ISA Supplement evaluated studies that
examined the relationship between
short- and long-term PM2.5 exposures
and SARS-CoV–2 infection and/or
COVID–19 death, as these studies are a
new area of research and were raised by
a number of public commenters in the
2020 review.
The evidence presented within the
2019 ISA, along with the targeted
identification and evaluation of new
scientific information in the ISA
Supplement, provides the scientific
basis for the reconsideration of the 2020
final decision on the primary PM2.5
standards. The subsections below
briefly summarize the nature of PM2.5related health effects (II.A.2.a), with a
focus on those health effects for which
the 2019 ISA concluded a ‘‘causal’’ or
‘‘likely to be causal’’ relationship, the
potential public health implications and
populations at risk (II.A.2.b), and PM2.5
concentrations in key studies reporting
health effects (II.A.2.c).
a. Nature of Effects
The evidence base available in the
reconsideration includes decades of
research on PM2.5-related health effects
(U.S. EPA, 2004b; U.S. EPA, 2009a; U.S.
EPA, 2019a), including the full body of
evidence evaluated in the 2019 ISA
(U.S. EPA, 2019a), along with the
targeted evaluation of recent evidence in
the ISA Supplement (U.S. EPA, 2022a).
In considering the available scientific
evidence, the sections below, and in
more detail in section II.B.1 of the
proposal (88 FR 5580, January 27, 2023),
summarize the relationships between
long-and short-term PM2.5 exposures
and mortality (II.A.2.a.i), cardiovascular
effects (II.A.2.a.ii), respiratory effects
(II.A.2.a.iii), cancer (II.A.2.a.iv), nervous
system effects (II.A.2.a.v) and other
effects (II.A.2.a.vi). For these outcomes,
the 2019 ISA concluded that the
evidence supports either a ‘‘causal’’ or
a ‘‘likely to be causal’’ relationship.52
52 In this reconsideration of the PM NAAQS, the
EPA considers the full body of health evidence,
placing the greatest emphasis on the health effects
for which the evidence has been judged in the 2019
ISA to demonstrate a ‘‘causal’’ or ‘‘likely to be
causal’’ relationship with PM2.5 exposures.
PO 00000
Frm 00024
Fmt 4701
Sfmt 4700
i. Mortality
Long-Term PM2.5 Exposures
In the 2012 review, the 2009 ISA
reported that the evidence was
‘‘sufficient to conclude that the
relationship between long-term PM2.5
exposures and mortality is causal’’ (U.S.
EPA, 2009a, p. 7–96). The strongest
evidence supporting this conclusion
was provided by epidemiologic studies,
particularly those examining two
seminal cohorts, the American Cancer
Society (ACS) cohort and the Harvard
Six Cities cohort. Analyses of the
Harvard Six Cities cohort included
evidence indicating that reductions in
ambient PM2.5 concentrations are
associated with reduced mortality risk
(Laden et al., 2006) and increases in life
expectancy (Pope et al., 2009). Further
support was provided by other cohort
studies conducted in North America
and Europe that reported positive
associations between long-term PM2.5
exposure and mortality (U.S. EPA,
2019a).
Cohort studies, which have become
available since the completion of the
2009 ISA and evaluated in the 2019 ISA,
continue to provide consistent evidence
of positive associations between longterm PM2.5 exposures and mortality.
These studies add support for
associations with all-cause and total
(non-accidental) mortality,53 as well as
with specific causes of mortality,
including cardiovascular disease and
respiratory disease (U.S. EPA, 2019a,
section 11.2.2). Several of these studies
conducted analyses over longer study
durations and periods of follow-up than
examined in the original ACS and
Harvard Six Cities cohort studies and
continue to report positive associations
between long-term exposure to PM2.5
and mortality (U.S. EPA, 2019a, section
11.2.2.1; Figures 11–18 and 11–19). In
addition to studies focusing on the ACS
and Harvard Six Cities cohorts,
additional studies examining other
cohorts also provide evidence of
consistent, positive associations
between long-term PM2.5 exposure and
mortality across a wide range of
demographic groups (e.g., age, sex,
occupation), spatial and temporal
extents, exposure assessment metrics,
and statistical techniques (U.S. EPA,
2019a, sections 11.2.2.1, 11.2.5; U.S.
EPA, 2022a, Table 11–8). This includes
some of the largest cohort studies
conducted to date, such as analyses of
the U.S. Medicare cohort that includes
53 The majority of these studies examined nonaccidental mortality outcomes, though some
Medicare studies lack cause-specific death
information and, therefore, examine total mortality.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
nearly 61 million enrollees and studies
that control for a range of individual
and ecological covariates, including
race, age, SES, smoking status, body
mass index, and annual weather
variables (e.g., temperature, humidity)
(U.S. EPA, 2019a).
In addition to those cohort studies
evaluated in the 2019 ISA, recent North
American cohort studies evaluated in
the ISA Supplement continue to
examine the relationship between longterm PM2.5 exposure and mortality and
report consistent, positive, and
statistically significant associations.
These recent studies also utilize large
and demographically diverse cohorts
that are generally representative of the
national populations in both the U.S.
and Canada. These ‘‘studies published
since the 2019 ISA support and extend
the evidence base that contributed to the
conclusion of a causal relationship
between long-term PM2.5 exposure and
mortality’’ (U.S. EPA, 2022a, section
3.2.2.2.1, Figure 3–19, Figure 3–20).
Furthermore, studies evaluated in the
2019 ISA and the ISA Supplement that
examined cause-specific mortality
expand upon previous research that
found consistent, positive associations
between PM2.5 exposure and specific
mortality outcomes, which include
cardiovascular and respiratory
mortality, as well as other mortality
outcomes. For cardiovascular-related
mortality, the evidence evaluated in the
ISA Supplement is consistent with the
evidence evaluated in the 2019 ISA with
recent studies reporting positive
associations with long-term PM2.5
exposure. When evaluating causespecific cardiovascular mortality, recent
studies reported positive associations
for a number of outcomes, such as
ischemic heart disease (IHD) and stroke
mortality (U.S. EPA, 2022a, Figure 3–
23). Moreover, recent studies also
provide some initial evidence that
individuals with pre-existing health
conditions, such as heart failure and
diabetes, are at an increased risk of
PM2.5-related health effects (U.S. EPA,
2022a, section 3.2.2.4) and that these
individuals have a higher risk of
mortality overall, which was previously
only examined in studies that used
stratified analyses rather than a cohort
of people with an underlying health
condition (U.S. EPA, 2022a, section
3.2.2.4). With regard to respiratory
mortality, epidemiologic studies
evaluated in the 2019 ISA and ISA
Supplement continue to provide
support for associations between longterm PM2.5 exposure and respiratory
mortality (U.S. EPA, 2019a, section
5.2.10; U.S. EPA, 2022a, Table 3–2).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
A series of epidemiologic studies
evaluated in the 2019 ISA tested the
hypothesis that past reductions in
ambient PM2.5 concentrations are
associated with increased life
expectancy or a decreased mortality rate
and report that reductions in ambient
PM2.5 are associated with improvements
in longevity (U.S. EPA, 2022a, section
11.2.2.5). Pope et al. (2009) conducted a
cross-sectional analysis using air quality
data from 51 metropolitan areas across
the U.S., beginning in the 1970s through
the early 2000s, and found that a 10 mg/
m3 decrease in long-term PM2.5
concentration was associated with a
0.61-year increase in life expectancy. In
a subsequent analysis, the authors
extended the period of analysis to
include 2000 to 2007, a time period
with lower ambient PM2.5
concentrations and found a decrease in
long-term PM2.5 concentration
continued to be associated with an
increase in life expectancy, though the
magnitude of the increase was smaller
than during the earlier time period (i.e.,
a 10 mg/m3 decrease in long-term PM2.5
concentration was associated with a
0.35-year increase in life expectancy)
(Correia et al., 2013). Additional studies
conducted in the U.S. or Europe
similarly report that reductions in
ambient PM2.5 are associated with
improvements in longevity (U.S. EPA,
2022a, section 11.2.2.5).
Since the literature cutoff date for the
2019 ISA, a few epidemiologic studies
were published that examined the
relationship between long-term PM2.5
exposure and life-expectancy (U.S. EPA,
2022a, section 3.2.1.3) and report results
that are consistent with and expand
upon the body of evidence from the
2019 ISA. For example, Bennett et al.
(2019) reported that PM2.5
concentrations above the lowest
observed concentration (2.8 mg/m3) were
associated with a 0.15 year decrease in
national life expectancy for women and
0.13 year decrease in national life
expectancy for men (U.S. EPA, 2022a,
section 3.2.2.2.4, Figure 3–25). Another
study compared participants living in
areas with PM2.5 concentrations >12 mg/
m3 to participants living in areas with
PM2.5 concentrations <12 mg/m3 and
reported that the number of years of life
lost due to living in areas with higher
PM2.5 concentrations was 0.84 years
over a 5-year period (Ward-Caviness et
al., 2020; U.S. EPA, 2022a, section
3.2.2.2.4).
Additionally, a number of
accountability studies, which are
epidemiologic studies that evaluate
whether an environmental policy or air
quality intervention resulted in
reductions in ambient air pollution
PO 00000
Frm 00025
Fmt 4701
Sfmt 4700
16225
concentrations and subsequent
reductions in mortality or morbidity,
have emerged and were evaluated in the
ISA Supplement (U.S. EPA, 2022a,
section 3.2.2.3). For example, Sanders et
al. (2020a) examined whether policy
actions (i.e., the first annual PM2.5
NAAQS implementation rule in 2005
for the 1997 annual PM2.5 standard with
a 3-year annual average of 15.0 mg/m3)
reduced PM2.5 concentrations and
mortality rates in Medicare beneficiaries
between 2000–2013, and found that
following implementation of the annual
PM2.5 NAAQS, annual PM2.5
concentrations decreased by 1.59 mg/m3
(95% CI: 1.39, 1.80) which
corresponded to a 0.93% reduction in
mortality rates among individuals 65
years and older ([95% CI: 0.10%,
1.77%) in non-attainment counties
relative to attainment counties.
The 2019 ISA also evaluated a small
number of studies that used alternative
methods for confounder control to
further assess relationship between
long-term PM2.5 exposure and mortality
(U.S. EPA, 2019a, section 11.2.2.4). In
addition, multiple epidemiologic
studies that implemented alternative
methods for confounder control and
were published since the literature
cutoff date of the 2019 ISA were
evaluated in the ISA Supplement (U.S.
EPA, 2022a, section 3.2.2.3). These
studies used a variety of statistical
methods including generalized
propensity score (GPS), inverse
probability weighting (IPW), and
difference-in-difference (DID) to reduce
uncertainties related to confounding
bias in the association between longterm PM2.5 exposure and mortality.
These studies reported consistent
positive associations between long-term
PM2.5 exposure and total mortality (U.S.
EPA, 2022a, section 3.2.2.3), and
provided further support for the
associations reported in the cohort
studies referenced above.
The 2019 ISA and ISA Supplement
also evaluated the degree to which
recent studies examining the
relationship between long-term PM2.5
exposure and mortality addressed key
policy-relevant issues and/or previously
identified data gaps in the scientific
evidence, including methods to estimate
exposure, methods to control for
confounding (e.g., co-pollutant
confounding), the shape of the C–R
relationship, as well as examining
whether a threshold exists below which
mortality effects do not occur. With
respect to exposure assessment, based
on its evaluation of the evidence, the
2019 ISA concludes that positive
associations between long-term PM2.5
exposures and mortality are robust
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16226
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
across recent analyses using various
approaches to estimate PM2.5 exposures
(e.g., based on monitors, models,
satellite-based methods, or hybrid
methods that combine information from
multiple sources) (U.S. EPA, 2019a,
section 11.2.5.1). Hart et al. (2015)
report that correction for bias due to
exposure measurement error increases
the magnitude of the hazard ratios
(confidence intervals widen but the
association remains statistically
significant), suggesting that failure to
correct for exposure measurement error
could result in attenuation or
underestimation of risk estimates.
The 2019 ISA additionally concludes
that positive associations between longterm PM2.5 exposures and mortality are
robust across statistical models that use
different approaches to control for
confounders or different sets of
confounders (U.S. EPA, 2019a, sections
11.2.3 and 11.2.5), across diverse
geographic regions and populations, and
across a range of temporal periods
including periods of declining PM
concentrations (U.S. EPA, 2019a,
sections 11.2.2.5 and 11.2.5.3).
Additional evidence further
demonstrates that associations with
mortality remain robust in copollutants
analyses (U.S. EPA, 2019a, section
11.2.3), and that associations persist in
analyses restricted to long-term
exposures (annual average PM2.5
concentrations) below 12 mg/m3 (Di et
al., 2017b) or 10 mg/m3 (Shi et al., 2016),
indicating that risks are not
disproportionately driven by the upper
portions of the air quality distribution.
Recent studies evaluated in the ISA
Supplement further assess potential
copollutant confounding and indicate
that while there is some evidence of
potential confounding of the PM2.5mortality association by copollutants in
some of the studies (i.e., those studies of
the Mortality Air Pollution Associations
in Low Exposure Environments
(MAPLE) cohort), this result is
inconsistent with other recent studies
evaluated in the 2019 ISA that were
conducted in the U.S. and Canada that
found associations in both single and
copollutant models (U.S. EPA, 2019a;
U.S. EPA, 2022a, section 3.2.2.4)
Additionally, a few studies use
statistical techniques to reduce
uncertainties related to potential
confounding to further inform
conclusions on causality for long-term
PM2.5 exposure and mortality, as further
detailed in section II.B.1.a.i of the
proposal (88 FR 5582, January 27, 2023),
studies by Greven et al. (2011), Pun et
al. (2017), and Eum et al. (2018)
completed sensitivity analyses as part of
their Medicare cohort study in which
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
they decompose ambient PM2.5 into
‘‘spatial’’ and ‘‘spatiotemporal’’
components in order to evaluate the
potential for bias due to unmeasured
spatial confounding. Pun et al. (2017)
observed positive associations for the
‘‘temporal’’ variation model and
approximately null associations for the
‘‘spatiotemporal’’ variation model for all
causes of death except for COPD
mortality. The difference in the results
of these two models for most causes of
death suggests the presence of
unmeasured confounding, though the
authors do not indicate anything about
the direction or magnitude of this bias.
It is important to note that the
‘‘temporal’’ and ‘‘spatiotemporal’’
coefficients are not directly comparable
to the results of other epidemiologic
studies when examined individually
and can only be used in comparison
with one another to evaluate the
potential for unmeasured confounding
bias. Eum et al. (2018) and Wu et al.
(2020) also attempted to address longterm trends and meteorological
variables as potential confounders and
found that not adjusting for temporal
trends could overestimate the
association, while effect estimates in
analyses that excluded meteorological
variables remained unchanged
compared to the main analyses. While
results of these analyses suggest the
presence of some unmeasured
confounding, they do not indicate the
direction or magnitude of the bias.54
An additional important
consideration in characterizing the
public health impacts associated with
PM2.5 exposure is whether C–R
relationships are linear across the range
of concentrations or if nonlinear
relationships exist along any part of this
range. Studies evaluated in the 2019 ISA
and the ISA Supplement examine this
issue, and continue to provide evidence
of linear, no-threshold relationships
between long-term PM2.5 exposures and
all-cause and cause-specific mortality
(U.S. EPA, 2019a, section 11.2.4; U.S.
EPA, 2022a, section 3.2.2.2.7, Table 3–
6). Across the studies evaluated in the
2019 ISA and the ISA Supplement, a
variety of statistical methods have been
used to assess whether there is evidence
of deviations in linearity (U.S. EPA,
54 In public comments on the 2019 draft PA, the
authors of the Pun et al. (2017) study further note
that ‘‘the presence of unmeasured
confounding. . .was expected given that we did not
control for several potential confounders that may
impact PM2.5-mortality associations, such as
smoking, socio-economic status (SES), gaseous
pollutants, PM2.5 components, and long-term time
trends in PM2.5’’ and that ‘‘spatial confounding may
bias mortality risks both towards and away from the
null’’ (Docket ID EPA–HQ–OAR–2015–0072–0065;
accessible in https://www.regulations.gov/).
PO 00000
Frm 00026
Fmt 4701
Sfmt 4700
2019a, Table 11–7; U.S. EPA, 2022a,
section 2.2.3.2). Studies have also
conducted cut-point analyses that focus
on examining risk at specific ambient
PM2.5 concentrations. Generally, the
evidence remains consistent in
supporting a no-threshold relationship,
and in supporting a linear relationship
for PM2.5 concentrations >8 mg/m3.
However, uncertainties remain about
the shape of the C–R curve at PM2.5
concentrations <8 mg/m3, with some
recent studies providing evidence for
either a sublinear, linear, or supralinear
relationship at these lower
concentrations (U.S. EPA, 2019a,
section 11.2.4; U.S. EPA, 2022a, section
2.2.3.2). There was also some limited
evidence indicating that the slope of the
C–R function may be steeper
(supralinear) at lower concentrations for
cardiovascular mortality (U.S. EPA,
2022a, section 3.1.1.2.6).
The biological plausibility of PM2.5attributable mortality is supported by
the coherence of effects across scientific
disciplines (i.e., animal toxicological,
controlled human exposure studies, and
epidemiologic) when evaluating
respiratory and cardiovascular
morbidity effects, which are some of the
largest contributors to total
(nonaccidental) mortality. The 2019 ISA
outlines the available evidence for
biologically plausible pathways by
which inhalation exposure to PM2.5
could progress from initial events (e.g.,
pulmonary inflammation, autonomic
nervous system activation) to endpoints
relevant to population outcomes,
particularly those related to
cardiovascular diseases such as
ischemic heart disease, stroke and
atherosclerosis (U.S. EPA, 2019a,
section 6.2.1), and to metabolic effects,
including diabetes (U.S. EPA, 2019a,
section 7.3.1). The 2019 ISA notes
‘‘more limited evidence from respiratory
morbidity’’ (U.S. EPA, 2019a, p. 11–101)
such as development of chronic
obstructive pulmonary disease (COPD)
(U.S. EPA, 2019a, section 5.2.1) to
support the biological plausibility of
mortality due to long-term PM2.5
exposures (U.S. EPA, 2019a, section
11.2.1).
Taken together, epidemiologic
studies, including those evaluated in the
2019 ISA and more recent studies
evaluated in the ISA Supplement,
consistently report positive associations
between long-term PM2.5 exposure and
mortality across different geographic
locations, populations, and analytic
approaches (U.S. EPA, 2019a; U.S. EPA,
2022a, section 3.2.2.4). As such, these
studies reduce key uncertainties
identified in previous reviews,
including those related to potential
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
copollutant confounding, and provide
additional information on the shape of
the C–R curve. As evaluated in the 2019
ISA, experimental and epidemiologic
evidence for cardiovascular effects, and
respiratory effects to a more limited
degree, supports the plausibility of
mortality due to long-term PM2.5
exposures. Overall, studies evaluated in
the 2019 ISA support the conclusion of
a causal relationship between long-term
PM2.5 exposure and mortality, which is
supported and extended by evidence
from recent epidemiologic studies
evaluated in the ISA Supplement (U.S.
EPA, 2022a, section 3.2.2.4).
ddrumheller on DSK120RN23PROD with RULES3
Short-Term PM2.5 Exposures
The 2009 ISA concluded that ‘‘a
causal relationship exists between shortterm exposure to PM2.5 and mortality’’
(U.S. EPA, 2009a). This conclusion was
based on the evaluation of both multiand single-city epidemiologic studies
that consistently reported positive
associations between short-term PM2.5
exposure and non-accidental mortality.
These associations were strongest, in
terms of magnitude and precision,
primarily at lags of 0 to 1 days.
Examination of the potential
confounding effects of gaseous
copollutants was limited, though
evidence from single-city studies
indicated that gaseous copollutants have
minimal effect on the PM2.5-mortality
relationship (i.e., associations remain
robust to inclusion of other pollutants in
copollutant models). The evaluation of
cause-specific mortality found that
effect estimates were larger in
magnitude, but also had larger
confidence intervals, for respiratory
mortality compared to cardiovascular
mortality. Although the largest mortality
risk estimates were for respiratory
mortality, the interpretation of the
results was complicated by the limited
coherence from studies of respiratory
morbidity. However, the evidence from
studies of cardiovascular morbidity
provided both coherence and biological
plausibility for the relationship between
short-term PM2.5 exposure and
cardiovascular mortality.
Multicity studies evaluated in the
2019 ISA and the ISA Supplement
provide evidence of primarily positive
associations between daily PM2.5
exposures and mortality, with percent
increases in total mortality ranging from
0.19% (Lippmann et al., 2013) to 2.80%
(Kloog et al., 2013) 55 at lags of 0 to 1
days in single-pollutant models.
55 As detailed in the Preface to the ISA, risk
estimates are for a 10 mg/m3 increase in 24-hour avg
PM2.5 concentrations, unless otherwise noted (U.S.
EPA, 2019a).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
Whereas many studies assign exposures
using data from ambient monitors, other
studies employ hybrid modeling
approaches, which estimate PM2.5
concentrations using data from a variety
of sources (i.e., from satellites, land use
information, and modeling, in addition
to monitors) and enable the inclusion of
less urban and more rural locations in
analyses (e.g., Kloog et al., 2013, Lee et
al., 2015, Shi et al., 2016).
Some studies have expanded the
examination of potential confounders
including long-term temporal trends,
weather, and co-occurring pollutants.
Mortality associations were found to
remain positive, although in some cases
were attenuated, when using different
approaches to account for temporal
trends or weather covariates (e.g., U.S.
EPA, 2019a, section 11.1.5.1). For
example, Sacks et al. (2012) examined
the influence of model specification
using the approaches for confounder
adjustment from models employed in
several multicity studies within the
context of a common data set (U.S. EPA,
2019a, section 11.1.5.1). These models
use different approaches to control for
long-term temporal trends and the
potential confounding effects of
weather. The authors report that
associations between daily PM2.5 and
cardiovascular mortality were similar
across models, with the percent increase
in mortality ranging from 1.5–2.0%
(U.S. EPA, 2019a, Figure 11–4). Thus,
alternative approaches to controlling for
long-term temporal trends and for the
potential confounding effects of weather
may influence the magnitude of the
association between PM2.5 exposures
and mortality but have not been found
to influence the direction of the
observed association (U.S. EPA, 2019a,
section 11.1.5.1). Taken together, the
2019 ISA and the ISA Supplement
conclude that recent multicity studies
conducted in the U.S., Canada, Europe,
and Asia continue to provide consistent
evidence of positive associations
between short-term PM2.5 exposures and
total mortality across studies that use
different approaches to control for the
potential confounding effects of weather
(e.g., temperature) (U.S. EPA, 2019a,
section 1.4.1.5.1; U.S. EPA, 2022a,
section 3.2.1.2).
With regard to copollutants, studies
evaluated in the 2019 ISA provide
additional evidence that associations
between short-term PM2.5 exposures and
mortality remain positive and relatively
unchanged in copollutant models with
both gaseous pollutants and PM10–2.5
(U.S. EPA, 2019a, section 11.1.4).
Additionally, the low (r < 0.4) to
moderate correlations (r = 0.4–0.7)
between PM2.5 and gaseous pollutants
PO 00000
Frm 00027
Fmt 4701
Sfmt 4700
16227
and PM10–2.5 increase the confidence in
PM2.5 having an independent effect on
mortality (U.S. EPA, 2019a, section
11.1.4). Consistent with the studies
evaluated in the 2019 ISA, studies
evaluated in the ISA Supplement that
used data from more recent years also
indicate that associations between shortterm PM2.5 exposure and mortality
remain unchanged in copollutant
models. However, the evidence
indicates that the association could be
larger in magnitude in the presence of
some copollutants such as oxidant gases
(Lavigne et al., 2018; Shin et al., 2021).
The generally positive associations
reported with mortality are supported
by a small group of studies employing
alternative methods for confounder
control or quasi-experimental statistical
approaches (U.S. EPA, 2019a, section
11.1.2.1). For example, two studies by
Schwartz et al. report associations
between PM2.5 instrumental variables
and mortality (U.S. EPA, 2019a, Table
11–2), including in an analysis limited
to days with 24-hour average PM2.5
concentrations <30 mg/m3 (Schwartz et
al., 2015; Schwartz et al., 2017). In
addition to the main analyses, these
studies conducted Granger-like
causality tests as sensitivity analyses to
examine whether there was evidence of
an association between mortality and
PM2.5 after the day of death, which
would support the possibility that
unmeasured confounders were not
accounted for in the statistical model.
Neither study reports evidence of an
association with PM2.5 after death (i.e.,
they do not indicate unmeasured
confounding). Yorifuji et al. (2016)
conducted a quasi-experimental study
to examine whether a specific regulatory
action in Tokyo, Japan (i.e., a diesel
emission control ordinance) resulted in
a subsequent reduction in daily
mortality (Yorifuji et al., 2016). The
authors reported a reduction in
mortality in Tokyo due to the ordinance,
compared to Osaka, which did not have
a similar diesel emission control
ordinance in place. In another study,
Schwartz et al. (2018) utilized three
statistical methods including
instrumental variable analysis, a
negative exposure control, and marginal
structural models to estimate the
association between PM2.5 and daily
mortality (Schwartz et al., 2018). Results
from this study continue to support a
relationship between short-term PM2.5
exposure and mortality. Additional
epidemiologic studies evaluated in the
ISA Supplement that employed
alternative methods for confounder
control to examine the association
between short-term PM2.5 exposure and
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16228
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
mortality also report consistent positive
associations in studies that examine
effects across multiple cities in the U.S.
(U.S. EPA, 2022a).
The positive associations for total
mortality reported across the majority of
studies evaluated are further supported
by cause-specific mortality analyses,
which generally report consistent,
positive associations with both
cardiovascular and respiratory mortality
(U.S. EPA, 2019a, section 11.1.3).
Recent multicity studies evaluated in
the ISA Supplement add to the body of
evidence indicating a relationship
between short-term PM2.5 exposure and
cause-specific mortality, with more
variability in the magnitude and
precision of associations for respiratory
mortality (U.S. EPA, 2022a; Figure 3–
14). For both cardiovascular and
respiratory mortality, there has been a
limited assessment of potential
copollutant confounding, though initial
evidence indicates that associations
remain positive and relatively
unchanged in models with gaseous
pollutants and PM10–2.5, which further
supports the copollutant analyses
conducted for total mortality. The strong
evidence for ischemic events and heart
failure, as detailed in the assessment of
cardiovascular morbidity (U.S. EPA,
2019a, Chapter 6), provides biological
plausibility for PM2.5-related
cardiovascular mortality, which
comprises the largest percentage of total
mortality (i.e., ∼33%) (NHLBI, 2017).
Although there is evidence for
exacerbations of COPD and asthma, the
collective body of respiratory morbidity
evidence provides limited biological
plausibility for PM2.5-related respiratory
mortality (U.S. EPA, 2019a, Chapter 5).
In the 2009 ISA, one of the main
uncertainties identified was the regional
and city-to-city heterogeneity in PM2.5mortality associations. Studies
evaluated in the 2019 ISA examine both
city-specific as well as regional
characteristics to identify the
underlying contextual factors that could
contribute to this heterogeneity (U.S.
EPA, 2019a, section 11.1.6.3). Analyses
focusing on effect modification of the
PM2.5 mortality relationship by PM2.5
components, regional patterns in PM2.5
components and city specific
differences in composition and sources
indicate some differences in the PM2.5
composition and sources across cities
and regions, but these differences do not
fully explain the observed
heterogeneity. Additional studies find
that factors related to potential exposure
differences, such as housing stock and
commuting, as well as city specific
factors (e.g., land use, port volume, and
traffic information), may also explain
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
some of the observed heterogeneity
(U.S. EPA, 2019a, section 11.1.6.3).
Collectively, studies evaluated in the
2019 ISA and the ISA Supplement
indicate that the heterogeneity in PM2.5
mortality risk estimates cannot be
attributed to one factor, but instead a
combination of factors including, but
not limited to, PM composition and
sources as well as community
characteristics that could influence
exposures (U.S. EPA, 2019a, section
11.1.12; U.S. EPA, 2022a, section
3.2.1.2.1).
A number of studies evaluated in the
2019 ISA and ISA Supplement
conducted systematic evaluations of the
lag structure of associations for the
PM2.5-mortality relationship by
examining either a series of single day
or multiday lags and these studies
continue to support an immediate effect
(i.e., lag 0 to 1 days) of short-term PM2.5
exposures on mortality (U.S. EPA,
2019a, section 11.1.8.1; U.S. EPA,
2022a, section 3.2.1.1). Recent studies
also conducted analyses comparing the
traditional 24-hour average exposure
metric with a subdaily metric (i.e., 1hour max) and provide evidence of a
similar pattern of associations for both
the 24-hour average and 1-hour max
metric, with the association larger in
magnitude for the 24-hour average
metric.
Multicity studies indicate that
positive and statistically significant
associations with mortality persist in
analyses restricted to short-term (24hour average PM2.5 concentrations)
PM2.5 exposures below 35 mg/m3 (Lee et
al., 2015),56 below 30 mg/m3 (Shi et al.,
2016), and below 25 mg/m3 (Di et al.,
2017a), indicating that risks associated
with short-term PM2.5 exposures are not
disproportionately driven by the peaks
of the air quality distribution.
Additional studies examined the shape
of the C–R relationship for short-term
PM2.5 exposure and mortality and
whether a threshold exists below which
mortality effects do not occur (U.S. EPA,
2019a, section 11.1.10). These studies
used various statistical approaches and
consistently demonstrate linear C–R
relationships with no evidence of a
threshold.
Moreover, recent studies evaluated in
the ISA Supplement provide additional
support for a linear, no-threshold C–R
relationship between short-term PM2.5
56 Lee et al. (2015) restrict exposures below 35 mg/
m3 only in areas with annual average
concentrations <12 mg/m3. Additionally, Lee et al.
(2015) also report that positive and statistically
significant associations between short-term PM2.5
exposures and mortality persist in analyses
restricted to areas with long-term concentrations
below 12 mg/m3.
PO 00000
Frm 00028
Fmt 4701
Sfmt 4700
exposure and mortality, with confidence
in the shape decreasing at
concentrations below 5 mg/m3 (Shi et al.,
2016; Lavigne et al., 2018). Recent
analyses provide initial evidence
indicating that PM2.5-mortality
associations persist and may be stronger
(i.e., a steeper slope) at lower
concentrations (e.g., Di et al., 2017a;
Figure 11–12 in U.S. EPA, 2019).
However, given the limited data
available at the lower end of the
distribution of ambient PM2.5
concentrations, the shape of the C–R
curve remains uncertain at these low
concentrations. Although difficulties
remain in assessing the shape of the
short-term PM2.5-mortality C–R
relationship, to date, studies have not
conducted systematic evaluations of
alternatives to linearity and recent
studies evaluated in the ISA
Supplement continue to provide
evidence of a no-threshold linear
relationship, with less confidence at
concentrations lower than 5 mg/m3.
Overall, epidemiologic studies
evaluated in the 2019 ISA and the ISA
Supplement build upon and extend the
conclusions of the 2009 ISA for the
relationship between short-term PM2.5
exposures and total mortality.
Supporting evidence for PM2.5-related
cardiovascular morbidity, and more
limited evidence from respiratory
morbidity, provide biological
plausibility for mortality due to shortterm PM2.5 exposures. The primarily
positive associations observed across
studies conducted in diverse geographic
locations is further supported by the
results from copollutant analyses
indicating robust associations, along
with evidence from analyses examining
the C–R relationship. Overall, studies
evaluated in the 2019 ISA support the
conclusion of a causal relationship
between short-term PM2.5 exposure and
mortality, which is further supported by
evidence from recent epidemiologic
studies evaluated in the ISA
Supplement (U.S. EPA, 2022a, section
3.2.1.4, p. 3–69).
ii. Cardiovascular Effects
Long-Term PM2.5 Exposures
The scientific evidence reviewed in
the 2009 ISA was ‘‘sufficient to infer a
causal relationship between long-term
PM2.5 exposure and cardiovascular
effects’’ (U.S. EPA, 2009a). The strongest
line of evidence comprised findings
from several large epidemiologic studies
of U.S. and Canadian cohorts that
reported consistent positive associations
between long-term PM2.5 exposure and
cardiovascular mortality (Pope et al.,
2004; Krewski et al., 2009; Miller et al.,
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
2007; Laden et al., 2006). Studies of
long-term PM2.5 exposure and
cardiovascular morbidity were limited
in number. Biological plausibility and
coherence with the epidemiologic
findings were provided by studies using
genetic mouse models of atherosclerosis
demonstrating enhanced atherosclerotic
plaque development and inflammation,
as well as changes in measures of
impaired heart function, following 4- to
6-month exposures to PM2.5
concentrated ambient particles (CAPs),
and by a limited number of studies
reporting CAPs-induced effects on
coagulation factors, vascular reactivity,
and worsening of experimentally
induced hypertension in mice (U.S.
EPA, 2009a).
Consistent with the evidence assessed
in the 2009 ISA, the 2019 ISA concludes
that recent studies, together with the
evidence available in previous reviews,
support a causal relationship between
long-term exposure to PM2.5 and
cardiovascular effects. Additionally,
recent epidemiologic studies published
since the completion of the 2019 ISA
and evaluated in the ISA Supplement
expands the body of evidence and
further supports such a conclusion (U.S.
EPA, 2022a). As discussed above
(section II.A.2.a.i), results from U.S. and
Canadian cohort studies evaluated in
the 2019 ISA conducted at varying
spatial and temporal scales and
employing a variety of exposure
assessment and statistical methods
consistently report positive associations
between long-term PM2.5 exposure and
cardiovascular mortality (U.S. EPA,
2019, Figure 6–19, section 6.2.10).
Positive associations between long-term
PM2.5 exposures and cardiovascular
mortality are generally robust in
copollutant models adjusted for ozone,
NO2, PM10–2.5, or SO2. In addition, most
of the results from analyses examining
the shape of the C–R relationship
between long-term PM2.5 exposures and
cardiovascular mortality support a
linear relationship and do not identify
a threshold below which mortality
effects do not occur (U.S. EPA, 2019a,
section 6.2.16, Table 6–52).
The body of literature examining the
relationship between long-term PM2.5
exposure and cardiovascular morbidity
has greatly expanded since the 2009
ISA, with positive associations reported
in several cohorts evaluated in the 2019
ISA (U.S. EPA, 2019a, section 6.2).
Though results for cardiovascular
morbidity are less consistent than those
for cardiovascular mortality (U.S. EPA,
2019a, section 6.2), studies in the 2019
ISA and the ISA Supplement provide
some evidence for associations between
long-term PM2.5 exposures and the
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
progression of cardiovascular disease.
Positive associations with
cardiovascular morbidity (e.g., coronary
heart disease, stroke, arrhythmias,
myocardial infarction (MI),
atherosclerosis progression) are
observed in several epidemiologic
studies (U.S. EPA, 2019a, sections 6.2.2
to 6.2.9; U.S. EPA, 2022a, section
3.1.2.2). Additionally, studies evaluated
in the ISA Supplement report positive
associations among those with preexisting conditions, among patients
followed after a cardiac event
procedure, and among those with a first
hospital admission for heart attacks
among older adults enrolled in
Medicare (U.S. EPA, 2022a, sections
3.1.1 and 3.1.2).
Recent studies published since the
literature cutoff date of the 2019 ISA
and evaluated in the ISA Supplement
further assessed the relationship
between long-term PM2.5 exposure and
cardiovascular effects by conducting
accountability analyses or by using
alternative methods for confounder
control in evaluating the association
between long-term PM2.5 exposure and
cardiovascular hospital admissions
(U.S. EPA, 2022a, section 3.1.2.3).
Studies that apply alternative methods
for confounder control increase
confidence in the relationship between
long-term PM2.5 exposure and
cardiovascular effects by using methods
that reduce uncertainties related to
potential confounding through
statistical and/or study design
approaches. For example, to control for
potential confounding Wei et al. (2021)
used a doubly robust additive model
(DRAM) and found an association
between long-term exposure to PM2.5
and cardiovascular effects, including
MI, stoke, and atrial fibrillation, among
the Medicare population. For example,
an accountability study by Henneman et
al. (2019) utilized a difference-indifference (DID) approach to determine
the relationship between coal-fueled
power plant emissions and
cardiovascular effects and found that
reductions in PM2.5 concentrations
resulted in reductions of cardiovascularrelated hospital admissions.
Furthermore, several recent
epidemiologic studies evaluated in the
ISA Supplement reported that the
association between long-term PM2.5
exposure with stroke persisted after
adjustment for NO2 but was attenuated
in the model with O3 and oxidant gases
represented by the redox weighted
average of NO2 and O3 (U.S. EPA, 2022a,
section 3.1.2.2.8). Overall, these studies
report consistent findings that long-term
PM2.5 exposure is related to increased
PO 00000
Frm 00029
Fmt 4701
Sfmt 4700
16229
hospital admissions for a variety of
cardiovascular disease outcomes among
large nationally representative cohorts
and provide additional support for a
relationship between long-term PM2.5
exposure and cardiovascular effects.
Positive associations reported in
epidemiologic studies are supported by
toxicological evidence evaluated in the
2019 ISA. The positive associations
reported in epidemiologic studies are
supported by toxicological evidence for
increased plaque progression in mice
following long-term exposure to PM2.5
collected from multiple locations across
the U.S. (U.S. EPA, 2019a, section
6.2.4.2). A small number of
epidemiologic studies also report
positive associations between long-term
PM2.5 exposure and heart failure,
changes in blood pressure, and
hypertension (U.S. EPA, 2019a, sections
6.2.5 and 6.2.7). Associations with heart
failure are supported by animal
toxicological studies demonstrating
decreased cardiac contractility and
function, and increased coronary artery
wall thickness following long-term
PM2.5 exposure (U.S. EPA, 2019a,
section 6.2.5.2). Similarly, a limited
number of animal toxicological studies
demonstrating a relationship between
long-term PM2.5 exposure and consistent
increases in blood pressure in rats and
mice are coherent with epidemiologic
studies reporting positive associations
between long-term exposure to PM2.5
and hypertension.
Additionally, a number of studies
evaluated in the ISA Supplement
focusing on morbidity outcomes,
including those that focused on
incidence of MI, atrial fibrillation (AF),
stroke, and congestive heart failure
(CHF), expand the evidence pertaining
to the shape of the C–R relationship
between long-term PM2.5 exposure and
cardiovascular effects. These studies use
statistical techniques that allow for
departures from linearity (U.S. EPA,
2022a, Table 3–3), and generally
support the evidence characterized in
the 2019 ISA showing linear, nothreshold C–R relationship for most
cardiovascular disease (CVD) outcomes.
However, there is evidence for a
sublinear or supralinear C–R
relationship for some outcomes (U.S.
EPA, 2022a, section 3.1.2.2.9).57
Longitudinal epidemiologic analyses
also report positive associations with
markers of systemic inflammation (U.S.
EPA, 2019a, section 6.2.11), coagulation
(U.S. EPA, 2019a, section 6.2.12), and
57 As noted above for mortality, uncertainty in the
shape of the C–R relationship increases near the
upper and lower ends of the distribution due to
limited data.
E:\FR\FM\06MRR3.SGM
06MRR3
16230
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
endothelial dysfunction (U.S. EPA,
2019a, section 6.2.13). These results are
coherent with animal toxicological
studies generally reporting increased
markers of systemic inflammation,
oxidative stress, and endothelial
dysfunction (U.S. EPA, 2019a, section
6.2.12.2 and 6.2.14).
In summary, the 2019 ISA concludes
that there is consistent evidence from
multiple epidemiologic studies
illustrating that long-term exposure to
PM2.5 is associated with mortality from
cardiovascular causes. Epidemiologic
studies evaluated in the ISA
Supplement provide additional
evidence of positive associations
between long-term PM2.5 exposure and
cardiovascular morbidity (U.S. EPA,
2022a, section 3.1.2.2). Associations
with coronary heart disease (CHD),
stroke and atherosclerosis progression
were observed in several additional
epidemiologic studies, providing
coherence with the mortality findings.
Results from copollutant models
generally support an independent effect
of PM2.5 exposure on mortality.
Additional evidence of the independent
effect of PM2.5 on the cardiovascular
system is provided by experimental
studies in animals, which support the
biological plausibility of pathways by
which long-term exposure to PM2.5
could potentially result in outcomes
such as CHD, stroke, CHF, and
cardiovascular mortality. Overall,
studies evaluated in the 2019 ISA
support the conclusion of a causal
relationship between long-term PM2.5
exposure and cardiovascular effects,
which is supported and extended by
evidence from recent epidemiologic
studies evaluated in the ISA
Supplement (U.S. EPA, 2022a, section
3.1.2.2).
Short-Term PM2.5 Exposures
The 2009 ISA concluded that ‘‘a
causal relationship exists between shortterm exposure to PM2.5 and
cardiovascular effects’’ (U.S. EPA,
2009a). The strongest evidence in the
2009 ISA was from epidemiologic
studies of emergency department (ED)
visits and hospital admissions for IHD
and heart failure (HF), with supporting
evidence from epidemiologic studies of
cardiovascular mortality (U.S. EPA,
2009a). Animal toxicological studies
provided coherence and biological
plausibility for the positive associations
reported with MI, ED visits, and
hospital admissions. These included
studies reporting reduced myocardial
blood flow during ischemia and studies
indicating altered vascular reactivity. In
addition, effects of PM2.5 exposure on a
potential indicator of ischemia (i.e., ST
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
segment depression on an
electrocardiogram) were reported in
both animal toxicological and
epidemiologic panel studies.58 Key
uncertainties from the last review
resulted from inconsistent results across
disciplines with respect to the
relationship between short-term
exposure to PM2.5 and changes in blood
pressure, blood coagulation markers,
and markers of systemic inflammation.
In addition, while the 2009 ISA
identified a growing body of evidence
from controlled human exposure and
animal toxicological studies,
uncertainties remained with respect to
biological plausibility.
Studies evaluated in the 2019 ISA
provide additional support for a causal
relationship between short-term PM2.5
exposure and cardiovascular effects.
This includes generally positive
associations observed in multicity
epidemiologic studies of emergency
department visits and hospital
admissions for IHD, heart failure (HF),
and combined cardiovascular-related
endpoints. In particular, nationwide
studies of older adults (65 years and
older) using Medicare records report
positive associations between PM2.5
exposures and hospital admissions for
HF (U.S. EPA, 2019a, section 6.1.3.1).
Moreover, recent multicity studies,
published after the literature cutoff date
of the 2019 ISA and evaluated in the
ISA Supplement, are consistent with
studies evaluated in the 2019 ISA that
report positive association between
short-term PM2.5 exposure and ED visits
and hospital admission for IHD, heart
attacks, and HF (U.S. EPA, 2022a,
section 3.1). Epidemiologic studies
conducted in single cities contribute
some support to the causality
determination, though associations
reported in single-city studies are less
consistently positive than in multicity
studies, and include a number of studies
reporting null associations (U.S. EPA,
2019a, sections 6.1.2 and 6.1.3). As a
whole, though, the recent body of IHD
and HF epidemiologic evidence
supports the evidence from previous
ISAs reporting mainly positive
associations between short-term PM2.5
concentrations and emergency
department visits and hospital
admissions.
Consistent with the evidence assessed
in the 2019 ISA, some studies evaluated
in the ISA Supplement report no
evidence of an association with stroke,
58 Some animal studies included in the 2009 ISA
examined exposures to mixtures, such as motor
vehicle exhaust or woodsmoke. In these studies, it
was unclear if the resulting cardiovascular effects
could be attributed specifically to the fine particle
component of the mixture.
PO 00000
Frm 00030
Fmt 4701
Sfmt 4700
regardless of stroke subtype.
Additionally, as in the 2019 ISA,
evidence evaluated in the ISA
Supplement continues to indicate an
immediate effect of PM2.5 on
cardiovascular-related outcomes
primarily within the first few days after
exposure, and that associations
generally persisted in models adjusted
for copollutants (U.S. EPA, 2022a,
section 3.1.1.2).
The ISA Supplement includes
additional epidemiologic studies,
published since the literature cutoff date
for the 2019 ISA, including
accountability analyses and
epidemiologic studies that employ
alternative methods for confounder
control to evaluate the association
between short-term PM2.5 exposure and
cardiovascular-related effects (U.S. EPA,
2022a, section 3.1.1.3). These studies
employ a number of statistical
approaches and report positive
associations, providing additional
support for a relationship between
short-term PM2.5 exposure and
cardiovascular effects, while also
reducing uncertainties related to
potential confounder bias.
A number of controlled human
exposure, animal toxicological, and
epidemiologic panel studies provide
evidence that PM2.5 exposure could
plausibly result in IHD or HF through
pathways that include endothelial
dysfunction, arterial thrombosis, and
arrhythmia (U.S. EPA, 2019a, section
6.1.1). The most consistent evidence
from recent controlled human exposure
studies is for endothelial dysfunction, as
measured by changes in brachial artery
diameter or flow mediated dilation.
Multiple controlled human exposure
studies that examined the potential for
endothelial dysfunction report an effect
of PM2.5 exposure on measures of blood
flow (U.S. EPA, 2019a, section 6.1.13.2).
However, these studies report variable
results regarding the timing of the effect
and the mechanism by which reduced
blood flow occurs (i.e., availability vs
sensitivity to nitric oxide). In addition,
some controlled human exposure
studies using CAPs report evidence for
small increases in blood pressure (U.S.
EPA, 2019a, section 6.1.6.3). Although
not entirely consistent, there is also
some evidence across controlled human
exposure studies for conduction
abnormalities/arrhythmia (U.S. EPA,
2019a, section 6.1.4.3), changes in heart
rate variability (HRV) (U.S. EPA, 2019a,
section 6.1.10.2), changes in hemostasis
that could promote clot formation (U.S.
EPA, 2019a, section 6.1.12.2), and
increases in inflammatory cells and
markers (U.S. EPA, 2019a, section
6.1.11.2). A recent study by Wyatt et al.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
(2020), evaluated in the ISA
Supplement, adds to the limited
evidence base of controlled human
exposure studies conducted at near
ambient PM2.5 concentrations. The
study, completed in healthy young
adults subject to intermittent exercise,
found some significant cardiovascular
effects (e.g., systematic inflammation
markers, including C-reactive protein
(CRP), and cardiac repolarization).
Thus, when taken as a whole, controlled
human exposure studies are coherent
with epidemiologic studies in that they
demonstrate that short-term exposures
to PM2.5 may result in the types of
cardiovascular endpoints that could
lead to emergency department visits and
hospital admissions for IHD or HF, as
well as mortality in some people.
Animal toxicological studies
published since the 2009 ISA and
evaluated in the 2019 ISA also support
a relationship between short-term PM2.5
exposure and cardiovascular effects. A
study demonstrating decreased cardiac
contractility and left ventricular
pressure in mice is coherent with the
results of epidemiologic studies that
report associations between short-term
PM2.5 exposure and heart failure (U.S.
EPA, 2019a, section 6.1.3.3). In
addition, and as with controlled human
exposure studies, there is generally
consistent evidence in animal
toxicological studies for indicators of
endothelial dysfunction (U.S. EPA,
2019a, section 6.1.13.3). Some studies in
animals also provide evidence for
changes in a number of other
cardiovascular endpoints following
short-term PM2.5 exposure including
conduction abnormalities and
arrhythmia (U.S. EPA, 2019a, section
6.1.4.4), changes in HRV (U.S. EPA,
2019a, section 6.1.10.3), changes in
blood pressure (U.S. EPA, 2019a,
section 6.1.6.4), and evidence for
systemic inflammation and oxidative
stress (U.S. EPA, 2019a, section
6.1.11.3).
In summary, evidence evaluated in
the 2019 ISA extends the consistency
and coherence of the evidence base
evaluated in the 2009 ISA and prior
assessments. Epidemiologic studies
reporting robust associations in
copollutant models are supported by
direct evidence from controlled human
exposure and animal toxicologic studies
reporting independent effects of PM2.5
exposures on endothelial dysfunction as
well as endpoints indicating impaired
cardiac function, increased risk of
arrhythmia, changes in HRV, increases
in BP, and increases in indicators of
systemic inflammation, oxidative stress,
and coagulation (U.S. EPA, 2019,
section 6.1.16). For some cardiovascular
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
effects, there are inconsistencies in
results across some animal
toxicological, controlled human
exposure, and epidemiologic panel
studies, though this may be due to
substantial differences in study design
and/or study populations. Overall, the
results from epidemiologic panel,
controlled human exposure, and animal
toxicological studies, in particular those
related to endothelial dysfunction,
impaired cardiac function, ST segment
depression, thrombosis, conduction
abnormalities, and changes in blood
pressure provide coherence and
biological plausibility for the consistent
results from epidemiologic studies
observing positive associations between
short-term PM2.5 exposures and IHD and
HF, and ultimately cardiovascular
mortality. Overall, studies evaluated in
the 2019 ISA support the conclusion of
a causal relationship between short-term
PM2.5 exposure and cardiovascular
effects, which is supported and
extended by evidence from recent
epidemiologic studies evaluated in the
ISA Supplement (U.S. EPA, 2022a,
section 3.1.1.4).
iii. Respiratory Effects
Long-Term PM2.5 Exposures
The 2009 ISA concluded that ‘‘a
causal relationship is likely to exist
between long-term PM2.5 exposure and
respiratory effects’’ (U.S. EPA, 2009a).
This conclusion was based mainly on
epidemiologic evidence demonstrating
associations between long-term PM2.5
exposure and changes in lung function
or lung function growth in children.
Biological plausibility was provided by
a single animal toxicological study
examining pre- and post-natal exposure
to PM2.5 CAPs, which found impaired
lung development. Epidemiologic
evidence for associations between longterm PM2.5 exposure and other
respiratory outcomes, such as the
development of asthma, allergic disease,
and COPD; respiratory infection; and
the severity of disease was limited, both
in the number of studies available and
the consistency of the results.
Experimental evidence for other
outcomes was also limited, with one
animal toxicological study reporting
that long-term exposure to PM2.5 CAPs
results in morphological changes in
nasal airways of healthy animals. Other
animal studies examined exposure to
mixtures, such as motor vehicle exhaust
and woodsmoke, and effects were not
attributed specifically to the particulate
components of the mixture.
Cohort studies evaluated in the 2019
ISA provided additional support for the
relationship between long-term PM2.5
PO 00000
Frm 00031
Fmt 4701
Sfmt 4700
16231
exposure and decrements in lung
function growth (as a measure of lung
development), indicating a robust and
consistent association across study
locations, exposure assessment
methods, and time periods (U.S. EPA,
2019a, section 5.2.13). This relationship
was further supported by a retrospective
study that reports an association
between declining PM2.5 concentrations
and improvements in lung function
growth in children (U.S. EPA, 2019a,
section 5.2.11). Epidemiologic studies
also examine asthma development in
children (U.S. EPA, 2019a, section
5.2.3), with prospective cohort studies
reporting generally positive
associations, though several are
imprecise (i.e., they report wide
confidence intervals). Supporting
evidence is provided by studies
reporting associations with asthma
prevalence in children, with childhood
wheeze, and with exhaled nitric oxide,
a marker of pulmonary inflammation
(U.S. EPA, 2019a, section 5.2.13).
Additionally, the 2019 ISA includes an
animal toxicological study showing the
development of an allergic phenotype
and an increase in a marker of airway
responsiveness supports the biological
plausibility of the development of
allergic asthma (U.S. EPA, 2019a,
section 5.2.13). Other epidemiologic
studies report a PM2.5-related
acceleration of lung function decline in
adults, while improvement in lung
function was observed with declining
PM2.5 concentrations (U.S. EPA, 2019a,
section 5.2.11). A longitudinal study
found declining PM2.5 concentrations
are also associated with an
improvement in chronic bronchitis
symptoms in children, strengthening
evidence reported in the 2009 ISA for a
relationship between increased chronic
bronchitis symptoms and long-term
PM2.5 exposure (U.S. EPA, 2019a,
section 5.2.11). A common uncertainty
across the epidemiologic evidence is the
lack of examination of copollutants to
assess the potential for confounding.
While there is some evidence that
associations remain robust in models
with gaseous pollutants, a number of
these studies examining copollutant
confounding were conducted in Asia,
and thus have limited generalizability
due to high annual pollutant
concentrations.
When taken together, the 2019 ISA
concludes that the epidemiologic
evidence strongly supports a
relationship with decrements in lung
function growth asthma development in
children, as well as increased bronchitis
symptoms in children with asthma.
Additionally, the epidemiologic
E:\FR\FM\06MRR3.SGM
06MRR3
16232
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
evidence strongly supports a
relationship with an acceleration of lung
function decline in adults, and with
respiratory mortality and cause-specific
respiratory mortality for COPD and
respiratory infection (U.S. EPA, 2019a,
p. 1–34). In support of the biological
plausibility of associations reported in
epidemiologic studies associated with
respiratory health effects, animal
toxicological studies evaluated in the
2019 ISA continue to provide direct
evidence that long-term exposure to
PM2.5 results in a variety of respiratory
effects, including pulmonary oxidative
stress, inflammation, and morphologic
changes in the upper (nasal) and lower
airways. Other results show that
changes are consistent with the
development of allergy and asthma, and
with impaired lung development.
Overall, the 2019 ISA concludes that
‘‘the collective evidence is sufficient to
conclude that a causal relationship is
likely to exist between long-term PM2.5
exposure and respiratory effects’’ (U.S.
EPA, 2019a, section 5.2.13).
Short-Term PM2.5 Exposures
The 2009 ISA (U.S. EPA, 2009a)
concluded that a ‘‘causal relationship is
likely to exist’’ between short-term
PM2.5 exposure and respiratory effects.
This conclusion was based mainly on
the epidemiologic evidence
demonstrating positive associations
with various respiratory effects.
Specifically, the 2009 ISA described
epidemiologic evidence as consistently
showing PM2.5-associated increases in
hospital admissions and ED visits for
COPD and respiratory infection among
adults or people of all ages, as well as
increases in respiratory mortality. These
results were supported by studies
reporting associations with increased
respiratory symptoms and decreases in
lung function in children with asthma,
though the epidemiologic evidence was
inconsistent for hospital admissions or
emergency department visits for asthma.
Studies examining copollutant models
showed that PM2.5 associations with
respiratory effects were robust to
inclusion of CO or SO2 in the model, but
often were attenuated (though still
positive) with inclusion of O3 or NO2. In
addition to the copollutant models,
evidence supporting an independent
effect of PM2.5 exposure on the
respiratory system was provided by
animal toxicological studies of PM2.5
CAPs demonstrating changes in some
pulmonary function parameters, as well
as inflammation, oxidative stress,
injury, enhanced allergic responses, and
reduced host defenses. Many of these
effects have been implicated in the
pathophysiology for asthma
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
exacerbation, COPD exacerbation, or
respiratory infection. In the few
controlled human exposure studies
conducted in individuals with asthma
or COPD, PM2.5 exposure mostly had no
effect on respiratory symptoms, lung
function, or pulmonary inflammation.
Available studies in healthy people also
did not clearly demonstrate respiratory
effects following short-term PM2.5
exposures.
Epidemiologic studies evaluated in
the 2019 ISA continue to provide strong
evidence for a relationship between
short-term PM2.5 exposure and several
respiratory-related endpoints, including
asthma exacerbation (U.S. EPA, 2019a,
section 5.1.2.1), COPD exacerbation
(U.S. EPA, 2019a, section 5.1.4.1), and
combined respiratory-related diseases
(U.S. EPA, 2019a, section 5.1.6),
particularly from studies examining ED
visits and hospital admissions. The
generally positive associations between
short-term PM2.5 exposure and asthma
and COPD as well as ED visits and
hospital admissions are supported by
epidemiologic studies demonstrating
associations with other respiratoryrelated effects such as symptoms and
medication use that are indicative of
asthma and COPD exacerbations (U.S.
EPA, 2019a, sections 5.1.2.2 and
5.4.1.2). The collective body of
epidemiologic evidence for asthma
exacerbation is more consistent in
children than in adults. Additionally,
epidemiologic studies examining the
relationship between short-term PM2.5
exposure and respiratory mortality
provide evidence of consistent positive
associations, demonstrating a
continuum of effects (U.S. EPA, 2019a,
section 5.1.9).
Epidemiologic studies evaluated in
the 2019 ISA expand the assessment of
potential copollutant confounding
evaluated in the 2009 ISA. There is
some evidence that PM2.5 associations
with asthma exacerbation, combined
respiratory-related diseases, and
respiratory mortality remain relatively
unchanged in copollutant models with
gaseous pollutants including O3, NO2,
SO2, and with more limited evidence for
CO, as well as other particle sizes (i.e.,
PM10–2.5) (U.S. EPA, 2019a, section
5.1.10.1).
Insight into whether there is an
independent effect of PM2.5 on
respiratory health is also partially
addressed by findings from animal
toxicological studies evaluated in the
2019 ISA. Specifically, short-term
exposure to PM2.5 enhanced asthmarelated responses in an animal model of
allergic airways disease and enhanced
lung injury and inflammation in an
animal model of COPD (U.S. EPA,
PO 00000
Frm 00032
Fmt 4701
Sfmt 4700
2019a, sections 5.1.2.4.4 and 5.1.4.4.3).
The experimental evidence provides
biological plausibility for some
respiratory-related endpoints, including
limited evidence of altered host defense
and greater susceptibility to bacterial
infection as well as consistent evidence
of respiratory irritant effects. However,
animal toxicological evidence for other
respiratory effects is inconsistent. A
recent study evaluated in the ISA
supplement by Wyatt et al. (2020) and
conducted at near ambient PM2.5
concentrations, adds to the limited
evidence base of controlled human
exposure studies. The study, completed
in healthy young adults subject to
intermittent exercise, found some
significant respiratory effects (including
decrease in lung function), however
these findings were inconsistent with
the controlled human exposure studies
evaluated in the 2019 ISA (U.S. EPA,
2019a, section 5.1.7.2, 5.1.2.3, and
6.1.11.2.1).
The 2019 ISA concludes that ‘‘[t]he
strongest evidence of an effect of shortterm PM2.5 exposure on respiratory
effects is provided by epidemiologic
studies of asthma and COPD
exacerbation. While animal
toxicological studies provide biological
plausibility for these findings, some
uncertainty remains with respect to the
independence of PM2.5 effects’’ (U.S.
EPA, 2019a, p. 5–155). When taken
together, the 2019 ISA concludes that
this evidence ‘‘is sufficient to conclude
that a causal relationship is likely to
exist between short-term PM2.5 exposure
and respiratory effects’’ (U.S. EPA,
2019a, p. 5–155).
iv. Cancer
The 2009 ISA concluded that the
overall body of evidence was
‘‘suggestive of a causal relationship
between relevant PM2.5 exposures and
cancer’’ (U.S. EPA, 2009a). This
conclusion was based primarily on
positive associations observed in a
limited number of epidemiologic
studies of lung cancer mortality. The
few epidemiologic studies that had
evaluated PM2.5 exposure and lung
cancer incidence or cancers of other
organs and systems generally did not
show evidence of an association.
Toxicological studies did not focus on
exposures to specific PM size fractions,
but rather investigated the effects of
exposures to total ambient PM, or other
source-based PM such as wood smoke.
Collectively, results of in vitro studies
were consistent with the larger body of
evidence demonstrating that ambient
PM and PM from specific combustion
sources are mutagenic and genotoxic.
However, animal inhalation studies
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
found little evidence of tumor formation
in response to chronic exposures. A
small number of studies provided
preliminary evidence that PM exposure
can lead to changes in methylation of
DNA, which may contribute to
biological events related to cancer.
Since the completion of the 2009 ISA,
additional cohort studies provide
evidence that long-term PM2.5 exposure
is positively associated with lung cancer
mortality and with lung cancer
incidence, and provide initial evidence
for an association with reduced cancer
survival (U.S. EPA, 2019a, section
10.2.5). Re-analyses of the ACS cohort
using different years of PM2.5 data and
follow up, along with various exposure
assignment approaches, provide
consistent evidence of positive
associations between long-term PM2.5
exposure and lung cancer mortality
(U.S. EPA, 2019a, Figure 10–3).
Additional support for positive
associations with lung cancer mortality
is provided by recent epidemiologic
studies using individual level data to
control for smoking status, as well as by
studies of people who have never
smoked (though such studies generally
report wide confidence intervals due to
the small number of lung cancer
mortality cases within this population),
and in additional analyses of cohorts
that relied upon proxy measures to
account for smoking status (U.S. EPA,
2019a, section 10.2.5.1.1). Although
studies that evaluate lung cancer
incidence, including studies of people
who have never smoked, are limited in
number, studies in the 2019 ISA
generally report positive associations
with long-term PM2.5 exposures (U.S.
EPA, 2019a, section 10.2.5.1.2). A subset
of the studies focusing on lung cancer
incidence also examined histological
subtype, providing some evidence of
positive associations for
adenocarcinomas, the predominate
subtype of lung cancer observed in
people who have never smoked (U.S.
EPA, 2019a, section 10.2.5.1.2).
Associations between long-term PM2.5
exposure and lung cancer incidence
were found to remain relatively
unchanged, though in some cases
confidence intervals widened, in
analyses that attempted to reduce
exposure measurement error by
accounting for length of time at
residential address or by examining
different exposure assignment
approaches (U.S. EPA, 2019a, section
10.2.5.1.2).
To date, relatively few studies have
evaluated the potential for copollutant
confounding of the relationship between
long-term PM2.5 exposure and lung
cancer mortality or incidence. A small
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
number of such studies have generally
focused on O3 and report that PM2.5
associations remain relatively
unchanged in copollutant models (U.S.
EPA, 2019a, section 10.2.5.1.3).
However, available studies have not
systematically evaluated the potential
for copollutant confounding by other
gaseous pollutants or by other particle
size fractions (U.S. EPA, 2019a, section
10.2.5.1.3).
Compared to total (non-accidental)
mortality (U.S. EPA, 2019a, section
10.2.4.1.4), fewer studies have examined
the shape of the C–R curve for causespecific mortality outcomes, including
lung cancer. Several studies of lung
cancer mortality and incidence have
reported no evidence of deviations from
linearity in the shape of the C–R
relationship (Lepeule et al., 2012;
Raaschou-Nielsen et al., 2013; Puett et
al., 2014), though authors provided only
limited discussions of results (U.S. EPA,
2019a, section 10.2.5.1.4).
In support of the biological
plausibility of an independent effect of
PM2.5 on lung cancer, the 2019 ISA
notes evidence from experimental and
epidemiologic studies demonstrating
that PM2.5 exposure can lead to a range
of effects indicative of mutagenicity,
genotoxicity, and carcinogenicity, as
well as epigenetic effects (U.S. EPA,
2019a, section 10.2.7). For example,
both in vitro and in vivo toxicological
studies have shown that PM2.5 exposure
can result in DNA damage (U.S. EPA,
2019a, section 10.2.2). Although such
effects do not necessarily equate to
carcinogenicity, the evidence that PM
exposure can damage DNA, and elicit
mutations, provides support for the
plausibility of epidemiologic
associations exhibited with lung cancer
mortality and incidence. Additional
supporting studies indicate the
occurrence of micronuclei formation
and chromosomal abnormalities (U.S.
EPA, 2019a, section 10.2.2.3), and
differential expression of genes that may
be relevant to cancer pathogenesis,
following PM2.5 exposures.
Experimental and epidemiologic studies
that examine epigenetic effects indicate
changes in DNA methylation, providing
some support that PM2.5 exposure
contributes to genomic instability (U.S.
EPA, 2019a, section 10.2.3). Overall,
there is limited evidence that long-term
PM2.5 exposure is associated with
cancers in other organ systems, though
there is some evidence that PM2.5
exposure may reduce survival in
individuals with cancer (U.S. EPA,
2019a, section 10.2.7; U.S. EPA, 2022a,
section 2.1.1.4.1).
Epidemiologic evidence for
associations between PM2.5 and lung
PO 00000
Frm 00033
Fmt 4701
Sfmt 4700
16233
cancer mortality and incidence, together
with evidence supporting the biological
plausibility of such associations,
contributes to the 2019 ISA’s conclusion
that the evidence ‘‘is sufficient to
conclude that a causal relationship is
likely to exist between long-term PM2.5
exposure and cancer’’ (U.S. EPA, 2019,
section 10.2.7).
v. Nervous System Effects
Reflecting the very limited evidence
available in the 2012 review, the 2009
ISA did not make a causality
determination for long-term PM2.5
exposures and nervous system effects
(U.S. EPA, 2009c). Since the 2012
review, this body of evidence has grown
substantially (U.S. EPA, 2019, section
8.2). Animal toxicological studies
assessed in in the 2019 ISA report that
long-term PM2.5 exposures can lead to
morphologic changes in the
hippocampus and to impaired learning
and memory. This evidence is
consistent with epidemiologic studies
reporting that long-term PM2.5 exposure
is associated with reduced cognitive
function (U.S. EPA, 2019a, section
8.2.5). Further, while the evidence is
limited, the presence of early markers of
Alzheimer’s disease pathology has been
demonstrated in rodents following longterm exposure to PM2.5 CAPs. These
findings support reported associations
with neurodegenerative changes in the
brain (i.e., decreased brain volume), allcause dementia, or hospitalization for
Alzheimer’s disease in a small number
of epidemiologic studies (U.S. EPA,
2019a, section 8.2.6). Additionally, loss
of dopaminergic neurons in the
substantia nigra, a hallmark of
Parkinson disease, has been reported in
mice (U.S. EPA, 2019a, section 8.2.4),
though epidemiologic studies provide
only limited support for associations
with Parkinson’s disease (U.S. EPA,
2019a, section 8.2.6). Overall, the lack of
consideration of copollutant
confounding introduces some
uncertainty in the interpretation of
epidemiologic studies of nervous system
effects, but this uncertainty is partly
addressed by the evidence for an
independent effect of PM2.5 exposures
provided by experimental animal
studies.
While the findings described above
are most relevant to older adults, several
studies of neurodevelopmental effects in
children have also been conducted.
Epidemiologic studies provided limited
evidence of an association between
PM2.5 exposure during pregnancy and
childhood on cognitive and motor
development (U.S. EPA, 2019, section
8.2.5.2). While some studies report
positive associations between long-term
E:\FR\FM\06MRR3.SGM
06MRR3
16234
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
exposure to PM2.5 during the prenatal
period and autism spectrum disorder
(ASD) (U.S. EPA, 2019, section 8.2.7.2),
the interpretation of these
epidemiologic studies is limited due to
the small number of studies, their lack
of control for potential confounding by
copollutants, and uncertainty related to
the critical exposure windows.
Biological plausibility is provided for
the ASD findings by a study in mice that
found inflammatory and morphologic
changes in the corpus collosum and
hippocampus, as well as
ventriculomegaly (i.e., enlarged lateral
ventricles) in young mice following
prenatal exposure to PM2.5 CAPs.
Taken together, the 2019 ISA
concludes that studies indicate longterm PM2.5 exposures can lead to effects
on the brain associated with
neurodegeneration (i.e.,
neuroinflammation and reductions in
brain volume), as well as cognitive
effects in older adults (U.S. EPA, 2019a,
Table 1–2). Animal toxicological studies
provide evidence for a range of nervous
system effects in adult animals,
including neuroinflammation and
oxidative stress, neurodegeneration,
cognitive effects, and effects on
neurodevelopment in young animals.
The epidemiologic evidence is more
limited, but studies generally support
associations between long-term PM2.5
exposure and changes in brain
morphology, cognitive decrements and
dementia. There is also initial, and
limited, evidence for
neurodevelopmental effects, particularly
ASD. The consistency and coherence of
the evidence supports the 2019 ISA’s
conclusion that ‘‘the collective evidence
is sufficient to conclude that a causal
relationship is likely to exist between
long-term PM2.5 exposure and nervous
system effects’’ (U.S. EPA, 2019a,
section 8.2.9).
vi. Other Effects
For other health effect categories that
were evaluated for their relationship
with PM2.5 exposures (i.e., short-term
PM2.5 exposure and nervous system
effects and short- and long-term PM2.5
exposure and metabolic effects,
reproduction and fertility, and
pregnancy and birth outcomes (U.S.
EPA, 2022a, Table ES–1), the currently
available evidence is ‘‘suggestive of, but
not sufficient to infer, a causal
relationship,’’ mainly due to
inconsistent evidence across specific
outcomes and uncertainties regarding
exposure measurement error, the
potential for confounding, and potential
modes of action (U.S. EPA, 2019a,
sections 7.14, 7.2.10, 8.1.6, and 9.1.5).
The causality determination for short-
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
term PM2.5 exposure and nervous
system effects in the 2019 ISA reflects
a revision to the causality determination
in the 2009 ISA from ‘‘inadequate to
infer a causal relationship,’’ while this
is the first-time assessments of causality
were conducted for long-term PM2.5
exposure and nervous system effects, as
well as short- and long-term PM2.5
exposure and metabolic effects reflect.
Recent studies evaluated in the 2019
ISA also further explored the
relationship between short-and longterm UFP exposure and health effects.
(i.e., cardiovascular effects and shortterm UFP exposures; respiratory effects
and short-term UFP exposures; and
nervous system effects and long- and
short-term exposures (U.S. EPA, 2022a,
Table ES–1). The currently available
evidence is ‘‘suggestive of, but not
sufficient to infer, a causal relationship’’
for short-term UFP exposure and
cardiovascular and respiratory effects
and for short- and long-term UFP
exposure and nervous system effects,
primarily due to uncertainties and
limitations in the evidence, specifically,
variability across studies in the
definition of UFPs and the exposure
metric used (U.S. EPA, 2019a, P.3.1;
U.S. EPA, 2022a, section 3.3.1.6.3). The
causality determinations for the other
health effect categories evaluated in the
2019 ISA are ‘‘inadequate to infer a
causal relationship.’’ Additionally, this
is the first time assessments of causality
were conducted for short- and long-term
UFP exposure and metabolic effects and
long-term UFP exposure and nervous
system effects (U.S. EPA, 2022a, Table
ES–1).
With the advent of the global COVID–
19 pandemic, a number of recent studies
evaluated in the ISA Supplement
examined the relationship between
ambient air pollution, specifically PM2.5,
and SARS–CoV–2 infections and
COVID–19 deaths, including a few
studies within the U.S. and Canada
(U.S. EPA, 2022a, section 3.3.2).59 Some
59 While there is no exact corollary within the
2019 ISA for these types of studies, the 2019 ISA
presented evidence that evaluates the potential
relationship between short- and long-term PM2.5
exposure and respiratory infection (U.S. EPA,
2022a, section 5.1.5 and 5.2.6). Studies assessed in
the 2019 ISA report some evidence of positive
associations between short-term PM2.5 and hospital
admissions and ED visits for respiratory infections,
however the interpretation of these studies is
complicated by the variability in the type of
respiratory infection outcome examined (U.S. EPA,
2022a, Figure 5–7). In the 2019 ISA, studies of longterm PM2.5 exposure were limited and while there
were some positive associations reported, there was
minimal overlap in respiratory infection outcomes
examined across studies. Exposure to PM2.5 has
been shown to impair host defense, specifically
altering macrophage function, providing a
biological pathway by which PM2.5 exposure could
lead to respiratory infection (U.S. EPA, 2022a,
PO 00000
Frm 00034
Fmt 4701
Sfmt 4700
studies examined whether daily changes
in PM2.5 can influence SARS–CoV–2
infection and COVID–19 death (U.S.
EPA, 2022a, section 3.3.2.1).
Additionally, several studies evaluated
whether long-term PM2.5 exposure
increases the risk of SARS–CoV–2
infection and COVID–19 death in North
America (U.S. EPA, 2022a, section
3.3.2.2). While there is initial evidence
of positive associations with SARS–
CoV–2 infection and COVID–19 death,
uncertainties remain due to
methodological issues that may
influence the results, including: (1) The
use of ecological study design; (2)
studies were conducted during the
ongoing pandemic when the etiology of
COVID–19 was still not well understood
(e.g., specifically, there are important
differences in COVID–19-related
outcomes by a variety of factors such as
race and SES); and (3) studies did not
account for crucial factors that could
influence results (e.g., stay-at-home
orders, social distancing, use of masks,
and testing capacity) (U.S. EPA, 2022a,
chapter 5). Taken together, while there
is initial evidence of positive
associations with SARS–CoV–2
infection and COVID–19 death,
uncertainties remain due to
methodological issues.
b. Public Health Implications and AtRisk Populations
The public health implications of the
evidence regarding PM2.5-related health
effects, as for other effects, are
dependent on the type and severity of
the effects, as well as the size of the
population affected. Such factors are
discussed below in the context of our
consideration of the health effects
evidence related to PM2.5 in ambient air.
This section also summarizes the
current information on population
groups at increased risk of the effects of
PM2.5 in ambient air.
The information available in this
reconsideration has not altered our
understanding of human populations at
risk of health effects from PM2.5
exposures. As recognized in the 2020
review, the 2019 ISA cites extensive
evidence indicating that ‘‘both the
general population as well as specific
populations and lifestages are at risk for
PM2.5-related health effects’’ (U.S. EPA,
2019a, p. 12–1). Factors that may
contribute to increased risk of PM2.5related health effects include lifestage
(children and older adults), pre-existing
diseases (cardiovascular disease and
sections 5.1.1 and 5.1.5.) There is some additional
evidence that PM2.5 exposure can lead to decreases
in an individual’s immune response, which can
subsequently facilitate replication of respiratory
viruses (Bourdrel et al., 2021).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
respiratory disease), race/ethnicity, and
SES.60
Children make up a substantial
fraction of the U.S. population, and
often have unique factors that contribute
to their increased risk of experiencing a
health effect due to exposures to
ambient air pollutants because of their
continuous growth and development.61
Children may be particularly at risk for
health effects related to ambient PM2.5
exposures compared with adults
because they have (1) a developing
respiratory system, (2) increased
ventilation rates relative to body mass
compared with adults, and (3) an
increased proportion of oral breathing,
particularly in boys, relative to adults
(U.S. EPA, 2019a, section 12.5.1.1).
There is strong evidence that
demonstrates PM2.5 associated health
effects in children, particularly from
epidemiologic studies of long-term
PM2.5 exposure and impaired lung
function growth, decrements in lung
function, and asthma development.
However, there is limited evidence from
stratified analyses that children are at
increased risk of PM2.5-related health
effects compared to adults.
Additionally, there is some evidence
that indicates that children receive
higher PM2.5 exposures than adults, and
dosimetric differences in children
compared to adults can contribute to
higher doses (U.S. EPA, 2019a, section
12.5.1.1).
In the U.S., older adults, often defined
as adults 65 years of age and older,
represent an increasing portion of the
population and often have pre-existing
diseases or conditions that may
compromise biological function. While
there is limited evidence to indicate that
older adults have higher exposures than
younger adults, older adults may receive
higher doses of PM2.5 due to dosimetric
differences. There is consistent evidence
from studies of older adults
demonstrating generally consistent
positive associations in studies
examining health effects from short- and
long-term PM2.5 exposure and
cardiovascular or respiratory hospital
admissions, emergency department
visits, or mortality (U.S. EPA, 2019a,
sections 6.1, 6.2, 11.1, 11.2, 12.5.1.2).
Additionally, several animal
toxicological, controlled human
exposure, and epidemiologic studies did
not stratify results by lifestage, but
instead focused the analyses on older
60 As described in the 2019 ISA, other factors that
have the potential to contribute to increased risk
include obesity, diabetes, genetic factors, smoking
status, sex, diet, and residential location (U.S. EPA,
2019, chapter 12).
61 Children, as used throughout this document,
generally refers to those younger than 18 years old.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
individuals, and can provide coherence
and biological plausibility for the
occurrence among this lifestage (U.S.
EPA, 2019a, section 12.5.1.2).
Individuals with pre-existing disease
may be considered at greater risk of an
air pollution-related health effect than
those without disease because they are
likely in a compromised biological state
that can vary depending on the disease
and severity. With regard to
cardiovascular disease, we first note that
cardiovascular disease is the leading
cause of death in the U.S., accounting
for one in four deaths, and
approximately 12% of the adult
population in the U.S. has a
cardiovascular disease (U.S. EPA,
2019a, section 12.3.1). Strong evidence
demonstrates that there is a causal
relationship between cardiovascular
effects and long- and short-term
exposures to PM2.5. Some of the
evidence supporting this conclusion is
from studies of panels or cohorts with
pre-existing cardiovascular disease,
which provide supporting evidence but
do not directly demonstrate an
increased risk (U.S. EPA, 2019a, section
12.3.1). Epidemiologic evidence
indicates that individuals with preexisting cardiovascular disease may be
at increased risk for PM2.5-associated
health effects compared to those
without pre-existing cardiovascular
disease. While the evidence does not
consistently support increased risk for
all pre-existing cardiovascular diseases,
there is evidence that certain preexisting cardiovascular diseases (e.g.,
hypertension) may be a factor that
increases PM2.5-related risk.
Furthermore, there is strong evidence
supporting a causal relationship for
long- and short-term PM2.5 exposure and
cardiovascular effects, particularly for
IHD (U.S. EPA, 2019a, chapter 6, section
12.3.1).
With regard to respiratory disease, we
first note that the most chronic
respiratory diseases in the U.S. are
asthma and COPD. Asthma affects a
substantial fraction of the U.S.
population and is the leading chronic
disease among children. COPD
primarily affects older adults and
contributes to compromised respiratory
function and underlying pulmonary
inflammation. The body of evidence
indicates that individuals with preexisting respiratory diseases,
particularly asthma and COPD, may be
at increased risk for PM2.5-related health
effects compared to those without preexisting respiratory diseases (U.S. EPA,
2019a, section 12.3.5). There is strong
evidence indicating PM2.5-associated
respiratory effects among those with
asthma, which forms the primary
PO 00000
Frm 00035
Fmt 4701
Sfmt 4700
16235
evidence base for the likely to be causal
relationship between short-term
exposures to PM2.5 and respiratory
health effects (U.S. EPA, 2019a, section
12.3.5). For asthma, epidemiologic
evidence demonstrates associations
between short-term PM2.5 exposures and
respiratory effects, particularly evidence
for asthma exacerbation, and controlled
human exposure and animal
toxicological studies demonstrate
support for the biological plausibility
for asthma exacerbation with PM2.5
exposures (U.S. EPA, 2019a, section
12.3.5.1). For COPD, epidemiologic
studies report positive associations
between short-term PM2.5 exposures and
hospital admissions and emergency
department visits for COPD, with
supporting evidence from panel studies
demonstration COPD exacerbation.
Epidemiologic evidence is supported by
some experimental evidence of COPDrelated effects, which provides support
for the biological plausibility for COPD
in response to PM2.5 exposures (U.S.
EPA, 2019a, section 12.3.5.2).
There is strong evidence for racial and
ethnic disparities in PM2.5 exposures
and PM2.5-related health risk, as
assessed in the 2019 ISA and with even
more evidence available since the
literature cutoff date for the 2019 ISA
and evaluated in the ISA Supplement.
There is strong evidence demonstrating
that Black and Hispanic populations, in
particular, have higher PM2.5 exposures
than non-Hispanic White populations
(U.S. EPA, 2019a, Figure 12–2; U.S.
EPA, 2022a, Figure 3–38). Black
populations or individuals that live in
predominantly Black neighborhoods
experience higher PM2.5 exposures, in
comparison to non-Hispanic White
populations. There is also consistent
evidence across multiple studies that
demonstrate increased risk of PM2.5related health effects, with the strongest
evidence for health risk disparities for
mortality (U.S. EPA, 2019a, section
12.5.4). There is also evidence of health
risk disparities for both Hispanic and
non-Hispanic Black populations
compared to non-Hispanic White
populations for cause-specific mortality
and incident hypertension (U.S. EPA,
2022a, section 3.3.3.2).
Socioeconomic status (SES) is a
composite measure that includes
metrics such as income, occupation, or
education, and can play a role in access
to healthy environments as well as
access to healthcare. SES may be a
factor that contributes to differential risk
from PM2.5-related health effects.
Studies assessed in the 2019 ISA and
ISA Supplement provide evidence that
lower SES communities are exposed to
higher concentrations of PM2.5
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16236
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
compared to higher SES communities
(U.S. EPA, 2019a, section 12.5.3; U.S.
EPA, 2022a, section 3.3.3.1.1). Studies
using composite measures of
neighborhood SES consistently
demonstrated a disparity in both PM2.5
exposure and the risk of PM2.5-related
health outcomes. There is some
evidence that supports associations
larger in magnitude between mortality
and long-term PM2.5 exposures for those
with low income or living in lower
income areas compared to those with
higher income or living in higher
income neighborhoods (U.S. EPA,
2019a, section 12.5.3; U.S. EPA, 2022a,
section 3.3.3.1.1). Additionally,
evidence supports conclusions that
lower SES is associated with causespecific mortality and certain health
endpoints (i.e., HI and CHF), but less so
for all-cause or total (non-accidental)
mortality (U.S. EPA, 2022a, section
3.3.3.1).
The magnitude and characterization
of a public health impact is dependent
upon the size and characteristics of the
populations affected, as well as the type
or severity of the effects. As summarized
above, lifestage (children and older
adults), race/ethnicity and SES are
factors that increase the risk of PM2.5related health effects. The American
Community Survey (ACS) for 2019
estimates that approximately 22% and
16% of the U.S. population are children
(age<18) and older adults (age 65+),
respectively. For all ages, non-Hispanic
Black and Hispanic populations
comprise approximately 12% and 18%
of the overall U.S. population in 2019.
Currently available information that
helps to characterize key features of
these population is included in the 2022
PA (U.S. EPA, 2022b, Table 3–2).
As noted above, individuals with preexisting cardiovascular disease and preexisting respiratory disease may also be
at increased risk of PM2.5-related health
effects. Currently available information
that helps to characterize key features of
populations with cardiovascular or
respiratory diseases or conditions is
included in the 2022 PA (U.S. EPA,
2022b, Table 3–3). The National Center
for Health Statistics data for 2018
indicate that, for adult populations,
older adults (e.g., those 65 years and
older) have a higher prevalence of
cardiovascular diseases compared to
younger adults (e.g., those 64 years and
younger). For respiratory diseases, older
adults also have a higher prevalence of
emphysema than younger adults, and
adults 44 years or older have a higher
prevalence of chronic bronchitis.
However, the prevalence for asthma is
generally similar across all adult age
groups.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
With respect to race, American
Indians or Alaskan Native populations
have the highest prevalence of all heart
disease and coronary heart disease,
while Black populations have the
highest prevalence of hypertension and
stroke. Hypertension has the highest
prevalence across all racial groups
compared to other cardiovascular
diseases or conditions, ranging from
approximately 22% to 32% of each
racial group. Overall, the prevalence of
cardiovascular diseases or conditions is
lowest for Asians compared to Whites,
Blacks, and American Indians or
Alaskan Natives. Asthma prevalence is
highest among Black and American
Indian or Alaska Native populations,
while the prevalence of chronic
bronchitis and emphysema is generally
similar across racial groups. Overall, the
prevalence of respiratory diseases is
lowest for Asians compared to Whites,
Blacks, and American Indians or
Alaskan Natives. With regard to
ethnicity, cardiovascular and respiratory
disease prevalence across all diseases or
conditions is generally similar between
Hispanic and non-Hispanic populations,
although non-Hispanics have a slightly
higher prevalence compared to
Hispanics.
Taken together, this information
indicates that the groups at increased
risk of PM2.5-related health effects
represent a substantial portion of the
total U.S. population. In evaluating the
primary PM2.5 standards, an important
consideration is the potential PM2.5related public health impacts in these
populations.
c. PM2.5 Concentrations in Key Studies
Reporting Health Effects
To inform conclusions on the
adequacy of the public health protection
provided by the current primary PM2.5
standards, the sections below
summarize the 2022 PA’s evaluation of
the PM2.5 exposures, specifically the
concentrations that have been examined
in controlled human exposure studies,
animal toxicological studies, and
epidemiologic studies. The 2022 PA
places the greatest emphasis on the
health outcomes for which the 2019 ISA
concludes that the evidence supports a
‘‘causal’’ or a ‘‘likely to be causal’’
relationship with short- or long-term
PM2.5 exposures (U.S. EPA, 2022b,
section 3.3.3). As described in greater
detail in section II.A.2 above, this
includes short- or long-term PM2.5
exposures and mortality, cardiovascular
effects, and respiratory effects and longterm PM2.5 exposures and cancer and
nervous system effects. While the
causality determinations in the 2019
ISA are informed by studies evaluating
PO 00000
Frm 00036
Fmt 4701
Sfmt 4700
a wide range of PM2.5 concentrations,62
the sections below summarize the
considerations in the 2022 PA regarding
the degree to which the evidence
assessed in the 2019 ISA and ISA
Supplement supports the occurrence of
PM-related health effects at
concentrations relevant to informing
conclusions on the primary PM2.5
standards. In so doing, the 2022 PA
focuses on the available studies that are
most directly informative to reaching
conclusions regarding the adequacy of
the current primary PM2.5 standards
(e.g., epidemiologic studies with annual
mean PM2.5 concentrations near or
below the level of the standard; and
controlled human exposure studies at
PM2.5 exposures that elicit consistent
effects, as well as examining PM2.5
exposures at concentrations that are at
or near the level of the standard).
i. PM2.5 Exposure Concentrations
Evaluated in Experimental Studies
Evidence for a particular PM2.5-related
health outcome is strengthened when
results from experimental studies
demonstrate biologically plausible
mechanisms through which adverse
human health outcomes could occur
(U.S. EPA, 2015, p. 20). Two types of
experimental studies are of particular
importance in understanding the effects
62 As described in more detail in section 5 of the
Preamble to the ISAs, judgments regarding causality
take into consideration a number of aspects when
evaluating the available scientific evidence (U.S.
EPA, 2015, Table I). In reaching conclusions
regarding causality, ‘‘evidence is evaluated for
major outcome categories or groups of related
endpoints (e.g., respiratory effects, vegetation
growth), integrating evidence from across
disciplines, and evaluating the coherence of
evidence across a spectrum of related endpoints’’
(U.S. EPA, 2015, p. 24). Furthermore, ‘‘[i]n drawing
judgments regarding causality for the criteria air
pollutants, the ISA focuses on evidence of effects
in the range of relevant pollutant exposures or
doses and not on determination of causality at any
dose. Emphasis is placed on evidence of effects at
doses (e.g., blood Pb concentration) or exposures
(e.g., air concentrations) that are relevant to, or
somewhat above, those currently experienced by
the population. The extent to which studies of
higher concentrations are considered varies by
pollutant and major outcome category, but generally
includes those with doses or exposures in the range
of one to two orders of magnitude above current or
ambient conditions to account for intra-species
variability and toxicokinetic or toxicodynamic
differences between experimental animals and
humans. Studies that use higher doses or exposures
may also be considered to the extent that they
provide useful information to inform understanding
of mode of action, inter-species differences, or
factors that may increase risk of effects for a
population and if biological mechanisms have not
been demonstrated to differ based on exposure
concentration. Thus, a causality determination is
based on weight-of-evidence evaluation for health
or welfare effects, focusing on the evidence from
exposures or doses generally ranging from recent
ambient concentrations to one or two orders of
magnitude above recent ambient concentrations’’
(U.S. EPA, 2015, p. 24).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
of PM exposures: controlled human
exposure and animal toxicological
studies. In such studies, investigators
expose human volunteers or laboratory
animals, respectively, to known
concentrations of air pollutants under
carefully regulated environmental
conditions and activity levels. Thus,
controlled human exposure and animal
toxicological studies can provide
information on the health effects of
experimentally administered pollutant
exposures under highly controlled
laboratory conditions (U.S. EPA, 2015,
p. 11).
Controlled human exposure studies
have reported that PM2.5 exposures
lasting from less than one hour up to
five hours can impact cardiovascular
function,63 and the most consistent
evidence from these studies is for
impaired vascular function (U.S. EPA,
2019a, section 6.1.13.2). In addition,
although less consistent, the 2019 ISA
notes that studies examining PM2.5
exposures also provide evidence for
increased blood pressure (U.S. EPA,
2019a, section 6.1.6.3), conduction
abnormalities/arrhythmia (U.S. EPA,
2019a, section 6.1.4.3), changes in heart
rate variability (U.S. EPA, 2019a, section
6.1.10.2), changes in hemostasis that
could promote clot formation (U.S. EPA,
2019a, section 6.1.12.2), and increases
in inflammatory cells and markers (U.S.
EPA, 2019a, section 6.1.11.2). The 2019
ISA concludes that, when taken as a
whole, controlled human exposure
studies demonstrate that short-term
exposure to PM2.5 may impact
cardiovascular function in ways that
could lead to more serious outcomes
(U.S. EPA, 2019a, section 6.1.16). Thus,
such studies can provide insight into
the potential for specific PM2.5
exposures to result in physiological
changes that could increase the risk of
more serious effects. Table 3–4 in the
2022 PA summarizes information from
the 2019 ISA and 2022 ISA supplement
on available controlled human exposure
studies that evaluate effects on markers
of cardiovascular function following
exposure to PM2.5 (U.S. EPA, 2022b).
Most of the controlled human exposure
studies in Table 3–4 of the 2022 PA
have evaluated average PM2.5
concentrations at or above about 100 mg/
m3, with exposure durations typically
up to about two hours. Statistically
significant effects on one or more
indicators of cardiovascular function are
63 In contrast, controlled human exposure studies
provide little evidence for respiratory effects
following short-term PM2.5 exposures (U.S. EPA,
2019a, section 5.1, Table 5–18). Therefore, this
section focuses on cardiovascular effects evaluated
in controlled human exposure studies of PM2.5
exposure.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
often, though not always, reported
following 2-hour exposures to average
PM2.5 concentrations at and above about
120 mg/m3, with less consistent
evidence for effects following exposures
to concentrations lower than 120 mg/m3.
Impaired vascular function, the effect
identified in the 2019 ISA as the most
consistent across studies (U.S. EPA,
2019a, section 6.1.13.2) is shown
following 2-hour exposures to PM2.5
concentrations at and above 149 mg/m3.
Mixed results are reported in the studies
that evaluated longer exposure
durations (i.e., longer than 2 hours) and
lower (i.e., near-ambient) PM2.5
concentrations (U.S. EPA, 2022b,
section 3.3.3.1). For example, significant
effects for some outcomes were reported
following 5-hour exposures to 24 mg/m3
in Hemmingsen et al. (2015b), but not
for other outcomes following 5-hour
exposures to 24 mg/m3 in Hemmingsen
et al. (2015a) and not following 24-hour
exposures to 10.5 mg/m3 in Bra¨uner et
al. (2008). Additionally, Wyatt et al.
(2020) found significant effects for some
cardiovascular (e.g., systematic
inflammation markers, cardiac
repolarization, and decreased
pulmonary function) effects following 4hour exposures to 37.8 mg/m3 in healthy
young participants (18–35 years, n=21)
who were subject to intermittent
moderate exercise. The higher
ventilation rate and longer exposure
duration in this study compared to most
controlled human exposure studies is
roughly equivalent to a 2-hour exposure
of 75–100 mg/m3 of PM2.5. Therefore,
dosimetric considerations may explain
the observed changes in inflammation
in young healthy individuals. Though
this study provides evidence of some
effects at lower PM2.5 concentrations,
overall, there is inconsistent evidence
for inflammation in other controlled
human exposure studies evaluated in
the 2019 ISA (U.S. EPA, 2019a, sections
5.1.7., 5.1.2.3.3, and 6.1.11.2.1; U.S.
EPA, 2022a, section 3.3.1).
While controlled human exposure
studies are important in establishing
biological plausibility, it is unclear how
the results from these studies alone and
the importance of the effects observed in
these studies, should be interpreted
with respect to adversity to public
health. More specifically, impaired
vascular function can signal an
intermediate effect along the potential
biological pathways for cardiovascular
effects following short-term exposure to
PM2.5 and show a role for exposure to
PM2.5 leading to potential worsening of
IHD and heart failure followed
potentially by ED visits, hospital
admissions, or mortality (U.S. EPA,
PO 00000
Frm 00037
Fmt 4701
Sfmt 4700
16237
2019a, section 6.1 and Figure 6–1).
However, just observing the occurrence
of impaired vascular function alone
does not clearly suggest an adverse
health outcome. Additionally,
associated judgments regarding
adversity or health significance of
measurable physiological responses to
air pollutants have been informed by
guidance, criteria or interpretative
statements developed within the public
health community, including the
American Thoracic Society (ATS) and
the European Respiratory Society (ERS),
which cooperatively updated the ATS
2000 statement What Constitutes an
Adverse Health Effect of Air Pollution
(ATS, 2000) with new scientific
findings, including the evidence related
to air pollution and the cardiovascular
system (Thurston et al., 2017).64 With
regard to vascular function, the ATS/
ERS statement considers the adversity of
both chronic and acute reductions in
endothelial function. While the ATS/
ERS statement concluded that chronic
endothelial and vascular dysfunction
can be judged to be a biomarker of an
adverse health effect from air pollution,
they also conclude that ‘‘the health
relevance of acute reductions in
endothelial function induced by air
pollution is less certain’’ (Thurston et
al., 2017). This is particularly
informative to our consideration of the
controlled human exposure studies
which are short-term in nature (i.e.,
generally ranging from 2- to 5-hours),
including those studies that are
conducted at near-ambient PM2.5
concentrations.
The 2022 PA also notes that it is
important to recognize that controlled
human exposure studies include a small
number of individuals compared to
epidemiologic studies. Additionally,
these studies tend to include generally
healthy adult individuals, who are at a
lower risk of experiencing health effects.
64 The ATS/ERS described its 2017 statement as
one ‘‘intended to provide guidance to policymakers,
clinicians and public health professionals, as well
as others who interpret the scientific evidence on
the health effects of air pollution for risk
management purposes’’ and further notes that
‘‘considerations as to what constitutes an adverse
health effect, in order to provide guidance to
researchers and policymakers when new health
effects markers or health outcome associations
might be reported in future.’’ The most recent
policy statement by the ATS, which once again
broadens its discussion of effects, responses and
biomarkers to reflect the expansion of scientific
research in these areas, reiterates that concept,
conveying that it does not offer ‘‘strict rules or
numerical criteria, but rather proposes
considerations to be weighed in setting boundaries
between adverse and nonadverse health effects,’’
providing a general framework for interpreting
evidence that proposes a ‘‘set of considerations that
can be applied in forming judgments’’ for this
context (Thurston et al., 2017).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16238
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
These studies, therefore, often do not
include children, older adults, or
individuals with pre-existing
conditions. As such, these studies are
somewhat limited in their ability to
inform at what concentrations effects
may be elicited in at-risk populations.
Nonetheless, to provide some insight
into what these controlled human
exposure studies may indicate regarding
short-term exposure to peak PM2.5
concentrations and how concentrations
relate to ambient PM2.5 concentrations,
analyses in the 2022 PA (U.S. EPA,
2022b, Figure 2–19) examine monitored
2-hour PM2.5 concentrations (the
exposure window most often utilized in
the controlled human exposure studies)
at sites meeting the current primary
PM2.5 standards to evaluate the degree to
which 2-hour ambient PM2.5
concentrations at such locations are
likely to exceed the 2-hour exposure
concentrations in the controlled human
exposure studies at which statistically
significant effects are reported in
multiple studies for one or more
indicators of cardiovascular function. At
sites meeting the current primary PM2.5
standards, most 2-hour concentrations
are below 10 mg/m3, and almost never
exceed 30 mg/m3. The extreme upper
end of the distribution of 2-hour PM2.5
concentrations is shifted higher during
the warmer months (April to
September), generally corresponding to
the period of peak wildfire frequency in
the U.S. At sites meeting the current
primary PM2.5 standards, the highest 2hour concentrations measured tend to
occur during the period of peak wildfire
frequency (i.e., 99.9th percentile of 2hour concentrations is 62 mg/m3 during
the warm season considered as a
whole). Most of the sites measuring
these very high concentrations are in the
northwestern U.S. and California (U.S.
EPA, 2022b, Appendix A, Figure A–1),
where wildfires have been relatively
common in recent years. When the
typical fire season is excluded from the
analysis, the extreme upper end of the
distribution is reduced (i.e., 99.9th
percentile of 2-hour concentrations is 55
mg/m3).65 Given these results, the 2022
PA concludes that PM2.5 exposure
concentrations evaluated in most of
these controlled human exposure
studies are well-above the 2-hour
ambient PM2.5 concentrations typically
measured in locations meeting the
current primary standards.
With respect to animal toxicological
studies, the 2019 ISA relies on animal
65 Similar analyses of 4-hour and 5-hour PM
2.5
concentrations are presented in Appendix A, Figure
A–2 and Figure A–3, respectively of the 2022 PA
(U.S. EPA, 2022b).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
toxicological studies to support the
plausibility of a wide range of PM2.5related health effects. While animal
toxicological studies often examine
more severe health outcomes and longer
exposure durations than controlled
human exposure studies, there is
uncertainty in extrapolating the effects
seen in animals, and the PM2.5
exposures and doses that cause those
effects, to human populations. The 2022
PA considers these uncertainties when
evaluating what the available animal
toxicological studies may indicate with
regard to the current primary PM2.5
standards.
As with controlled human exposure
studies, most animal toxicological
studies evaluated in the 2019 ISA have
examined effects following exposure to
PM2.5 well above the concentrations
likely to be allowed by the current PM2.5
standards. Such studies have generally
examined short-term exposures to PM2.5
concentrations ranging from 100 to
>1,000 mg/m3 and long-term exposures
to concentrations from 66 to >400 mg/m3
(e.g., see U.S. EPA, 2019a, Table 1–2).
Two exceptions are animal toxicological
studies reporting impaired lung
development following long-term
exposures (i.e., 24 hours per day for
several months prenatally and
postnatally) to an average PM2.5
concentration of 16.8 mg/m3 (Mauad et
al., 2008) and increased carcinogenic
potential following long-term exposures
(i.e., 2 months) to an average PM2.5
concentration of 17.7 mg/m3 (Cangerana
Pereira et al., 2011). These two studies
report serious effects following longterm exposures to PM2.5 concentrations
similar to the ambient concentrations
reported in some PM2.5 epidemiologic
studies (U.S. EPA, 2019a, Table 1–2),
though still above the ambient
concentrations likely to occur in areas
meeting the current primary PM2.5
standards. However, noting uncertainty
in extrapolating the effects seen in
animals, and the PM2.5 exposures and
doses that cause those effects to human
populations, animal toxicological
studies are of limited utility in
informing decisions on the public
health protection provided by the
current or alternative primary PM2.5
standards. Therefore, the animal
toxicological studies are most useful in
providing further evidence to support
the biological mechanisms and
plausibility of various adverse effects.
ii. Ambient PM2.5 Concentrations in
Locations of Epidemiologic Studies
As summarized in section II.A.2.a
above, epidemiologic studies examining
associations between daily or annual
average PM2.5 exposures and mortality
PO 00000
Frm 00038
Fmt 4701
Sfmt 4700
or morbidity represent a large part of the
evidence base supporting several of the
2019 ISA’s ‘‘causal’’ and ‘‘likely to be
causal’’ determinations. The 2022 PA
considers the ambient PM2.5
concentrations present in areas where
epidemiologic studies have evaluated
associations with mortality or
morbidity, and what such
concentrations may indicate regarding
the adequacy of the primary PM2.5
standards. The use of information from
epidemiologic studies to inform
conclusions on the primary PM2.5
standards is complicated by the fact that
such studies evaluate associations
between distributions of ambient PM2.5
and health outcomes, and do not
identify the specific exposures that can
lead to the reported effects. Rather,
health effects can occur over the entire
distribution of ambient PM2.5
concentrations evaluated, and
epidemiologic studies conducted to date
do not identify a population-level
threshold below which it can be
concluded with confidence that PM2.5associated health effects do not occur.
Therefore, the 2022 PA evaluates the
PM2.5 air quality distributions over
which epidemiologic studies support
health effect associations (U.S. EPA,
2022b, section 3.3.3.2). In the absence of
discernible thresholds, the 2022 PA
considers the study-reported ambient
PM2.5 concentrations reflecting
estimated exposure with a focus around
the middle portion of the PM2.5 air
quality distribution, where the bulk of
the observed data reside and which
provides the strongest support for
reported health effect associations. The
section below, as well as in more detail
in section II.B.3.b.i of the proposal (88
FR 5594, January 27, 2023), describes
the consideration of the key
epidemiologic studies and observations
from these studies, as evaluated in the
2022 PA (U.S. EPA, 2022b, section
3.3.3.2).
As an initial matter, in considering
the PM2.5 air quality distributions
associated with mortality or morbidity
in the key epidemiologic studies, the
2022 PA recognizes that in previous
reviews, the decision framework used to
judge adequacy of the existing PM2.5
standards, and what levels of any
potential alternative standards should
be considered, placed significant weight
on epidemiologic studies that assessed
associations between PM2.5 exposure
and health outcomes that were most
strongly supported by the body of
scientific evidence. In doing so, the
decision framework recognized that
while there is no specific point in the
air quality distribution of any
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
epidemiologic study that represents a
‘‘bright line’’ at and above which effects
have been observed and below which
effects have not been observed, there is
significantly greater confidence in the
magnitude and significance of observed
associations for the part of the air
quality distribution corresponding to
where the bulk of the health events in
each study have been observed,
generally at or around the mean
concentration. This is the case both for
studies of daily PM2.5 exposures and for
studies of annual average PM2.5
exposures (U.S. EPA, 2022b, section
3.3.3.2.1).
As discussed further in the 2022 PA,
studies of daily PM2.5 exposures
examine associations between day-today variation in PM2.5 concentrations
and health outcomes, often over several
years (U.S. EPA, 2022b, section
3.3.3.2.1). While there can be
considerable variability in daily
exposures over a multi-year study
period, most of the estimated exposures
reflect days with ambient PM2.5
concentrations around the middle of the
air quality distributions examined (i.e.,
‘‘typical’’ days rather than days with
extremely high or extremely low
concentrations). Similarly, for studies of
annual PM2.5 exposures, most of the
health events occur at estimated
exposures that reflect annual average
PM2.5 concentrations around the middle
of the air quality distributions
examined. In both cases, epidemiologic
studies provide the strongest support for
reported health effect associations for
this middle portion of the PM2.5 air
quality distribution, which corresponds
to the bulk of the underlying data, rather
than the extreme upper or lower ends of
the distribution. Consistent with this, as
noted in the 2022 PA (U.S. EPA, 2022b,
section 3.3.1.1), several epidemiologic
studies report that associations persist
in analyses that exclude the upper
portions of the distributions of
estimated PM2.5 exposures, indicating
that ‘‘peak’’ PM2.5 exposures are not
disproportionately responsible for
reported health effect associations.
Thus, in considering PM2.5 air quality
data from epidemiologic studies,
consistent with approaches in the 2012
and 2020 reviews (78 FR 3161, January
15, 2013; U.S. EPA, 2011, sections 2.1.3
and 2.3.4.1; 85 FR 82716–82717,
December 18, 2020; U.S. EPA, 2020b,
sections 3.1.2 and 3.2.3), the 2022 PA
evaluates study-reported means (or
medians) of daily and annual average
PM2.5 concentrations as indicators for
the middle portions of the air quality
distributions, over which studies
generally provide strong support for
reported associations and for which
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
confidence in the magnitude and
significance of associations observed in
the epidemiologic studies is greatest (78
FR 3101, January 15, 2013). In addition
to the overall study means, the 2022 PA
also focuses on concentrations
somewhat below the means (e.g., 25th
and 10th percentiles), when such
information is available from the
epidemiologic studies, which again is
consistent with approaches used in
previous reviews. In so doing, the 2022
PA notes, as in previous reviews, that a
relatively small portion of the health
events are observed in the lower part of
the air quality distribution and
confidence in the magnitude and
significance of the associations begins to
decrease in the lower part of the air
quality distribution. Furthermore,
consistent with past reviews, there is no
single percentile value within a given
air quality distribution that is most
appropriate or ‘‘correct’’ to use to
characterize where our confidence in
associations becomes appreciably lower.
However, and as detailed further in the
2022 PA, the range from the 25th to 10th
percentiles is a reasonable range to
consider as a region where there is
appreciably less confidence in the
associations observed in epidemiologic
studies compared to the means (U.S.
EPA, 2022b, p. 3–69).66
In evaluating the overall studyreported means, and concentrations
somewhat below the means from
epidemiologic studies, the 2022 PA
focuses on the form, averaging time and
level of the current primary annual
PM2.5 standard. Consistent with the
approaches used in the 2012 and 2020
reviews (78 FR 3161–3162, January 15,
2013; 85 FR 82716–82717, December 18,
2020), the annual standard has been
utilized as the primary means of
providing public health protection
against the bulk of the distribution of
short- and long-term PM2.5 exposures.
Thus, the evaluation of the studyreported mean concentrations from key
epidemiologic studies lends itself best
to evaluating the adequacy of the annual
PM2.5 standard (rather than the 24-hour
standard with its 98th percentile form).
This is true for the study-reported
means from both long-term and shortterm exposure epidemiologic studies,
recognizing that the overall mean PM2.5
concentrations reported in studies of
short-term (24-hour) exposures reflect
66 As detailed in the 2011 PA, we note the
interrelatedness of the distributional statistics and
a range of one standard deviation around the mean
which represents approximately 68% of normally
distributed data, and in that one standard deviation
below the mean falls between the 25th and 10th
percentiles (U.S. EPA, 2011, p. 2–71; U.S. EPA,
2005, p. 5–22).
PO 00000
Frm 00039
Fmt 4701
Sfmt 4700
16239
averages across the study population
and over the years of the study. Thus,
mean concentrations from short-term
exposure studies reflect long-term
averages of 24-hour PM2.5 exposure
estimates. In this manner, the
examination of study-reported means in
key epidemiologic studies in the 2022
PA aims to evaluate the protection
provided by the annual PM2.5 standard
against the exposures where confidence
is greatest for associations with
mortality and morbidity. In addition,
the protection provided by the annual
standard is evaluated in conjunction
with that provided by the 24-hour
standard, with its 98th percentile form,
which aims to provide supplemental
protection against the short-term
exposures to peak PM2.5 concentrations
that can occur in areas with strong
contributions from local or seasonal
sources, even when overall ambient
mean PM2.5 concentrations in an area
remain relatively low.
In focusing on the annual standard,
and in evaluating the range of studyreported exposure concentrations for
which the strongest support for adverse
health effects exists, the 2022 PA
examines exposure concentrations in
key epidemiologic studies to determine
whether the current primary annual
PM2.5 standard provides adequate
protection against these exposure
concentrations. This means, as in past
reviews, application of a decision
framework based on assessing means
reported in key epidemiologic studies
must also consider how the study means
were computed and how these values
compare to the annual standard metric
(including the level, averaging time and
form) and the use of the monitor with
the highest PM2.5 design value in an area
for compliance. In the 2012 review, it
was recognized that the key
epidemiologic studies computed the
study mean using an average across
monitor-based PM2.5 concentrations. As
such, the Agency noted that this
decision framework applied an
approach of using maximum monitor
concentrations to determine compliance
with the standard, while selecting the
standard level based on consideration of
composite monitor concentrations.
Further, the Agency included analyses
(Hassett-Sipple et al., 2010; Frank, 2012)
that examined the differences in these
two metrics (i.e., maximum monitor
concentrations and composite monitor
concentrations) across the U.S. and in
areas included in the key epidemiologic
studies and found that the maximum
design value in an area was generally
higher than the monitor average across
that area, with the difference varying
E:\FR\FM\06MRR3.SGM
06MRR3
16240
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
based on location and concentration.
This information was taken into account
in the Administrator’s final decision in
selecting a level for the primary annual
PM2.5 standard the 2012 review and
discussed more specifically in her
considerations on adequate margin of
safety.
Consistent with the approach taken in
2012, in assessing how the overall mean
(or median) PM2.5 concentrations
reported in key epidemiologic studies
can inform conclusions on the primary
annual PM2.5 standard, the 2022 PA
notes that the relationship between
mean PM2.5 concentrations and the area
design value continues to be an
important consideration in evaluating
the adequacy of the current or potential
alternative annual PM2.5 standard levels
in this reconsideration. In a given area,
the area design value is based on the
monitor in an area with the highest
PM2.5 concentrations and is used to
determine compliance with the
standard. The highest PM2.5
concentrations spatially distributed in
the area would generally occur at or
near the area design value monitor and
the distribution of PM2.5 concentrations
would generally be lower in other
locations and at monitors in that area.
As such, when an area is meeting a
specific annual standard level, the
annual average exposures in that area
are expected to be at concentrations
lower than that level and the average of
the annual average exposures across that
area are expected (i.e., a metric similar
to the study-reported mean values) to be
lower than that level.67
Another important consideration is
that there are a substantial number of
different types of epidemiologic studies
available since the 2012 review,
included in both the 2019 ISA and the
ISA Supplement, that make
understanding the relationship between
the mean PM2.5 concentrations and the
area design value even more important
(U.S. EPA, 2019a; U.S. EPA, 2022a).
While the key epidemiologic studies in
the 2012 review were all monitor-based
studies, the newer studies include
hybrid modeling approaches, which
have emerged in the epidemiologic
literature as an alternative to approaches
that only use ground-based monitors to
estimate exposure. As assessed in the
2019 ISA and ISA Supplement, a
67 In setting a standard level that would require
the design value monitor to meet a level equal to
the study-reported mean PM2.5 concentrations
would generally result in lower concentrations of
PM2.5 across the entire area, such that even those
people living near an area design value monitor
(where PM concentrations are generally highest)
will be exposed to PM2.5 concentrations below the
air quality conditions reported in the epidemiologic
studies.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
substantial number of epidemiologic
studies used hybrid model-based
methods in evaluating associations
between PM2.5 exposure and health
effects (U.S. EPA, 2019a; U.S. EPA,
2022a). Hybrid model-based studies
employ various fusion techniques that
combine ground-based monitored data
with air quality modeled estimates and/
or information from satellites to
estimate PM2.5 exposures.68
Additionally, hybrid modeling
approaches tend to broaden the areas
captured in the exposure assessment,
and in so doing, tend to report lower
mean PM2.5 concentrations than
monitor-based approaches because they
include more suburban and rural areas
where concentrations are lower. While
these studies provide a broader
estimation of PM2.5 exposures compared
to monitor-based studies (i.e., PM2.5
concentrations are estimated in areas
without monitors), the hybrid modeling
approaches result in study-reported
means that are more difficult to relate to
the annual standard metric and to the
use of maximum monitor design values
to assess compliance. In addition, and to
further complicate the comparison,
when looking across these studies,
variations exist in how exposure is
estimated between such studies, which
in turn affects how the study means are
calculated. Two important variations
across studies include: (1) Variability in
spatial scale used (i.e., averages
computed across the nation (or large
portions of the country) versus a focus
on only CBSAs) and (2) variability in
exposure assignment methods (i.e.,
averaging across all grid cells [nonpopulation weighting], averaging across
a scaled-up area like a ZIP code [aspects
of population weighting applied], and/
or applying population weighting). To
elaborate further on the variability in
exposure assignment methods, studies
that use hybrid modeling approaches
can estimate PM2.5 concentrations at
different spatial resolutions, including
at 1 km x 1 km grid cells, at 12 km x
12 km grid cells, or at the census tract
level. Mean reported PM2.5
concentrations can then be estimated
either by averaging up to a larger spatial
resolution that corresponds to the
spatial resolution for which health data
exists (e.g., ZIP code level) and therefore
apply aspects of population weighting.
These values are then averaged across
all study locations at the larger spatial
resolution (e.g., averaged across all ZIP
codes in the study) over the study
period, resulting in the study-reported
68 More detailed information about hybrid model
methods and performance is described in section
2.3.3.2 of the 2022 PA (U.S. EPA, 2022b).
PO 00000
Frm 00040
Fmt 4701
Sfmt 4700
mean 24-hour average or average annual
PM2.5 concentration. Other studies that
use hybrid modeling methods to
estimate PM2.5 concentrations may use
each grid cell to calculate the studyreported mean 24-hour average or
average annual PM2.5 concentration. As
such, these types of studies do not apply
population weighting in their mean
concentrations. In studies that use each
grid cell to report a mean PM2.5
concentration and do not apply aspects
of population weighting, the study mean
may not reflect the exposure
concentrations used in the
epidemiologic study to assess the
reported association. The impact of the
differences in methods is an important
consideration when comparing mean
concentrations across studies (U.S. EPA,
2022b, section 3.3.3.2.1). Thus, the 2022
PA also considers the methods used to
estimate PM2.5 concentrations, which
vary from traditional methods using
monitoring data from ground-based
monitors 69 to those using more complex
hybrid modeling approaches and how
these methods calculate the studyreported mean PM2.5 concentration.70
Given the emergence of the hybrid
model-based epidemiologic studies
since the 2012 review, the 2022 PA
explores the relationship between the
approaches used in these studies to
estimate PM2.5 concentrations and the
impact that the different methods have
on the study-reported mean PM2.5
concentrations. The 2022 PA further
seeks to understand how the approaches
and resulting mean concentrations
compare across studies, as well as what
the resulting mean values represent
relative to the annual standard. In so
doing, the 2022 PA presents analyses
that compare the area annual design
values, composite monitor PM2.5
concentrations, and mean
concentrations from two hybrid
modeling approaches, including
evaluation of the means when
population weighting is applied and
when population weighting is not
69 In those studies that use ground-based monitors
alone to estimate long- or short-term PM2.5
concentrations, approaches include: (1) PM2.5
concentrations from a single monitor within a city/
county; (2) average of PM2.5 concentrations across
all monitors within a city/county or other defined
study area (e.g., CBSA); or (3) population-weighted
averages of exposures. Once the study location
average PM2.5 concentration is calculated, the
study-reported long-term average is derived by
averaging daily/annual PM2.5 concentrations across
all study locations over the entire study period.
70 Detailed information on the methods by which
mean PM2.5 concentrations are calculated in key
monitor- and hybrid model-based U.S. and
Canadian epidemiologic studies are presented in
Tables 3–6 through 3–9 in the 2022 PA (U.S. EPA,
2022b).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
applied (U.S. EPA, 2022b, section
2.3.3.1).
In the air quality analyses comparing
composite monitored PM2.5
concentrations with annual PM2.5 design
values in U.S. CBSAs, maximum annual
PM2.5 design values were approximately
10% to 20% higher than annual average
composite monitor concentrations (i.e.,
averaged across multiple monitors in
the same CBSA) (sections I.D.5.a above
and U.S. EPA, 2022b, section 2.3.3.1,
Figure 2–28 and Table 2–3). The
difference between the maximum
annual design value and average
concentration in an area can be smaller
or larger than this range (10–20%),
depending on a variety of factors such
as the number of monitors, monitor
siting characteristics, the distribution of
ambient PM2.5 concentrations, and how
the average concentrations are
calculated (i.e., averaged across
monitors versus across modeled grid
cells). Results of this analysis suggest
that there will be a distribution of
concentrations across an area and the
maximum annual average monitored
concentration in an area (at the design
value monitor, used for compliance
with the standard), will generally be 10–
20% higher than the average PM2.5
concentration across the other monitors
in the area. Thus, in considering how
the annual standard levels would relate
to the study-reported means from key
monitor-based epidemiologic studies,
the 2022 PA generally concludes that an
annual standard level that is no more
than 10–20% higher than monitor-based
study-reported mean PM2.5
concentrations would generally
maintain air quality exposures to be
below those associated with the studyreported mean PM2.5 concentrations,
exposures for which the strongest
support for adverse health effects
occurring is available.
The 2022 PA also evaluates data from
two hybrid modeling approaches
(DI2019 and HA2020) that have been
used in several recent epidemiologic
studies (U.S. EPA, 2022b, section
2.3.3.2.4).71 The analysis shows that the
means differ when PM2.5 concentrations
are estimated in urban areas only
(CBSAs) versus when the averages were
calculated with all or most grid cells
nationwide, likely because areas
included outside of CBSAs tend to be
more rural and have lower estimated
PM2.5 concentrations. The 2022 PA
recognizes the importance of this
variability in the means since the study
areas included in the calculation of the
71 More details on the evaluation of the two
hybrid modeling approaches is provided in section
2.3.3.2.4 of the 2022 PA (U.S. EPA, 2022b).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
mean, and more specifically whether a
study is focused on nationwide,
regional, or urban areas, will affect the
calculation of the study mean based on
how many rural areas, with lower
estimated PM2.5 concentrations, are
included in the study area. While the
determination of what spatial scale to
use to estimate PM2.5 concentrations
does not inherently affect the quality of
the epidemiologic study, the spatial
scale can influence the calculated
reported long-term mean concentration
across the study area and period. The
results of the analysis show that,
regardless of the hybrid modeling
approach assessed, the annual average
PM2.5 concentrations in CBSA-only
analyses are 4–8% higher than for
nationwide analyses, likely as a result of
higher PM2.5 concentrations in more
densely populated areas, and exclusion
of more rural areas (U.S. EPA, 2022b,
Table 2–4). When evaluating
comparisons between surfaces that
estimate exposure using aspects of
population weighting versus surfaces
that do not calculate means using
population weighting, surfaces that
calculate long-term mean PM2.5
concentrations with populationweighted averages have higher average
annual PM2.5 concentrations, compared
to annual PM2.5 concentrations in
analyses that do not apply population
weighting.72 Analyses show that average
maximum annual design values are 40
to 50% higher when compared to
annual average PM2.5 concentrations
estimated without population weighting
versus 15% to 18% higher when
compared to average annual PM2.5
concentrations estimated with
population weighting applied (similar to
the differences observed for the
composite monitor comparison values
for the monitor-based epidemiologic
studies) (U.S. EPA, 2022b, section
2.3.3.2.4). Given these results, it is
worth noting that for the studies using
the hybrid modeling approaches, the
choice of methodology employed in
calculating the study-reported means
(i.e., using population weighting or not),
and not a difference in estimates of
exposure in the study itself, can
produce substantially different studyreported mean values, where
approaches that do not apply
population weighting leading to much
lower estimated mean PM2.5
concentrations.
72 The annual PM
2.5 concentrations for the
population-weighted averages ranged from 8.2–10.2
mg/m3, while those that do not apply population
weighting ranged from 7.0–8.6 mg/m3. Average
maximum annual design values ranged from 9.5 to
11.7 mg/m3.
PO 00000
Frm 00041
Fmt 4701
Sfmt 4700
16241
Based on these results, and similar to
conclusions for the monitor-based
studies, the 2022 PA generally
concludes that study-reported mean
concentrations in the studies that
employ hybrid modeling approaches
and calculate a population-weighted
mean are associated with air quality
conditions that would be achieved by
meeting annual standard levels that are
15–18% higher than study-reported
means. Therefore, an annual standard
level that is no more than 15–18%
higher than the study-reported means
would generally maintain air quality
exposures to be below those associated
with the study-reported mean PM2.5
concentrations, exposures for which we
have the strongest support for adverse
health effects occurring. For the studies
that utilize hybrid modeling approaches
but do not incorporate population
weighting in calculating the mean, the
annual design values associated with
these air quality conditions are expected
to be much higher (i.e., 40–50% higher)
and this larger difference makes it more
difficult to consider how these studies
can be used to determine the adequacy
of the protection afforded by the current
or potential alternative annual
standards. Additionally, as noted above
in studies that utilize hybrid modeling
approaches and that do not incorporate
population weighting in calculating the
mean (e.g., use each grid cell to
calculate a mean PM2.5 concentration),
the study mean does not reflect the
exposure concentrations used in the
epidemiologic study to assess the
reported association.
The 2022 PA notes that while these
analyses can be useful to informing the
understanding of the relationship
between study-reported mean
concentrations and the level of the
annual standard, some limitations of
this analysis must be recognized (U.S.
EPA, 2022a, section 3.3.3.2.1). First, the
comparisons used only two hybrid
modeling approaches. Although these
two hybrid modeling surfaces have been
used in a number of recent
epidemiologic studies, they represent
just two of the many hybrid modeling
approaches that have been used in
epidemiologic studies to estimate PM2.5
concentrations. These methods continue
to evolve, with further development and
improvement to prediction models that
estimate PM2.5 concentrations in
epidemiologic studies. In addition to
differences in hybrid modeling
approaches, epidemiologic studies also
use different methods to assign a
population weighted average PM2.5
concentration to their study population,
and the assessment presented in the
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16242
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
2022 PA does not evaluate all of the
potential methods that could be used.
Additionally, while some of these
epidemiologic studies also provide
information on the broader distributions
of exposure estimates and/or health
events and the PM2.5 concentrations
corresponding to the lower percentiles
of those data (e.g., 25th and/or 10th), the
air quality analysis in the 2022 PA
focuses on mean PM2.5 concentrations
and a similar comparison for lower
percentiles of data was not assessed.
Therefore, any direct comparison of
study-reported PM2.5 concentrations
corresponding to lower percentiles and
annual design values is more uncertain
than such comparisons with the mean.
Finally, air quality analysis presented in
the 2022 PA and detailed above in
section I.D.5 included two hybrid
modeling-based approaches that used
U.S.-based air quality information for
estimating PM2.5 concentrations. As
such, the analyses are most relevant to
interpreting the study-reported mean
concentrations from U.S. epidemiologic
studies and do not provide additional
information about how the mean
exposures concentrations reported in
epidemiologic studies in other countries
would compare to annual design values
observed in the U.S. In addition, while
information from Canadian studies can
be useful in assessing the adequacy of
the annual standard, differences in the
exposure environments and population
characteristics between the U.S. and
other countries can affect the studyreported mean value and its relationship
with the annual standard level. Sources
and pollutant mixtures, as well as PM2.5
concentration gradients, may be
different between countries, and the
exposure environments in other
countries may differ from those
observed in the U.S. Furthermore,
differences in population characteristics
and population densities can also make
it challenging to directly compare
studies from countries outside of the
U.S. to a design value in the U.S.
As with the experimental studies
discussed above, the 2022 PA focuses
on epidemiologic studies assessed in the
2019 ISA and ISA Supplement that have
the potential to be most informative in
reaching decisions on the adequacy of
the primary PM2.5 standards. The 2022
PA focuses on epidemiologic studies
that provide strong support for ‘‘causal’’
or ‘‘likely to be causal’’ relationships
with PM2.5 exposures in the 2019 ISA.
Further, the 2022 PA also focuses on the
health effect associations that are
determined in the 2019 ISA and ISA
Supplement to be consistent across
studies, coherent with the broader body
of evidence (e.g., including animal and
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
controlled human exposure studies),
and robust to potential confounding by
co-occurring pollutants and other
factors.73 In particular the 2022 PA
considers the U.S. and Canadian
epidemiologic studies to be more useful
for reaching conclusions on the current
standards than studies conducted in
other countries, given that the results of
the U.S. and Canadian studies are more
directly applicable for quantitative
considerations, whereas studies
conducted in other countries reflect
different populations, exposure
characteristics, and air pollution
mixtures. Additionally, epidemiologic
studies outside of the U.S. and Canada
generally reflect higher PM2.5
concentrations in ambient air than are
currently found in the U.S., and are less
relevant to informing questions about
adequacy of the current standards.74
However, and as noted above, the 2022
PA also recognizes that while
information from Canadian studies can
be useful in assessing the adequacy of
the annual standard, there are still
important differences between the
exposure environments in the U.S. and
Canada and interpreting the data (e.g.,
mean concentrations) from the Canadian
studies in the context of a U.S.-based
standard may present challenges in
directly and quantitatively informing
questions regarding the adequacy of the
73 As described in the Preamble to the ISAs (U.S.
EPA, 2015), ‘‘the U.S. EPA emphasizes the
importance of examining the pattern of results
across various studies and does not focus solely on
statistical significance or the magnitude of the
direction of the association as criteria of study
reliability. Statistical significance is influenced by
a variety of factors including, but not limited to, the
size of the study, exposure and outcome
measurement error, and statistical model
specifications. Statistical significance may be
informative; however, it is just one of the means of
evaluating confidence in the observed relationship
and assessing the probability of chance as an
explanation. Other indicators of reliability such as
the consistency and coherence of a body of studies
as well as other confirming data may be used to
justify reliance on the results of a body of
epidemiologic studies, even if results in individual
studies lack statistical significance. Traditionally,
statistical significance is used to a larger extent to
evaluate the findings of controlled human exposure
and animal toxicological studies. Understanding
that statistical inferences may result in both false
positives and false negatives, consideration is given
to both trends in data and reproducibility of results.
Thus, in drawing judgments regarding causality, the
U.S. EPA emphasizes statistically significant
findings from experimental studies, but does not
limit its focus or consideration to statistically
significant results in epidemiologic studies.’’
74 This emphasis on studies conducted in the U.S.
or Canada is consistent with the approach in the
2012 and 2020 reviews of the PM NAAQS (U.S.
EPA, 2011, section 2.1.3; U.S. EPA, 2020b, section
3.2.3.2.1) and with approaches taken in other
NAAQS reviews. However, the importance of
studies in the U.S., Canada, and other countries in
informing an ISA’s considerations of the weight of
the evidence that informs causality determinations
is recognized.
PO 00000
Frm 00042
Fmt 4701
Sfmt 4700
current or potential alternative the
levels of the annual standard. Lastly, the
2022 PA emphasizes multicity/
multistate studies that examine health
effect associations, as such studies are
more encompassing of the diverse
atmospheric conditions and population
demographics in the U.S. than studies
that focus on a single city or State.
Figures 3–4 through 3–7 in the 2022 PA
summarize the study details for the key
U.S. and Canadian epidemiologic
studies (U.S. EPA, 2022b, section
3.3.3.2.1).75
The key epidemiologic studies
identified in the 2022 PA indicate
generally positive and statistically
significant associations between
estimated PM2.5 exposures (short- or
long-term) and mortality or morbidity
across a range of ambient PM2.5
concentrations (U.S. EPA, 2022b,
section 3.3.3.2.1), report overall mean
(or median) PM2.5 concentrations, and
include those for which the years of
PM2.5 air quality data used to estimate
exposures overlap entirely with the
years during which health events are
reported.76 Additionally, for studies that
estimate PM2.5 exposure using hybrid
modeling approaches, the 2022 PA also
considers the approach used to estimate
PM2.5 concentrations and the approach
used to validate hybrid model
predictions when evaluating those
studies as key epidemiologic studies 77
and focuses on those studies that use
recent methods based on surfaces that
are with fused with monitored PM2.5
75 The cohorts examined in the studies included
in Figure 3–4 to Figure 3–7 of the 2022 PA include
large numbers of individuals in the general
population, and often also include those
populations identified as at-risk (i.e., children,
older adults, minority populations, and individuals
with pre-existing cardiovascular and respiratory
disease).
76 For some studies of long-term PM
2.5 exposures,
exposure is estimated from air quality data
corresponding to only part of the study period,
often including only the later years of the health
data, and are not likely to reflect the full ranges of
ambient PM2.5 concentrations that contributed to
reported associations. While this approach can be
reasonable in the context of an epidemiologic study
that is evaluating health effect associations with
long-term PM2.5 exposures, under the assumption
that spatial patterns in PM2.5 concentrations are not
appreciably different during time periods for which
air quality information is not available (e.g., Chen
et al., 2016), the 2022 PA focuses on the
distribution of ambient PM2.5 concentrations that
could have contributed to reported health
outcomes. Therefore, the 2022 PA identifies studies
as key epidemiologic studies when the years of air
quality data and health data overlap in their
entirety.
77 Such studies are identified as those that use
hybrid modeling approaches for which recent
methods and models were used (e.g., recent
versions and configurations of the air quality
models); studies that are fused with PM2.5 data from
national monitoring networks (i.e., FRM/FEM data);
and studies that reported a thorough model
performance evaluation for core years of the study.
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
concentration data (U.S. EPA, 2022b,
section 3.3.3.2.1).
Figure 1 below (U.S. EPA, 2022b,
Figure 3–8) highlights the overall mean
(or median) PM2.5 concentrations
reported in key U.S. studies that use
ground-based monitors alone to estimate
long- or short-term PM2.5 exposure.78
For the small subset of studies with
available information on the broader
distributions of underlying data, Figure
1 below also identifies the study-period
PM2.5 concentrations corresponding to
ddrumheller on DSK120RN23PROD with RULES3
78 Canadian studies that use ground-based
monitors estimate long- or short-term PM2.5
exposures are found in Figure 3–9 of the 2022 PA,
including concentrations corresponding to the 25th
and 10th percentiles of estimated exposures or
health events, when available (U.S. EPA, 2022b).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
the 25th and 10th percentiles of health
events 79 (see Appendix B, Section B.2
of the 2022 PA for more information).
Figure 2 (U.S. EPA, 2022a, Figure 3–14)
presents overall means of predicted
PM2.5 concentrations for key U.S.
model-based epidemiologic studies that
apply aspects of population-weighting,
and the concentrations corresponding to
the 25th and 10th percentiles of
estimated exposures or health events 80
79 That is, 25% of the total health events occurred
in study locations with mean PM2.5 concentrations
(i.e., averaged over the study period) below the 25th
percentiles identified in Figure 3–8 of the 2022 PA
and 10% of the total health events occurred in
study locations with mean PM2.5 concentrations
below the 10th percentiles identified.
80 For most studies in Figure 2 below (Figure 3–
14 in the 2022 PA), 25th percentiles of exposure
PO 00000
Frm 00043
Fmt 4701
Sfmt 4700
16243
when available (see Appendix B, section
B.3 for additional information).81
estimates are presented. The exception is Di et al.
(2017b), for which Figure 2 (U.S. EPA, 2022b,
Figure 3–14) presents the short-term PM2.5 exposure
estimates corresponding to the 25th and 10th
percentiles of deaths in the study population (i.e.,
25% and 10% of deaths occurred at concentrations
below these concentrations). In addition, the
authors of Di et al. (2017b) provided populationweighted exposure values. The 10th and 25th
percentiles of these population-weighted exposure
estimates are 7.9 and 9.5 mg/m3, respectively.
81 Overall mean (or median) PM
2.5 concentrations
reported in key Canadian studies that use modelbased approaches to estimate long- or short-term
PM2.5 concentrations and the concentrations
corresponding to the 25th and 10th percentiles of
estimated exposures or health events, when
available are found in Figure 3–9 of the 2022 PA
(U.S. EPA, 2022b).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
& Mortality
Franklin 2008 (US: 25 Cities)
■
Klemm 2003 (US: Harvard 6 City)
PO 00000
summary Statistics
0 10th percentile
• 25th percentile
■ Mean
■
.■
Krall 2013 (US: 72 Cities)
■
Dai 2014 (US: 75 Cities)
Frm 00044
•
0
zanobettl and Schwartz 2009 (us: 112 Cltles)
■
■
Liu 2019 (US: 107 cities; ST Exposure)•
■
ST exposure ostro 2016 (us: 8 California counties)
& Morbidity Zanobetti 2009 {US: 26cities)
Fmt 4701
Bell 2014 (US: 4 Counties in MA & CT)
■
Dominici 2006 (US: 204 Urban Counties)
Bell 2008 (US: 202 Counties)
•
0'
Sfmt 4725
Bravo 2017 (US: 418 Counties)
■
■
■
Peng 2009 (US: 119 Urban Counties)
E:\FR\FM\06MRR3.SGM
■
LT exposure Zeger 2008 (us: 421 Eastern Region counties)
■
Zeger 2008 {US: 62 Western Region Counties)
■
Hart 2015 (US: Nationwide)
Eum 2018 (us: Eastern Geographic Region; LT Exposure)•
Kioumourtzogtou 2016 (us: 207 Cities)
■
Eum 2018 {US: 3 Geographic Regions overall; LT Exposure)•
06MRR3
i
Eum 2018 (US: Central Geographic Region; LT Exposure)•
I
■
■
- - - - - - - - - ··---·----------·-·····--· ..--.-·---~-~------·-·-f-~~-- '-·--·-
& Morbidity Gharibvand 2016 (US: Nationwide)
•
■
■
Eum 2018 {US: Western Geographic Region; LT Exposure)•
Zeger 2008 (US: 185 Central Region counties)
LT Exposure Mcconnell 2010 {us: 13 California communities)
•
■
Bell 2015 (us: 70 urban counties)
& Mortality
■
••
•
~
-~
'
. ",-
'
■
•
■
10.0 10.5 11.0 11.5 12.0 12.5 13.0 13.5 14.0 14.5 15.0 15.5 16.0 16.5
Overall PM 2 . 5 Concentration for the Study Period (µg/m3}
ER06MR24.029
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
Jkt 262001
•
ST Exposure Franklin 2007 (US: 27 Cities)
16244
19:03 Mar 05, 2024
BILLING CODE 6560–50–P
VerDate Sep<11>2014
Figure 1. Monitor-based PM2.s Concentrations in Key U.S. Epidemiologic Studies. (Asterisks denote studies included in the ISA
Supplement)
ddrumheller on DSK120RN23PROD with RULES3
19:03 Mar 05, 2024
■ Mean
Jkt 262001
■
•
ST expc:>sure & Morbidity deSouza2921 (us: Nationwide; sr ExpoS1Jre)"
•
LT eJstates and 2 MSAs; LT Elq)osure)
II
Hart_ 2015. {US: Nationwide; LT Exposure)
•
Dominici 2019 (us: Nationwide; LT Exposure}'"
PO 00000
Frm 00045
Fmt 4701
Sfmt 4700
•
■
Wang 2020 (us: Nationwide; LT Exposure)*
Overal I PM2 .,; Concentration for the Study Period (tt9/rrr)
12
11
10
9
06MRR3
16245
Continued
82 This is generally consistent with, but slightly
below, the lowest study-reported mean PM2.5
concentration from monitor-based studies available
epidemiologic studies. For studies that
use monitors to estimate PM2.5
exposures, overall mean PM2.5
concentrations range between 9.9 mg/
m3 82 to 16.5 mg/m3 (Figure 1 above and
wu 2020a (us: Nationwide; LT Bq>osure)*
•
•
Wang 2017 (US: 7 SE States; LT Exposure)
E:\FR\FM\06MRR3.SGM
across a wide range of ambient PM2.5
concentrations (U.S. EPA, 2022b,
section 3.3.3.2.1). The 2022 PA makes a
number of observations with regard to
the study-reported PM2.5 concentrations
in the key U.S. and Canadian
epidemiologic studies.
The 2022 PA first considers the PM2.5
concentrations from the key U.S.
wvatt 202oc (us:sao us counties; ST ExP9Sum)*
•
•
0
DI 2017b (US: Nationwide; LT Exposure)
8
7
6
5
■
•
0
ST Exposure & Mortality Di 2011a (US: Natioowide; STExposure)
■
Lee 2015b (US: 3 SE Statl!S; ST Bq>osute}
0 10th percentile
• 25th percentile
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
BILLING CODE 6560–50–C
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ~ - - - - - - - - - - - - - - - - - - - - - - - - - - ~ summarystatistics
Based on its evaluation of studyreported mean concentrations, the 2022
PA notes that key epidemiologic studies
conducted in the U.S. or Canada report
generally positive and statistically
significant associations between
estimated PM2.5 exposures (short- or
long-term) and mortality or morbidity
VerDate Sep<11>2014
ER06MR24.030
Figure 2. Hybrid Model-Predicted PM2.s Concentrations in Key U.S. Epidemiologic Studies that Apply Aspects of PopulationWeighting. (Asterisks denote studies included in the ISA Supplement)
ddrumheller on DSK120RN23PROD with RULES3
16246
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
U.S. EPA, 2022b, Figure 3–8). For key
U.S. epidemiologic studies that use
hybrid model-predicted exposures and
apply aspects of population-weighting,
mean PM2.5 concentrations range from
9.3 mg/m3 to just above 12.2 mg/m3
(Figure 2 above and U.S. EPA, 2022b,
Figure 3–14). In studies that average up
from the grid cell level to the ZIP code,
postal code, or census tract level, mean
PM2.5 concentrations range from 9.8 mg/
m3 to 12.2 mg/m3. The one study that
population-weighted the grid cell prior
to averaging up to the ZIP code or
census tract level reported mean PM2.5
concentrations of 9.3 mg/m3. Based on
air quality analyses noted above, these
hybrid modelled epidemiologic studies
are expected to report means similar to
those from monitor-based studies.
Other key U.S. epidemiologic studies
that use hybrid modeling approaches
estimate mean PM2.5 exposure by
averaging each grid cell across the entire
study area, whether that be the nation
or a region of the country. These studies
do not weight the estimated exposure
concentrations based on population
density or location of health events. As
such, the study mean reported in these
studies may not reflect the exposure
concentrations used in the
epidemiologic study to assess the
reported association. As a result, these
reported mean concentrations are the
most different (and much lower) than
the means reported in monitor-based
studies. Due to the methodology
employed in calculating the studyreported means and not necessarily a
difference in estimates of exposure,
these epidemiologic studies are
expected to report some of the lowest
mean values. For these studies, the
reported mean PM2.5 concentrations
range from 8.1 mg/m3 to 11.9 mg/m3 (U.S.
EPA, 2022b, Figure 3–14). As noted
above, for studies that utilize hybrid
modeling approaches but do not
incorporate population weighting into
the reported mean calculation, the
associated annual design values would
be expected to be much higher (i.e., 40–
50% higher) than the study-reported
means. This larger difference between
design values and study-reported mean
concentrations makes it more difficult to
consider how these studies can be used
to determine the adequacy of the
protection afforded by the current or
potential alternative annual standards
(U.S. EPA, 2022b, section 3.3.3.2.1).
In addition to the mean PM2.5
concentrations, a subset of the key U.S.
epidemiologic studies report PM2.5
concentrations corresponding to the
in the 2020 PA, which was 10.7 mg/m3 (U.S. EPA,
2020a, Figure 3–7).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
25th and 10th percentiles of health data
or exposure estimates to provide insight
into the concentrations that comprise
the lower quartile of the air quality
distributions. In studies that use
monitors to estimate PM2.5 exposures,
25th percentiles of health events
correspond to PM2.5 concentrations (i.e.,
averaged over the study period for each
study city) at or above 11.5 mg/m3 and
10th percentiles of health events
correspond to PM2.5 concentrations at or
above 9.8 mg/m3 (i.e., 25% and 10% of
health events, respectively, occur in
study locations with PM2.5
concentrations below these values)
(Figure 1 above and U.S. EPA, 2022b,
Figure 3–8). Of the key U.S.
epidemiologic studies that use hybrid
modeling approaches and apply
population-weighting to estimate longterm PM2.5 exposures, the ambient PM2.5
concentrations corresponding to 25th
percentiles of estimated exposures are
9.1 mg/m3 (Figure 2 and U.S. EPA,
2022b, Figure 3–14). In key U.S.
epidemiologic studies that use hybrid
modeling approaches and apply
population-weighting to estimate shortterm PM2.5 exposures, the ambient
concentrations corresponding to 25th
percentiles of estimated exposures, or
health events, are 6.7 mg/m3 (Figure 2
and U.S. EPA, 2022b, Figure 3–14). In
key U.S. epidemiologic studies that use
hybrid modeling approaches and do not
apply population-weighting to estimate
PM2.5 exposures, the ambient
concentrations corresponding to 25th
percentiles of estimated exposures, or
health events, range from 4.6 to 9.2 mg/
m3 (U.S. EPA, 2022b, Figure 3–14).83 In
the key epidemiologic studies that apply
hybrid modeling approaches with
population-weighting and with
information available on the 10th
percentile of health events, the ambient
PM2.5 concentration corresponding to
that 10th percentile range from 4.7 mg/
m3 to 7.3 mg/m3 (Figure 2 and U.S. EPA,
2022b, Figure 3–14).
The 2022 PA next considers the PM2.5
concentrations from the key Canadian
epidemiologic studies. Generally, the
study-reported mean concentrations in
Canadian studies are lower than those
reported in the U.S. studies for both
monitor-based and hybrid model
methods. For the majority of key
Canadian epidemiologic studies that use
monitor-based exposure, mean PM2.5
concentrations generally ranged from
83 In the one study that reports 25th percentile
exposure estimates of 4.6 mg/m3 (Shi et al., 2016),
the authors report that most deaths occurred at or
above the 75th percentile of annual exposure
estimates (i.e., 10 mg/m3). The short-term exposure
estimates accounting for most deaths are not
presented in the published study.
PO 00000
Frm 00046
Fmt 4701
Sfmt 4700
7.0 mg/m3 to 9.0 mg/m3 (U.S. EPA,
2022b, Figure 3–9). For these studies,
25th percentiles of health events
correspond to PM2.5 concentrations at or
above 6.5 mg/m3 and 10th percentiles of
health events correspond to PM2.5
concentrations at or above 6.4 mg/m3
(U.S. EPA, 2022b, Figure 3–9). For the
key Canadian epidemiologic studies that
use hybrid model-predicted exposure,
the mean PM2.5 concentrations are
generally lower than in U.S. modelbased studies (U.S. EPA, 2022b, Figure
3–10), ranging from approximately 6.0
mg/m3 to just below 10.0 mg/m3 (U.S.
EPA, 2022b, Figure 3–11). The majority
of the key Canadian epidemiologic
studies that used hybrid modeling were
completed at the nationwide scale,
while four studies were completed at
the regional geographic spatial scale. In
addition, all the key Canadian
epidemiologic studies apply aspects of
population weighting, where all grid
cells within a postal code are averaged,
individuals are assigned exposure at the
postal code resolution, and study mean
PM2.5 concentrations are based on the
average of individual exposures. The
majority of studies estimating exposure
nationwide range between just below
6.0 mg/m3 to 8.0 mg/m3 (U.S. EPA,
2022b, Figure 3–11). One study by
Erickson et al. (2020) presents an
analysis related immigrant status and
length of residence in Canada versus
non-immigrant populations, which
accounts for the four highest mean PM2.5
concentrations which range between 9.0
mg/m3 and 10.0 mg/m3 (U.S. EPA, 2022b,
Figure 3–11). The four studies that
estimate exposure at the regional scale
report mean PM2.5 concentrations that
range from 7.8 mg/m3 to 9.8 mg/m3 (U.S.
EPA, 2022b, Figure 3–11). Three key
Canadian epidemiologic studies report
information on the 25th percentile of
health events. In these studies, the
ambient PM2.5 concentration
corresponding to the 25th percentile is
approximately 8.0 mg/m3 in two studies,
and 4.3 mg/m3 in a third study (U.S.
EPA, 2022b, Figure 3–11).
In addition to the expanded body of
evidence from the key U.S.
epidemiologic studies discussed above,
there are also a subset of epidemiologic
studies that have emerged that further
inform an understanding of the
relationship between PM2.5 exposure
and health effects, including studies
with the highest exposures excluded
(restricted analyses), epidemiologic
studies that employed statistical
approaches that attempt to more
extensively account for confounders and
are more robust to model
misspecification (i.e., used alternative
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
methods for confounder control),84 and
accountability studies (U.S. EPA, 2019a,
U.S. EPA, 2021a, U.S. EPA, 2022a).
Restricted analyses are studies that
examine health effect associations in
analyses with the highest exposures
excluded, restricting analyses to daily
exposures less than the 24-hour primary
PM2.5 standard and annual exposures
less than the annual PM2.5 standard. The
2022 PA presents a summary of
restricted analyses evaluated in the 2019
ISA and ISA Supplement (U.S. EPA,
2022b, Table 3–10). The restricted
analyses can be informative in assessing
the nature of the association between
long-term exposures (e.g., annual
average concentrations <12.0 mg/m3) or
short-term exposures (e.g., daily
concentrations <35 mg/m3) when
looking only at exposures to lower
concentrations, including whether the
association persists in such restricted
analyses compared to the same analyses
for all exposures, as well as whether the
association is stronger, in terms of
magnitude and precision, than when
completing the same analysis for all
exposures. While these studies are
useful in supporting the confidence and
strength of associations at lower
concentrations, these studies also have
inherent uncertainties and limitations,
including uncertainty in how studies
exclude concentrations (e.g., are they
excluded at the modeled grid cell level,
the ZIP code level) and in how
concentrations in studies that restrict air
quality data relate to design values for
the annual and 24-hour standards.
Further, these studies often do not
report descriptive statistics (e.g., mean
PM2.5 concentrations, or concentrations
at other percentiles) that allow for
additional consideration of this
information. As such, while these
studies can provide additional
supporting evidence for associations at
lower concentrations, the 2022 PA notes
that there are also limitations in how to
interpret these studies when evaluating
the adequacy of the current or potential
alternative standards.
Restricted analyses provide additional
information on the nature of the
association between long- or short-term
84 As noted in the ISA Supplement (U.S. EPA,
2022a, p. 1–3): ‘‘In the peer-reviewed literature,
these epidemiologic studies are often referred to as
alternative methods for confounder control. For the
purposes of this Supplement, this terminology is
not used to prevent confusion with the main
scientific conclusions (i.e., the causality
determinations) presented within an ISA. In
addition, as is consistent with the weight-ofevidence framework used within ISAs and
discussed in the Preamble to the Integrated Science
Assessments, an individual study on its own cannot
inform causality, but instead represents a piece of
the overall body of evidence.’’
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
exposures when analyses are restricted
to lower PM2.5 concentrations and
indicate that effect estimates are
generally greater in magnitude in the
restricted analyses for long- and shortterm PM2.5 exposure compared to the
main analyses. In two U.S. studies that
report mean PM2.5 concentrations in
restricted analyses and that estimate
effects associated with long-term
exposure to PM2.5, the effect estimates
are greater in the restricted analyses
than in the main analyses. Di et al.
(2017a) and Dominici et al. (2019) report
positive and statistically significant
associations in analyses restricted to
concentrations less than 12.0 mg/m3 for
all-cause mortality and effect estimates
are greater in the restricted analyses
than effect estimates reported in main
analyses. In addition, both studies
report mean PM2.5 concentrations of 9.6
mg/m3. While none of the U.S. studies of
short-term exposure present mean PM2.5
concentrations for the restricted
analyses, these studies generally have
mean 24-hour average PM2.5
concentrations in the main analyses
below 12.0 mg/m3, and report increases
in the effect estimates in the restricted
analyses compared to the main analyses.
Additionally, in the one Canadian study
of long-term PM2.5 exposure, Zhang et
al. (2021) conducted analyses where
annual PM2.5 concentrations were
restricted to concentrations below 10.0
mg/m3 and 8.8 mg/m3, which presumably
have lower mean concentrations than
the mean of 7.8 mg/m3 reported in the
main analyses, though restricted
analysis mean PM2.5 concentrations are
not reported. Effect estimates for nonaccidental mortality are greater in
analyses restricted to PM2.5
concentrations less than 10.0 mg/m3, but
less in analyses restricted to <8.8 mg/m3.
The second type of studies that have
recently emerged and further inform the
consideration of the relationship
between PM2.5 exposure and health
effects in the 2022 PA are those that
employ alternative methods for
confounder control. Alternative
methods for confounder control seek to
mimic randomized experiments through
the use of study design and statistical
methods to more extensively account for
confounders and are more robust to
model misspecification. The 2022 PA
presents a summary of the studies that
employ alternative methods for
confounder control, and employ a
variety of statistical methods, which are
evaluated in the 2019 ISA and ISA
Supplement (U.S. EPA, 2022b, Table 3–
11). These studies reported consistent
results among large study populations
across the U.S. and can further inform
PO 00000
Frm 00047
Fmt 4701
Sfmt 4700
16247
the relationship between long- and
short-term PM2.5 exposure and total
mortality. Studies that employ
alternative methods for confounder
control to assess the association
between long-term exposure to PM2.5
and mortality reduce uncertainties
related to confounding and provide
additional support for the associations
reported in the broader body of cohort
studies that examined long-term PM2.5
exposure and mortality.
Lastly, there is a subset of
epidemiologic studies that assess
whether long-term reductions in
ambient PM2.5 concentrations result in
corresponding reductions in health
outcomes. These include studies that
evaluate the potential for improvements
in public health, including reductions
in mortality rates, increases in life
expectancy, and reductions in
respiratory disease as ambient PM2.5
concentrations have declined over time.
Some of these studies, accountability
studies, provide insight on whether the
implementation of environmental
policies or air quality interventions
result in changes/reductions in air
pollution concentrations and the
corresponding effect on health
outcomes.85 The 2022 PA presents a
summary of these studies, which are
assessed in the 2019 ISA and ISA
Supplement (U.S. EPA, 2022b, Table 3–
12). These studies lend support for the
conclusion that improvements in air
quality are associated with
improvements in public health.
More specifically, of the
accountability studies that account for
changes in PM2.5 concentrations due to
a policy or the implementation of an
intervention and whether there was
evidence of changes in associations with
mortality or cardiovascular effects as a
result of changes in annual PM2.5
concentrations, Corrigan et al. (2018),
Henneman et al. (2019) and Sanders et
al. (2020a) present analyses with
starting PM2.5 concentrations (or
concentrations prior to the policy or
intervention) below 12.0 mg/m3.
Henneman et al. (2019) explored
changes in modeled PM2.5
concentrations following the retirement
of coal fired power plants in the U.S.,
and found that reductions from mean
annual PM2.5 concentrations of 10.0 mg/
m3 in 2005 to mean annual PM2.5
concentrations of 7.2 mg/m3 in 2012
from coal-fueled power plants resulted
in corresponding reductions in the
number of cardiovascular-related
85 Given the nature of these studies, the majority
tend to focus on time periods in the past during
which ambient PM2.5 concentrations were
substantially higher than those measured more
recently (e.g., see U.S. EPA, 2022b, Figure 2–16).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16248
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
hospital admissions, including for all
cardiovascular disease, acute MI, stroke,
heart failure, and ischemic heart disease
in those aged 65 and older. Corrigan et
al. (2018) examined whether there was
a change in the cardiovascular mortality
rate before (2000–2004) and after (2005–
2010) implementation of the first annual
PM2.5 NAAQS implementation based on
mortality data from the National Center
for Health Statistics and reported 1.10
(95% confidence interval (CI): 0.37,
1.82) fewer cardiovascular deaths per
year per 100,000 people for each 1 mg/
m3 reduction in annual PM2.5
concentrations. When comparing
whether counties met the annual PM2.5
standard (attainment counties), there
were 1.96 (95% CI: 0.77, 3.15) fewer
cardiovascular deaths for each 1 mg/m3
reduction in annual PM2.5
concentrations between the two periods
for attainment counties, whereas in nonattainment counties (e.g., counties that
did not meet the annual PM2.5 standard),
there were 0.59 (95% CI: ¥ 0.54, 1.71)
fewer cardiovascular deaths between the
two periods. And lastly, Sanders et al.
(2020a) examined whether policy
actions (i.e., the first annual PM2.5
NAAQS implementation rule in 2005
for the 1997 annual PM2.5 standard with
a 3-year annual average of 15 mg/m3)
reduced PM2.5 concentrations and
mortality rates in Medicare beneficiaries
between 2000–2013. They report
evidence of changes in associations with
mortality (a decreased mortality rate of
∼0.5 per 1,000 in attainment and nonattainment areas) due to changes in
annual PM2.5 concentrations in both
attainment and non-attainment areas.
Additionally, attainment areas had
starting concentrations below 12.0 mg/
m3 prior to implementation of the
annual PM2.5 NAAQS in 2005. In
addition, following implementation of
the annual PM2.5 NAAQS, annual PM2.5
concentrations decreased by 1.59 mg/m3
(95% CI: 1.39, 1.80) which
corresponded to a reduction in mortality
rates among individuals 65 years and
older (0.93% [95% CI: 0.10%, 1.77%])
in non-attainment counties relative to
attainment counties. In a life expectancy
study, Bennett et al. (2019) reports
increases in life expectancy in all but 14
counties (1325 of 1339 counties) that
have exhibited reductions in PM2.5
concentrations from 1999 to 2015. These
studies provide support for
improvements in public health
following the implementation of
policies, including in areas with PM2.5
concentrations below the level of the
current annual standard, as well as
increases in life expectancy in areas
with reductions in PM2.5 concentrations.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
d. Uncertainties in the Health Effects
Evidence
The 2022 PA recognizes that there are
a number of uncertainties and
limitations associated with the available
health effects evidence. Although the
epidemiologic studies clearly
demonstrate associations between longand short-term PM2.5 exposures and
health outcomes, several uncertainties
and limitations in the health effects
evidence remain. Epidemiologic studies
evaluating short-term PM2.5 exposure
and health effects have reported
heterogeneity in associations between
cities and geographic regions within the
U.S. Heterogeneity in the associations
observed across epidemiologic studies
may be due in part to exposure error
related to measurement-related issues,
the use of central fixed-site monitors to
represent population exposure to PM2.5,
and a limited understanding of factors
including exposure error related to
measurement-related issues, variability
in PM2.5 composition regionally, and
factors that result in differential
exposures (e.g., topography, the built
environment, housing characteristics,
personal activity patterns).
Heterogeneity is expected when the
methods or the underlying distribution
of covariates vary across studies (U.S.
EPA, 2019a, p. 6–221). Studies assessed
in the 2019 ISA and ISA Supplement
have advanced the state of exposure
science by presenting innovative
methodologies to estimate PM exposure,
detailing new and existing measurement
and modeling methods, and further
informing our understanding of the
influence of exposure measurement
error due to exposure estimation
methods on the associations between
PM2.5 and health effects reported in
epidemiologic studies (U.S. EPA, 2019a,
section 1.2.2; U.S. EPA, 2022a). Data
from PM2.5 monitors continue to be
commonly used in health studies as a
surrogate for PM2.5 exposure, and often
provide a reasonable representation of
exposures throughout a study area (U.S.
EPA, 2019a, section 3.4.2.2; U.S. EPA,
2022a, section 3.2.2.2.2). However, an
increasing number of studies employ
hybrid modeling methods to estimate
PM2.5 exposure using data from several
sources, often including satellites and
models, in addition to ground-based
monitors. These hybrid models typically
have good cross-validation, especially
for PM2.5, and have the potential to
reduce exposure measurement error and
uncertainty in the health effect
estimates from epidemiologic models of
long-term exposure (U.S. EPA, 2019a,
section 3.5; U.S. EPA, 2022a, section
2.3.3).
PO 00000
Frm 00048
Fmt 4701
Sfmt 4700
While studies using hybrid modeling
methods have reduced exposure
measurement error and uncertainty in
the health effect estimates, these studies
use a variety of approaches to estimate
PM2.5 concentrations and to assign
exposure to assess the association
between health outcomes and PM2.5
exposure. This variability in
methodology has inherent limitations
and uncertainties, as described in more
detail in section 2.3.3.1.5 of the 2022
PA, and the performance of the
modeling approaches depends on the
availability of monitoring data which
varies by location. Factors that likely
contribute to poorer model performance
often coincide with relatively low
ambient PM2.5 concentrations, in areas
where predicted exposures are at a
greater distance to monitors, and under
conditions where the reliability and
availability of key datasets (e.g., air
quality modeling) are limited. Thus,
uncertainty in hybrid model predictions
becomes an increasingly important
consideration as lower predicted
concentrations are considered.
Regardless of whether a study uses
monitoring data or a hybrid modeling
approach when estimating PM2.5
exposures, one key limitation that
persists is associated with the
interpretation of the study-reported
mean PM2.5 concentrations and how
they compare to design values, the
metric that describes the air quality
status of a given area relative to the
NAAQS.86 As discussed above in
section II.B.3.b, the overall mean PM2.5
concentrations reported by key
epidemiologic studies reflect averaging
of short- or long-term PM2.5 exposure
estimates across location (i.e., across
multiple monitors or across modeled
grid cells) and over time (i.e., over
several years). For monitor-based
studies, the comparison is somewhat
more straightforward than for studies
that use hybrid modeling methods, as
the monitors used to estimate exposure
in the epidemiologic studies are
generally the same monitors that are
used to calculate design values for a
given area. It is expected that areas
meeting a PM2.5 standard with a
particular level would be expected to
have average PM2.5 concentrations (i.e.,
averaged across space and over time in
the area) somewhat below that standard
level., but the difference between the
maximum annual design value and
86 For the annual PM
2.5 standard, design values
are calculated as the annual arithmetic mean PM2.5
concentration, averaged over 3 years. For the 24hour standard, design values are calculated as the
98th percentile of the annual distribution of 24hour PM2.5 concentrations, averaged over three
years (Appendix N of 40 CFR part 50).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
average concentration in an area can be
smaller or larger than analyses
presented above in section I.D.5.a, likely
depending on factors such as the
number of monitors, monitor siting
characteristics, and the distribution of
ambient PM2.5 concentrations. For
studies that use hybrid modeling
methods to estimate PM2.5
concentrations, the comparison between
study-reported mean PM2.5
concentrations and design values is
more complicated given the variability
in the modeling methods, temporal
scales (i.e., daily versus annual), and
spatial scales (i.e., nationwide versus
urban) across studies. Analyses above in
section I.D.5.b and detailed more in the
2022 PA (U.S. EPA, 2022b, section
2.3.3.2.4) present a comparison between
two hybrid modeling surfaces, which
explored the impact of these factors on
the resulting mean PM2.5 concentrations
and provided additional information
about the relationship between mean
concentrations from studies using
hybrid modeling methods and design
values. However, the results of those
analyses only reflect two surfaces and
two types of approaches, so uncertainty
remains in understanding the
relationship between estimated modeled
PM2.5 concentrations and design values
more broadly across hybrid modeling
studies. Moreover, this analysis was
completed using two hybrid modeling
methods that estimate PM2.5
concentrations in the U.S., thus an
additional uncertainty includes
understanding the relationship between
modeled PM2.5 concentrations and
design values reported in Canada.
In addition, where PM2.5 and other
pollutants (e.g., ozone, nitrogen dioxide,
and carbon monoxide) are correlated, it
can be difficult to distinguish whether
attenuation of effects in some studies
results from copollutant confounding or
collinearity with other pollutants in the
ambient mixture (U.S. EPA, 2019a,
section 1.5.1; U.S. EPA, 2022a, section
2.2.1). Studies evaluated in the 2019
ISA and ISA Supplement further
examined the potential confounding
effects of both gaseous and particulate
copollutants on the relationship
between long- and short-term PM2.5
exposure and health effects. As noted in
the Appendix to the 2019 ISA (U.S.
EPA, 2019a, Table A–1), copollutant
models are not without their limitations,
such as instances for which correlations
are high between pollutants resulting in
greater copollutant confounding bias in
results. However, the studies continue
to provide evidence indicating that
associations with PM2.5 are relatively
unchanged in copollutants models (U.S.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
EPA, 2019a, section 1.5.1; U.S. EPA,
2022a, section 2.2.1).
Another area of uncertainty is
associated with other potential
confounders, beyond copollutants.
Some studies have expanded the
examination of potential confounders to
not only include copollutants, but also
systematic evaluations of the potential
impact of inadequate control from longterm temporal trends and weather (U.S.
EPA, 2019a, section 11.1.5.1). Analyses
examining these covariates further
confirm that the relationship between
PM2.5 exposure and mortality is unlikely
to be biased by these factors. Other
studies have explored the use of
alternative methods for confounder
control to more extensively account for
confounders and are more robust to
model misspecification that can further
inform the causality determination for
long-term and short-term PM2.5 and
mortality and cardiovascular effects
(U.S. EPA, 2019a, section 11.2.2.4; U.S.
EPA, 2022a, sections 3.1.1.3, 3.1.2.3,
3.2.1.2, and 3.2.2.3). These studies
indicate that bias from unmeasured
confounders can occur in either
direction, although controlling for these
confounders did not result in the
elimination of the association, but
instead provided additional support for
associations between long-term PM2.5
exposure and mortality when
accounting for additional confounders
(U.S. EPA, 2022a, section 3.2.2.2.6).
Another important limitation
associated with the evidence is that,
while epidemiologic studies indicate
associations between PM2.5 and health
effects, the currently available evidence
does not identify particular PM2.5
concentrations that do not elicit health
effects. Rather, health effects can occur
over the entire distribution of ambient
PM2.5 concentrations evaluated, and
epidemiologic studies conducted to date
do not identify a population-level
threshold below which it can be
concluded with confidence that PM2.5related effects do not occur.
Overall, evidence assessed in the 2019
ISA and ISA Supplement continues to
indicate a linear, no-threshold C–R
relationship for PM2.5 concentrations >8
mg/m3. However, uncertainties remain
about the shape of the C–R curve at
PM2.5 concentrations <8 mg/m3, with
some recent studies providing evidence
for either a sublinear, linear, or
supralinear relationship at these lower
concentrations (U.S. EPA, 2019a,
section 11.2.4; U.S. EPA, 2022a, section
2.2.3.2).
There are also a number of
uncertainties and limitations associated
with the experimental evidence (i.e.,
controlled human exposure studies and
PO 00000
Frm 00049
Fmt 4701
Sfmt 4700
16249
animal toxicological studies). With
respect to controlled human exposure
studies, the PA recognizes that these
studies include a small number of
individuals compared to epidemiologic
studies. Additionally, these studies tend
to include generally healthy adult
individuals, who are at a lower risk of
experiencing health effects. These
studies, therefore, often do not include
populations that are at increased risk of
PM2.5-related health effects, including
children, older adults, or individuals
with pre-existing conditions. As such,
these studies are somewhat limited in
their ability to inform at what
concentrations effects may be elicited in
at-risk populations. With respect to
animal toxicological studies, while
these studies often examine more severe
health outcomes and longer exposure
durations and higher exposure
concentrations than controlled human
exposure studies, there is uncertainty in
extrapolating the effects seen in
animals, and the PM2.5 exposures and
doses that cause those effects, to human
populations.
Consideration of health effects are
informed by the epidemiologic,
controlled human exposure, and animal
toxicological studies. The evaluation
and integration of the scientific
evidence in the ISA focuses on
evaluating the findings from the body of
evidence across disciplines, including
evaluating the strengths and weaknesses
in the overall collection of studies
across disciplines. Integrating evidence
across disciplines can strengthen causal
inference, such that a weak inference
from one line of evidence can be
addressed by other lines of evidence,
and coherence of these lines of evidence
can add support to a cause-effect
interpretation of the association.
Evaluation and integration of the
evidence also includes consideration of
uncertainties that are inherent in the
scientific findings (U.S. EPA, 2015, pp.
13–15), some of which are described
above.
3. Summary of Exposure and Risk
Estimates
Beyond the consideration of the
scientific evidence, discussed above in
section II.B, the EPA also considers the
extent to which new or updated
quantitative analyses of PM2.5 air
quality, exposure, or health risks could
inform conclusions on the adequacy of
the public health protection provided by
the current primary PM2.5 standards.
Additionally, the 2022 PA includes an
at-risk analysis that assesses PM2.5attributable risk associated with PM2.5
air quality that has been adjusted to
simulate air quality scenarios of policy
E:\FR\FM\06MRR3.SGM
06MRR3
16250
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
interest (e.g., ‘‘just meeting’’ the current
or potential alternative standards).
Drawing on the summary in section II.C
of the proposal, the sections below
provide a brief overview of key aspects
of the assessment design (II.A.3.a), key
limitations and uncertainties (II.A.3.b),
and exposure/risk estimates (II.A.3.c).
ddrumheller on DSK120RN23PROD with RULES3
a. Key Design Aspects
Risk assessments combine data from
multiple sources and involve various
assumptions and uncertainties. Input
data for these analyses includes C–R
functions from epidemiologic studies
for each health outcome and ambient
annual or 24-hour PM2.5 concentrations
for the study areas utilized in the risk
assessment (U.S. EPA, 2022b, section
3.4.1). Additionally, quantitative and
qualitative methods were used to
characterize variability and uncertainty
in the risk estimates (U.S. EPA, 2022b,
section 3.4.1.7).
Concentration-response functions
used in the risk assessment are from
large, multicity U.S. epidemiologic
studies that evaluate the relationship
between PM2.5 exposures and mortality.
Epidemiologic studies and
concentration-response studies that
were used in the risk assessment to
estimate risk were identified using
criteria that take into account factors
such as study design, geographic
coverage, demographic populations, and
health endpoints (U.S. EPA, 2022b,
section 3.4.1.1).87 The risk assessment
focuses on all-cause or nonaccidental
mortality associated with long-term and
short-term PM2.5 exposures, for which
the 2019 ISA concluded that the
evidence provides support for a ‘‘causal
relationship’’ (U.S. EPA, 2022b, section
3.4.1.2).88
As described in more detail in the
2022 PA, the risk assessment first
estimated health risks associated with
air quality for 2015 adjusted to simulate
‘‘just meeting’’ the current primary
PM2.5 standards (i.e., the annual
standard with its level of 12.0 mg/m3
and the 24-hour standard with its level
of 35 mg/m3). Air quality modeling was
then used to simulate air quality just
meeting an alternative standard with a
level of 10.0 mg/m3 (annual) and 30 mg/
m3 (24-hour). In addition to the modelbased approach, for the subset of 30
87 Additional detail regarding the selection of
epidemiologic studies and specification of C–R
functions is provided in the 2022 PA (U.S. EPA,
2022b, Appendix C, section C.1.1).
88 While the 2019 ISA also found that evidence
supports the determination of a ‘‘causal
relationship’’ between long- and short-term PM2.5
exposures and cardiovascular effects,
cardiovascular mortality was not included as a
health outcome as it will be captured in the
estimates of all-cause mortality.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
areas controlled by the annual standard
linear interpolation and extrapolation
were employed to simulate just meeting
alternative annual standards with levels
of 11.0 (interpolated between 12.0 and
10.0 mg/m3), 9.0 mg/m3, and 8.0 mg/m3
(both extrapolated from 12.0 and 10.0
mg/m3) (U.S. EPA, 2022b, section
3.4.1.3). The 2022 PA notes that there is
greater uncertainty regarding whether a
revised 24-hour standard (i.e., with a
lower level) is needed to further limit
‘‘peak’’ PM2.5 concentration exposure
and whether a lower 24-hour standard
level would most effectively reduce
PM2.5-associated health risks associated
with ‘‘typical’’ daily exposures. The risk
assessment estimates health risks
associated with air quality adjusted to
meet a revised 24-hour standard with a
level of 30 mg/m3, in conjunction with
estimating the health risks associated
with meeting a revised annual standard
with a level of 10.0 mg/m3 (U.S. EPA,
2022b, section 3.4.1.3). More details on
the air quality adjustment approaches
used in the risk assessment are
described in section 3.4.1.4 and
Appendix C of the 2022 PA (U.S. EPA,
2022b).
When selecting U.S. study areas for
inclusion in the risk assessment, the
available ambient monitors, geographic
diversity, and ambient PM2.5 air quality
concentrations were taken into
consideration (U.S. EPA, 2022b, section
3.4.1.4). When these factors were
applied, 47 urban study areas were
identified, which include nearly 60
million people aged 30–99, or
approximately 30% of the U.S
population in this age range (U.S. EPA,
2022b, section 3.4.1.5, Appendix C,
section C.1.3). Of the 47 study areas,
there were 30 study areas where just
meeting the current standards is
controlled by the annual standard,89 11
study areas where just meeting the
current standards is controlled by the
daily standard,90 and 6 study areas
where the controlling standard differed
depending on the air quality adjustment
approach (U.S. EPA, 2022b, section
3.4.1.5).91
89 For these areas, the annual standard is the
‘‘controlling standard’’ because when air quality is
adjusted to simulate just meeting the current or
potential alternative annual standards, that air
quality also would meet the 24-hour standard being
evaluated.
90 For these areas, the 24-hour standard is the
controlling standard because when air quality is
adjusted to simulate just meeting the current or
potential alternative 24-hour standards, that air
quality also would meet the annual standard being
evaluated. Some areas classified as being controlled
by the 24-hour standard also violate the annual
standard.
91 In these 6 areas, the controlling standard
depended on the air quality adjustment method
used and/or the standard scenarios evaluated.
PO 00000
Frm 00050
Fmt 4701
Sfmt 4700
In addition to the overall risk
assessment, the 2022 PA also includes
an at-risk analysis and estimates
exposures and health risks of specific
populations identified as at-risk that
would be allowed under the current and
potential alternative standards to further
inform the Administrator’s conclusions
regarding the adequacy of the public
health protection provided by the
current primary PM2.5 standards. In so
doing, the 2022 PA evaluates exposure
and PM2.5 mortality risk for older adults
(e.g., 65 years and older), stratified for
White, Black, Asian, Native American,
Non-Hispanic, and Hispanic individuals
residing in the same study areas
included in the overall risk assessment.
This analysis utilizes a recent
epidemiologic study that provides raceand ethnicity-specific risk coefficients
(Di et al., 2017b).
b. Key Limitations and Uncertainties
Uncertainty in risk estimates (e.g., in
the size of risk estimates) can result
from a number of factors, including the
assumptions about the shape of the C–
R function with mortality at low
ambient PM concentrations, the
potential for confounding and/or
exposure measurement error in the
underlying epidemiologic studies, and
the methods used to adjust PM2.5 air
quality. More specifically, the use of air
quality modeling to adjust PM2.5
concentrations are limited as they rely
on model predictions, are based on
emission changes scaled by fixed
percentages, and use only two of the full
set of possible emission scenarios and
linear interpolation/extrapolation to
adjust air quality that may not fully
capture potential non-linearities
associated with real-world changes in
air quality. Additionally, the selection
of case study areas is limited to urban
areas predominantly located CA and in
the Eastern U.S. that are controlled by
the annual standard. While the risk
assessment does not report quantitative
uncertainty in the risk estimates as
exposure concentrations are reduced, it
does provide information on the
distribution of concentrations associated
with the risk estimates when evaluating
progressively lower alternative annual
standards. Based on these data, as lower
alternative annual standards are
evaluated, larger proportions of the
distributions in risk occur at or below
10 mg/m3 (at concentrations below or
near most of the study-reported means
from the key U.S. epidemiologic
studies) and at or below 8 mg/m3 (the
concentration at which the ISA reports
increasing uncertainty in the shape of
the C–R curve based on the body of
epidemiologic evidence).
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
Similarly, the at-risk analysis is also
subject to many of these same
uncertainties noted above. Additionally,
the at-risk analysis included C–R
functions from only one study (Di et al.,
2017b), which reported associations
between long-term PM2.5 exposures and
mortality, stratified by race/ethnicity, in
populations age 65 and older, as
opposed to the multiple studies used in
the overall risk assessment to convey
risk estimate variability. These and
other sources of uncertainty in the
overall risk assessment and the at-risk
analyses are characterized in more
depth in the 2022 PA (U.S. EPA, 2022b,
section 3.4.1.7, section 3.4.1.8,
Appendix C, section C.3).
c. Summary of Risk Estimates
Although limitations in the
underlying data and approaches lead to
some uncertainty regarding estimates of
PM2.5-associated risk, the risk
assessment estimates that the current
primary PM2.5 standards could allow a
substantial number of PM2.5-associated
deaths in the U.S. For example, when
air quality in the 47 study areas is
adjusted to simulate just meeting the
current standards, the risk assessment
estimates up to 45,100 deaths in 2015
are attributable to long-term PM2.5
exposures associated with just meeting
the current annual and 24-hour PM2.5
standards (U.S. EPA, 2022b, section
3.4.2.1). Additionally, as described in
more detail in the 2022 PA, the at-risk
analysis suggests that a lower annual
standard level (i.e., below 12 mg/m3 and
down as low as 8 mg/m3) will help to
reduce PM2.5 exposure and may also
help to mitigate exposure and risk
disparities in populations identified as
particularly at-risk for adverse effects
from PM exposures (i.e., minority
populations).
Compared to the current annual
standard, meeting a revised annual
standard with a lower level is estimated
to reduce PM2.5-associated health risks
in the 30 study areas controlled by the
annual standard by about 7–9% for a
level of 11.0 mg/m3, 15–19% for a level
of 10.0 mg/m3, 22–28% for a level of 9.0
mg/m3, and 30–37% for a level of 8.0 mg/
m3) (U.S. EPA, 2022b, Table 3–17).
Meeting a revised annual standard with
a lower level may also help to mitigate
exposure and risk disparities in
populations identified as particularly atrisk for adverse effects from PM
exposures (i.e., minority populations) in
simulated scenarios just meeting
alternative annual standards. However,
though reduced, disparities by race and
ethnicity persist even at an alternative
annual standard level of 8 mg/m3, the
lowest alternative annual standard
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
included in the risk assessment (U.S.
EPA, 2022b, section 3.4.2.4).
Revising the level of the 24-hour
standard to 30 mg/m3 is estimated to
lower PM2.5-associated risks across a
more limited population and number of
areas than revising the annual standard
(U.S. EPA, 2022, section 3.4.2.4). Risk
reduction predictions are largely
confined to areas located in the western
U.S., several of which are also likely to
experience risk reductions upon
meeting a revised annual standard. In
the 11 areas controlled by the 24-hour
standard, when air quality is simulated
to just meet the current 24-hour
standard, PM2.5 exposures are estimated
to be associated with as many as 2,570
deaths annual. Compared to just
meeting the current standard, air quality
just meeting an alternative 24-hour
standard level of 30 mg/m3 is associated
with reductions in estimated risk of 9–
13% (U.S. EPA, 2022b, section 3.4.2.3).
B. Conclusions on the Primary PM2.5
Standards
In drawing conclusions on the
adequacy of the current primary PM2.5
standards, in view of the advances in
scientific knowledge and additional
information now available, the
Administrator has considered the
evidence base, information, and policy
judgments that were the foundation of
the 2012 and 2020 reviews and reflects
upon the body of evidence and
information newly available in this
reconsideration. In so doing, the
Administrator has taken into account
both evidence-based and risk-based
considerations, as well as advice from
the CASAC and public comments.
Evidence-based considerations draw
upon the EPA’s integrated assessment of
the scientific evidence of health effects
related to PM2.5 exposure presented in
the 2019 ISA and ISA Supplement
(summarized in the proposal in sections
II.B (88 FR 5580, January 27, 2023) and
II.D.2.a (88 FR 5609, January 27, 2023),
and also in section II.A.2 above) to
address key policy-relevant questions in
the reconsideration. Similarly, the riskbased considerations draw upon the
assessment of population exposure and
risk (summarized in the proposal in
sections II.C (88 FR 5605, January 27,
2023) and II.D.2.b (88 FR 5614, January
27, 2023), and also in section II.A.3
above) in addressing policy-relevant
questions focused on the potential for
PM2.5 exposures associated with
mortality under air quality conditions
just meeting the current and potential
alternative standards.
The approach to reviewing the
primary standards is consistent with
requirements of the provisions of the
PO 00000
Frm 00051
Fmt 4701
Sfmt 4700
16251
CAA related to the review of the
NAAQS and with how the EPA and the
courts have historically interpreted the
CAA. As discussed in section I.A above,
these provisions require the
Administrator to establish primary
standards that, in the Administrator’s
judgment, are requisite (i.e., neither
more nor less stringent than necessary)
to protect public health with an
adequate margin of safety. Consistent
with the Agency’s approach across all
NAAQS reviews, the EPA’s approach to
informing these judgments is based on
a recognition that the available health
effects evidence generally reflects a
continuum that includes ambient air
exposures for which scientists generally
agree that health effects are likely to
occur through lower levels at which the
likelihood and magnitude of response
become increasingly uncertain. The
CAA does not require the Administrator
to establish a primary standard at a zerorisk level or at background
concentration levels, but rather at a
level that reduces risk sufficiently so as
to protect public health, including the
health of sensitive groups, with an
adequate margin of safety.
The decisions on the adequacy of the
current primary PM2.5 standards
described below is a public health
policy judgment by the Administrator
that draws on the scientific evidence for
health effects, quantitative analyses of
population exposures and/or health
risks, and judgments about how to
consider the uncertainties and
limitations that are inherent in the
scientific evidence and quantitative
analyses. The four basic elements of the
NAAQS (i.e., indicator, averaging time,
form, and level) have been considered
collectively in evaluating the public
health protection afforded by the
current standards.
Section II.B.2 below briefly
summarizes the basis for the
Administrator’s proposed decision,
drawing from section II.D.3 of the
proposal (88 FR 5617, January 27, 2023).
The advice and recommendations of the
CASAC and public comments on the
proposed decision are addressed below
in sections II.B.1 and II.B.3,
respectively. The Administrator’s final
conclusions in this reconsideration
regarding the adequacy of the current
primary PM2.5 standards and whether
any revisions are appropriate are
described in section II.B.4.
1. CASAC Advice
As part of its review of the 2019 draft
PA, the CASAC provided advice on the
adequacy of the public health protection
afforded by the current primary PM2.5
standards. Its advice is documented in
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16252
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
a letter sent to the EPA Administrator
(Cox, 2019b). In this letter, the
committee recommended retaining the
current 24-hour PM2.5 standard but did
not reach consensus on whether the
scientific and technical information
support retaining or revising the current
annual standard. In particular, though
the CASAC agreed that there is a longstanding body of health evidence
supporting relationships between PM2.5
exposures and various health outcomes,
including mortality and serious
morbidity effects, individual CASAC
members ‘‘differ[ed] in their
assessments of the causal and policy
significance of these associations’’ (Cox,
2019b, p. 8 of consensus responses).
Drawing from this evidence, ‘‘some
CASAC members’’ expressed support
for retaining the current annual
standard while ‘‘other members’’
expressed support for revising that
standard in order to increase public
health protection (Cox, 2019b, p.1 of
letter). These views are summarized
below.
The CASAC members who supported
retaining the current annual standard
expressed the view that substantial
uncertainty remains in the evidence for
associations between PM2.5 exposures
and mortality or serious morbidity
effects. These committee members
asserted that ‘‘such associations can
reasonably be explained in light of
uncontrolled confounding and other
potential sources of error and bias’’
(Cox, 2019b, p. 8 of consensus
responses). They noted that associations
do not necessarily reflect causal effects,
and they contended that recent
epidemiologic studies assessed in the
2019 ISA that report positive
associations at lower estimated
exposure concentrations mainly confirm
what was anticipated or already
assumed in setting the 2012 NAAQS. In
particular, they concluded that such
studies have some of the same
limitations as prior studies and do not
provide new information calling into
question the existing standard. They
further asserted that ‘‘accountability
studies provide potentially crucial
information about whether and how
much decreasing PM2.5 causes decreases
in future health effects’’ (Cox, 2019b, p.
10 of consensus responses), and they
cited recent reviews (i.e., Henneman et
al., 2017; Burns et al., 2019) to support
their position that in such studies,
‘‘reductions of PM2.5 concentrations
have not clearly reduced mortality
risks’’ (Cox, 2019b, p. 8 of consensus
responses). Thus, the committee
members who supported retaining the
current annual standard advise that,
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
‘‘while the data on associations should
certainly be carefully considered, this
data should not be interpreted more
strongly than warranted based on its
methodological limitations’’ (Cox,
2019b, p. 8 of consensus responses).
These members of the CASAC further
concluded that the quantitative risk
assessment included in the 2019 draft
PA does not provide a valid basis for
revising the current standards. This
conclusion was based on concerns that
(1) ‘‘the risk assessment treats regression
coefficients as causal coefficients with
no justification or validation provided
for this decision;’’ (2) the estimated
regression concentration-response
functions ‘‘have not been adequately
adjusted to correct for confounding,
errors in exposure estimates and other
covariates, model uncertainty, and
heterogeneity in individual biological
(causal) [concentration-response]
functions;’’ (3) the estimated
concentration-response functions ‘‘do
not contain quantitative uncertainty
bands that reflect model uncertainty or
effects of exposure and covariate
estimation errors;’’ and (4) ‘‘no
regression diagnostics are provided
justifying the use of proportional
hazards . . . and other modeling
assumptions’’ (Cox, 2019b, p. 9 of
consensus responses). These committee
members also contended that details
regarding the derivation of
concentration-response functions,
including specification of the beta
values and functional forms, were not
well-documented, hampering the ability
of readers to evaluate these design
details. Thus, these members ‘‘think that
the risk characterization does not
provide useful information about
whether the current standard is
protective’’ (Cox, 2019b, p. 11 of
consensus responses).
Drawing from their evaluation of the
evidence and the risk assessment in the
2019 draft PA, these committee
members concluded that ‘‘the Draft PM
PA does not establish that new scientific
evidence and data reasonably call into
question the public health protection
afforded by the . . . 2012 PM2.5 annual
standard’’ (Cox, 2019b, p.1 of letter).
In contrast, ‘‘[o]ther members of
CASAC conclude[d] that the weight of
the evidence, particularly reflecting
recent epidemiology studies showing
positive associations between PM2.5 and
health effects at estimated annual
average PM2.5 concentrations below the
current standard, does reasonably call
into question the adequacy of the 2012
annual PM2.5 [standard] to protect
public health with an adequate margin
of safety’’ (Cox, 2019b, p.1 of letter). The
committee members who supported this
PO 00000
Frm 00052
Fmt 4701
Sfmt 4700
conclusion noted that the body of health
evidence for PM2.5 not only includes the
repeated demonstration of associations
in epidemiologic studies, but also
includes support for biological
plausibility established by controlled
human exposure and animal toxicology
studies. They pointed to recent studies
demonstrating that the associations
between PM2.5 and health effects occur
in a diversity of locations, in different
time periods, with different
populations, and using different
exposure estimation and statistical
methods. They concluded that ‘‘the
entire body of evidence for PM health
effects justifies the causality
determinations made in the Draft PM
ISA’’ (Cox, 2019b, p. 8 of consensus
responses).
The members of the CASAC who
supported revising the current annual
standard particularly emphasized recent
findings of associations with PM2.5 in
areas with average long-term PM2.5
concentrations below the level of the
annual standard and studies that show
positive associations even when
estimated exposures above 12 mg/m3 are
excluded from analyses. They found it
‘‘highly unlikely’’ that the extensive
body of evidence indicating positive
associations at low estimated exposures
could be fully explained by
confounding or by other non-causal
explanations (Cox, 2019b, p. 8 of
consensus responses). They additionally
concluded that ‘‘the risk
characterization does provide a useful
attempt to understand the potential
impacts of alternate standards on public
health risks’’ (Cox, 2019b, p. 11 of
consensus responses). These CASAC
members concluded that the available
evidence reasonably calls into question
the protection provided by the current
primary PM2.5 standards and supports
revising the annual standard to increase
that protection (Cox, 2019b).
As a part of this reconsideration, the
CASAC reviewed the 2021 draft PA
(developed to support the
reconsideration as described in section
I.C.5 above). As a part of their review of
the 2021 draft PA, the CASAC provided
advice on the adequacy of the current
primary PM2.5 standards. The range of
views summarized here generally
reflects differing judgments as to the
relative weight to place on various types
of evidence, the risk-based information,
and the associated uncertainties, as well
as differing judgments about the
importance of various PM2.5-related
health effects from a public health
perspective.
In its comments on the 2021 draft PA,
the CASAC stated that: ‘‘[o]verall the
CASAC finds the Draft PA to be well-
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
written and appropriate for helping to
‘bridge the gap’ between the agency’s
scientific assessments and quantitative
technical analyses, and the judgments
required of the Administrator in
determining whether it is appropriate to
retain or revise the National Ambient
Air Quality Standards (NAAQS)’’
(Sheppard, 2022a, p. 1 of consensus
letter). The CASAC also stated that the
‘‘[d]raft PA adequately captures and
appropriately characterizes the key
aspects of the evidence assessed and
integrated in the 2019 ISA and Draft ISA
Supplement of PM2.5-related health
effects’’ (Sheppard, 2022b, p. 2 of
consensus letter). The CASAC also
stated that ‘‘[t]he interpretation of the
risk assessment for the purpose of
evaluating the adequacy of the current
primary PM2.5 annual standard is
appropriate given the scientific findings
presented’’ (Sheppard, 2022a, p. 2 of
consensus letter).
With regard to the adequacy of the
current primary annual PM2.5 standard,
‘‘all CASAC members agree that the
current level of the annual standard is
not sufficiently protective of public
health and should be lowered’’
(Sheppard, 2022a, p. 2 of consensus
letter). Additionally, ‘‘the CASAC
reached consensus that the indicator,
form, and averaging time should be
retained, without revision’’ (Sheppard,
2022a, p. 2 of consensus letter). With
regard to the level of the primary annual
PM2.5 standard, the CASAC had
differing recommendations for the
appropriate range for an alternative
level. The majority of the CASAC
‘‘judge[d] that an annual average in the
range of 8–10 mg/m3’’ was most
appropriate, while the minority of the
CASAC members stated that ‘‘the range
of the alternative standard of 10–11 mg/
m3 is more appropriate’’ (Sheppard,
2022a, p. 16 of consensus responses).
The CASAC did highlight, however, that
‘‘the alternative standard level of 10 mg/
m3 is within the range of acceptable
alternative standards recommended by
all CASAC members, and that an annual
standard below 12 mg/m3 is supported
by a larger and coherent body of
evidence’’ (Sheppard, 2022a, p. 16 of
consensus responses).
In reaching conclusions on a
recommended range of 8–10 mg/m3 for
the primary annual PM2.5 standard, the
majority of the CASAC placed weight on
various aspects of the available
scientific evidence and quantitative risk
assessment information discussed in the
2021 draft PA (Sheppard, 2022a, p. 16
of consensus responses). In particular,
these members cited recent U.S.- and
Canadian-based epidemiologic studies
that show positive associations between
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
PM2.5 exposure and mortality with
study-reported mean concentrations
below 10 mg/m3. Further, these members
also noted that the lower portions of the
air quality distribution (i.e.,
concentrations below the mean) provide
additional information to support
associations between health effects and
PM2.5 concentrations lower than the
reported long-term mean concentration.
In addition, the CASAC members
recognized that the available evidence
has not identified a threshold
concentration, below which an
association no longer remains, pointing
to the conclusion in the draft ISA
Supplement that the ‘‘evidence remains
clear and consistent in supporting a nothreshold relationship, and in
supporting a linear relationship for
PM2.5 concentrations >8 mg/m3’’
(Sheppard, 2022a, p. 16 of consensus
responses). Finally, these CASAC
members placed weight on the at-risk
analysis as providing support for
protection of at-risk demographic
groups, including minority populations.
In recommending a range of 10–11 mg/
m3 for the primary annual PM2.5
standard, the minority of the CASAC
emphasized that there were few key
epidemiologic studies that reported
positive and statistically significant
health effects associations for PM2.5 air
quality distributions with overall mean
concentrations below 9.6 mg/m3
(Sheppard, 2022a, p. 17 of consensus
responses). In so doing, the minority of
the CASAC specifically noted the
variability in the relationship between
study-reported means and area annual
design values based on the methods
utilized in the studies, noting that
design values are generally higher than
area average exposure levels. Further,
the minority of the CASAC stated that
‘‘uncertainties related to copollutants
and confounders make it difficult to
justify a recommendation below 10–11
mg/m3’’ (Sheppard, 2022a, p. 17 of
consensus responses). Finally, the
minority of the CASAC placed less
weight on the risk assessment results,
noting large uncertainties, including the
approaches used for adjusting air
quality to simulate just meeting the
current and alternative standards.
With regard to the current primary 24hour PM2.5 standard, in their review of
the 2021 draft PA, the CASAC did not
reach consensus regarding the adequacy
of the public health protection provided
by the current standard. As described
further below, the majority of the
CASAC members concluded ‘‘that the
available evidence calls into question
the adequacy of the current 24-hour
standard’’ (Sheppard, 2022a, p. 3 of
consensus letter), while the minority of
PO 00000
Frm 00053
Fmt 4701
Sfmt 4700
16253
the CASAC members agreed with ‘‘the
EPA’s preliminary conclusion [in the
draft PA] to retain the current 24-hour
PM2.5 standard without revision’’
(Sheppard, 2022a, p. 4 of consensus
letter). The CASAC recommended that
in future reviews, the EPA should also
consider alternative forms for the
primary 24-hour PM2.5 standard.
Specifically, the CASAC ‘‘suggests
considering a rolling 24-hour average
and examining alternatives to the 98th
percentile of the 3-year average,’’
pointing to concerns that computing 24hour average PM2.5 concentrations using
the current midnight-to-midnight
timeframe could potentially
underestimate the effects of high 24hour exposures, especially in areas with
wood-burning stoves and wintertime
stagnation (Sheppard, 2022a, p. 18 of
consensus responses).
As noted above, the majority of the
CASAC favored revising the level of the
primary 24-hour PM2.5 standard,
suggesting that a range of 25–30 mg/m3
would be adequately protective. In so
doing, the majority of the CASAC
placed weight on the available
epidemiologic evidence, including
epidemiologic studies that restricted
analyses to 24-hour PM2.5
concentrations below 25 mg/m3. These
members also placed weight on results
of controlled human exposure studies
with exposures close to the current
standard, which they note provide
support for the epidemiologic evidence
to lower the standard. These members
noted the limitations in using controlled
human exposure studies alone in
considering the adequacy of the 24-hour
standard, recognizing that controlled
human exposure studies preferentially
recruit less susceptible individuals and
have a typical exposure duration shorter
than 24 hours. These members also
placed ‘‘greater weight on the scientific
evidence than on the values estimated
by the risk assessment,’’ citing their
concerns that the risk assessment ‘‘may
not adequately capture areas with
wintertime stagnation and residential
wood-burning where the annual
standard is less likely to be protective’’
(Sheppard, 2022a, p. 17 of consensus
responses). Furthermore, these CASAC
members ‘‘also are less confident that
the annual standard could adequately
protect against health effects of shortterm exposures’’ (Sheppard, 2022a, p.
17 of consensus responses).
The minority of the CASAC agreed
with the EPA’s preliminary conclusion
in the 2021 draft PA to retain the
current primary 24-hour PM2.5 standard.
In so doing, the minority of the CASAC
placed greater weight on the risk
assessment, noting that the risk
E:\FR\FM\06MRR3.SGM
06MRR3
16254
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
assessment accounts for both the level
and the form of the current standard and
the manner by which attainment with
the standard is determined. Further, the
minority of the CASAC stated that the
‘‘risk assessment indicates that the
annual standard is the controlling
standard across most of the urban study
areas evaluated and revising the level of
the 24-hour standard is estimated to
have minimal impact on the PM2.5associated risks’’ and therefore, ‘‘the
annual standard can be used to limit
both long- and short-term PM2.5
concentrations’’ (Sheppard, 2022a, p. 18
of consensus responses). Further, the
minority of the CASAC placed more
weight on the controlled human
exposure studies, which show ‘‘effects
at PM2.5 concentrations well above those
typically measured in areas meeting the
current standards’’ and which suggest
that ‘‘the current standards are
providing adequate protection against
these exposures’’ (Sheppard, 2022a, p.
18 of consensus responses).
While the CASAC members expressed
differing opinions on the appropriate
revisions to the current standards, they
did ‘‘find that both primary standards,
24-hour and annual, are critical to
protect public health given the evidence
on detrimental health outcomes at both
short-term and long-term exposures
including peak events’’ (Sheppard,
2022a, p. 13 of consensus responses).
The comments from the CASAC also
took note of uncertainties that remain in
this reconsideration of the primary
PM2.5 standards and they identified a
number of additional areas for future
research and data gathering and
dissemination that would inform future
reviews of the primary PM2.5 NAAQS
(Sheppard, 2022a, pp. 14–15 of
consensus responses).
2. Basis for the Proposed Decision
In reaching his proposed decisions to
revise the level of the primary annual
PM2.5 standard from its current level of
12.0 mg/m3 to within the range of 9.0 to
10.0 mg/m3, and to retain the current
primary 24-hour PM2.5 standard (88 FR
5558, January 27, 2023), the
Administrator carefully considered the
assessment of the current evidence and
conclusions reached in the 2019 ISA
and ISA Supplement; the currently
available exposure and risk information,
including associated limitations and
uncertainties, described in detail in the
2022 PA; the considerations and staff
conclusions and associated rationales
presented in the 2022 PA; the advice
and recommendations from the CASAC;
and public comments that had been
offered up to that point (88 FR 5558,
January 27, 2023).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
In reaching his proposed conclusions
on whether the currently available
scientific evidence and quantitative
risk-based information support or call
into question the adequacy of the public
health protection afforded by the
current primary PM2.5 standards, and as
is the case with NAAQS reviews in
general, the extent to which the current
primary PM2.5 standards are judged to
be adequate will depend on a variety of
factors, including science policy and
public health policy judgments to be
made by the Administrator on the
strength and uncertainties of the
scientific evidence. The factors relevant
to judging the adequacy of the standards
also include the interpretation of, and
decisions as to the weight to place on,
different aspects of the results of the risk
assessment for the study areas included
and the associated uncertainties. Thus,
in reaching proposed conclusions of the
current standards, the Administrator
recognized that such a determination
depends in part on judgments regarding
aspects of the evidence and risk
estimates, and judgments about the
degree of protection that is requisite to
protect public health with an adequate
margin of safety.
The Administrator’s full rationale for
his proposed conclusions is presented
in section II.D.3 of proposal (88 FR
5658, January 27, 2023), but is also
briefly summarized here. In reaching the
proposed decision to revise the annual
standard level to 9–10 mg/m3, the
Administrator placed weight on the full
body of scientific information. He noted
that the 2019 ISA finds that exposure to
PM2.5 causes mortality and
cardiovascular effects and is likely to
cause respiratory effects, cancer, and
nervous system effects as detailed
further in section II.B.1 of the proposal.
As detailed further in section II.B.4 of
the proposal, he additionally noted that
the 2019 ISA identifies at-risk
populations at greater risk of health
effects from exposure to PM2.5,
including children, older adults, people
with pre-existing respiratory or
cardiovascular disease, minority
populations, and low socioeconomic
status (SES) populations.
The Administrator also recognized
that epidemiologic studies provide the
strongest scientific evidence when
evaluating the adequacy of the level of
the annual standard. He noted that there
is no specific point in the air quality
distribution of any epidemiologic study
that represents a ‘bright line’ at and
above which effects have been observed
and below which effects have not been
observed. In his proposed decision, he
noted previous decision-making
frameworks, which placed weight on
PO 00000
Frm 00054
Fmt 4701
Sfmt 4700
values at or near the study-reported
mean PM2.5 concentrations, which is
where the most confidence in the
reported association of the
epidemiologic study exists. He further
noted that there are a number of
epidemiologic studies available in this
reconsideration that use new PM2.5
exposure estimation techniques (e.g.,
hybrid modeling) that were not used in
epidemiologic studies that were
available in previous reviews. These
recent epidemiologic studies that use
new exposure estimation techniques
report long-term mean PM2.5
concentrations that are well below
corresponding design values, which is
an important consideration in reaching
decisions on the level of the annual
PM2.5 standard.
In reaching his proposed decision, the
Administrator noted that a level of 9–10
mg/m3 would near or below the reported
25th percentiles in key U.S. based
epidemiologic studies, while also
recognizing that he has less confidence
in the magnitude and significance of the
association at even lower percentiles
(e.g., 10th percentile), where even fewer
health events are observed. The
Administrator also noted that a
proposed level of 9–10 mg/m3 would be
near the mean PM2.5 reported in
Canadian based studies, though he also
recognized that there are a number of
factors associated with the studies in
Canada (e.g., exposure environments)
that make it more difficult to compare
mean concnetrations from Canadian
studies to design values, which
determine compliance with the standard
in the U.S.
The Administrator took note of
additional pieces of scientific evidence,
which were not available in previous
reviews, including restricted analyses,
which support that the association seen
in epidemiologic studies does not just
occur from the peaks of the exposure
distribution. Additionally, he notes that
a level of 9–10 mg/m3 would be below
the starting concentration in newly
available accountability studies, though
he did note that it is more difficult to
interpret these studies in the context of
selecting the level of the annual PM2.5
standard.
Further, the Administrator took into
consideration the advice of the CASAC,
noting that all members included 10 mg/
m3 in their recommended range, and
that the proposed range of 9–10 mg/m3
for the level of the primary annual PM2.5
standard was within the range
recommended by the majority of the
CASAC.
In reaching the proposed conclusion
of a range between 9–10 mg/m3, the
Administrator noted that a level as high
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
as 11 mg/m3 might not provide an
adequate margin of safety, given that 11
mg/m3 was well above many of the
epidemiologic study-reported mean
PM2.5 concentrations. Additionally, the
Administrator noted the uncertainties
associated with the scientific and
quantitative information supporting a
level as low as 8 mg/m3, which call into
question the potential public health
improvements of a standard below 9 mg/
m3. The Administrator specifically
noted the lack of key U.S. studies with
mean concentrations below 9.3 mg/m3
and he further noted that the risk
assessment suggests that the risk
remaining under a standard of 8 mg/m3
would occur at very low concentrations
(e.g., mainly 7 mg/m3 and below).
As such, the Administrator’s
proposed decision noted that the
current PM2.5 annual standard did not
adequately provide requisite protection
against exposures to PM2.5 and that a
proposed range of 9–10 mg/m3 would
provide an adequate margin of safety.
In his proposed decision to retain the
current primary 24-hour PM2.5 standard
with a level of 35 mg/m3, the
Administrator first considered the
scientific information related to shortterm exposures to PM2.5 and health
effects. He noted that the controlled
human exposure studies are the
strongest line of evidence for informing
his conclusions regarding the adequacy
of the current 24-hour standard. In so
doing, the Administrator recognized
that controlled human exposure studies
are conducted with healthy adult
volunteers and that these studies do not
include individuals who may be at
increased risk of PM2.5-related health
effects (i.e., children, older adults,
people with pre-existing diseases). He
also noted that the effects observed in
the controlled human exposure studies
(e.g., changes in vascular function) are
not effects that are judged to be clearly
adverse. He recognized the most
consistent evidence of effects in these
studies occurs at higher concentrations
(e.g., >120 mg/m3) following 1–5 hour
exposures, and that one study observed
effects at concentrations as low as 38 mg/
m3 following 4-hour exposures.
However, the Administrator reiterated
that these studies do not tell us at
exactly what concentrations an adverse
effect might occur, especially for at-risk
populations. As noted above in section
II.A.2.c, controlled human exposure
studies tend to include generally
healthy adult individuals who are at a
lower risk of experiencing health effects,
and often do not include at-risk
populations (e.g., children, older adults,
or individuals with pre-existing
conditions). As such, the Administrator
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
recognized that these studies are
somewhat limited in their ability to
inform at what concentrations effects
may be elicited in in at-risk populations.
The Administrator also considered air
quality analyses in the 2022 PA that
demonstrate that there will be very few,
if any, days with PM2.5 concentrations at
levels evaluated in controlled human
exposure studies that are associated
with effects in areas that meet the
current primary 24-hour PM2.5 standard.
The Administrator also noted that as,
in previous PM NAAQS reviews, the
protection provided by the suite of
standards (e.g., annual and 24-hour
standards) is evaluated together. He
noted that the annual standard is the
controlling standard in most areas of the
country. He also considered air quality
analyses in the 2022 PA that suggest
that revision of the annual standard to
a level between 9–10 mg/m3 would also
control 24-hour PM2.5 concentrations in
most areas to, or below, 30 mg/m3.
Finally, the Administrator noted the
agreement with the advice from the
minority of CASAC and additionally
noted the limited rationale and evidence
provided by the majority CASAC’s
recommendation to support revision of
the 24-hour standard. As such, the
Administrator proposed to retain the
current 24-hour standard with its level
of 35 mg/m3.
Additionally, the Administrator
proposed to conclude that it is
appropriate to retain all other elements
(i.e., indicator, averaging time, and
form) of the annual and 24-hour
standards.
3. Comments on the Proposed Decision
With respect to the adequacy of the
primary annual PM2.5 standard, a
number of commenters, primarily those
from industry and industry groups, nongovernmental organizations, and some
State and local governments, disagree
with the EPA’s proposed decision to
revise the level of the primary annual
PM2.5 standard. These commenters
generally expressed the view that the
current standards provide the requisite
degree of public health protection and
should be retained, consistent with the
2020 final decision. In supporting their
view, these commenters assert that the
scientific evidence available in this
reconsideration is essentially
unchanged since the 2020 final decision
and that the additional scientific
evidence and quantitative risk
information available for the
reconsideration does not support
strengthening the primary annual PM2.5
standard. These commenters also assert
that uncertainties associated with the
available scientific evidence have not
PO 00000
Frm 00055
Fmt 4701
Sfmt 4700
16255
changed since the 2020 final decision,
and they note that these uncertainties
were essential factors in the thenAdministrator’s decision to retain the
primary annual PM2.5 standard. These
commenters argue that, while the
current Administrator acknowledges
these uncertainties, he does not place
enough weight on them in reaching his
conclusions regarding the current
standard. The commenters specifically
highlight uncertainties related to
exposure misclassification,
confounding, and other sources of
potential bias, which they claim
supports retaining the current level of
the annual standard. These commenters
also note that these uncertainties were
emphasized by the minority of the
CASAC in their review of the 2021 draft
PA, and the commenters further suggest
that the lack of consensus from the
CASAC on the appropriate level for the
primary annual PM2.5 standard show
that the research is unclear. The
commenters contend that there is not
support in this reconsideration for
deviating from the then-Administrator’s
decision in 2020.
In contrast, other commenters,
primarily from public health and
environmental organizations, some State
and local elected representatives, and
some State and local government
agencies agree with the EPA’s proposed
decision that the primary annual PM2.5
standard is not adequate. These
commenters support revising the level
of the primary annual PM2.5 standard
and emphasize that the available
scientific evidence, in particular
epidemiologic studies, along with the
CASAC’s advice in their review for the
2021 draft PA, provide strong support
for the proposed decision. In particular,
these commenters agree with the EPA’s
conclusions about the strength of the
scientific evidence, including
uncertainties, and they emphasize that
the CASAC reached consensus in their
review of the 2021 draft PA that the
current primary annual PM2.5 standard
is not adequate. Some of these
commenters also note that a revised
primary annual PM2.5 standard would
result in significant public health
benefits by reducing morbidity and
mortality associated with PM2.5
exposure, especially for at-risk
populations.
The EPA agrees with commenters that
the primary annual PM2.5 standard is
not adequate. The EPA recognizes the
longstanding body of health evidence
supporting relationships between PM2.5
exposures (short- and long-term) and
both mortality and serious morbidity
effects. The evidence available in this
reconsideration (i.e., the studies
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16256
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
assessed in the 2019 ISA and ISA
Supplement summarized above in
section II.A.2.a) reaffirms, and in some
cases strengthens, the conclusions from
the 2009 ISA regarding the health effects
of PM2.5 exposures. As noted above,
epidemiologic studies demonstrate
generally positive and often statistically
significant associations between PM2.5
exposures and health effects. Such
studies report associations between
estimated PM2.5 exposures and nonaccidental, cardiovascular, or
respiratory mortality; cardiovascular or
respiratory hospitalizations or
emergency room visits; and other
mortality/morbidity outcomes (e.g., lung
cancer mortality or incidence, asthma
development). Recent experimental
evidence, as well as evidence from
epidemiologic panel studies,
strengthens support for potential
biological pathways through which
PM2.5 exposures could lead to the
serious effects reported in many
population-level epidemiologic studies,
including support for pathways that
could lead to cardiovascular,
respiratory, nervous system, and cancerrelated effects. Moreover, these recent
epidemiologic studies strengthen
support for health effect associations at
PM2.5 concentrations lower than in
those evaluated in epidemiologic
studies available at the time of previous
reviews.
Additionally, as discussed in more
detail in section I.C.5.b above, the ISA
Supplement focused on studies that
were most likely to inform decisions on
the appropriate standard, but not to
reassess areas that, based on the
assessment of available science
published since the cutoff date of the
2019 ISA and through 2021, were
judged unlikely to have new
information that would be useful for the
Administrator’s decision making. The
ISA Supplement included U.S. and
Canadian epidemiologic studies for
health effect categories where the 2019
ISA concluded a causal relationship
(i.e., short- and long-term PM2.5
exposure and cardiovascular effects and
mortality), as well as U.S. and Canadian
epidemiologic studies that employed
alternative methods for confounder
control or conducted accountability
analyses (i.e., studies that examined the
effect of a policy on reducing PM2.5
concentrations). These studies,
summarized in section II.A.2.a above,
examine both short- and long-term PM2.5
exposure and cardiovascular effects and
mortality. Additionally, studies that
employ alternative methods for
confounder control, as described in
II.A.2.a above and in Table 3–11 and of
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
the 2022 PA (U.S. EPA, 2022b), use a
variety of statistical methods to control
for confounding bias. These studies
consistently report positive associations,
which further supports the broader body
of epidemiologic evidence for both
cardiovascular effects and mortality.
In addition, there are epidemiologic
studies that provide supplemental
information for consideration in
reaching conclusions that the current
suite of PM2.5 standards is not adequate.
These studies include analyses that
restrict annual average PM2.5
concentrations to concentrations below
12 mg/m3 and provide support for
positive and statistically significant
associations with mortality and
cardiovascular morbidity at mean PM2.5
concentrations below the current level
of the primary annual PM2.5 standard
(described above in section II.A.2.c.ii
and in Table 3–10 of the 2022 PA (U.S.
EPA, 2022b)). Recent accountability
studies that have starting annual PM2.5
concentrations at or below 12 mg/m3
suggest public health improvements
may occur at concentrations below 12
mg/m3. These studies indicate positive
and statistically significant associations
with mortality and morbidity (e.g.,
cardiovascular hospital admissions) and
reductions in PM2.5 concentrations in
ambient air (described above in section
II.A.2.c.ii and in Table 3–12 of the 2022
PA (U.S. EPA, 2022b)).
Thus, in considering the available
scientific evidence to inform
conclusions on the adequacy of the
primary PM2.5 standards, the
Administrator recognizes that the 2019
ISA and the ISA Supplement together
provides a strong scientific foundation
for concluding that the current primary
PM2.5 standards are not adequate.
In addition to the scientific evidence
above, the risk assessment estimates that
the current primary annual PM2.5
standard could allow a substantial
number of deaths in the U.S. Although
the Administrator recognizes that while
the risk estimates can help to place the
evidence for specific health effects into
a broader public health context, they
should be considered along with the
inherent uncertainties and limitations of
such analyses when informing
judgments about the potential for
additional public health protection
associated with PM2.5 exposures and
related health effects. The Administrator
takes into consideration these
uncertainties, which are described in
more detail in section II.A.3.b above, but
notes that the general magnitude of risk
estimates supports the potential for
significant public health impacts,
particularly for lower alternative annual
standard levels.
PO 00000
Frm 00056
Fmt 4701
Sfmt 4700
In the CASAC’s review of the 2019
draft PA, the CASAC did not reach
consensus on whether the current
annual standard is adequate, with the
majority of the CASAC recommending
that the annual standard be retained and
the minority of the CASAC
recommending that the standard be
revised. In their review of the 2021 draft
PA, the CASAC unanimously
recommended that the current annual
standard is not sufficiently protective of
public health (Sheppard, 2022a, p. 2 of
consensus letter).
The EPA disagrees with the
commenters who state that the available
scientific and quantitative information
available in this reconsideration does
not provide support for the current
Administrator to reach a different
decision than the then-Administrator
reached in the 2020 final action. The
EPA agrees with these commenters that
there are uncertainties associated with
the currently available scientific
evidence. The EPA has considered these
uncertainties extensively both in
reaching conclusions in the 2022 PA
(U.S. EPA, 2022b, sections 3.4.3, 3.6.1,
and 4.6.3) and in the proposal (88 FR
5604, 5609, January 27, 2023), and the
EPA addresses more detailed public
comments about these uncertainties,
including those related to copollutant
confounding, unmeasured confounding,
and temporal and spatiotemporal
confounding, in the Response to
Comments document. However, we
disagree with the commenters that the
evidence does not provide support for
the Administrator’s conclusion that the
current primary annual PM2.5 standard
is not adequate to protect public health
with an adequate margin of safety, and
should be revised. As described above,
epidemiologic studies in the 2019 ISA
and the ISA Supplement support and
extend the evidence evaluated in the
2009 ISA, through studies conducted in
diverse populations and geographic
locations, using various statistical
models and approaches to control for
potential confounders, and using a
variety of exposure assessment
methodologies. Therefore, the
consistent, positive associations
reported across studies (U.S. EPA,
2019a, Figures 11–1 and 11–18; U.S.
EPA, 2022a) are unlikely to be to be the
result of unmeasured confounding and
other biases are unlikely to account for
the consistent positive associations
observed across epidemiologic studies.
Additionally, this reconsideration
includes epidemiologic studies that
were not before the then-Administrator
for consideration in reaching his final
decisions at the time of the 2020
decision and that specifically evaluate
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
confounding using alternative methods
for confounder control). These recent
epidemiologic studies provide support
for the current Administrator’s
conclusion that the suite of primary
PM2.5 standards are not adequate. While
confounding was an uncertainty noted
by the then-Administrator in the 2020
decision, he recognized ‘‘that
methodological study designs to address
confounding, such as causal inference
methods, are an emerging field of
study’’ (85 FR 82710, December 18,
2020). The ISA Supplement considered
studies that employed statistical
approaches that attempt to more
extensively account for confounders and
are more robust to model
misspecification (i.e., used alternative
methods for confounder control),92
given that such studies were highlighted
by the CASAC in their review of the
2019 draft PA and identified in public
comments on the 2020 proposal. Since
the literature cutoff date for the 2019
ISA, multiple studies that employ
alternative methods for confounder
control have become available for
consideration in the ISA Supplement
and, subsequently, in this
reconsideration. For example, one study
before the Administrator in this
reconsideration that was not available in
the 2019 ISA is Schwartz et al. (2021),
which used a causal modeling approach
focused on exposure changes and
controls for measured confounders by
design in order to evaluate the
association between long-term PM2.5
exposure and mortality in the Medicare
population. The study authors found
significant associations of PM2.5 with
increased mortality rates using a causal
modeling approach robust to omitted
confounding. The results of this study
and other studies in the ISA
Supplement that employ alternative
methods to control for confounders lend
support to the robustness of positive
associations between PM2.5 exposure
and multiple morbidity and mortality
endpoints exhibited across
epidemiologic studies, and also indicate
that unmeasured confounding and other
biases are unlikely to account for the
92 As noted in the ISA Supplement: ‘‘In the peerreviewed literature, these epidemiologic studies are
often referred to as causal inference studies or
studies that used causal modeling methods. For the
purposes of this Supplement, this terminology is
not used to prevent confusion with the main
scientific conclusions (i.e., the causality
determinations) presented within an ISA. In
addition, as is consistent with the weight-ofevidence framework used within ISAs and
discussed in the Preamble to the Integrated Science
Assessments, an individual study on its own cannot
inform causality, but instead represents a piece of
the overall body of evidence’’ (U.S. EPA, 2022a, p.
1–3).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
consistent positive associations
observed across epidemiologic studies
(U.S. EPA, 2022b, sections 3.1.1.3,
3.1.2.3, 3.2.1.3, and 3.2.2.3).
Further, the EPA disagrees with the
commenters who argue that the
Administrator did not appropriately
consider the strengths and limitations of
the health evidence in reaching his
decision to revise the current primary
annual PM2.5 standard in this
reconsideration. In reaching his
proposed decision, the Administrator
considered the entire body of evidence
and how to appropriately weigh the
uncertainties associated with the health
evidence (88 FR 5617, January 27,
2023). Such an approach is consistent
with setting standards that are neither
more nor less stringent than necessary,
recognizing that ‘‘Congress provided
that the Administrator is to use his
judgment in setting air quality standards
precisely to permit him to act in the face
of uncertainty,’’ the Administrator must
set standards on ‘‘the frontiers of
scientific and medical knowledge’’ and
‘‘Congress directed the Administrator to
err on the side of caution in making the
necessary decisions.’’ Lead Indus. Ass’n,
Inc. v. EPA, 647 F.2d 1130, 1155 & n.50
(D.C. Cir. 1980) (quoting H.R. Rep. No.
95–294, at 50). As such, a determination
of identifying a specific level at which
the standard should be set necessarily
requires the Administrator’s judgement
(e.g., weighing the uncertainties and
margin of safety).
Additionally, the EPA disagrees with
the commenters that contend that there
is no basis in this reconsideration for
deviating from the previous
Administrator’s decision in 2020. It is
well-established that in CAA section
109 Congress specifically left the
determination of the requisite NAAQS
to the judgment of the Administrator
and, moreover, that ‘‘decisions about the
appropriate NAAQS level must
‘necessarily . . . rest largely on policy
judgments.’ ’’ Mississippi v. EPA, 744
F.3d 1344, 1357 (D.C. Cir. 2013)
(quoting Lead Industries Ass’n v. EPA,
647 F.2d 1130, 1147 (D.C. Cir. 1980)).
As the Court of Appeals for the D.C.
Circuit has noted, ‘‘Every time EPA
reviews a NAAQS, it (presumably) does
so against contemporary policy
judgments and the existing corpus of
scientific knowledge.’’ Id., at 1343.
In this reconsideration, both the
existing corpus of scientific knowledge
as well as the Administrator’s policy
judgments about how to interpret and
weigh that evidence to protect public
health with an adequate margin of safety
have changed. The expansion of the air
quality criteria to encompass additional
studies, information and analyses in the
PO 00000
Frm 00057
Fmt 4701
Sfmt 4700
16257
ISA Supplement and 2022 PA, as well
as the additional consideration of the
scientific record by the CASAC and the
public provided the Administrator with
significant additional information on
which to base his decision.93 In
addition, in this reconsideration, the
Administrator is reaching different
judgments about how to weigh the
epidemiologic evidence, including the
uncertainties in the scientific evidence,
and how to ensure an adequate margin
of safety to protect against uncertain
harms, compared to the approach in the
2020 final decision. For example, as
discussed in greater detail above in
section II.A.1 and in the 2020 notice of
final rulemaking (85 FR 82717,
December 18, 2020), in considering the
epidemiologic evidence as part of his
decision to retain the current primary
annual PM2.5 standard in the 2020
decision, the then-Administrator placed
weight on the mean of the studyreported means (or medians) (i.e., 13.5
mg/m3) from key U.S. epidemiologic
studies that are monitor-based being
above the level of the current primary
annual PM2.5 standard of 12.0 mg/m3. By
contrast, in this reconsideration, the
current Administrator has taken an
approach more similar to how the EPA
has considered study-reported mean
PM2.5 concentrations relative to the level
of the primary annual PM2.5 standard in
other recent PM NAAQS reviews. In so
doing, in reaching his decision to revise
the level of the primary annual PM2.5
standard to 9.0 mg/m3, he is using an
approach that places weight on selecting
a level for the standard that is below the
study-reported mean PM2.5
concentrations reported in key U.S.
epidemiologic studies, including recent
epidemiologic studies that use hybrid
model-based methods, as well as being
near or below the 25th percentile PM2.5
concentrations in those key U.S.
epidemiologic studies that report these
concentrations.
As such and further detailed in
section II.B.4 below, in considering the
adequacy of the current primary PM
standards in this reconsideration, the
Administrator has carefully considered
the: (1) Policy-relevant evidence and
conclusions contained in the 2019 ISA
and 2022 ISA Supplement; (2) the
quantitative information presented and
93 The EPA notes that, in considering the
additional scientific evidence available in this
reconsideration, one member of the CASAC who
reviewed both the 2019 draft PA and the 2021 draft
PA found that the available scientific and
quantitative information available in this
reconsideration supported revising the level of the
primary annual PM2.5 standard, whereas he
recommended retaining the standard during the
review of the 2019 draft PA.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16258
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
assessed in the 2022 PA; (3) the
evaluation of this evidence, the
quantitative information, and the
rationale and conclusions presented in
the 2022 PA; (4) the advice and
recommendations from the CASAC; and
(5) public comments. The Administrator
concludes that the current suite of
primary PM2.5 standards are not
adequate to protect public health with
an adequate margin of safety.
The four basic elements of the
NAAQS (indicator, averaging time,
form, and level) are considered
collectively in evaluating the health
protection afforded by a standard. The
EPA received relatively few comments
on the averaging time and form for the
primary PM2.5 standards, but those who
did provide comments on these
elements were primarily from public
health and environmental organizations,
State and local elected representatives,
and State and local government
agencies. Some commenters assert that
the current 24-hour averaging time for
the primary 24-hour PM2.5 standard
does not adequately protect against
short-term peaks. These commenters
further state that the 24-hour averaging
time protects against chronic exposures
but does not adequately protect against
serious acute risks from certain sources
such as prescribed burning. Also, a few
commenters explicitly recommend that
a subdaily averaging time would be
more appropriate, although none of the
commenters recommended a specific
averaging time for consideration.
Additionally, some commenters cite to
the CASAC’s advice in their review of
the 2021 draft PA that future reviews of
the PM NAAQS should include
evaluation of alternative forms and
averaging times of the current primary
24-hour PM2.5 standard.
The EPA disagrees with commenters
that the current primary 24-hour PM2.5
standard, with its 24-hour averaging
time, does not adequately protect
against short-term peaks and disagrees
that that there is sufficient information
to conclude that a subdaily averaging
time would be more appropriate than a
24-hour averaging time. The EPA has
reviewed the currently available
scientific evidence and finds that it does
not indicate that alternative averaging
times would be more appropriate for the
primary PM2.5 standards. Accordingly,
the EPA concludes that it is appropriate
to retain both the annual and 24-hour
averaging times for standards meant to
protect against long- and short-term
PM2.5.
As noted in the proposal, the 2019
ISA and ISA Supplement found that the
scientific evidence continues to provide
strong support for health effect
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
associations with both long-term (e.g.,
annual or multi-year) and short-term
(e.g., mostly 24-hour) exposures to
PM2.5. Epidemiologic studies continue
to provide strong support for health
effects associated with short-term PM2.5
exposures based on 24-hour PM2.5
averaging periods, and we note that
subdaily effect estimates are less
consistent and, in some cases, smaller in
magnitude (88 FR 5618, January 27,
2023). Controlled human exposure and
panel-based studies of subdaily
exposures typically examine subclinical
effects rather than the more serious
population-level effects that have been
reported to be associated with 24-hour
exposures (e.g., mortality,
hospitalizations). Collectively, the 2019
ISA concludes that epidemiologic
studies do not indicate that subdaily
averaging periods are more closely
associated with health effects than the
24-hour average exposure metric (U.S.
EPA, 2019a, section 1.5.2.1).
Additionally, the EPA notes that while
recent controlled human exposure
studies provide consistent evidence for
cardiovascular effects following PM2.5
exposures for less than 24 hours (i.e.,
<30 minutes to 5 hours), exposure
concentrations in these studies are wellabove the ambient concentrations
typically measured in locations meeting
the current standards (U.S. EPA, 2022a,
section 3.3.3.1). Therefore, this
information does not indicate that a
revision to the averaging time is needed
to provide additional protection against
subdaily PM2.5 exposures, beyond that
provided by the current primary
standards. This conclusion is also
supported by the advice given to EPA by
the CASAC in their review of the 2021
draft PA, which reached consensus that
averaging times for the standards should
be retained, without revision (Sheppard,
2022a, p. 2 of consensus letter).94 For all
of these reasons, the Administrator
concludes that the currently available
evidence does not support considering
alternatives to the annual and 24-hour
averaging times for standards meant to
protect against long- and short-term
PM2.5 exposures.
Multiple commenters, primarily from
public health and environmental
organizations, recommend revising the
form of the primary 24-hour PM2.5
standard to a 99th percentile to provide
increased public health protection
against peak PM2.5 exposures,
94 In providing advice on the 2019 draft PA, the
CASAC did not weigh in specifically on the
averaging time of the primary 24-hour PM2.5
standard but did recommend that the standard be
retained because the available evidence does not
call into question its adequacy (Cox, 2019b, p. 3 of
consensus letter).
PO 00000
Frm 00058
Fmt 4701
Sfmt 4700
particularly for at-risk populations.
These commenters express concern that
the current 98th percentile form allows
7 exceedances per year and contend that
a 99th percentile form that would allow
half that number is more appropriate.
Commenters also cite to the CASAC’s
advice in their review of the 2021 draft
PA, which recommended that the EPA
consider alternative percentiles for the
form of the primary 24-hour PM2.5
standard in the future.
The EPA disagrees that the current
98th percentile form does not provide
the requisite public health protection
against peak PM2.5 exposures and
concludes that the 98th percentile,
averaged over three years, remains
appropriate for the primary 24-hour
PM2.5 standard. As noted in previous
reviews and in the proposal, the EPA
has set both an annual standard and a
24-hour standard to provide protection
from health effects associated with both
long- and short-term exposures to PM2.5
(62 FR 38667, July 18, 1997; 88 FR 5620,
January 27, 2023). With respect to the
form of the 24-hour standard, as
described just above, the epidemiologic
studies continue to provide strong
support for health effect associations
with short-term (e.g., mostly 24-hour)
PM2.5 exposures and controlled human
exposure studies provide evidence for
health effects following single shortterm ‘‘peak’’ PM2.5 exposures (88 FR
5619, January 27, 2023). Both the 98th
and the 99th percentile form provide a
very high degree of control of peak
concentrations. As the commenters
point out, a 99th percentile would
reduce the number of allowable
exceedances to four days per year. The
EPA anticipates, however, that such a
revision to the form would make the
attainment status of an area more
subject to change from unpredictable
nonanthropogenic factors, such as
meteorological events. The EPA has
often noted that frequent shifts in
attainment status that are unrelated to
long-term air quality trends is
inconsistent with providing a stable
target for air quality planning and risk
management programs, which in turn
provides for the most effective public
health protection in the long run (78 FR
3127, January 15, 2013; 80 FR 65351,
October 26, 2015). Thus, the EPA’s
interest in an appropriate degree of
stability is to ensure that the State air
quality programs are effective in
controlling pollution and that the public
health protections of the standard are
achieved. As discussed above, while
recent controlled human exposure
studies provide consistent evidence for
cardiovascular effects following PM2.5
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
exposures for less than 24 hours (i.e., <
30 minutes to 5 hours), exposure
concentrations in these studies are wellabove the ambient concentrations
typically measured in locations meeting
the current standards (U.S. EPA, 2022a,
section 3.3.3.1), and the 98th percentile
form is very effective at limiting
occurrences of exposures of concern.
Taking into consideration the available
scientific information and quantitative
information, the EPA therefore
concludes that the 98th percentile form
provides an appropriate balance
between limiting the occurrence of peak
24-hour PM2.5 concentrations and
identifying a stable target for risk
management programs. This conclusion
is also supported by the advice given to
the EPA by the CASAC in their review
of the 2021 draft PA, where they
reached consensus that the form for the
standards should be retained, without
revision (Sheppard, 2022a, p. 2 of
consensus letter).95
Additionally, the EPA recognizes the
CASAC’s advice in their review of the
2021 draft PA, where they
recommended ‘‘that in future reviews,
the EPA provide a more comprehensive
assessment of the 24-hour standard that
includes the form as well as the level’’
(Sheppard, 2022a, p. 4 of consensus
letter). This advice is reflected in the
proposal by the EPA, which noted ‘‘that
it would be appropriate to gather
additional air quality and scientific
information and further consider these
issues in future reviews’’ (88 FR 5619,
January 27, 2023). The EPA will
consider the information provided by
the commenters regarding the form of
the 24-hour PM2.5 standard in the next
review of the PM NAAQS.
A number of commenters who
support revising the level of the primary
annual PM2.5 standard, particularly
those who support a revised level of 8
mg/m3, disagree with how the EPA has
emphasized the mean PM2.5
concentrations reported in key
epidemiologic studies to inform
conclusions on the level of the primary
PM2.5 standard. These commenters
argue that, in this reconsideration, the
EPA is arbitrarily emphasizing
uncertainties in key epidemiologic
studies in the focus on mean
concentrations. Many of these
commenters recommend that the EPA
consider the full distribution of PM2.5
concentrations from the key
epidemiologic studies in reaching
conclusions on the appropriate level for
95 The CASAC did not provide advice or
recommendations regarding the forms of the
primary PM2.5 standards in their review of the 2019
draft PA (Cox, 2019b).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
the primary annual PM2.5 standards, in
particular concentrations below the
mean, such as the 25th percentile. In
supporting this view, commenters point
to the CASAC’s advice in their review
of the 2021 draft PA, where the majority
of the CASAC stated that the ‘‘use of the
mean to define where the data provide
the most evidence is conservative since
robust data clearly indicate effects
below the mean in concentrationresponse functions’’ (Sheppard, 2022a,
p. 16 of consensus responses), and that
‘‘[e]pidemiologic studies require
consideration of distribution around the
mean of exposure to identify effects and
thus lower levels than the mean must be
considered as part of the range where
the data provide higher confidence’’
(Sheppard, 2022a, p. 13 of consensus
responses).
As an initial matter, consistent with
some previous approaches and as
detailed by the Administrator in
reaching conclusions on the level of the
primary annual PM2.5 standard in
section II.B.4 below, the EPA considers
the long-term study-reported mean
PM2.5 concentrations from key
epidemiologic studies and sets the level
of the standard to somewhat below the
lowest long-term mean PM2.5
concentration. Additionally, as
discussed further below, the EPA also
considers the available information from
a subset of epidemiologic studies that
report exposure estimates or health
events at the 25th and 10th percentiles
of PM2.5 concentrations. The
Administrator gives some weight to the
25th percentile data, although he
recognizes that his confidence in the
magnitude and significance in the
reported concentrations, and their
ability to inform decisions on the
appropriate level of the annual
standard, decreases with reduced data
(below the mean) and diminishes
further at percentiles that are even
further below the mean and the 25th
percentile. Therefore, the Administrator
places weight on the reported 25th
percentiles concentrations, rather than
the reported 10th percentile
concentrations, for the subset of studies
that report lower percentile PM2.5
concentrations in reaching his
conclusions regarding the appropriate
level for the primary annual PM2.5
standard.
In considering the available scientific
evidence to reach decisions on the
adequacy of the suite of primary PM2.5
standards, the EPA notes that in
previous PM NAAQS reviews
(including the 1997, 2006 and 2012
reviews), evidence-based approaches
were used that focused on identifying
standard levels near or somewhat below
PO 00000
Frm 00059
Fmt 4701
Sfmt 4700
16259
long-term mean concentrations reported
in key epidemiologic studies. These
approaches were supported by the
CASAC in previous reviews and were
supported in this reconsideration by the
CASAC in their review of the 2021 draft
PA.96
In considering the available scientific
evidence, the EPA notes the strength of
the epidemiologic evidence which
includes multiple studies that
consistently report positive associations
for short- and long-term PM2.5 exposures
and mortality and cardiovascular
effects. Some available studies also use
a variety of statistical methods to
control for confounding bias and report
similar associations, which further
supports the broader body of
epidemiologic evidence for both
mortality and cardiovascular effects.
Additionally, the EPA notes that recent
epidemiologic studies strengthen
support for health effect associations at
PM2.5 concentrations lower than in
those evaluated in epidemiologic
studies available at the time of previous
reviews.
While these epidemiologic studies
evaluate associations between
distributions of ambient PM2.5
concentrations and health outcomes,
they do not identify the specific
exposures that led to the reported
effects. As such, there is no specific
point in the air quality distribution of
any epidemiologic study that represents
a ‘‘bright line’’ at and above which
effects have been observed and below
which effects have not been observed.
Studies of daily PM2.5 exposures
examine associations between day-today variation in PM2.5 concentrations
and health outcomes, often over several
years. While there can be considerable
variability in daily exposures over a
multi-year study period, most of the
estimated exposures reflect days with
96 The Administrator notes that, in their review of
the 2021 draft PA, a majority of members of the
CASAC noted that there are some limitations for
this approach ‘‘for the purpose of informing the
adequacy of the standards’’ (Sheppard, 2022a, p. 8
of consensus responses) and advised that future
reviews should include evaluation of other metrics,
including the distribution of concentrations
reported in epidemiologic studies and in analyses
restricting concentrations to below the current
standard level. The Administrator also notes that,
in their review of the 2019 draft PA, the CASAC
lacked consensus on the inferences to be drawn
from the epidemiologic evidence, with a majority of
CASAC having concerns about confounding, error
and bias and concluding that newer studies did not
provide a basis for revising the current standards,
while a minority concluded that the evidence,
including more recent studies showing associations
in areas with average long-term PM2.5
concentrations below the current annual standard,
supported their conclusion that the current
standards are inadequate (Cox, 2019b, pp. 8–9 of
consensus responses).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16260
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ambient PM2.5 concentrations around
the middle of the air quality
distributions examined (i.e., ‘‘typical’’
days rather than days with extremely
high or extremely low concentrations).
Similarly, for studies of annual PM2.5
exposures, most of the health events
occur at estimated exposures that reflect
annual average PM2.5 concentrations
around the middle of the air quality
distributions examined. In both cases,
epidemiologic studies provide the
strongest support for reported health
effect associations for this middle
portion of the PM2.5 air quality
distribution, which corresponds to the
bulk of the underlying data, rather than
the extreme upper or lower ends of the
distribution. Therefore, in the absence
of discernible thresholds, long-term
study-reported means—that is, the
study-reported ambient PM2.5
concentrations in the epidemiologic
studies that reflect estimated exposures
with a focus around the middle portion
of the PM2.5 air quality distribution
where the bulk of the observed data
reside—provide the strongest support
for reported health effect associations in
epidemiologic studies.
Based on the air quality criteria for
this reconsideration, as described in the
2019 ISA, ISA Supplement, 2022 PA
and the proposal, the EPA believes it is
appropriate to continue to use the mean
PM2.5 concentrations from the key
epidemiologic studies to inform
conclusions regarding the appropriate
level for the primary annual PM2.5
standard.
There are a large number of key
epidemiologic studies available in this
reconsideration to inform conclusions
regarding the level of the primary
annual PM2.5 standard. For the key U.S.
epidemiologic studies, the studyreported mean PM2.5 concentrations
range from 9.9–16.5 mg/m3 for monitorbased studies (Figure 1 above) and range
from 9.3–12.2 mg/m3 for hybrid
modeling-based studies (Figure 2
above).
In addition to the study-reported
mean PM2.5 concentrations, the EPA
agrees with the CASAC’s advice in their
review of the 2021 draft PA and public
comments that information on other
percentiles below the mean can also be
informative, and the EPA notes that the
CASAC advised that for the purpose of
informing the adequacy of the
standards, future reviews should
include an evaluation of other metrics,
including the distribution of
concentrations reported in
epidemiologic studies (Sheppard,
2022a, p. 9 of consensus responses). As
such, in reaching conclusions in this
reconsideration, the EPA takes note of
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
the additional study-reported PM2.5
concentrations below the means (e.g.,
25th and 10th percentiles) that are
available from a limited subset of key
U.S. epidemiologic studies. As shown in
Figures 1 and 2 above, six key U.S.
epidemiologic studies report
information on other percentiles (e.g.,
10th and 25th percentiles of PM2.5
concentrations or 10th and 25th
percentiles of PM2.5 concentrations
associated with health events) that are
below the mean.97 Three of the studies
are monitor-based and three are hybrid
model-based.
The key U.S. epidemiologic studies
that report percentiles below the mean
that are monitor based are older studies.
These studies included smaller numbers
of people than the newer hybrid modelbased studies. For the three older,
monitor-based studies, because the
cohorts were smaller in size, a relatively
smaller portion of the health events
were observed in the lower part of the
air quality distribution. As such, our
confidence in the magnitude and
significance of the associations begins to
decrease in the lower part of the air
quality distribution of those older,
monitor-based studies.
The three newer, hybrid model-based
studies have larger cohort sizes than the
older, monitor-based studies and, as
noted by commenters, have more health
events in the lower part of the air
quality distribution. For these reasons,
the EPA notes that we have more
confidence in the reported association at
concentrations lower than the reported
mean in these more recent hybrid
model-based studies, particularly at the
25th percentile compared to the 10th
percentile. While the cohort sizes in the
more recent, hybrid model-based
studies are larger than the older,
monitor-based studies, the EPA notes
that the 10th percentiles are well below
the middle portion of the air quality
distribution for which we have the
greatest confidence, and as noted above,
our confidence in the magnitude and
significance of associations in the lower
parts of the air quality distribution
begins to decrease. While we have more
confidence in the lower percentiles
because of the larger cohort sizes in the
more recent hybrid model-based
studies, we also have more confidence
in the 25th percentiles than in the 10th
percentiles, which are further from the
means and closer to the lower end of the
air quality distribution.
In considering how the six studies
that report percentiles lower than the
97 The Wang et al. (2017) study only reports the
25th percentile of the estimated PM2.5
concentrations, not the 10th percentile.
PO 00000
Frm 00060
Fmt 4701
Sfmt 4700
mean can be used to inform conclusions
regarding the level of the primary
annual PM2.5 standard, the EPA first
notes that the three monitor-based
epidemiologic studies (Bell et al., 2008;
Franklin et al., 2007; Zanobetti and
Schwartz, 2009) report 25th percentile
concentrations that are at or above 11.5
mg/m3. For two of the more recent
hybrid model-based studies (Di et al.,
2017b; Wang et al., 2017), the 25th
percentile of estimated PM2.5
concentrations are just above 9 mg/m3,
while one study (Di et al., 2017a) reports
a PM2.5 concentrations corresponding to
25th percentiles of health events of just
below 7 mg/m3. For the Di et al. (2017a)
study, the 25th percentile PM2.5
concentration (6.7 mg/m3) is based on
the PM2.5 concentration at which the
25th percentile of deaths occur in the
study, while the reported mean (11.6 mg/
m3) is based on estimated PM2.5
exposure concentrations. Additionally,
the 25th percentiles of the other two
recently available hybrid model-based
studies (Di et al., 2017b; Wang et al.,
2017) are based on estimated PM2.5
concentrations. As such, the PM2.5
concentration at which the 25th
percentile of health events occur may be
different from the estimated 25th
percentile PM2.5 concentration in this
study (Di et al., 2017a), creating an
uncertain basis for comparison with the
studies by Di et al. (2017b) and Wang et
al. (2017). The 25th percentiles from
these studies, in particular those that are
more recently available, help to inform
the Administrator’s judgments regarding
the appropriate level for the primary
annual PM2.5 standard.
Some commenters disagree with the
EPA’s consideration of the relationship
between mean PM2.5 concentrations
reported in the key epidemiologic
studies and design values to inform
conclusions on the appropriate level for
the primary annual PM2.5 standards.
Commenters contend that setting the
level of the primary annual standard
below the design values in the
epidemiologic studies, rather than
below the study-reported mean
concentrations, might keep overall mean
PM2.5 concentrations throughout an area
below the study-reported means but
allow PM2.5 concentrations in some
parts of the area, including near the
‘‘design value monitor’’ to remain above
the study-reported mean PM2.5
concentrations, which are the
concentrations where the evidence of
health effects is strongest. Commenters
contend that such a decision framework
would not result in a standard that
would provide requisite protection with
an adequate margin of safety,
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
particularly for at-risk populations.
These commenters further support this
view by citing the CASAC’s advice in
their review of the 2021 draft PA, where
the majority of CASAC stated that ‘‘even
if a design value is somewhat higher
than the area average, it reflects actual
exposure levels and thus any portion of
the population living near the design
value monitor does experience
exposures at that level and consequent
health effects of exposure to that higher
concentration’’ (Sheppard, 2022a, p. 14
of consensus responses). Additionally,
these commenters suggest that the EPA
should not deviate from the approach
taken in the 2012 review, which was to
set the standard at a level ‘‘somewhat
below’’ the lowest mean PM2.5
concentration in the key epidemiologic
studies.
To the extent that commenters are
suggesting that the EPA is setting the
level of the primary annual PM2.5
standard below the design values in the
epidemiologic studies, rather than
below the study-reported mean PM2.5
concentrations, we disagree with the
commenters. In reaching conclusions on
the level of the primary annual PM2.5
standard, the EPA considers the longterm study-reported mean PM2.5
concentrations from key epidemiologic
studies and sets the level of the standard
to somewhat below the lowest long-term
mean PM2.5 concentration, not below
the design values in the epidemiologic
studies. Additionally, the EPA also
considers the available information from
a subset of epidemiologic studies that
report exposure estimates or health
events at the 25th and 10th percentiles
of PM2.5 concentrations. The EPA
particularly considers the 25th
percentile data, while recognizing that
our confidence in the magnitude and
significance in the reported
concentrations, and the ability of the
lower percentile PM2.5 concentrations to
inform decisions on the appropriate
level of the annual standard, decreases
with reduced data (below the mean) and
diminishes further at percentiles that
are even further below the mean and the
25th percentile.
However, the EPA notes that it is
important to understand, and to not
ignore, the relationship between the
study-reported mean PM2.5
concentrations reported in key
epidemiologic studies and the area
design value. As an initial matter, the
NAAQS consists of all four elements of
the standard (indicator, averaging time,
form, and level) and setting a standard
that is requisite to protect public health
includes consideration of all four
elements together. Following
implementation of the NAAQS, the
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
design value is the metric used to
determine compliance with the standard
and is the statistic that describes the air
quality status of a given location relative
to the level of the primary annual PM2.5
NAAQS. The design value is different
from the study-reported mean PM2.5
concentrations. This is because the
study-reported mean PM2.5
concentrations are an annual average
PM2.5 concentration, similar to the level
of the standard, but the epidemiologic
studies do not report statistics that take
into account the other elements of the
standard (i.e., averaging time and form).
Therefore, when considering the
appropriate revisions to the annual
PM2.5 standard, the EPA must consider
the protection provided by a revised
standard taking into account all of the
elements of the standard, not just the
annual average PM2.5 concentration
alone.
In considering the annual standard,
and in assessing the range of studyreported exposure concentrations for
which we have the strongest support for
adverse health effects observed in
epidemiologic studies, the EPA focuses
on whether the current primary annual
PM2.5 standard provides adequate
protection against these exposure
concentrations or if the level of the
standard should be revised to provide
the appropriate public health
protection. This means that, as in some
previous reviews, it is important to
consider how the study means were
computed and how these concentrations
compare to the annual standard metric
(including the level, averaging time and
form) which must be met at the monitor
with the highest PM2.5 design value in
an area for compliance with the
NAAQS. This approach is based on the
application of a decision framework
based on assessing means (as well as the
lower distribution of reported PM2.5
concentration, as noted above) reported
in key epidemiologic studies. In the
2012 review, the available key
epidemiologic studies computed the
mean PM2.5 concentrations using an
average across monitor-based PM2.5
concentrations. As such, at that time,
the decision framework used an
approach based on maximum monitor
concentrations to determine compliance
with the standard, while selecting the
standard level based on consideration of
composite monitor concentrations (i.e.,
selecting the standard level of 12.0 mg/
m3 was just below the long-term studyreported mean PM2.5 concentrations in
key epidemiologic studies). Further, the
EPA conducted analyses that examined
the differences in these two metrics (i.e.,
maximum monitor concentrations,
PO 00000
Frm 00061
Fmt 4701
Sfmt 4700
16261
which is how compliance with the
standard is assessed and composite
monitor concentrations, which is how
key epidemiologic studies report their
mean concentrations) across the U.S.
and in areas included in the key
epidemiologic studies and found that
the maximum design value in an area
was generally higher than the monitor
average across that area, with the
amount of difference between the two
metrics varying based on location and
concentration (Hassett-Sipple et al.,
2010; Frank, 2012). This information
was taken into account by the thenAdministrator’s final decision in
selecting a level of 12.0 mg/m3 for the
primary annual PM2.5 standard in the
2012 review and discussed more
specifically in her considerations on
adequate margin of safety.
The relationship between the mean
PM2.5 concentrations and the area
design value continues to be an
important consideration in evaluating
the adequacy of the current or potential
alternative annual standard levels in
this reconsideration. Again, in a given
area, the area design value is based on
the monitor in an area with the highest
PM2.5 concentrations and is used to
determine compliance with the
standard, including the averaging time
and form of the standard (i.e., an annual
average over 3-years must not exceed
the level of the of the annual PM2.5
standard). The highest PM2.5
concentrations spatially distributed in
the area would generally occur at or
near the area design value monitor and
the distribution of PM2.5 concentrations
would generally be lower in other
locations and at monitors in that area.
As such, when an area is meeting a
specific annual standard level (e.g., 9.0
mg/m3), we would expect the annual
average exposures (i.e., a metric similar
to the study-reported mean values) in
that area to be at concentrations lower
than that level (e.g., lower than 9.0 mg/
m3).
However, as described in section
II.A.2.c.ii, we note that there are a
substantial number of different types of
epidemiologic studies available since
the 2012 review, as assessed in both the
2019 ISA and the ISA Supplement, that
make understanding the relationship
between the mean PM2.5 concentrations
and the area design value an even more
important consideration in this
reconsideration (U.S. EPA, 2019a; U.S.
EPA, 2022a). While the key
epidemiologic studies in the 2012
review were all monitor-based studies,
the recent epidemiologic studies in this
reconsideration include hybrid
modeling approaches that have emerged
in the epidemiologic literature as an
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16262
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
alternative to approaches that only use
ground-based monitors to estimate PM2.5
exposure. As assessed in the 2019 ISA
and ISA Supplement, a substantial
number of epidemiologic studies used
hybrid model-based methods in
evaluating associations between PM2.5
exposure and health effects. Hybrid
model-based studies employ various
fusion techniques that combine groundbased monitored data with air quality
modeled estimates and/or information
from satellites to estimate PM2.5
exposures. While these studies provide
a broader estimation of PM2.5 exposures
compared to monitor-based studies (i.e.,
PM2.5 concentrations are estimated in
areas without monitors), the hybrid
modeling approaches result in studyreported means that are more difficult to
relate to the annual standard metric and
to the maximum monitor design values
used to assess compliance. In addition,
to further complicate the comparison,
when looking across these studies, we
find variations in how exposure is
estimated between such studies, and
thus, how the study means are
calculated. Two important variations
across studies include: (1) Variability in
spatial scale used (i.e., averages
computed across the national (or large
portions of the country) versus a focus
on only CBSAs); and (2) variability in
exposure assignment methods (i.e.,
averaging across all grid cells, averaging
across a scaled-up area like a ZIP code,
and population weighting). The
differences in these approaches can
result in studies reporting different
study means, even though the
association between PM2.5 exposure and
health effects outcomes are similar.
To emphasize the importance of the
differences between the studies, we
revisit the simplified example in the
State of Georgia from the 2022 PA that
evaluates monitors and hybrid modeling
approaches, noting that this example is
useful to exhibit how the differences in
the methods used to estimate exposure
can lead to differences in the reported
mean concentrations (U.S. EPA, 2022b,
p. 3–71). In this example, for all
monitors within the Atlanta-Sandy
Springs-Roswell CBSA, the average
PM2.5 concentration is 9.3 mg/m3, while
the area design value (based on the
highest monitored PM2.5 concentration
in the area) is 10.4 mg/m3. This
comparison helps to illustrate the fact
that composite monitor values tend to
be somewhat lower than the highest
area monitor values, consistent with the
key points made in the 2012 review.
This example also illustrates how
monitors are sited to represent the
higher concentrations within the area
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
and that the area’s annual design value,
which is used for compliance with the
standard, is calculated based on the
highest monitor in the area. Next, in this
example, mean PM2.5 concentrations
were calculated using similar
approaches to those used in hybrid
modeling-based epidemiologic studies
to compute study-reported means,
including (1) the average concentration
across the entire State of Georgia; (2) the
population-weighted average across the
entire State; (3) the average
concentration across the Atlanta-Sandy
Springs-Roswell CBSA; and (4) the
population-weighted average across the
Atlanta-Sandy Springs-Roswell CBSA.
At the urban level (e.g., Atlanta-Sandy
Springs-Roswell CBSA), the average
PM2.5 concentration when taking the
mean of all grid cells is 9.2 mg/m3,
whereas the population-weighted mean
is 9.6 mg/m3. Across Georgia, the average
PM2.5 concentration using the hybrid
approach and averaged across each grid
cell is 8.3 mg/m3, which is lower than
the population-weighted statewide
average of 9.1 mg/m3. While this is a
simple example completed in one State
and one CBSA, it suggests that the
lowest mean values tend to result from
the approaches that use concentrations
from all or most grid cells (e.g., did not
apply population weighting), both urban
and rural, across the study area to
compute the mean. Higher mean values
are observed when the approach focuses
on the urban areas alone or when the
approach incorporates population
weighting. Overall, this example
suggests that the means from studies
using hybrid modeling approaches are
generally lower than the means from
monitor-based approaches, and means
from both approaches are lower than the
annual design values for the same area.
Population weighting tends to increase
the calculated mean concentration,
likely because more densely populated
areas also tend to have higher PM2.5
concentrations. In other words, this
simplified example exhibits how not all
reported mean PM2.5 concentrations
from key epidemiologic studies are the
same; some reported means are from
monitored studies and some reported
means are from hybrid modeling
studies, while some reported means
include only urban areas, and other
reported means include both urban and
rural areas, and some reported means
include aspects of population weighting
while others do not.
As detailed above in section I.D.5, in
the air quality analyses comparing
composite monitored PM2.5
concentrations with annual PM2.5 design
values in U.S. CBSAs, maximum annual
PO 00000
Frm 00062
Fmt 4701
Sfmt 4700
PM2.5 design values were approximately
10% to 20% higher than annual average
composite monitor concentrations (i.e.,
averaged across multiple monitors in
the same CBSA). Based on these results,
this analysis suggests that there will be
a distribution of concentrations and the
maximum annual average monitored
concentration in an area (at the design
value monitor, used for compliance
with the standard), will generally be 10–
20% higher than the average across the
other monitors in the area. Thus, in
considering how the annual standard
levels would relate to the study-reported
means from monitor-based studies, we
can generally conclude that an annual
standard level that is no more than 10–
20% higher than monitor-based studyreported mean PM2.5 concentrations
would generally maintain air quality
exposures to be below those associated
with the study-reported mean PM2.5
concentrations, exposures for which we
have the strongest support for adverse
health effects occurring.
Air quality analyses described in
section I.D.5 above also consider
information from the epidemiologic
studies that utilized the hybrid
modeling approaches. Analyses show
that average maximum annual design
values are 40–50% higher when
compared to annual average PM2.5
concentrations estimated without
population weighting and are 15–18%
higher when compared to average
annual PM2.5 concentrations with
population weighting applied. Given
these results, it is worth noting that for
the studies using the hybrid modeling
approaches, the choice of methodology
employed in calculating the studyreported means (i.e., using population
weighting versus not applying aspects of
population weighting), and not a
difference in estimates of exposure in
the study itself, can produce
substantially different study-reported
mean values, with the approach that
does not employ population weighting
producing a much lower reported mean
PM2.5 concentration. Therefore, the
impact of the differences in methods is
an important consideration when
comparing mean concentrations across
studies.
Because of the differences in the
methods employed by the key
epidemiologic studies, and as
demonstrated by the example and air
quality analyses above, the application
of any decision framework that
considers the study-reported mean
PM2.5 concentrations, and evaluates
whether the current annual standard
provides adequate protection against
these reported exposure concentrations,
is more complicated than the
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
approaches used in past reviews. As
such, the EPA disagrees with
commenters who argue that the EPA’s
consideration of the relationship
between mean PM2.5 concentrations
reported in key epidemiologic studies
and design values is not appropriate and
should be ignored.
In considering the information from
the epidemiologic studies, while the
EPA does not dispute the reported
associations of epidemiologic studies in
hybrid modeling studies that report
long-term mean concentrations and do
not apply aspects of population
weighting, using the reported long-term
mean concentration from these studies
in informing an appropriate level of the
annual PM2.5 standard is more
uncertain. Given this, hybrid modeling
studies that do not apply aspects of
population weighting provide less
information on conclusions regarding
the appropriate level of the primary
annual PM2.5 standard. In support of
this, some commenters also noted this
consideration and suggested that the
Administrator place lower weight on
U.S. studies that did not use population
weighting.
In considering the relationship
between study-reported mean PM2.5
concentrations and the design values,
the EPA agrees with commenters that
setting the level of the primary annual
standard below the design values, rather
than below the study-reported mean
concentrations, might allow PM2.5
concentrations in some part of the area
near the design value monitor to remain
above the study-reported mean
PM2.5concentration, where evidence of
health effects is strongest. As discussed
in the proposal and in section II.B.4
below, the Administrator specifically
notes that that the highest PM2.5
concentrations spatially distributed in
the area would generally occur at or
near the area design value monitor and
that PM2.5 concentrations will be equal
to or lower at other monitors in the area.
Furthermore, since monitoring strategies
aim to site monitors in areas with higher
PM2.5 concentrations, monitored areas
will generally have higher
concentrations compared to areas
without monitors. Therefore, by setting
the level of the standard to 9.0 mg/m3
and just below the lowest studyreported mean PM2.5 concentration (e.g.,
9.3 mg/m3), the highest possible design
value in a given area would be just
below the study-reported mean PM2.5
concentration, the concentration where
we have the most confidence in the
reported health effect association, and
we anticipate that, based on our
assessment of air quality data, the
distribution of PM2.5 concentrations
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
would decrease even further with
distance from the highest monitor (i.e.,
the ‘‘design value monitor’’) (see, for
example, U.S. EPA, 2022a, section
2.3.3.2.4 and pp. 3–71 to 3–77). The
Administrator further notes that when
an epidemiologic study reports a mean
PM2.5 concentration that reflects the
average of annual average monitor-based
concentrations across an area, the area
design value will generally be higher
than the study-reported mean.
Similarly, he observes that when a study
reports a mean that reflects the average
of annual average concentrations
estimated at across an area using a
hybrid modeling approach, the area
design value will generally be higher.
As such, by evaluating the difference
between the study-reported mean PM2.5
concentrations and design values, the
Administrator seeks to set the level of
the standard below the lowest studyreported mean, while ensuring that the
primary annual PM2.5 standard,
including its averaging time and form,
provides protection against the
exposures associated with health effects
observed in the key epidemiologic
studies.
Additionally, the EPA disagrees with
commenters who contend that the
approach taken may allow PM2.5 near
the design value monitor to remain
above the study-reported mean PM2.5
concentrations. In following this
approach of setting the annual standard
level somewhat below the lowest
reported mean PM2.5 concentration,
setting a standard level that requires the
design value monitor (which is the
highest monitor in an area) to be just
below the lowest study-reported mean
across key studies will generally result
in distributions of even lower
concentrations of PM2.5 across the entire
area, such that even those people living
near an area design value monitor
(where PM2.5 concentrations are
generally highest) will be exposed to
PM2.5 concentrations below the PM2.5
concentrations reported in the
epidemiologic studies where there is the
highest confidence of an association. In
their review of the 2021 draft PA, the
majority of the CASAC had some
concerns about the approach for
comparing study means and design
values, questioning whether such an
approach would provide adequate
protection for people who live in areas
with higher concentrations, such as
those living in areas with higher
concentrations (e.g., near the design
value monitor) (Sheppard, 2022a, p. 8 of
consensus responses). The minority of
the CASAC, in considering the
relationship between the study-reported
PO 00000
Frm 00063
Fmt 4701
Sfmt 4700
16263
mean PM2.5 concentration and design
values, stated that ‘‘the form of the
standard and the way attainment with
the standard is determined (i.e., highest
design value in the CBSA) are important
factors when determining the
appropriate level for the standard’’ and
noted that that design values are
generally higher than area average
exposure levels (Sheppard, 2022a, p. 17
of consensus responses). For all of the
reasons discussed above, and consistent
with the minority of the CASAC’s
advice in their review of the 2021 draft
PA, we disagree with the commenters
that areas near the design value
monitors would be expected to
experience PM2.5 concentrations above
the study-reported mean concentrations.
Several commenters assert that
epidemiologic studies that restrict PM2.5
concentration to below 12 mg/m3
provide additional support for revising
the level of the primary annual PM2.5
standard to 8 mg/m3. Some commenters
disagree with the EPA’s assertion that
the studies that employ restricted
analyses do not provide enough
information to understand how the
studies were restricted to certain PM2.5
concentrations, with commenters
providing additional information on the
methods for restricted analyses. The
commenters state that for the long-term
studies at issue here, the study authors
simply examined their database that
linked subjects to long-term PM2.5
concentrations above 12 mg/m3,
removed those data from the analysis,
and reran the analysis. Additionally,
one commenter provided an explanation
of how the restricted analyses were
conducted in studies for which he was
an author. The commenter notes that for
each year a subject was in the study,
annual PM2.5 concentrations were
assigned at the ZIP code level. If they
moved, they were assigned the ZIP code
level PM2.5 concentration for the new
ZIP code. The commenter notes that
these restricted analyses only included
subjects whose annual PM2.5 exposure
never exceeded that restricted
concentration for any year of follow-up
in the study. The commenter suggested
that the EPA may be concerned as to
how PM2.5 concentrations in restricted
analyses related to a design value since
these are exposures for individuals who
may have relocated during the study but
argue that that is not the point. The
commenters assert that while the
analyses were restricted to people never
exposed above certain concentrations
over longer periods of time, the actual
PM2.5 exposure was one year of
exposure in most of these studies.
Commenters also suggest that, since the
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16264
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
EPA has deviated from its approach
from the 2012 review for considering
study-reported mean PM2.5
concentrations, the EPA should dismiss
its concerns regarding being able to
relate the mean PM2.5 concentrations
from these studies to design values.
First, the EPA agrees with
commenters that studies that employ
restricted analyses can be used for
informing conclusions regarding the
appropriate level of the primary annual
PM2.5 standard. However, the EPA
disagrees that the information provided
by the commenters provides a sufficient
basis for an annual standard level of 8
mg/m3. Restricted analyses provide
additional support for effects at lower
concentrations, exhibiting associations
for mean concentrations presumably
below the mean concentrations for the
main analyses. However, even though
commenters note that any individual
with exposures over the restricted
analyses is excluded from restricted
analyses, uncertainties remain with
regard to how the mean PM2.5
concentrations in restricted analyses
compare to design values, particularly
in light of the removal of entire ZIP
codes from analyses. Design values are
calculated based on all measured PM2.5
concentrations. When an analysis is
restricted below a certain level, some
parts of the air quality distribution are
removed, but comparing the restricted
mean to a design value is not possible
because these are two different metrics.
For example, in a study that restricts
concentrations below 12 mg/m3, that
represents only part of the air quality
distribution, whereas a design value for
that study area would include all PM2.5
concentrations, not just the ones below
12 mg/m3. Therefore, in contrast to
means from the main (unrestricted)
analysis, it is not possible to compare
mean concentrations from restricted
analyses to design values. Further, it is
unclear how one could evaluate such a
relationship between design values and
mean PM2.5 concentrations from studies
that use restricted analyses because the
standard is set based on all of its
elements (indicator, averaging time,
form, and level) and removing PM2.5
concentrations from the calculation of
the design value for such a comparison
would result in a metric that is no
longer a design value that would
provide the intended protection of the
standard. This leads to greater
uncertainty in how to use the mean
PM2.5 concentrations from these studies
that use restricted analyses in a similar
decision framework as the
epidemiologic studies that report longterm mean PM2.5 concentrations for
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
health effect associations for the full
distribution of PM2.5 concentrations.
As described in reaching his
conclusions in the section below, the
Administrator judges that, despite these
uncertainties and limitations, studies
that use restricted analyses can provide
supplemental information for
consideration in reaching conclusions
regarding both the adequacy and level of
the standard. He notes two studies (Di
et al., 2017b and Dominici et al., 2019)
are available in this reconsideration that
report means in their restricted analyses
(restricting annual average PM2.5
exposure below 12 mg/m3) and used
population-weighted approaches to
estimate PM2.5 exposures and these
studies report mean PM2.5
concentrations of 9.6 mg/m3. He
recognizes that these studies are just one
line of evidence for consideration and
that along with the broader evidence
base, including the key epidemiologic
studies, these studies provide support
that the level of the primary annual
PM2.5 standard should be set below 10
mg/m3.
We disagree with the commenters that
concerns about relating the mean PM2.5
concentrations from restricted analyses
to design values are not valid. As an
initial matter, restricted analyses were
not available and did not inform the
2012 decision to revise the annual PM2.5
standard level to 12.0 mg/m3. The
approach in 2012 in revising the annual
standard was to set the level to
somewhat below the mean of key
epidemiologic studies. As noted above,
while the EPA believes that restricted
analyses can help inform conclusions
regarding the adequacy and the level of
the primary annual PM2.5 standard, in
the context of placing the studies in a
decision framework to inform the
appropriate level of the annual PM2.5
standard, the EPA has not deviated from
its approach from the 2012 review.
Given that restricted analyses are new
since the 2012 review, the EPA
disagrees with commenters that
uncertainties associated with these
studies should not be considered, and
that these studies should be used in a
similar manner to their main analyses in
taking an approach to set a level of the
standard somewhat below the lowest
long-term reported mean PM2.5
concentration. Specifically, as detailed
above there are uncertainties and
limitations associated with relating the
mean PM2.5 concentrations from these
studies to design values for studies that
use restricted analyses, and many of
these studies did not expressly report a
mean PM2.5 concentration for the
restricted analysis which makes it
impossible to make such a comparison.
PO 00000
Frm 00064
Fmt 4701
Sfmt 4700
Several commenters contend that in
considering the accountability studies,
the EPA inappropriately reached
conclusions regarding the level of the
primary annual PM2.5 standard based on
the starting PM2.5 concentrations of
these studies, rather than the ending
concentrations (i.e., concentrations after
a policy was implemented). The
commenters assert that these studies
provide support for revising the level of
the primary annual PM2.5 standard to
below the proposed range of 9–10 mg/m3
to protect public health with an
adequate margin of safety.
Accountability studies examine the
effect of a policy on reducing PM2.5
concentrations in ambient air and
evaluate whether such reductions were
observed to also lead to reductions in
PM2.5- associated health outcomes (e.g.,
mortality). Additionally, accountability
studies can reduce uncertainties related
to residual confounding of temporal and
spatial factors (U.S. EPA, 2022a, p. 3–
25). Prior to implementation of the
policies, three accountability studies
newly available in this reconsideration
and assessed in the ISA Supplement,
report mean PM2.5 concentrations below
the level of the current annual standard
level (12.0 mg/m3) and ranged from 10.0
mg/m3 to 11.1 mg/m3 (Sanders et al.,
2020b; Corrigan et al., 2018; and
Henneman et al., 2019). These studies
suggest that public health improvements
may occur following the
implementation of a policy that reduces
annual average PM2.5 concentrations
below the level of the current standard
of 12.0 mg/m3, and potentially below the
lowest ‘‘starting’’ concentrations in
these studies of 10.0 mg/m3. However,
while the small number of studies may
provide limited information related to
informing the adequacy and level of the
annual PM2.5 standard, we note that
accountability studies are only one line
of evidence, and that these studies
provide supplemental information for
consideration in addition to the full
body of evidence. Further, the EPA does
not believe it would be appropriate to
determine the level of the standard by
reference to ending concentrations in
accountability studies. Accountability
studies are most informative in
demonstrating that public health
improvements may occur following the
implementation of a policy that reduces
annual average PM2.5 concentrations
below the level of the current standard
of 12.0 mg/m3, and potentially below the
lowest ‘‘starting’’ concentrations in
these studies of 10.0 mg/m3. However,
the EPA finds the available information
from accountability studies is too
limited to support a conclusion that the
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
appropriate level at which to set the
primary annual PM2.5 standard would
be equal to the ending concentrations of
those studies, as the commenters
suggest. These studies demonstrate that
there are reductions in health outcomes
when PM2.5 concentrations are reduced
in these studies from the starting
concentration to the ending
concentration, but do not provide
support for health effect associations at
or below the ending concentrations that
would warrant a more stringent
standard.
Commenters disagree with the
Administrator placing less weight on
the epidemiologic studies conducted in
Canada when reaching conclusions
regarding the level of the primary
annual PM2.5 standard. These
commenters argue that the Canadian
epidemiologic studies provide support
for setting the level at the lowest end of
the proposed range (i.e., 8 mg/m3)
because they report mean PM2.5
concentrations, in some cases, below 8
mg/m3. Commenters disagree with the
EPA’s reasoning for placing less weight
on the Canadian epidemiologic studies,
suggesting it conflicts with the
approaches in previous PM NAAQS
reviews and arguing that the findings of
the Canadian epidemiologic studies can
be directly translated into a primary
annual PM2.5 standard. Additionally,
while the commenters disagree with the
EPA’s approach for considering the
study-reported mean PM2.5
concentrations and design values in
general, they note that the CASAC, in
their review of the 2021 PA, noted that
‘‘while there may be no design value in
Canada, there are data that indicate
what a U.S. design value would be if an
area average like that found in the
Canadian studies were to occur in the
U.S.’’ (Sheppard, 2022a, p. 13 of
consensus responses). The commenters
contend that the EPA failed to
acknowledge this advice from the
CASAC, specifically noting that the
majority of the CASAC highlighted
Canadian epidemiologic studies as a
part of their rationale for revising the
level of the primary annual PM2.5
standard to within the range of 8–10 mg/
m3.
In considering the information from
the epidemiologic studies in reaching
his conclusions, the Administrator
considered the full body of evidence,
including studies conducted in the U.S.
and Canada. However, as described in
the proposal and in section II.B.4 below,
the Administrator also recognizes that
the exposure environments in the U.S.
are different from those in Canada. In
particular, the U.S. population density
is approximately 43 people per square
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
kilometer in the contiguous U.S.98
compared to Canada, which has one of
the lowest population densities on the
Earth with 4.2 people per square
kilometer (Statistics Canada, 2023). This
difference in population density
between the U.S. and Canada was not as
apparent, and did not need to be
highlighted, in the 2012 review given
that the available Canadian
epidemiologic studies used populationweighting and focused on urban areas
where monitors were available and
population densities were more
comparable with those in the U.S. Given
this, the study-reported mean
concentrations from U.S. and Canadian
studies in the 2012 review were very
similar. The recent epidemiologic
evidence available in this
reconsideration, however, includes
studies that utilize approaches that
highlight the importance of considering
the differences between the two
exposure environments in the U.S.
versus Canada. When focusing on the
recently available Canadian monitorbased epidemiologic studies in this
reconsideration, the information
indicates that these studies, unlike the
studies available in the 2012 review, do
not apply population weighting (e.g.,
Lavigne et al., 2018; Liu et al., 2019). As
noted in responding to other public
comments above, the absence of
population weighting is an important
consideration that limits the utility of
these studies in informing the
appropriate level of the primary annual
PM2.5 standard. In addition, there are
recently available studies in the 2019
ISA and ISA Supplement that expand
the geographical extent of the
epidemiologic study areas by estimating
exposure concentrations in areas where
there are no monitors. To do this, these
studies use either a statistical
extrapolation of monitored values or use
air quality modeling and other forms of
data (e.g., hybrid model-based
approaches). For these Canadian
studies, the EPA notes two important
considerations in using the information
to directly translate to policy decisions
regarding the level of the annual
standard in the U.S. The first is that in
incorporating a larger portion of Canada
into these recent studies, more rural
areas are included, and as such, the
population densities and exposure
environment differences become more
important. The second is that in
analyses that evaluate and validate
hybrid models, there is less certainty in
PM2.5 exposure estimates in more rural
98 All of the key U.S. epidemiologic studies
considered in this reconsideration focus on all or
subsections of the continental U.S.
PO 00000
Frm 00065
Fmt 4701
Sfmt 4700
16265
areas, which are further from air quality
monitors and where PM2.5
concentrations in the ambient air tend
to be lower (U.S. EPA, 2022b, pp. 2–51
and 2–63). Additionally, it is unclear
what portion of the PM2.5 concentrations
from rural areas are contributing to the
study reported mean. Given this, studies
that incorporate more rural areas into
the epidemiologic studies highlight the
importance of considering the
differences between the population
exposures in the studies themselves and
in the U.S. versus Canadian study areas,
as well as the influence these
differences have on the interpretation of
the epidemiologic study results. For
these reasons, while the Canadian
epidemiologic studies provide
additional support for associations
between PM2.5 concentrations and
health effects, the long-term means from
Canadian epidemiologic studies are a
less certain basis for informing the
EPA’s selection of the annual standard
level, given that it is a U.S.-based
standard.
With respect to the CASAC’s advice
in their review of the 2021 draft PA, the
EPA recognizes that the majority of the
CASAC pointed to the Canadian studies
as supporting their recommendation to
revise the annual standard level to
within the range of 8–10 mg/m3.
However, the EPA also notes that the
CASAC did not advise the EPA to revise
the annual standard to a level that was
below the study-reported means in the
key Canadian epidemiologic studies.
Indeed, the CASAC noted that some of
the Canadian studies showed
associations below 8 mg/m3, but did not
recommend that the Administrator
consider levels below 8 mg/m3 for the
annual standard. Further, based on the
CASAC’s advice, the Administrator is
not excluding Canadian studies from his
consideration in this reconsideration,
but he is considering them in light of
the limitations and challenges presented
and in the context of the full body of
available scientific evidence.
Lastly, the EPA disagrees with
commenters that the findings of the
Canadian epidemiologic studies can be
directly translated into a primary annual
PM2.5 standard based on the evaluation
of the relationship between U.S. studyreported mean PM2.5 concentrations and
U.S. design values. It is unclear whether
the relationship between U.S. studyreported mean PM2.5 concentrations and
U.S. design values (which, in the case
of U.S. hybrid model-based studies,
indicates that design values are 15–18%
greater than area mean PM2.5
concentrations) would apply to the
Canadian epidemiologic studies and
their reported mean PM2.5
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16266
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
concentrations, given that these studies
generally report lower PM2.5
concentrations than the U.S.-based
studies. As such, interpreting the studyreported mean concentrations from the
Canadian studies in the context of a
U.S.-based standard may present
challenges in directly and quantitatively
informing decisions regarding potential
alternative levels of the annual
standard, particularly noting the
different in exposure relationships in
the U.S. versus Canada given the large
difference in population densities
between the two countries. Further, as
mentioned above, while the CASAC
advised the EPA to consider the
Canadian studies as relevant evidence
and found that placing weight on the
Canadian studies supported their
recommendation to revise the annual
standard level to within the range of 8–
10 mg/m3, the lower end of their
recommended range for the level of the
annual standard did not extend below
the lower study-reported means from
those studies.
Commenters who supported retaining
and revising the primary annual PM2.5
standard both raised concerns regarding
how the EPA used the scientific
evidence and quantitative risk
assessment related to disparities in
PM2.5 exposure and risk in informing
conclusions on the standard.
Commenters who supported retaining
the standard assert that the available
scientific evidence that demonstrates
disparities for minority populations do
not support revising the standard,
noting that these studies are in areas
that tend to have large minority
populations and more sources of PM.
These commenters contend that because
the studies conclude that minority
populations experience more effects
than others living in the same area that
something other than PM2.5
concentrations in ambient air is causing
the disproportionate impact on minority
populations, providing proximity to a
source as an example. The commenters
note that it is unclear how a national
standard will reduce exposure
disparities for population groups living
in the same area, and further assert that
studies of exposure disparities among
minority populations were considered
in reaching the 2020 final decision to
retain the standards.
Conversely, commenters who support
revising the standard assert that the atrisk analyses conducted in the 2022 PA
provide support for revising the primary
annual PM2.5 standard to a level of 8 mg/
m3. In particular, these commenters
state that the at-risk analysis
demonstrated that while disparities in
mortality risk remain at a standard level
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
of 9.0 mg/m3, disparities in exposure are
significantly reduced for an alternative
standard level of 8.0 mg/m3 (U.S. EPA,
2022b, p. 3–162).
As discussed in section I above, the
primary (health-based) NAAQS are
established at a level that is requisite to
protect public health, including the
health of sensitive or at-risk groups,
with an adequate margin of safety.99 In
so doing, decisions on the NAAQS are
based on an explicit and comprehensive
assessment of the current scientific
evidence and associated risk analyses.
More specifically, the EPA expressly
considers the available information
regarding health effects among at-risk
populations in decisions on the primary
NAAQS. Where populations with
disparities in exposure and risk are
among the at-risk populations, the
decision on the standards is based on
providing requisite protection for these
and other at-risk populations and
lifestages.
The Administrator expressly
considered the available information
regarding health effects among at-risk
populations in reaching the proposed
decisions that the current primary
annual PM2.5 standard is not requisite to
protect public health with an adequate
margin of safety, and should be revised.
The 2019 ISA and ISA Supplement
identified children, older adults, people
with pre-existing diseases
(cardiovascular disease and respiratory
disease), minority populations, and low
SES populations as at-risk populations.
The Administrator is thus, in his final
decision, establishing primary PM2.5
standards which, in his judgment, will
provide protection for these at-risk
populations, including minority
populations, with an adequate margin of
safety.
With respect to the risk assessment,
while the EPA notes that the analyses
support the conclusion that the primary
PM2.5 standards are not adequate, as
detailed further in the proposal and
above in section II.A.3, the EPA also
cautions against an over-interpretation
of the absolute results. The quantitative
risk assessment provides estimates of
PM2.5-attributable mortality based on
input data that include C–R functions
99 The legislative history of section 109 indicates
that a primary standard is to be set at ‘‘the
maximum permissible ambient air level . . . which
will protect the health of any [sensitive] group of
the population,’’ and that for this purpose
‘‘reference should be made to a representative
sample of persons comprising the sensitive group
rather than to a single person in such a group.’’ S.
Rep. No. 91–1196, 91st Cong., 2d Sess. 10 (1970);
see also, e.g., Am. Lung Ass’n v. EPA, 134 F.3d 388,
389 (D.C. Cir. 1998) (‘‘If a pollutant adversely affects
the health of these sensitive individuals, EPA must
strengthen the entire national standard’’).
PO 00000
Frm 00066
Fmt 4701
Sfmt 4700
from epidemiologic studies that do not
quantitatively account for uncertainties
in associations between PM2.5 exposure
and health effects at lower
concentrations and are based on an air
quality adjustment approach that
incorporates proportional decreases in
PM2.5 concentrations to meet lower
alternative standard levels. As a result,
simulated air quality improvements
used in the risk assessment will always
lead to proportional decreases in risk
(i.e., each additional mg/m3 reduction
produces additional benefits with no
clear stopping point), without
considering the substantially greater
uncertainties associated with the
relationship between PM2.5 exposures
and health effects at lower
concentrations.
The same is true for the new at-risk
analysis in the risk assessment
presented in the 2022 PA that is based
on a recent epidemiologic study that is
available in this reconsideration that
provides mortality risk coefficients for
older adults (i.e., 65 years and older)
based on PM2.5 exposure and stratified
by racial and ethnic demographics.
Generally, the results of at-risk analyses
can vary greatly depending on the
inputs to the analyses, including the
representativeness of the populations
and demographics captured by the
study areas that are a part of the
analyses, as well as the available C–R
functions from epidemiologic studies
that stratify by race and ethnicity and
the air quality adjustment approaches
that are used to simulate air quality at
different standard levels. In fact, for this
at-risk analysis, the results are even
more uncertain than similar estimates
from the overall risk assessment due to
additional sources of uncertainty
specific to the at-risk analysis, such as
using C–R functions derived from
smaller epidemiologic sample sizes
along with the sources of uncertainty
that apply to the overall risk assessment
(U.S. EPA, 2022b, section 3.4.1.8).
Additionally, in characterizing at-risk
populations, the at-risk analysis only
used one of the air quality adjustment
approaches used in the overall risk
assessment, which decreases the
potential representativeness of the PM2.5
concentrations across the study areas
(U.S. EPA, 2022b, section 3.4.1.8).
Lastly, this at-risk analysis relies on the
stratified risk coefficients from only one
epidemiologic study.100 For these
reasons, the Administrator places little
100 Additional information on all available at-risk
epidemiologic studies in this reconsideration are
available in section 3.4 and Appendix C of the 2022
PA (U.S. EPA, 2022b, section 3.4, Figure 3–17, and
Appendix C, section C.3.2).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
weight on the absolute results of the risk
assessment, including the at-risk
analysis, for purposes of selecting the
level of the annual standard that is
requisite.
While there are substantial
uncertainties in the absolute results of
the quantitative risk assessment, the
EPA also notes that recent scientific
evidence evaluated in the ISA
Supplement, which built upon the 2019
PM ISA conclusions, found that the
evidence ‘‘[c]ontinue[s] to support
disparities in PM2.5 exposure and health
risks by race and ethnicity’’ while
studies of SES ‘‘provide additional
support indicating there may be
disparities in PM2.5 exposure and health
risk by SES’’ (U.S. EPA, 2022a, p. 5–4).
Thus, in light of the statutory
requirement to provide protection for atrisk populations, it is not surprising that
the stratified population results of the
risk assessment suggest that meeting a
revised standard would result in higher
risk reductions for minority and low
SES populations.
In conclusion, the EPA recognizes
that the at-risk analysis was based on
one epidemiologic study that stratified
by race/ethnicity for older adults (e.g.,
65+ years old) and that there is
increasing uncertainty in quantitative
estimates of stratified risk estimates at
the lower end of the range of standard
levels assessed. Moreover, the EPA finds
that the goal of the NAAQS is to provide
the requisite protection to at-risk
groups, and where minority populations
are included among the at-risk groups,
providing requisite protection to
minority populations will also result in
protecting the public health of other
populations. Thus, in setting the
NAAQS to protect the health of at-risk
groups with an adequate margin of
safety, the Administrator is selecting the
standard that will provide requisite
protection, including for minority
populations and other at-risk
populations, which also generally
results in protecting the public health of
other populations and reducing risk
disparities.
A number of commenters, primarily
from industries and industry groups and
some States, support the EPA’s
proposed decision to retain the primary
24-hour PM2.5 standard. Many of these
commenters contend that the available
scientific evidence and quantitative
information has not significantly
changed since the 2020 final decision
and note that important uncertainties
remain. The commenters agree with the
EPA’s conclusions regarding the
controlled human exposure studies and
their relationship to short-term peak
PM2.5 concentrations in ambient air.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
These commenters also noted the
primary annual and 24-hour PM2.5
standards work together to provide
public health protection, with the 98th
percentile form of the 24-hour standard
effectively limiting peak daily
concentrations. The commenters agree
with the EPA that the current suite of
standards maintain subdaily
concentrations below the higher
concentrations in controlled human
exposure studies where more consistent
health effects are observed. Commenters
also agree with the EPA’s conclusions
that the epidemiologic studies are not
useful for informing decisions on the
level of the primary 24-hour PM2.5
standard because the standard focuses
on reducing peak exposures with its
98th percentile form, while the
epidemiologic studies often focus on the
mean or median as the percentile for
which associations with short-term
exposures are observed. These
commenters also agree with the EPA’s
focus on U.S.-based studies because of
differences compared to Canadian
studies. The commenters also generally
agree with the Administrator’s judgment
that it was appropriate to place less
weight on the risk assessment, noting
that the annual standard is controlling
in most areas of the country and
revising the annual standard would
have the most potential to reduce risk
related to PM2.5 exposures and would
reduce both average (annual) and peak
(daily) PM2.5 concentrations. Finally,
these commenters note that the CASAC
did not reach consensus on whether the
current primary 24-hour PM2.5 standard
should be revised, and they agree with
the minority of the CASAC’s
recommendation in their review of the
2021 draft PA that the primary 24-hour
primary PM2.5 standard should be
retained. These commenters also note
the CASAC’s support in their review of
the 2019 draft PA for retaining the
primary 24-hour PM2.5 standard.
A number of commenters, primarily
from public health and environmental
organizations and some States, oppose
the EPA’s proposed decision to retain
the primary 24-hour PM2.5 standard.
These commenters support revising the
level of the primary 24-hour PM2.5
standard, contending that a more
stringent standard is necessary to
provide requisite public health
protection with an adequate margin of
safety, particularly for at-risk groups. In
so doing, these commenters place
weight on the same aspects of the
available scientific evidence as the
majority of the CASAC in their review
of the 2021 draft PA, and generally
advocate for revising the level of the
PO 00000
Frm 00067
Fmt 4701
Sfmt 4700
16267
standard to within the range of 25–30
mg/m3 as recommended by the majority
of the CASAC. Some of these
commenters support a level no higher
than 25 mg/m3 and others support a
level of 20 mg/m3. These commenters
generally cite to the available scientific
evidence, including evidence of
disproportionate exposures and risks for
certain at-risk groups, and the CASAC’s
advice in support for their
recommendation. Some of these
commenters also suggest that decisions
regarding the primary 24-hour PM2.5
standard should not be related to
decisions on the primary annual PM2.5
standard.
As an initial matter, the EPA disagrees
with commenters who suggest that
decisions regarding the primary 24-hour
PM2.5 standard should not be related to
decisions on the primary annual PM2.5
standard. In reviewing the adequacy of
the public health protection afforded by
the primary PM2.5 standards, the
Administrator’s consistent past practice
has been to evaluate the combination of
the annual and 24-hour standards
together. In 2012, the thenAdministrator concluded that the most
effective and efficient way to reduce
total population risk associated with
both long- and short-term PM2.5
exposures was to set a generally
controlling annual standard, and to
provide supplemental protection by
means of a 24-hour standard set at the
appropriate level. In so doing, the thenAdministrator explicitly recognized that
potential air quality changes associated
with meeting a revised annual standard
(with a level of 12 mg/m3) would result
in lowering risks associated with both
long- and short-term PM2.5 exposures by
lowering the overall distribution of air
quality concentrations, and that
retaining a 24-hour standard at the
appropriate level would ensure an
adequate margin of safety against shortterm effects in areas with high peak-tomean ratios (78 FR 3163, January 15,
2013). In this reconsideration, also, the
Administrator considers it appropriate
to rely on the annual standard
(arithmetic mean, averaged over three
years) for targeting protection against
both long- and short-term PM2.5
exposures, noting that the annual
standard is typically controlling, while
the 24-hour standard (98th percentile,
averaged over three years) can provide
supplemental protection against the
occurrence of peak 24-hour PM2.5
concentrations (U.S. EPA, 2022b,
section 3.6.3). Further, the
Administrator notes that, as in the 2012
review, changes in PM2.5 air quality to
meet a revised annual standard would
E:\FR\FM\06MRR3.SGM
06MRR3
16268
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
affect the entire distribution of long- and
short-term concentrations, thus likely
resulting not only in lower short- and
long-term PM2.5 concentrations near the
middle of the air quality distribution,
but also in fewer and lower short-term
peak PM2.5 concentrations.101 Thus, the
Administrator continues to conclude it
is appropriate to consider whether the
annual and 24-hour standards together
provide requisite protection of public
health, rather than considering each
standard in isolation.
Regarding the appropriate basis for
determining the level of the 24-hour
standard, a number of commenters who
support revising the primary 24-hour
PM2.5 standard to a lower level contend
that the EPA should not rely on the
controlled human exposure studies in
evaluating the adequacy of the public
health protection afforded by the
primary 24-hour PM2.5 standard. These
commenters support this view by citing
the CASAC comments in their review of
the 2019 draft PA which advised that
controlled human exposure studies have
limitations that may impact their ability
to inform conclusions on the adequacy
of the public health protection afforded
by the primary 24-hour PM2.5 standard.
Commenters noted that these studies do
not include the most vulnerable
populations and often involve exposure
to only one pollutant to elicit a
response, and therefore are not
representative of real-world exposures.
Other commenters support the EPA’s
use of the controlled human exposure
studies to inform the adequacy of the
public health protection and note that
the 24-hour standard must at least
provide protection against the health
effects observed in controlled human
exposure studies. Some of the
commenters cite the Wyatt et al. (2020)
study that demonstrated cardiovascular
effects following 2-hour exposures to
120 mg/m3 and 4-hour exposures to 37.8
mg/m3. Some of these commenters
contend that the current primary 24hour PM2.5 standard allows PM2.5
exposures comparable to those observed
to elicit effects in the controlled human
exposure studies, and therefore, the EPA
must revise the level of the current
standard to protect public health. To
support this view, some commenters
101 Similarly, the Administrator recognizes that
changes in air quality to meet a 24-hour standard,
would result not only in fewer and lower peak 24hour PM2.5 concentrations, but also in lower annual
average PM2.5 concentrations. However, as noted in
2012, an approach that relied on setting the level
of the 24-hour standard such that the 24-hour
standard was generally controlling would be less
effective and result in less uniform protection
across the U.S. than an approach that focuses on
setting a generally controlling annual standard (78
FR 3163, January 15, 2013).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
submitted an analysis of monitoring
data from 2017–2020, which compares
the number of days per year where
maximum daily PM2.5 concentrations
exceed 120 mg/m3 and 37.8 mg/m3.
Additionally, other commenters assert
that the EPA should focus less on peak
PM2.5 concentrations ‘‘typically
measured’’ in areas meeting the current
primary PM2.5 standards even if they do
not exceed the concentrations in the
controlled human exposure studies
because, in their view, the standard
needs to protect against atypical
exposures to atypical peak PM2.5
concentrations. These commenters
conclude that, when considered
together, the controlled human exposure
studies and the epidemiologic studies
warrant strengthening the level of the
primary 24-hour PM2.5 standard.
The EPA generally disagrees with
commenters who contend that it is
inappropriate to rely on the controlled
human exposures studies in evaluating
the adequacy of the public health
protection afforded by the primary 24hour PM2.5 standard. The Agency
considers these studies informative both
for establishing biological plausibility
and for determining an appropriate level
for the 24-hour standard. When looking
to the experimental studies, the EPA
finds that the 2019 ISA and ISA
Supplement included controlled human
exposure studies that report statistically
significant effects on one or more
indicators of cardiovascular function
following 2-hour exposures to PM2.5
concentrations at and above 120 mg/m3
(and at and above 149 mg/m3 for
vascular impairment, the effect shown
to be most consistent across studies). As
noted in the 2019 ISA, these studies are
important in establishing biological
plausibility for PM2.5 exposures causing
more serious health effects, such as
those seen in short-term exposure
epidemiologic studies, and they provide
support that more adverse effects may
be experienced following longer
exposure durations and/or exposure to
higher concentrations. Additionally, one
controlled human exposure study
assessed in the ISA Supplement reports
evidence of some effects for
cardiovascular markers at lower PM2.5
concentrations, 4-hour exposures to 37.8
mg/m3 (Wyatt et al., 2020). However,
there is inconsistent evidence for
inflammation in other controlled human
exposure studies evaluated in the 2019
ISA. The EPA notes that although the
controlled human exposure studies do
not provide a threshold below which no
effects occur, the observed effects in
these controlled human exposures
studies are ones that signal an
intermediate effect in the body, likely
PO 00000
Frm 00068
Fmt 4701
Sfmt 4700
due to short-term exposure to PM2.5, and
typically would not, by themselves, be
judged as adverse (88 FR 5620, January
27, 2023) 102 103
The EPA notes that the majority of the
CASAC, in their review of the 2021 draft
PA, commented that these controlled
human exposure studies generally do
not include populations with
substantially increased risk from
exposure to PM2.5, such as children,
older adults, or those with more severe
underlying illness, and often involve
exposure to only one pollutant to elicit
a response. However, both the majority
and the minority of the CASAC
explained that, even taking into
consideration their limitations, the
controlled human exposure studies
provide some support for assessing the
adequacy of the 24-hour standard.104
The EPA agrees with the CASAC that
the controlled human exposure studies
generally do not include populations
with substantially increased risk from
exposure to PM2.5, like children, older
adults, or those with pre-existing severe
illness, like cardiovascular effects. As
such, and as an initial note, these
102 Judgments regarding adversity or health
significance of measurable physiological responses
to air pollutants have been informed by guidance,
criteria or interpretative statements developed
within the public health community, including the
American Thoracic Society (ATS) and the European
Respiratory Society (ERS), which cooperatively
updated the ATS 2000 statement What Constitutes
an Adverse Health Effect of Air Pollution (ATS,
2000) with new scientific findings, including the
evidence related to air pollution and the
cardiovascular system (Thurston et al., 2017).
103 The ATS/ERS described its 2017 statement as
one ‘‘intended to provide guidance to policymakers,
clinicians and public health professionals, as well
as others who interpret the scientific evidence on
the health effects of air pollution for risk
management purposes’’ and further notes that
‘‘considerations as to what constitutes an adverse
health effect, in order to provide guidance to
researchers and policymakers when new health
effects markers or health outcome associations
might be reported in future.’’ The most recent
policy statement by the ATS, which once again
broadens its discussion of effects, responses and
biomarkers to reflect the expansion of scientific
research in these areas, reiterates that concept,
conveying that it does not offer ‘‘strict rules or
numerical criteria, but rather proposes
considerations to be weighed in setting boundaries
between adverse and nonadverse health effects,’’
providing a general framework for interpreting
evidence that proposes a ‘‘set of considerations that
can be applied in forming judgments’’ for this
context (Thurston et al., 2017).
104 In their review of the 2021 draft PA, the
majority of the CASAC advised that ‘‘evidence of
effects from controlled human exposure studies
with exposures close to the current standard
support epidemiologic evidence for lowering the
standard’’ (Sheppard, 2022a, p. 4 of consensus
letter). The minority of the CASAC also advised that
it was appropriate to place ‘‘more emphasis on the
controlled human exposure studies, showing effects
at PM2.5 concentrations well above those typically
measured in areas meeting the current standards’’
(Sheppard, 2022a, p. 4 of consensus letter), in
evaluating adequacy of the 24-hour standard.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
studies are therefore somewhat limited
in their ability to inform at what
concentrations effects may be elicited in
at-risk populations. In spite of this
limitation, the EPA also agrees with the
CASAC, that even taking into
consideration the limitations of the
controlled human exposure studies,
these studies can provide some support
for evaluating the adequacy of the 24hour standard. However, the EPA
further notes that while the controlled
human exposure studies are important
in establishing biological plausibility,
the health outcomes observed in these
controlled human exposure studies are
often ‘‘intermediate’’ outcomes (i.e., not
always clearly adverse) and therefore it
is unclear how the importance of the
effects observed in the studies should be
interpreted with respect to adversity to
public health. The EPA finds that it is
appropriate to consider these study
limitations in assessing the information
provided by controlled human exposure
studies in evaluating the adequacy of
the primary 24-hour PM2.5 standard.
The EPA agrees with commenters that
the primary 24-hour PM2.5 standard
must at least provide protection against
the health effects consistently observed
in controlled human exposure studies.
As discussed in the proposal, the EPA
looks at whether the exposures that
elicit a response following exposure to
PM2.5 in the controlled human exposure
studies occur under recent air quality
conditions in areas meeting the current
standards. Based on these air quality
analyses, the EPA concludes that these
types of exposures very rarely occur
when the current standards are being
met.
The EPA did receive multiple
comments questioning these results and
the approach in the EPA’s analyses. For
example, some commenters submitted
an analysis of monitoring data from
2017–2020, which compares the number
of days per year where maximum daily
PM2.5 concentrations exceed 120 mg/m3
and 37.8 mg/m3 and evaluate the number
of days subset by groups of monitors
with 4-year average PM2.5
concentrations close to the levels of
combinations of current and proposed
annual (+/¥ 0.2 mg/m3) and 24-hour (+/
¥2 mg/m3) PM2.5 standards. To support
their view that the primary PM2.5
standards should be revised, the
commenters describe decreases in days
per monitor per year with 2-hour
maximum concentrations greater than
120 mg/m3 and 4-hour maximum
concentrations greater than 37.8 mg/m3
when comparing monitors that achieve
close to 10 and 30 mg/m3 versus
monitors that meet close to 8 mg/m3 and
25 mg/m3. The commenters noted
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
decreases in the number of days per
monitor per year with 2-hour maximum
concentrations over 120 mg/m3 and 4hour max concentration over 37.8 mg/m3
were also seen when comparing
monitors close to achieving 24-hour
standards with levels of 35 mg/m3 versus
25 mg/m3.
First, the EPA notes that this analysis
submitted by commenters was limited
to a very small number of monitors and
did not include a national perspective.
Second, the EPA notes that this analysis
focused on number of days (rather than
the number of times) where there was a
2-hour maximum concentration over
120 mg/m3 or a 4-hour max
concentration over 37.8 mg/m3. In order
to evaluate the protection provided by
the current 24-hour standard against
peak exposures, including exposures
with 2-hour concentrations greater than
120 mg/m3 and 4-hour concentrations
greater than 37.8 mg/m3, the EPA
considers it more informative and
appropriate from a public health
perspective to assess the number of
times a subdaily exposure of concern
occurs in a year, rather than the number
of days on which they occur because the
former identifies more potential
exposures of concern and provides more
information about the scale and scope of
the occurrences of those exposures.
Lastly, the analyses allowed monitors
somewhat above the standards to be
included. Therefore, it is unclear
whether the exceedances of the 2-hour
or 4-hour benchmarks would still have
occurred if the area had actually been
meeting the current primary PM2.5
standards. However, in considering the
analyses submitted by the commenters,
the EPA conducted new analyses 105
that looked at all individual monitors
across the U.S. and evaluated the
percentage of times the monitors
experienced a 2-hour maximum
concentration over 120 mg/m3 or a 4hour max concentration over 37.8 mg/m3
when that monitor was meeting the
current standards. Further, given that
the Administrator concludes that the
level of the current primary annual
PM2.5 is not adequate and that it should
be revised to 9.0 mg/m3, the new
analysis evaluates the percentage of
times during a recent 3-year period (i.e.
2019–2021) that individual monitors
experienced a 2-hour maximum
concentration over 120 mg/m3 or a 4105 Jones et al. (2023). Comparison of Occurrence
of Scientifically Relevant Air Quality Observations
Between Design Value Groups. Memorandum to the
Rulemaking Docket for the Review of the National
Ambient Air Quality Standards for Particulate
Matter (EPA–HQ–OAR–2015–0072). Available at:
https://www.regulations.gov/docket/EPA-HQ-OAR2015-0072.
PO 00000
Frm 00069
Fmt 4701
Sfmt 4700
16269
hour max concentration over 37.8 mg/m3
when that monitor was meeting the
current primary 24-hour PM2.5 standard
with its level of 35 mg/m3 and a revised
primary annual PM2.5 standard of 9.0
mg/m3.
In evaluating the results from the new
analyses, it is important to keep in mind
that the 2019 ISA and ISA Supplement
concluded that the most consistent
evidence from the controlled human
exposures studies is for impaired
vascular function following 2-hour
exposures to average PM2.5
concentrations at and above about 120
mg/m3, with less consistent evidence for
effects following exposures to
concentrations lower than 120 mg/m3.
The new analyses show that across all
monitors, on average, only 0.029 percent
of 2-hour observations reach PM2.5
concentrations higher than 120 mg/m3 in
areas meeting the current 24-hour
standard and a revised annual standard
of 9.0 mg/m3. Further, recognizing that
one purpose of the 24-hour standard is
to protect against exposure in areas with
high peak-to-mean ratios, when
assessing the monitors individually
across the U.S. under these same
conditions, the monitors reporting the
highest PM2.5 concentrations have only
0.47 percent of 2-hour observations
reach PM2.5 concentrations higher than
120 mg/m3.
Additionally, the analyses also
evaluated the frequency of reporting a 4hour maximum concentration over 37.8
mg/m3 when monitors were meeting the
current 24-hour standard and a revised
annual standard of 9.0 mg/m3. For this
part of the analysis, the EPA finds that
across all monitors, on average, only
0.41 percent of 4-hour observations
reach PM2.5 concentrations higher than
37.8 mg/m3 in areas meeting the current
24-hour standard and a revised annual
standard of 9.0 mg/m3. Further, when
assessing the monitors individually
across the U.S. under these same
conditions, the monitors reporting the
highest PM2.5 concentrations have only
2.6 percent of 4-hour observations reach
PM2.5 concentrations higher than 37.8
mg/m3. Thus, the EPA disagrees with
commenters that the current primary 24hour PM2.5 standard typically allows
PM2.5 exposures at or above those
observed to cause health effects in
controlled human exposure studies.
Furthermore, the EPA notes that in light
of the small number of occurrences and
the intermediate nature of the effects
observed in Wyatt et al. (2020) at
concentrations of 37.8 mg/m3 (i.e.,
effects that typically would not, by
themselves, be judged as adverse), there
is substantial basis to doubt whether
further improvements in public health
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16270
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
would be achieved by further reducing
these exposures. In drawing this
conclusion, the EPA notes the lack of
evidence of effects from controlled
human exposure studies at levels below
the current 24-hour standard and the
fact that the results of Wyatt et al. (2020)
are inconsistent with other currently
available studies, and this study only
observes intermediate effects.
In response to commenters that cited
the majority of the CASAC’s view that,
in general, ‘‘[t]here is . . . less
confidence that the annual standard
could adequately protect against health
effects of short-term exposures’’
(Sheppard, 2022a, p. 4 of consensus
letter), the EPA disagrees with the
majority of CASAC, noting that the
results of the EPA’s analysis suggest that
high peak concentrations are extremely
infrequent in areas meeting an annual
standard of 9.0 mg/m3, occurring less
than 0.029–0.41 percent of the time (for
2-hour concentrations >120 mg/m3 and
4-hour concentrations >37.8 mg/m3,
respectively). This suggests that in most
locations, even the upper tail of the
distribution would be controlled quite
well under a revised annual standard.
With regard to the likelihood that the
current standards would allow peak
concentrations that are clearly of
concern from a health perspective,
therefore, the EPA concludes that such
occurrences are extremely infrequent—
and will be even less frequent under the
improved air quality conditions
associated with meeting a revised
annual PM2.5 standard of 9.0 mg/m3.
A number of commenters who
support revising the primary 24-hour
PM2.5 standard to a lower level suggest
that the available epidemiologic
evidence provides support for such a
revision. To support their view, the
commenters note that the currently
available evidence, including a number
of epidemiologic studies that
demonstrate associations between shortterm PM2.5 exposures and health effects,
provides support for causal
relationships for short-term PM2.5
exposures and health effects as
described in the 2019 ISA and ISA
Supplement. The commenters further
note that the available epidemiologic
studies include diverse populations that
are broadly representative of the U.S.
population, including at-risk
populations, which they assert is an
advantage over the controlled human
exposure studies and the risk
assessment, which are not as broadly
representative.
These commenters highlight a number
of specific epidemiologic studies that
they suggest provide support for
revising the level of the 24-hour
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
standard. Additionally, commenters
contend that there are epidemiologic
studies using restricted analyses that
show that positive and statistically
significant associations between shortterm PM2.5 exposure and mortality
persist at daily mean concentrations
below 25 mg/m3. The commenters also
cite several studies that provide no
evidence of a threshold. These
commenters also point to the CASAC
advice in their review of the 2021 draft
PA, where the majority of the CASAC
cited epidemiologic studies using
restricted analyses as offering support
for revision. The commenters argue that
the EPA cannot base discretion on
uncertainties related to the methods
used in restricted analyses in the
epidemiologic studies. In so doing,
these commenters disagree with the
EPA that it is important to take into
consideration that these studies do not
consider the form or averaging time of
the 24-hour standard. Finally, the
commenters claim that while the EPA
stated that the study-reported means
from epidemiologic studies that use
restricted analyses are more useful for
identifying impacts from typical 24hour exposures than for peak 24-hour
exposures, the commenters assert that
the studies also indicate that there are
health risks at relatively high
concentrations below the current level
of the primary 24-hour PM2.5 standard
that must be addressed.
As noted by the commenters,
epidemiologic studies that show
positive and statistically significant
associations between short-term PM2.5
exposure and mortality provide support
for the causal determination in the 2019
ISA. The EPA also agrees that the
available epidemiologic studies include
diverse populations that are broadly
representative of the U.S. population,
including at-risk populations. Further,
the EPA agrees that studies evaluated in
the 2019 ISA and the ISA Supplement
continue to provide evidence of linear,
no-threshold concentration-response
relationships, but with less certainty in
the shape of the curve at lower
concentrations (i.e., below about 8 mg/
m3), with some recent studies providing
evidence for either a sublinear, linear, or
supralinear relationship at these lower
concentrations (U.S. EPA, 2019a,
section 11.2.4; U.S. EPA, 2022a, section
2.2.3.2).
However, findings of positive,
significant associations in short-term
epidemiologic studies do not directly
indicate that short-term effects would
occur in areas meeting the 24-hour
standard and therefore, do not directly
address the question of whether the
current 24-hour standard is adequate.
PO 00000
Frm 00070
Fmt 4701
Sfmt 4700
While short-term epidemiologic studies
evaluate associations between
distributions of ambient PM2.5 and
health outcomes, they do not identify
the specific exposures (i.e., a specific
24-hour concentration) that can lead to
the reported effects. Short-term
epidemiologic studies evaluate the
association between day-to-day
variation in daily (24-hour) PM2.5
exposure and health endpoints (e.g.,
mortality) to understand how these
changes in air pollution concentrations
are associated with changes in health
outcomes. But these studies do not
report daily concentrations; rather, they
report the long-term mean concentration
of the 24-hour PM2.5 concentrations over
the entire multi-year period of the
study, and typically report their results
as a relative risk (e.g., for each 10 mg/
m3 increase in PM2.5, the risk of
mortality or cardiovascular hospital
admissions increases by a certain
percentage, across the full range of the
24-hour PM2.5 concentrations in the
study). This means that there is no
specific point in the air quality
distribution of any epidemiologic study
that represents a ‘‘bright line’’ at and
above which effects have been observed
and below which effects have not been
observed. Nor, as noted above, do these
studies allow for any direct inferences
about health impacts associated with
the short-term ‘‘peak’’ exposures that
the primary 24-hour standard is
designed to protect against. While there
can be considerable variability in daily
exposures over a multi-year study
period, most of the estimated exposures
in these epidemiologic studies reflect
days with ambient PM2.5 concentrations
around the mean or middle of the air
quality distributions examined (i.e.,
‘‘typical’’ days rather than days with
extremely high or extremely low
concentrations). This is true of longterm epidemiologic studies as well. The
difference between epidemiologic
studies examining associations with
long-term exposures and short-term
exposures is comparing different levels
of exposure over different exposure
durations (i.e., long-term studies
exposures are defined as those that are
annual or multi-year, while short-term
exposures are defined as those that are
mostly 24-hour) (U.S. EPA, 2019a,
section P.3.1). Thus, in both cases, and
in the absence of a discernible
threshold, epidemiologic studies of
short-term and long-term exposures
provide the strongest support and
confidence for reported health effect
associations around the middle portion
of the PM2.5 air quality distribution (e.g.,
the study-reported mean PM2.5
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
concentration), which corresponds to
the bulk of the underlying data, rather
than at the extreme upper or lower ends
of the distribution. However, the
difference between the annual standard
and the 24-hour standard, aside from
averaging times, is that the form of the
annual standard is a mean PM2.5
concentration, which is based on the
bulk of the air quality data, while the
form of the 24-hour standard is a 98th
percentile form, which is based on peak
concentrations. Both long-term and
short-term epidemiologic studies are
informative for determining the
appropriate level of the annual PM2.5
standard, which is designed to control
‘‘typical’’ daily exposures and risks,
because these studies most often report
long-term mean (or median) PM2.5
concentrations that are representative of
‘‘typical’’ exposures that are associated
with health effects. In contrast, while
the short-term epidemiologic studies
examine health effects associated with
shorter exposure durations (e.g., mostly
24-hour exposures), these studies are
less informative for determining the
appropriate level of the 24-hour PM2.5
standard because these studies do not
report the 98th percentile PM2.5
concentrations,106 which is more
directly comparable to the form of the
24-hour standard. Additionally, if the
98th percentile of data were reported,
the EPA would consider the peak
concentrations observed in these studies
(which by definition rarely occur) in
conjunction with other supporting
evidence. However, as already noted,
there is an absence of new information
in this reconsideration (either from
controlled human exposure studies or
epidemiologic studies) suggesting that
peak concentrations just below the level
of the current 24-hour standard (with its
level of 35 mg/m3) are associated with
adverse effects. Instead, the evidence
links risk to more typical daily
exposures near the middle of the air
quality distribution—exposures most
effectively controlled through a
strengthening of the annual standard. As
noted in the 2012 final rule, ‘‘reducing
the annual standard is the most efficient
106 In the 2022 PA, the EPA has identified a
number of key areas for additional research and
data collection for PM2.5, based on the uncertainties
and limitations that remain in the scientific
evidence and technical information. In addition to
research and data collection, the EPA specifically
highlights additional information that could be
reported in the epidemiologic studies that may help
inform future reviews of the primary PM2.5
standards, including additional descriptive
statistics in the upper percentiles of the air quality
distribution (i.e., from the 95th to the 99th
percentile), as well as the number of days of
concentrations and/or health events within each of
these percentiles (U.S. EPA, 2022a, section 3.7).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
way to reduce the risks from short-term
exposures . . . as the bulk of the risk
comes from the large number of days
across the bulk of the air quality
distribution, not the relatively small
number of days with peak
concentrations’’ (78 FR 3156, January
15, 2013).
As noted above, in evaluating the
adequacy of the current standards, the
EPA has consistently considered the
annual standard (based on arithmetic
mean concentrations) and 24-hour
standard (based on 98th percentile
concentrations) together in evaluating
the public health protection provided by
the standards against the full
distribution of short- and long-term
PM2.5 exposures. Moreover, the EPA has
previously noted that the annual
standard is generally controlling in most
parts of the country, providing an
effective and efficient way to reduce
total population risk to both long- and
short-term PM2.5 exposures, while the
24-hour standard, with its 98th
percentile form, provides supplemental
protection, particularly for areas with
high peak-to-mean ratios of 24-hour
PM2.5 concentrations (78 FR 3158,
January 15, 2013). In such areas, annual
average PM2.5 concentrations could be
quite low, and the 24-hour standard
provides a means of ensuring control of
episodic peaks possibly associated with
strong local or seasonal sources, or
PM2.5-related effects that may be
associated with shorter-than daily
exposure periods. The approach taken
in evaluating the adequacy and
alternative levels of the annual standard
has been to evaluate the long-term mean
PM2.5 concentrations of both long-term
and short-term key epidemiologic
studies, where we have the most
confidence in the reported health effects
association, while also giving some
consideration to lower percentiles of the
air quality distribution (e.g., 25th
percentiles). However, using a similar
approach to evaluate the adequacy of
the current and any potential alternative
levels of the 24-hour standard with
short-term epidemiologic studies, as the
majority of CASAC and some
commenters are suggesting, presents
challenges.
Short-term epidemiologic studies,
including those that use restricted
analyses, often report metrics that
include mean PM2.5 concentrations,
with some studies also reporting lower
percentiles, such as the 25th percentile.
As previously noted above, for studies
of daily PM2.5 exposure, which examine
associations between day-to-day
variation in PM2.5 concentrations and
health outcomes, often over several
years, most of the estimated exposures
PO 00000
Frm 00071
Fmt 4701
Sfmt 4700
16271
reflect days with ambient PM2.5
concentrations around the middle of the
air quality distributions examined (i.e.,
the mean or median). However, there is
not a metric or statistic reported in
short-term epidemiologic studies that
allows for a direct comparison to the
current 24-hour PM2.5 standard and its
98th percentile form. While a 98th
percentile of PM2.5 concentrations is a
metric that might be more closely
compared to the 24-hour standard level,
98th percentile PM2.5 concentrations
were not reported in key epidemiologic
studies. Consistent with the
Administrator’s final decision in 2012,
the EPA notes that even if 98th
percentile values were reported, it
would be inappropriate to focus on
these concentrations without also
considering the impact of a revised
annual standard on short-term
concentrations, since many areas would
be expected to experience decreasing
short- and long-term PM2.5
concentrations in response to a revised
annual standard (78 FR 3156, January
15, 2013). Furthermore, in light of the
scarcity of days at the very upper end
of the distribution, and to avoid placing
undue reliance on the peak
concentrations observed in these studies
(which by definition rarely occur), the
EPA finds that such values would need
to be considered in conjunction with
other supporting evidence. In addition,
as described above, the other lines of
evidence available for consideration by
the EPA do not indicate that the current
primary 24-hour standard requires
revision to protect public health with an
adequate margin of safety. The EPA
notes again the lack of corroborating
evidence from controlled human
exposure studies. While the EPA agrees
with the CASAC that the controlled
human exposure studies are limited in
their ability to speak to the
concentrations at which effects may be
elicited in at-risk populations, as
discussed above the lowest
concentration associated with effects is
37.8 mg/m3 and the effects observed
were ‘‘intermediate’’ outcomes that are
not by themselves considered adverse.
We also note that, as detailed in section
II.A.2.a above, the study that observed
intermediate effects at concentrations of
37.8 mg/m3 was evaluated in the ISA
Supplement and the results of this study
were inconsistent with the controlled
human exposure studies evaluated in
the 2019 ISA. Additionally, as noted
above, the EPA finds that across all
monitors, on average, only 0.41 percent
of 4-hour observations reach PM2.5
concentrations higher than 38 mg/m3 in
areas meeting the current 24-hour
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16272
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
standard and a revised annual standard
of 9.0 mg/m3. Given the rarity of these
occurrences and the fact that the effects
associated with exposures to this PM2.5
concentration have not been found to be
adverse in and of themselves, the EPA
finds it reasonable to conclude that this
pattern of air quality will protect at-risk
populations, even though such
populations were not in the study
groups. The EPA concludes that further
evidence would be needed at specific
short-term (i.e., hourly or daily)
concentrations below the level of the
current 24-hour standard to support any
revision to the current 24-hour standard.
With regard to the data that are
available from the short-term
epidemiologic studies (which, as noted,
do not include 98th percentile values),
the EPA considers it inappropriate to
utilize the study-reported means from
the short-term epidemiologic evidence
to assess the adequacy of the 24-hour
standard, with its 98th percentile form,
considering that the study-reported
mean concentrations do not provide
meaningful insight regarding the
frequency or health significance of peak
concentrations occurring during the
study period. As indicated in the 2022
PA, the study-reported means of shortterm epidemiologic studies do not serve
a purpose in determining a level at
which we can confidently attribute
effects to the impact of ‘‘peak’’
exposures. The 24-hour standard is
intended to provide supplemental
protection against short-term peak
exposures and while there is a general
relationship between mean
concentrations and 98th percentile
concentrations in individual locations,
such relationships vary by location and
there is not an established relationship
that can be relied upon to predict 98th
percentile concentrations based on
mean PM2.5 concentrations reported in
multi-city epidemiologic studies.
Instead, mean concentrations from
short-term epidemiologic studies are
more useful in addressing questions
regarding the effects of ‘‘typical’’ or
average 24-hour exposures, which are
addressed through the annual standard.
For this reason, the EPA does consider
the mean concentrations of short-term
studies (as well as the means from the
long-term studies) in evaluating the
level of the annual standard, which the
EPA recognizes as the generally
controlling standard for both long- and
short-term exposures. However, the EPA
does not agree with commenters that it
is appropriate to use means from shortterm epidemiologic studies as the basis
for a decision-making framework to
determine the adequacy of the current
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
24-hour standard, with its 98th
percentile form.
As described in the proposal (88 FR
5613, January 27, 2023), the 2022 PA
also noted the epidemiologic studies
that restrict 24-hour average PM2.5
concentrations to values of less than 35
mg/m3, and in some cases less than 25
mg/m3, and annual average PM2.5
concentrations less than 12 mg/m3.
Restricted analyses use a subset of data
from their main analyses and conduct
an epidemiologic study with health
events that occur at concentrations
below a certain concentration (e.g., 25
mg/m3). While some of these studies do
not report the mean PM2.5 concentration
for the restricted analysis, the mean of
the restricted analysis is presumably
less than the mean PM2.5 concentration
in the main analysis. Restricted analyses
from long-term and short-term exposure
epidemiologic studies are informative in
providing support that the health effects
associations are not driven by just the
upper peaks of the PM2.5 air quality
distributions and provide support for
revision to the level of the annual PM2.5
standard. Short-term restricted analyses
also report positive associations
between short-term PM2.5 exposure and
morbidity and mortality. As an example,
in a restricted analysis evaluating the
association between short-term
exposures and PM2.5 concentrations less
than 25 mg/m3, Di et al. (2017a) removed
6.3 percent of the data from their main
analyses, (i.e., all PM2.5 concentrations
greater than 25 mg/m3), and still found
a positive and significant association
between short-term PM2.5 exposure and
mortality. This study provides
additional support that the association
between short-term exposure to PM2.5
and mortality in the main epidemiologic
analysis is not driven by the upper
peaks of the PM2.5 air quality
distribution, which in turn supports the
conclusion that lowering the entire
distribution of air quality concentrations
through a revised annual standard is an
appropriate means of protecting against
adverse effects from short-term
exposure, as discussed further below.
In their review of the 2021 draft PA,
the majority of the CASAC highlighted
three U.S.-based epidemiologic studies
that restricted 24-hour average PM2.5
concentrations below 25 mg/m3 as a part
of their rationale for recommending that
the EPA revise the level of the primary
24-hour PM2.5 standard. Similarly, in
evaluating positive associations in
restricted analyses, some commenters
also suggest that because an association
exists at 24-hour concentrations below
25 mg/m3, the 24-hour standard level
should be set at the concentration at
which the analysis was restricted (e.g.,
PO 00000
Frm 00072
Fmt 4701
Sfmt 4700
25 mg/m3). However, the EPA notes that
neither the CASAC nor public
commenters provided any detail
regarding, how, in their view, these
studies demonstrate that the level of the
current 24-hour standard is not
adequate, and/or how these studies
demonstrate what revised level of the
24-hour standard would provide
requisite public health protection with
an adequate margin of safety. The EPA
considers that such an approach would
have several important limitations.
First, the approach assumes that a
specific point on the air quality
distribution (e.g., the point at which the
analysis was restricted) is where health
effects are exhibited and where we have
the most confidence in the reported
association. However, in addition to the
limitations associated with the shortterm epidemiologic studies outlined
above, the EPA does not agree that it
would be appropriate to identify the
requisite level of the primary 24-hour
PM2.5 standard based on the specific
concentration at which the analyses
restrict their studies. The choice to
restrict the data at a particular
concentration is in effect arbitrary, and
does not establish that any particular
effects are attributable to that
concentration as opposed to other
concentrations within the restricted
analysis.
Further, these restricted analyses do
not report the PM2.5 concentration at the
98th percentile of data or other metrics
relating to the upper end of the
distribution that could provide
information about health risks
associated with peak exposures. For
example, the CASAC does not provide
a discussion of what the comparable
98th percentile concentration is in the
distribution of remaining 24-hour PM2.5
concentrations of restricted analyses
(because such data is not reported by
the study authors) and what degree of
confidence the Administrator should
place on those upper percentile values
(e.g., 98th percentile values). In order to
identify a level of the 24-hour standard
based on associations between the
‘‘upper end’’ of exposures, either in the
unrestricted or the restricted analyses,
and adverse health effects, it would be
necessary to have both greater detail on
the distribution of air quality in the
study and greater confidence in the
reported association at the peak
concentrations such as the 98th
percentile—in other words, a better
understanding of how specific 24-hour
concentrations correspond to the
frequency and total number of observed
health events in the study.
Further, the EPA notes that when
resulting analyses based on the
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
restricted dataset continue to find
positive associations between the
remaining air quality distribution and
health effects, it suggests that the
relationship was in fact not driven
primarily by the upper tail (now
removed from the dataset) but rather by
lower portions of the distribution of air
quality. In other words, we have no
confidence that the remaining upper
end of the air quality distribution is
driving the remaining associations
reported in the restricted analyses, as
opposed to the vast array of health
events at and around the mean PM2.5
concentration. In fact, it is reasonable to
conclude that to effectively address the
health effects observed in the study, it
is necessary to control not just the peak
concentrations but to reduce the bulk of
the exposures (occurring near the
mean), a task more effectively achieved,
as noted above through a tightening of
the annual standard, which has the
effect of shifting the entire distribution
of PM2.5 concentrations downward (both
peaks and means). Therefore, while the
EPA agrees that both short- and longterm epidemiologic studies that
completed restricted analyses and
reported the resulting study means
could be used to inform conclusions
regarding the adequacy of the annual
standard, given that the resulting study
means (when reported) could be
evaluated in the context of the decision
framework described above for
informing decisions on the level of the
annual standard, the EPA considers that
current short-term epidemiologic
studies that restrict analyses are subject
to the same limitations outlined above
for current short-term epidemiologic
studies in how they can be used in a
decision-making framework to inform
the adequacy and alternative level of the
primary 24-hour PM2.5 standard. As
such, while the available short-term
epidemiologic studies that restrict their
analyses are useful for informing
conclusions regarding the strength of
the associations for health outcomes,
they are not, as currently designed, as
useful for informing conclusions
regarding the adequacy of the current
primary 24-hour PM2.5 standard. In
reaching this conclusion, the EPA notes
that the majority of the CASAC did not
address the limitations of these studies
outlined in the 2021 draft PA,
particularly in the context of the 24hour standard with its 98th percentile
form. Among the future research needs
identified by the EPA in the 2022 final
PA, the Agency noted a number of gaps
in the currently available information
reported in the epidemiologic studies of
short-term exposure, including
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
‘‘descriptive statistics of PM2.5
concentrations at individual percentiles
from the 95th percentile to the 99th
percentile, as well as the number of
days of concentrations and/or health
events within each of these percentiles’’
and other descriptive statistics and
details regarding analytical design in
studies employing restricted analyses
(U.S. EPA, 2022b, pp. 3–225 to 3–226).
Such information could significantly
improve the EPA’s ability to draw
conclusions from these studies with
regard to the adequacy of the current
primary 24-hour PM2.5 standard.
Due to the limitations and
uncertainties outlined above, in
reaching his decision on the primary 24hour PM2.5 standard, the Administrator
judges that the information from
currently available short-term
epidemiologic studies, including those
that use restricted analyses, is
inadequate to inform decisions
regarding the adequacy of the current
24-hour standard. Additionally,
consistent with the final decision in
2012, the EPA continues to view an
approach that focuses on setting a
generally controlling annual standard as
the most effective and efficient way to
reduce total population risk associated
with both long- and short-term PM2.5
exposures. Potential air quality changes
associated with meeting an annual
standard level of 9.0 mg/m3 will result
in lowering risk associated with both
long- and short-term PM2.5 exposure by
lowering the overall air quality
distribution. As discussed above,
reducing the annual standard is the
most efficient way to reduce the risks
from short-term exposures identified in
the epidemiologic studies, as the
available evidence suggests the bulk of
the risk comes from the large number of
days across the bulk of the air quality
distribution, not the relatively small
number of days with peak
concentrations. However, as in the 2012
review, the Administrator recognizes
that an annual standard alone would not
be expected to offer sufficient protection
with an adequate margin of safety
against the effects of short-term PM2.5
exposures in all parts of the country,
particularly in areas with high peak-tomean ratios, and concludes that it is
appropriate to continue to provide
supplemental protection by means of a
24-hour standard. In so doing, the
Administrator concludes that retaining
the level of the primary 24-hour PM2.5
standard of 35 mg/m3 will provide
requisite protection against short-term
peak PM2.5 concentrations, in
conjunction with a revised annual
standard level of 9.0 mg/m3.
PO 00000
Frm 00073
Fmt 4701
Sfmt 4700
16273
4. Administrator’s Conclusions
This section summarizes the
Administrator’s considerations and
conclusions related to the adequacy of
the current primary PM2.5 standards and
presents his decision to revise the
primary annual PM2.5 standard to a level
of 9.0 mg/m3 and retain the primary 24hour PM2.5 standard. In establishing
primary standards under the Act that
are ‘‘requisite’’ to protect public health
with an adequate margin of safety, the
Administrator is seeking to establish
standards that are neither more nor less
stringent than necessary for this
purpose. He recognizes that the
requirement to provide an adequate
margin of safety was intended to
address uncertainties associated with
inconclusive scientific and technical
information and to provide a reasonable
degree of protection against hazards that
research has not yet identified.
However, the Act does not require that
primary standards be set at a zero-risk
level; rather, the NAAQS must be
sufficiently protective, but not more
stringent than necessary.
Given these requirements, the
Administrator’s final decision in this
reconsideration is a public health policy
judgment drawing upon scientific and
technical information examining the
health effects of PM2.5 exposures,
including how to consider the range and
magnitude of uncertainties inherent in
that information. This public health
policy judgment is based on an
interpretation of the scientific and
technical information that neither
overstates nor understates its strengths
and limitations, nor the appropriate
inferences to be drawn, and is informed
by the Administrator’s consideration of
advice from the CASAC and public
comments received on the proposal.
The initial issue to be addressed in
the reconsideration of the primary PM2.5
standards is whether, in view of the
advances in scientific knowledge and
other information reflected in the 2019
ISA, ISA Supplement, and 2022 PA, the
current standards are requisite to protect
public health with an adequate margin
of safety. In considering the adequacy of
the current suite of primary PM2.5
standards, the Administrator has
considered the large body of evidence
presented and assessed in the 2019 ISA
and ISA Supplement, the conclusions
presented in the 2022 PA, the views
expressed by the CASAC, and public
comments. The Administrator has taken
into account both evidence- and riskbased considerations in developing final
conclusions on the adequacy of the
current primary PM2.5 standards. The
Administrator has additionally
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16274
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
considered the associated public health
policy judgments and judgments about
the uncertainties inherent in the
scientific evidence and quantitative
analyses that are integral to the
conclusions on the adequacy of the
current primary PM2.5 standards.
In evaluating the adequacy of the
current standards, the Administrator
first recognizes the longstanding body of
health evidence supporting
relationships between PM2.5 exposures
(short- and long-term) and mortality and
serious morbidity effects. The evidence
available in this reconsideration (i.e.,
that assessed in the 2019 ISA and ISA
Supplement) and summarized above in
section II.A.2.a reaffirms, and in some
cases strengthens, the conclusions from
the 2009 ISA regarding the health effects
of PM2.5 exposures. Recent
epidemiologic studies demonstrate
generally positive and often statistically
significant associations between PM2.5
exposures and a number of health
effects, including non-accidental,
cardiovascular, or respiratory mortality;
cardiovascular or respiratory
hospitalizations or emergency room
visits; and other mortality/morbidity
outcomes (e.g., lung cancer mortality or
incidence, asthma development). Recent
controlled human exposure and animal
toxicological studies, as well as
evidence from epidemiologic panel
studies, strengthens support for
potential biological pathways through
which PM2.5 exposures could lead to the
serious effects reported in many
population-level epidemiologic studies,
including support for pathways that
could lead to cardiovascular,
respiratory, nervous system, and cancerrelated effects. In considering the
available scientific evidence, and
consistent with approaches employed in
past NAAQS reviews, the Administrator
places the most weight on evidence
supporting ‘‘causal’’ or ‘‘likely to be
causal’’ relationship with long or shortterm PM2.5 exposures. In addition, the
Administrator also takes note of those
populations identified to be at greater
risk of PM2.5-related health effects, as
characterized in the 2019 ISA and ISA
Supplement, and the potential public
health implications.
In evaluating what existing or revised
standards may be requisite to protect
public health, as described above in
section II.A.2, the Administrator’s
approach recognizes that the current
annual standard (based on arithmetic
mean concentrations) and 24-hour
standard (based on 98th percentile
concentrations), together, are intended
to provide public health protection
against the full distribution of short- and
long-term PM2.5 exposures. This
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
approach recognizes that changes in
PM2.5 air quality designed to meet either
the annual or the 24-hour standard
would likely result in changes to both
long-term average and short-term peak
PM2.5 concentrations.
Further, consistent with the approach
adopted in 2012, the Administrator
concludes that the most effective and
efficient way to reduce total population
risk associated with both long- and
short-term PM2.5 exposures is to set a
generally controlling annual standard,
and to provide supplemental protection
against the occurrence of peak 24-hour
PM2.5 concentrations by means of a 24hour standard set at the appropriate
level. In reaching this conclusion, the
Administrator explicitly recognizes that
air quality changes associated with
meeting a revised annual standard
would result in lowering risks
associated with both long- and shortterm PM2.5 exposures by lowering the
overall distribution of air quality
concentrations, leading to not only in
lower short- and long-term PM2.5
concentrations near the middle of the
air quality distribution, but also in fewer
and lower short-term peak PM2.5
concentrations. Similarly, the
Administrator recognizes that changes
in air quality to meet a 24-hour
standard, would result not only in fewer
and lower peak 24-hour PM2.5
concentrations, but also in lower annual
average PM2.5 concentrations. However,
as noted in 2012, he also recognizes that
an approach that relies on setting the
level of the 24-hour standard such that
the 24-hour standard is generally
controlling would be less effective and
result in less uniform protection across
the U.S. than an approach that focuses
on setting a generally controlling annual
standard. Thus, he concludes that
relying on a revised annual standard as
the controlling standard will reduce
aggregate risks associated with both
long- and short-term exposures more
consistently than a generally controlling
24-hour standard. He further concludes
that retaining a 24-hour standard at the
appropriate level will ensure an
adequate margin of safety against shortterm effects in areas with high peak-tomean ratios.
In light of his focus on the annual
standard as the generally controlling
standard, in considering whether the
primary PM2.5 standards are adequate,
the Administrator first considers
information available to inform his final
conclusions regarding the primary
annual PM2.5 standard. In so doing, he
notes that in this reconsideration, a
large number of key U.S. epidemiologic
studies report positive and statistically
significant associations for air quality
PO 00000
Frm 00074
Fmt 4701
Sfmt 4700
distributions with overall mean PM2.5
concentrations that are well below the
current level of the annual standard of
12.0 mg/m3. He further recognizes that
there is additional scientific evidence
assessed in the 2019 ISA and newly
assessed in this reconsideration in the
ISA Supplement that can provide
supplemental information to inform his
decisions. In addition to the key U.S.
epidemiologic studies, the
Administrator also recognizes that key
Canadian epidemiologic studies also
demonstrate positive and statistically
significant associations at
concentrations below 12 mg/m3. He also
recognizes that epidemiologic studies
that restrict annual average PM2.5
concentrations to below 12 mg/m3 also
provide support for positive and
statistically significant associations at
lower mean PM2.5 concentrations, as do
accountability studies that also suggest
public health improvements may occur
at concentrations below 12 mg/m3.
With regard to the available scientific
evidence to inform his final decisions
on the adequacy of the current 24-hour
standard, the Administrator finds that
there is less information available to
support decisions on the 24-hour
standard than that summarized above
for the annual standard. The
Administrator first notes that controlled
human exposure studies, including
those newly available in this
reconsideration, demonstrate effects
following short-term PM2.5 exposures at
concentrations higher than the current
24-hour standard. The Administrator
also considers air quality analyses
conducted in the 2022 PA and in
responding to public comments, as
described above in section II.B.3, that
evaluate PM2.5 concentrations in
ambient air for similar durations to the
controlled human exposure studies. As
noted above, these air quality analyses
indicate that the current 24-hour
standard, particularly in conjunction
with the revised level of the annual
standard, provides a high degree of
protection against subdaily PM2.5
concentrations that have been shown to
elicit effects in controlled human
exposure studies. The Administrator
considers a limited number of available
epidemiologic studies that report
associations with health effects when
the analyses are restricted to daily PM2.5
concentrations below 35 mg/m3. As
described above, although these studies
are useful in demonstrating that health
effects are associated with exposure to
daily PM2.5 concentrations in the lower
part of the air quality distribution, they
do not provide information about health
effects associated with the short-term
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
‘‘peak’’ exposures that the 24-hour
standard is designed to protect against.
Accordingly, these studies have limited
relevance in informing a decision about
the appropriate level of the 24-hour
standard.
In addition to the scientific evidence,
the Administrator also considers the
information from the risk assessment. In
so doing, he notes that the risk
assessment estimates that the current
primary annual PM2.5 standard could
allow a substantial number of deaths in
the U.S. With respect to the 24-hour
standard, the Administrator recognizes
that there are only a small number of
study areas where the 24-hour standard
is controlling and changes in the 24hour standard level are estimated to
have a much smaller impact on public
health. The Administrator recognizes
that while the risk estimates can help to
place the evidence for specific health
effects into a broader public health
context, they should be considered
along with the inherent uncertainties
and limitations of such analyses when
informing judgments about the potential
for additional public health protection
associated with PM2.5 exposure and
related health effects. While the
Administrator recognizes that these
uncertainties are important, he also
notes that the general magnitude of the
risk estimates provide support for
significant public health impacts,
particularly for lower alternative annual
standard levels.
In reaching his final conclusions
regarding the adequacy of the primary
PM2.5 standards, the Administrator also
considers the CASAC’s advice and
recommendations, as well as public
comments. With respect to the CASAC’s
advice, the Administrator recognizes
that, in their review of the 2021 draft
PA, the CASAC reached consensus that
the current primary annual PM2.5
standard is not adequate and that it is
not sufficiently protective of public
health. The Administrator also takes
note of the CASAC’s advice in their
review of the 2019 draft PA, where the
CASAC did not reach consensus on the
adequacy of the primary annual PM2.5
standard, with the minority
recommending revision and the
majority recommending the standard be
retained. Furthermore, he recognizes
that in reviewing the 2019 draft PA, the
CASAC reached consensus regarding
the adequacy of the primary 24-hour
PM2.5 standard, concluding that the
standard should be retained.
Conversely, in their review of the 2021
draft PA, the majority of the CASAC
advised that the current primary 24hour PM2.5 standard is not adequate and
recommended revising the level of the
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
standard, while the minority of the
CASAC concluded that the standard
was adequate and should be retained.
However, in considering the advice of
the CASAC collectively in the context of
this reconsideration, the Administrator
recognizes that the 2021 draft PA
included scientific evidence and
quantitative risk information that was
not available in the 2019 draft PA, and
therefore, the advice and
recommendations of the CASAC in their
review of the 2021 draft PA are based
on consideration of the full body of
scientific evidence available in this
reconsideration, including the evidence
evaluated in the 2019 ISA and the ISA
Supplement.
The Administrator recognizes that
much of the scientific evidence
available in this reconsideration was
also available in the 2019 ISA and was
considered by the then-Administrator
when he decided that the current
primary PM2.5 standards are requisite to
protect public health with an adequate
margin of safety. However, as described
in section I.C.5.b above, in reaching his
decision to reconsider the 2020 final
decision, the Administrator also
recognized that there were a number of
studies published since the literature
cutoff date of the 2019 ISA that were
raised by some members of the CASAC
in their review of the 2019 draft PA, in
public comments on the 2020 proposal,
and in the petitions for reconsideration.
As such, the expansion of the air quality
criteria in this reconsideration to
encompass both the 2019 ISA and the
additional scientific evidence evaluated
in the ISA Supplement, along with
evidence and updated quantitative
analyses in the 2022 PA also provided
an expanded record for the CASAC’s
review and public comments as a part
of this reconsideration. Taken together,
the 2019 ISA, ISA Supplement, and
2022 PA, along with the CASAC’s
advice and recommendations and
public comments, provide the
Administrator with additional
information for consideration in
reaching his final conclusions in this
reconsideration. As a result, the record
before him notably expands upon and
strengthens the basis for the conclusions
of the 2019 ISA while reducing some
uncertainties that were identified in the
2020 final action.
In considering the available
information in this reconsideration, the
current Administrator reached different
conclusions regarding the appropriate
weight to place on certain aspects of the
evidence than the then-Administrator in
the 2020 final decision. For example, in
reaching his conclusions on the primary
annual PM2.5 standard in 2020, the then-
PO 00000
Frm 00075
Fmt 4701
Sfmt 4700
16275
Administrator concluded that it was
appropriate to place more weight on
epidemiologic studies that used groundbased monitors and to place less weight
on the studies that used hybrid modelbased approaches, citing to increased
uncertainties associated with this new
and emerging approach to estimating
exposure. In placing more weight on the
key U.S. monitor-based studies, the
then-Administrator noted that the
majority of these studies had mean
concentrations at or above the level of
the annual standard (12.0 mg/m3).
However, unlike the approach for
considering such studies in the 2012
review, the then-Administrator
concluded that it was appropriate to
consider the study-reported means
collectively, and in so doing, he placed
weight on the average of the studyreported means (or medians) across the
U.S. monitor-based studies of 13.5 mg/
m3, and noted that this concentration
was above the level of the standard (85
FR 82717, December 18, 2020). In
contrast, in this reconsideration, the
current Administrator judges that it is
appropriate to consider the individual
study-reported mean PM2.5
concentrations from not only the U.S.
monitor-based epidemiologic studies,
but also the U.S. hybrid model-based
epidemiologic studies, which are an
advancement in the available science
since the completion of the 2009 ISA.
The current Administrator also adopts
an approach similar to some previous
approaches for the PM NAAQS in
which he judges it most appropriate to
set the level of the standard to
somewhat below the lowest long-term
study-reported mean PM2.5
concentration reported in key U.S.
epidemiologic studies, which is 9.3 mg/
m3. The study that reports the long-term
mean PM2.5 concentration of 9.3 mg/m3
is newly available in this
reconsideration and is evaluated in the
ISA Supplement. In the 2019 ISA, the
lowest long-term study-reported mean
PM2.5 concentrations for U.S.-based
studies that use ground-based monitors
and hybrid model-based approaches are
9.9 mg/m3 and 10.7 mg/m3, respectively.
In judging that it is appropriate to
consider both monitor- and hybrid
model-based epidemiologic studies and
that it is appropriate to adopt an
approach to set the level of the standard
to somewhat below the lowest long-term
mean PM2.5 concentration, the current
Administrator judges that the available
scientific evidence—evaluated in both
the 2019 ISA and in the ISA
Supplement—provide support for his
conclusion that that current primary
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16276
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
PM2.5 standard is not adequate and
should be revised.
In addition to adopting a different
approach than the previous
Administrator for considering the longterm mean PM2.5 concentrations from
key U.S. epidemiologic studies (one
more consistent with the approach of
the EPA in other prior reviews), the
current Administrator both has
information newly available in this
reconsideration before him and is
reaching different conclusions about
how to weigh the evidence before him
in reaching his final conclusions. For
example, in reaching his final decision
in 2020, the then-Administrator was
concerned about placing too much
weight on epidemiologic studies to
inform his conclusions on the adequacy
of the primary PM2.5 standards, noting
that the epidemiologic studies do not
identify particular PM2.5 concentrations
that cause effects and cannot alone
identify a specific level at which to set
the standard. In so doing, the thenAdministrator placed greater weight on
the uncertainties and limitations
associated with the epidemiologic
studies, including exposure
measurement error, potential
confounding by copollutants, increased
uncertainty of associations at lower
PM2.5 concentrations, and heterogeneity
of effects across different cities or
regions (85 FR 82716, December 18,
2020). The Administrator recognizes
that in reaching these judgments, the
then-Administrator took into
consideration the views of some
members of the CASAC, who, in their
advice on the 2019 draft PA, expressed
the view that the current PM NAAQS
should be retained because reported
associations between short- and longterm PM2.5 exposures and adverse
health outcomes ‘‘can reasonably be
explained in light of uncontrolled
confounding and other potential sources
of error and bias’’ (Cox, 2019b, p. 8 of
consensus responses).
In this reconsideration, the current
Administrator notes that the ISA
Supplement evaluates additional
studies that employed statistical
approaches that attempted to more
extensively account for confounders and
are more robust to model
misspecification (i.e., used alternative
methods for confounder control, which
are sometimes referred to as causal
modeling or causal inference methods)
that build upon those studies available
and evaluated in the 2019 ISA (U.S.
EPA, 2019, sections 11.1.2.1 and
11.2.2.4). These studies report
consistent positive associations between
long-term and short-term PM2.5
exposure and total mortality and
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
cardiovascular effects (U.S. EPA, 2022a,
section 3.2.2.3). In considering the
epidemiologic evidence evaluated in the
2019 ISA, along with the newly
available studies evaluated in the ISA
Supplement, the current Administrator
also recognizes that there are
uncertainties and limitations associated
with the epidemiologic studies, but
judges that it is appropriate to place less
weight on these uncertainties than the
then-Administrator placed on them in
reaching his final decision in 2020,
given the strength of the longstanding
large body of epidemiologic evidence,
employing a variety of study designs,
that demonstrates associations between
long- and short-term PM2.5 exposures
and health effects across multiple U.S.
cities and in diverse populations,
including in studies examining
populations and lifestages that may be
at comparatively higher risk of
experiencing a PM2.5-related health
effect (e.g., older adults, children).
In reaching this final decision, the
Administrator recognizes he is differing
not only with the prior Administrator
but also with the advice some members
of the CASAC provided during their
review of the 2019 draft PA.
Specifically, taking into consideration
the strength of the evidence providing
support for causality determinations,
the advice of other members of the
CASAC and the need to protect public
health with an adequate margin of
safety, the current Administrator
disagrees with these members of CASAC
regarding the weight to be given to
epidemiologic evidence ‘‘based on its
methodological limitations’’ (Cox,
2019b, p. 8 of consensus responses),
such as the possibility ‘‘that such
associations could reasonably be
explained by uncontrolled confounding
and other potential sources of error and
bias’’ (Cox, 2019b, p. 8 of consensus
responses).
As another example of information
that was not available to the CASAC in
providing advice to the Administrator in
reaching his final decision in 2020, the
then-Administrator noted in his final
decision that, while some members of
the CASAC and public commenters
highlighted a number of accountability
studies that examined past reductions in
ambient PM2.5 concentrations and the
degree to which those reductions have
resulted in public health improvements,
the small number of available
accountability studies did not examine
air quality with starting concentrations
meeting the primary annual PM2.5
standard of 12.0 mg/m3. The thenAdministrator took into consideration
the absence of such accountability
studies, as part of his consideration of
PO 00000
Frm 00076
Fmt 4701
Sfmt 4700
the full body of scientific evidence, in
reaching his judgment that there was
considerable uncertainty in the
potential for increased public health
protection from further reductions in
ambient PM2.5 concentrations beyond
those achieved under the existing
primary PM2.5 NAAQS (85 FR 82717,
December 18, 2020). However, there are
several accountability studies available
since the literature cutoff date of the
2019 ISA and evaluated in the ISA
Supplement in this reconsideration that
have starting concentrations (or
concentrations prior to the policy or
intervention) below 12.0 mg/m3
(Corrigan et al, 2018; Henneman et al.,
2019; Sanders et al., 2020a). The current
Administrator concludes that, while the
number of available accountability
studies is limited, he recognizes that
these studies provide supplemental
information for consideration for
informing decisions on the appropriate
level of the primary annual PM2.5
standard along with the full body of
evidence.
As EPA has frequently noted
throughout this document, the extent to
which the current primary PM2.5
standards are judged to be adequate
depends in part on science policy and
public health policy judgments to be
made by the Administrator on the
strength and uncertainties of the
scientific evidence, such as how to
consider epidemiologic evidence and
the need for an adequate margin of
safety in setting the standards. Thus, it
would be pure speculation to guess
whether the then-Administrator would
have reached the same or different
conclusions in the 2020 final decision
had the record before him included the
newly available information in this
reconsideration.107 However, the
current Administrator concludes that,
for the reasons explained herein that, in
his judgment, based on the record before
him in this reconsideration, it is
necessary and appropriate to revise the
primary annual PM2.5 NAAQS to
provide requisite protection of public
health with an adequate margin of
safety.
Based on the available scientific
evidence and quantitative information,
as well as consideration of the CASAC’s
advice and public comments, the
Administrator concludes that the
107 The EPA notes that, in considering the
additional scientific evidence available in this
reconsideration, one member of the CASAC who
reviewed both the 2019 draft PA and the 2021 draft
PA found that the available scientific and
quantitative information available in this
reconsideration supported revising the level of the
primary annual PM2.5 standard to within the range
of 10–11 mg/m3, whereas he recommended retaining
the standard during the review of the 2019 draft PA.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
current primary annual PM2.5 standard
is not adequate to protect public health
with an adequate margin of safety. In
addition, he finds the available
information insufficient to call into
question the adequacy of the public
health protection afforded by the
current primary 24-hour PM2.5 standard.
In considering how to revise the
current suite of primary PM2.5 standards
in order to achieve the requisite
protection for public health, with an
adequate margin of safety, against longand short-term PM2.5 exposures the
Administrator considers the four basic
elements of the NAAQS (indicator,
averaging time, form, and level)
collectively. With respect to indicator,
the Administrator recognizes that the
scientific evidence in this
reconsideration, as in previous reviews,
continues to provide strong support for
health effects associated with PM2.5
mass. He notes the 2022 PA conclusion
that the available information continues
to support the PM2.5 mass-based
indicator and remains too limited to
support a distinct standard for any
specific PM2.5 component or group of
components, and too limited to support
a distinct standard for the ultrafine
fraction of PM (U.S. EPA, 2022b, section
3.6.3.2.1). In its advice on the adequacy
of the current primary PM2.5 standards
in their review of the 2021 draft PA, the
CASAC reached consensus that the
PM2.5 mass-based indicator should be
retained, without revision (Sheppard,
2022a, p. 2 of consensus letter).108
Additionally, there was no information
in the public comments that provided a
rationale for an alternative indicator.
For all of these reasons, the
Administrator concludes that it is
appropriate to retain PM2.5 mass as the
indicator for the primary standards for
fine particles.
Consistent with his proposed
conclusions regarding averaging time,
the Administrator notes that the
scientific evidence continues to provide
strong support for health effect
associations with both long- and shortterm PM2.5 exposures (88 FR 5618,
January 27, 2023). Epidemiologic
studies continue to provide strong
support for health effects associated
with short-term PM2.5 exposures based
on 24-hour averaging periods, and
associations in epidemiologic studies
with subdaily estimates are less
consistent and, in some cases, smaller in
magnitude (88 FR 5618, January 27,
2023). Taken together, the 2019 ISA
concludes that epidemiologic studies do
not indicate that subdaily averaging
periods are more closely associated with
health effects than the 24-hour average
exposure metric (U.S. EPA, 2019a,
section 1.5.2.1). In addition, controlled
human exposure and panel-based
studies of subdaily exposures typically
examine subclinical effects rather than
the more serious population-level
effects that have been reported to be
associated with 24-hour exposures (e.g.,
mortality, hospitalizations). While
recent controlled human exposure
studies provide consistent evidence for
cardiovascular effects following PM2.5
exposures for less than 24 hours (i.e.,
<30 minutes to 5 hours), air quality
analyses have shown that the current
averaging times can effectively protect
against the exposure concentrations in
these studies. This information does not
indicate that a revision to the averaging
time is necessary to provide additional
protection against subdaily PM2.5
exposures, beyond that provided by the
current primary annual and 24-hour
PM2.5 standards. The Administrator also
notes that this conclusion is also
support by the CASAC’s advice in their
review of the 2021 draft PA where they
reached consensus that averaging times
for the primary PM2.5 standards should
be retained, without revision (Sheppard,
2022a, p. 2 of consensus letter).109 The
Administrator also considers the
relatively few public comments received
that support a subdaily averaging time,
but concludes that the currently
available information does not provide
support for an alternate averaging time.
Consistent with his proposed decision,
the Administrator concludes that it is
appropriate to retain the annual and 24hour averaging times for the primary
PM2.5 standards to protect against longand short-term PM2.5 exposures.
With regard to form, the
Administrator first notes that the EPA
has set both an annual standard and a
24-hour standard to provide protection
from health effects associated with both
long- and short-term exposures to PM2.5
(62 FR 38667, July 18, 1997; 88 FR 5620,
January 27, 2023). With regard to the
form of the annual standard, the
Administrator recognizes that a large
majority of the recently available
epidemiologic studies continue to report
associations between health effects and
annual average PM2.5 concentrations.
These studies of annual average PM2.5
concentrations provide support for
retaining the current form of the annual
108 The CASAC did not provide advice or
recommendations regarding the indicator of the
primary PM2.5 standards in their review of the 2019
draft PA (Cox, 2019b).
109 The CASAC did not provide advice or
recommendations regarding the averaging times of
the primary PM2.5 standards in their review of the
2019 draft PA (Cox, 2019b).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
PO 00000
Frm 00077
Fmt 4701
Sfmt 4700
16277
standard to provide protection against
long- and short-term PM2.5 exposures. In
its review of the 2021 draft PA, the
CASAC reached consensus that the form
of the annual standard (i.e., annual
mean, averaged over 3 years) should be
retained, without revision (Sheppard,
2022a, p. 2 of consensus letter).110 The
Administrator also notes that there were
no public comments that recommended
an alternative form for the primary
annual PM2.5 standard.
With regard to the form of the 24-hour
standard (98th percentile, averaged over
three years), epidemiologic studies
continue to provide strong support for
health effect associations with shortterm (e.g., mostly 24-hour) PM2.5
exposures, and controlled human
exposure studies provide evidence for
health effects following single shortterm ‘‘peak’’ PM2.5 exposures (88 FR
5618, January 27, 2023). Therefore, the
Administrator concludes that the
evidence supports retaining a standard
focused on providing supplemental
protection against short-term peak
exposures and supports a 98th
percentile form for a 24-hour standard,
in combination with a primary annual
PM2.5 standard with its annual mean
averaged over three years form. As
described in the proposal and in
responding to comments in section
II.B.3 above, the Administrator further
notes that the 98th percentile, averaged
over three years, form also provides an
appropriate balance between limiting
the occurrence of peak 24-hour PM2.5
concentrations and identifying a stable
target for risk management programs
(U.S. EPA, 2022b, section 3.6.3.2.3).
Furthermore, the Administrator notes
that the multi-year percentile form (i.e.,
averaged over three years) offers greater
stability to the air quality management
process by reducing the possibility that
statistically unusual indicator values
will lead to transient violations of the
standard. This conclusion is also
supported by the CASAC’s advice in
their review of the 2021 draft PA, where
they reached consensus that the form for
the primary PM2.5 standards should be
retained, without revision (Sheppard,
2022a, p. 2 of consensus letter).111
The Administrator also recognizes
that the CASAC recommended that in
future reviews, the EPA also consider
alternative forms for the primary 24hour PM2.5 standard (Sheppard, 2022a,
110 The CASAC did not provide advice or
recommendations regarding the forms of the
primary PM2.5 standards in their review of the 2019
draft PA (Cox, 2019b).
111 The CASAC did not provide advice or
recommendations regarding the forms of the
primary PM2.5 standards in their review of the 2019
draft PA (Cox, 2019b).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16278
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
p. 18 of consensus responses). Based on
the CASAC’s advice, the proposal
solicited comment on alternatives to the
current form for consideration in future
reviews (88 FR 5619, January 27, 2023).
The Administrator recognizes that there
were a limited number of public
comments related to the form of the
primary PM2.5 standards as discussed in
section II.D.3 above and in the Response
to Comments document, and notes that,
the EPA will consider the information
provided by the commenters regarding
the form of the 24-hour PM2.5 standard
in the next review of the PM NAAQS.
Consistent with his proposed decision,
in considering the information
summarized above, the Administrator
concludes that it is appropriate to retain
the forms of the current annual and 24hour PM2.5 standards.
In considering how to revise the
current suite of PM2.5 standards to
provide the requisite public health
protection with an adequate margin of
safety, the Administrator next evaluates
the appropriate levels of the primary
PM2.5 standards, beginning with the
annual PM2.5 standard. In having
carefully considered public comments
related to the primary annual PM2.5
standard, the Administrator believes
that the fundamental conclusions
regarding the scientific evidence and
quantitative information that supported
his proposed conclusions (as described
in the 2019 ISA, ISA Supplement, 2022
PA, and the proposal) remain valid. In
considering the level at which the
primary annual PM2.5 standard should
be set, the Administrator considers the
entire body of evidence and
information, giving appropriate weight
to each part of that body of evidence
and information. He continues to place
the greatest weight in this
reconsideration on the available
scientific evidence that provides
support for associations between health
effects and long- and short-term PM2.5
exposures. In conjunction with his
decisions to retain the current indicator,
averaging time, and form as described
above, the Administrator is revising the
level of the primary annual PM2.5
standard to 9.0 mg/m3. In so doing, he
is selecting a primary annual PM2.5
standard that, together with the primary
24-hour PM2.5 standard, provides
requisite public health protection with
an adequate margin of safety, based on
his judgments about and interpretation
of the scientific evidence and
quantitative risk information.
The Administrator’s decision to revise
the level of the primary annual PM2.5
standard to 9.0 mg/m3 builds upon his
conclusion that the overall body of
scientific evidence and quantitative risk
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
information calls into question the
adequacy of public health protection
afforded by the current standard,
particularly for at-risk populations.
Consistent with his consideration of the
available information in reaching his
proposed decisions, the Administrator’s
final decision on the level of the
primary annual PM2.5 standard places
the greatest emphasis on key U.S.
epidemiologic studies that report
associations between long- and shortterm PM2.5 exposures and mortality and
morbidity. As in the proposal, and as
discussed further below, he views
additional epidemiologic studies (i.e.,
studies that employ alternative methods
for confounding control, studies that
employ restricted analyses, and
accountability studies), the controlled
human exposure studies, and the risk
assessment as providing supplemental
information in support of his decision to
revise the current annual standard, but
recognizes that some of these lines of
evidence and information provide a
more limited basis for selecting a
particular standard level among a range
of options. See Mississippi, 744 F. 3d at
1351–52 (studies can legitimately
support a decision to revise the
standard, but not provide sufficient
information to justify their use in setting
the level of a revised standard).
Given his consideration of the
scientific evidence, quantitative risk
information, advice from the CASAC,
and public comments, the
Administrator judges that a primary
annual PM2.5 standard with a level of
9.0 mg/m3 is requisite to protect public
health with an adequate margin of
safety. He notes that the determination
of what constitutes an adequate margin
of safety is expressly left to the
judgment of the EPA Administrator. See
Lead Industries Association v. EPA, 647
F.2d at 1161–62; Mississippi, 744 F.3d
at 1353. He further notes that in
evaluating how particular standards
address the requirement to provide an
adequate margin of safety, it is
appropriate to consider such factors as
the nature and severity of the health
effects, the size of the at-risk
populations, and the kind and degree of
the uncertainties present. In considering
the need for an adequate margin of
safety, the Administrator notes that a
primary annual PM2.5 standard with a
level of 9.0 mg/m3 would be expected to
provide substantial improvements in
public health compared to the current
annual standard, including for at-risk
groups such as children, older adults,
people with preexisting conditions,
minority populations, and low SES
populations.
PO 00000
Frm 00078
Fmt 4701
Sfmt 4700
Consistent with his conclusions on
the need for revision of the current
annual standard, in reaching a decision
on level, the Administrator places the
most weight on information from
epidemiologic studies. In so doing, the
Administrator notes that these studies
provide consistent evidence of positive
and statistically significant associations
between long- and short-term exposure
to PM2.5 and mortality and morbidity
(88 FR 5624, January 27, 2023). The
Administrator recognizes that placing
weight on the information from the
epidemiologic studies allows for
examination of the entire population,
including those that may be at
comparatively higher risk of
experiencing a PM2.5-related health
effects (e.g., children, older adults,
minority populations) (88 FR 5624,
January 27, 2023). The Administrator
also recognizes that recent
epidemiologic studies continue to
support a no-threshold relationship,
meaning that there is no ‘‘bright line’’
below which no effects have been
found. These studies also support a
linear relationship between health
effects and PM2.5 exposures at PM2.5
concentrations greater than 8 mg/m3,
though uncertainties remain about the
shape of the C–R curve at PM2.5
concentrations less than 8 mg/m3, with
some recent studies providing evidence
for either a sublinear, linear, or
supralinear relationship at these lower
concentrations (U.S. EPA, 2019a,
section 11.2.4; U.S. EPA, 2022a, section
2.2.3.2; 88 FR 5625, January 27, 2023).
As at the time of proposal, the
Administrator notes that some recent
epidemiologic studies have adopted a
broad range of approaches to examine
confounding and the results of those
examinations support the robustness of
reported associations seen in
epidemiologic studies. These include
studies that employ alternative methods
for confounder control and studies that
evaluate the uncertainty related to
exposure measurement error, both of
which continue to support associations
between PM2.5 exposures and health
effects while taking approaches to
address uncertainties.
In considering the epidemiologic
evidence, the Administrator judges that,
in reaching his decision on an
appropriate level for the annual
standard that will protect public health
with an adequate margin of safety, in
the absence of any discernible
population-level thresholds, and in
recognizing the need to weigh
uncertainties associated with the
epidemiologic evidence, it is most
appropriate to examine where the
evidence of associations observed in the
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
epidemiologic studies is strongest and,
conversely, to place less weight where
he has less confidence in the
associations observed in the
epidemiologic studies. As at the time of
proposal, the Administrator notes that
in previous reviews, evidence-based
approaches noted that the evidence of
an association in any epidemiologic
study is ‘‘strongest at and around the
long-term average where the data in the
study are most concentrated’’ (78 FR
3140, January 15, 2013). Given this,
these approaches focused on identifying
standard levels near or somewhat below
long-term mean concentrations reported
in key epidemiologic studies. These
approaches were supported by previous
CASAC advice as well as the CASAC’s
advice in their review of the 2021 draft
PA as a part of this reconsideration.
Additionally, the Administrator
acknowledges that in the 2020 final
action, the then-Administrator decided
to retain the standard based in part on
concerns about placing reliance on the
epidemiologic studies and his judgment
that even if he did rely on them, the
majority of the studies had means or
medians, as well as the mean of all of
the key study-reported means or
medians, above the level of the current
annual standard. However, after
considering the evidence, the advice of
CASAC, and public comments the
Administrator judges that this approach
is insufficient to protect public health
with an adequate margin of safety. The
Administrator’s decision to reach a
different judgment about the
appropriate level of the annual standard
reflects the updated and expanded
scientific record available to the
Administrator in this reconsideration, as
well as the additional advice from the
CASAC and the public comments based
on this newly available information. In
addition, the Administrator observes the
decision in this action to place weight
on the epidemiologic studies, and to
revise the annual primary standard to a
level below the lowest long-term mean
in the U.S.-based epidemiologic studies,
is consistent with the EPA’s past
practice in PM NAAQS reviews.
In this reconsideration, the
Administrator is considering the
scientific record which has been
expanded and updated since the 2020
final action, as well as the additional
advice from the CASAC and the public
comments that are based on the newly
available information that expands upon
the information previously available. In
addition, the Administrator is exercising
his judgment about how to interpret and
weigh the expanded evidence in a way
that is more consistent with the
approaches used in prior PM NAAQS
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
reviews. As a result, the Administrator
has concluded on reconsideration that
the level of the primary annual standard
is not adequate and should be revised to
protect public health with an adequate
margin of safety.
Consistent with his proposed
decisions, in reaching conclusions on
the level of the primary annual PM2.5
standard, the Administrator considers
the long-term 112 study-reported mean
PM2.5 concentrations from key long- and
short-term epidemiologic studies and
sets the level of the standard to
somewhat below the lowest long-term
mean PM2.5 concentration.113 He notes
that in previous PM NAAQS reviews
(including the 1997, 2006 and 2012
reviews), evidence-based approaches
focused on identifying standard levels
near or somewhat below long-term
mean concentrations reported in key
long- and short-term epidemiologic
studies. These approaches were
supported by the CASAC in previous
reviews and were supported in this
reconsideration by the CASAC in their
review of the 2021 draft PA. In
considering the available scientific
evidence to inform such an approach,
the Administrator notes the strength of
the epidemiologic evidence which
includes multiple studies that
consistently report positive associations
for short- and long-term PM2.5 exposure
and mortality and cardiovascular
effects. Some available studies also use
a variety of statistical methods to
control for confounding bias and report
similar associations, which further
supports the broader body of
epidemiologic evidence for both
mortality and cardiovascular effects.
Additionally, he notes that recent
epidemiologic studies available for
consideration in reaching his final
decision strengthen support for health
effect associations at PM2.5
concentrations lower than in those
evaluated in epidemiologic studies
available at the time of previous
reviews. The Administrator does
recognize, however, that while these
epidemiologic studies evaluate
associations between distributions of
ambient PM2.5 concentrations and
health outcomes, they do not identify
the specific exposures that led to the
reported effects. As such, he notes that
there is no specific point in the air
quality distribution of any
112 ‘‘Long-term’’ represents PM
2.5 exposures and
concentrations that are annual or multi-year.
113 As described in section II.A.2.c above, key
epidemiologic studies are those that report overall
mean (or median) PM2.5 concentrations and for
which the years of PM2.5 air quality data used to
estimate exposures overlap entirely with the years
during which health events are reported.
PO 00000
Frm 00079
Fmt 4701
Sfmt 4700
16279
epidemiologic study that represents a
‘‘bright line’’ at and above which effects
have been observed and below which
effects have not been observed. The
Administrator further notes that the
epidemiologic studies provide the
strongest support for reported health
effect associations for this middle
portion of the PM2.5 air quality
distribution, which corresponds to the
bulk of the underlying data, rather than
the extreme upper or lower ends of the
distribution, and concludes that the
long-term study-reported means from
both long- and short-term studies
provide the strongest support for
reported health effect associations in
epidemiologic studies. For these
reasons, as described in the proposal
and in responding to public comments
in section II.B.3 above, the
Administrator concludes that it is
appropriate to continue to employ an
approach that focuses on the mean
PM2.5 concentrations from the key
epidemiologic studies to inform his
conclusions regarding the appropriate
level for the primary annual PM2.5
standard.
In adopting such an approach, the
Administrator considers the long-term
mean concentrations reported in two
types of key epidemiologic studies: (1)
Monitor-based studies 114
(epidemiologic studies that used
ground-based monitors to estimate
exposure, similar to approaches used in
past reviews), and (2) hybrid modelingbased studies 115 (epidemiologic studies
that used hybrid modeling approaches
and apply aspects of population
weighting to estimate exposures). In
reaching conclusions regarding the level
of a standard that would provide
requisite protection with an adequate
margin of safety, the Administrator
recognizes that he must use his
judgment regarding the appropriate
weight to place on the available
evidence and technical information,
including uncertainties. As shown in
Figures 1 and 2 above, for the key U.S.
monitor-based epidemiologic studies,
114 Reported mean PM
2.5 concentrations in
monitor-based studies are averaged across monitors
in each study area with multiple monitors, referred
to as a composite monitor concentration, in contrast
to the highest concentration monitored in the study
area, referred to as a maximum monitor
concentration (i.e., the ‘‘design value’’
concentration), which is used to determine whether
an area meets a given standard.
115 Studies that use hybrid modeling approaches
employ methods to estimate ambient PM2.5
concentrations across large geographical areas,
including areas without monitors, and thus, when
compared to monitor-based studies, require
additional information to inform the relationship
between the estimated PM2.5 concentrations across
an area and the maximum monitor design values
used to assess compliance.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16280
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
the study-reported mean concentrations
range from 9.9–16.5 mg/m3, and for the
key U.S. hybrid modeling-based
epidemiologic studies, the mean
concentrations range from 9.3–12.2 mg/
m3. The Administrator also recognizes
that, in their review of the 2021 draft
PA, both the majority and minority of
the CASAC emphasized the
epidemiologic studies in support of
their recommendations for the level of
the annual standard, but they weighed
the studies in different ways (Sheppard,
2022a, p. 16–17 of consensus
responses).
Based on this information, and in
considering the CASAC’s advice in their
review of the 2021 draft PA, the
Administrator judges that it is
appropriate to set the level of the
primary PM2.5 standard at least as low
as the lowest mean PM2.5 concentration
from these key U.S.-based
epidemiologic studies, which is 9.3 mg/
m3. The Administrator additionally
notes that setting the annual standard
level at 9.0 mg/m3, which is below the
lowest study-reported mean PM2.5
concentration of 9.3 mg/m3, would be
expected to shift the distribution of
PM2.5 concentrations in an area such
that the area’s highest monitor would
generally be at or below 9.0 mg/m3
annually, when meeting the annual
standard. In this situation, the resulting
average or mean PM2.5 concentration for
the entire area (measured across a
number of monitors) would be even
further below the study-reported
means,116 and will provide adequate
protection not only in areas where the
highest allowable concentrations would
be expected (i.e., near design value
monitors) but also in other parts of the
area where PM2.5 concentrations would
be expected to be maintained even
lower.
As noted above, however, the
Administrator must exercise his
judgment regarding the appropriate
weight to place on the available
scientific evidence and quantitative
information, including uncertainties, in
determining what level of the annual
standard is sufficient to protect public
health with an adequate margin of
safety. In so doing, he considers other
information available in this
reconsideration to inform his
judgments, including study-reported
PM2.5 concentrations at lower
percentiles in key epidemiologic
studies, supplemental information from
116 Analyses in the 2022 PA suggest that the
highest monitored value would be expected to be
greater than the study-reported mean values by 10–
20% for monitor-based studies and 15–18% for
hybrid modeling studies that apply aspects of
population weighting.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
other types of epidemiologic studies,
study-reported PM2.5 concentrations
from key Canadian epidemiologic
studies, and the results from the
quantitative risk assessment.
In weighing the evidence in
considering the requisite level of the
annual standard, the Administrator also
takes into account additional
information from the key long- and
short-term U.S. epidemiologic studies
available that provide study-reported
PM2.5 concentrations below the mean
and, in particular, the subset of
epidemiologic studies that report 25th
and 10th percentile concentrations.
Consistent with his proposed
conclusions, as well as the CASAC’s
advice in their review of the 2021 draft
PA and public comments, the
Administrator judges that it is
appropriate to place some weight on
these lower percentiles in reaching his
conclusions on the level of the primary
annual standard. There are six key U.S.
epidemiologic studies that report
information on other percentiles (e.g.,
10th and 25th percentiles of PM2.5
concentrations or 10th and 25th
percentiles of PM2.5 concentrations
associated with health events) that are
below the mean.117 In considering the
information from these studies, the
Administrator first notes that the three
older, monitor-based studies that report
lower percentiles of PM2.5
concentrations have smaller cohort sizes
than the three hybrid model-based
studies. Thus, the Administrator
recognizes that the older, monitor-based
studies had a relatively smaller portion
of the health events that were observed
in the lower part of the air quality
distribution because of the generally
smaller size of the cohorts. He further
notes that the recent hybrid modelbased studies have larger cohort sizes
than the older, monitor-based studies,
and therefore, have more health events
in the lower part of the air quality
distribution. Because of the larger
cohort sizes and having a larger portion
of health events that are observed across
the air quality distribution, the
Administrator has more confidence in
the magnitude and significance of the
associations in the lower parts of the air
quality distribution for the recent,
hybrid model-based studies compared
to the older, monitor-based studies.
Given this, the Administrator judges
that it is appropriate to place weight on
the 25th percentile concentrations
reported in the recently available hybrid
model-based studies in reaching his
117 The Wang et al. (2017) study only reports the
25th percentile of the estimated PM2.5
concentrations, not the 10th percentile.
PO 00000
Frm 00080
Fmt 4701
Sfmt 4700
conclusions regarding the appropriate
level for the primary annual PM2.5
standard. However, the Administrator
also recognizes that his confidence in
the magnitude and significance in the
reported concentrations, and their
ability to inform decisions on the
appropriate level of the annual
standard, starts to diminish at
percentiles that are even further below
the mean and the 25th percentile. For
these reasons, the Administrator places
weight on the reported 25th percentiles
concentrations in the recent hybrid
model-based studies, rather than the
reported 10th percentile concentrations,
in reaching his conclusions regarding
the appropriate level for the primary
annual PM2.5 standard.
In considering the information from
these studies, as described in section
II.A.2.c and in responding to public
comments in section II.B.3 above, the
Administrator notes that there are two
hybrid model-based studies with large
cohort sizes that apply population
weighting and report lower percentile
values. These studies are Di et al.
(2017b) and Wang et al. (2017) and the
reported 25th percentile concentration
is 9.1 mg/m3 for both studies.118 In
considering these studies, the
Administrator concludes that it is
appropriate to place weight on the 25th
percentile concentrations of these newer
hybrid model-based studies (of 9.1 mg/
m3) such that setting the level of the
standard near these 25th percentile
concentrations would provide requisite
protection. The Administrator observes
that an annual standard level of 9.0 mg/
m3 would be near the reported 25th
percentile concentrations in these
studies.
As at the time of proposal, the
Administrator also takes note of the
study-reported long-term mean PM2.5
concentrations in long- and short-term
Canadian epidemiologic studies, which
ranged from 6.9 to 13.3 mg/m3 for
monitor-based studies and 5.9 to 9.8 mg/
m3 for hybrid model-based studies.
While the Administrator notes that
these studies provide additional support
for associations between PM2.5
concentrations and health effects, he is
also mindful that there are important
differences between the exposure
environments in the U.S. and Canada
and that interpreting the data (e.g.,
study-reported mean concentrations)
118 There is a third hybrid model-based study, as
described in the 2022 PA and in section II.B.3 above
in responding to public comments, but it is not
referenced here because it reports a 25th percentile
PM2.5 concentration based on the 25th percentile of
health events that occur in the study (Di et al.,
2017a) rather than report the 25th percentile based
on air quality concentrations.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
from the Canadian studies in the context
of a U.S.-based standard may present
challenges in directly and quantitatively
informing decisions regarding potential
alternative levels of the annual
standard. For example, in terms of
people per square kilometer, the U.S.
population density is nearly 10 times in
the contiguous U.S. compared to
Canada. As described in more detail in
responding to public comments in
section II.B.3 above, in this
reconsideration, the Administrator
recognizes that this difference in
population density between the U.S.
and Canada is more apparent than in
previous reviews because the studies
available in this reconsideration use
different approaches than those
previously available. In the 2012 review,
the available Canadian epidemiologic
studies used population-weighting and
focused on urban areas where monitors
were available and population densities
were more comparable with those in the
U.S., and at that time, the U.S. and
Canadian studies reported similar mean
PM2.5 concentrations. However, in this
reconsideration, the Administrator takes
note that for the new Canadian
epidemiologic studies: (1) The Canadian
monitor-based studies available in this
reconsideration do not apply population
weighting as the previously available
studies did; and (2) some of the studies
now use hybrid modeling approaches
for estimating exposure. The
Administrator recognizes that these
differences are important to consider in
reaching conclusions on how these
Canadian epidemiologic studies should
be interpreted regarding decisions on
the requisite level of the primary annual
PM2.5 standard. Specifically, the
Administrator notes that the more
recent Canadian studies that use hybrid
modeling incorporate larger portions of
the country, and therefore include more
rural areas. The more rural areas that are
included in the study using the hybrid
modeling approaches, the more
important it is to consider how the
population densities and exposure
environments differ between the U.S.
and Canada. Additionally, the
Administrator notes that for hybrid
modeling-based studies there is less
certainty in PM2.5 exposure estimates in
more rural areas, which are further from
air quality monitors and where PM2.5
concentrations in the ambient air tend
to be lower. For these hybrid modelbased studies, the portion of the rural
areas that are contributing to the studyreported mean PM2.5 concentrations in
these studies is unclear. For these
reasons, the Administrator concludes
that it is important to consider the
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
differences between the population
exposures in the U.S. and Canadian
study areas and how these differences
influence the interpretation of the
epidemiologic study results.
Thus, the Administrator considers the
Canadian studies to inform his
judgments on what level for the annual
standard is requisite in light of the
limitations and challenges presented.
The Administrator also recognizes that
the majority of the CASAC in their
review of the 2021 draft PA, as well as
a number of public commenters, place
weight on the Canadian epidemiologic
studies in recommending that the level
of the primary annual PM2.5 standard be
revised to 8–10 mg/m3. The
Administrator further notes while the
majority of the CASAC advised the EPA
to consider the Canadian studies in
revising the annual standard level to
within the range of 8.0–10.0 mg/m3, they
did not advise the EPA to set the annual
standard level below the study-reported
means from those studies. Given these
considerations, the Administrator
judges that it is appropriate to set the
level of annual standard within the
range of 8–10 mg/m3 to be consistent
with the majority of the CASAC’s advice
in their consideration of these studies.
The Administrator also recognizes
that information from epidemiologic
studies that included analyses that
restrict annual average PM2.5
concentrations to concentrations below
the level of the current annual standard
can be useful for informing conclusions
regarding the appropriate level of the
primary annual PM2.5 standard. In so
doing, he particularly notes the two key
U.S. epidemiologic studies (Di et al.,
2017b and Dominici et al., 2019) that
restrict annual average PM2.5
concentrations to less than 12 mg/m3
and report positive and statistically
significant associations with all-cause
mortality and mean PM2.5
concentrations of 9.6 mg/m3. He also
considers these results along with the
uncertainties and limitations associated
with studies that restricted analyses
below certain PM2.5 concentrations. As
described in responding to comments in
section II.B.3 above, uncertainties
associated with how the studies exclude
PM2.5 concentrations from the analyses
(e.g., at what spatial resolution are
concentrations being excluded), make it
difficult to understand how to interpret
the results of the restricted analyses in
the context of the approach employed in
this reconsideration, which takes into
consideration the relationship between
mean PM2.5 concentrations and design
values.
The Administrator also recognizes
that, in their review of the 2021 draft
PO 00000
Frm 00081
Fmt 4701
Sfmt 4700
16281
PA, the CASAC noted that
epidemiologic studies that restrict
analyses below certain PM2.5
concentrations represent one area for
which the evidence has expanded in
this reconsideration, stating that these
studies provide support for mortality
effects at concentrations below the
current PM NAAQS (Sheppard, 2022a,
p. 5 of consensus responses). In their
recommendations on alternative levels
for the primary annual PM2.5 standard,
the majority of the CASAC cited to
studies that restrict PM2.5 concentrations
to below 12 mg/m3 as a part of their
rationale for supporting a level within
the range of 8–10 mg/m3 (Sheppard,
2022a p. 16 of consensus responses).
Additionally, the Administrator notes
that some members of the CASAC, in
their review of the 2019 draft PA,
concluded that the epidemiologic
studies that restrict analyses below 12
mg/m3 and show positive associations
with health effects, along with other
aspects of the scientific evidence,
provide support for their conclusion
that the primary annual PM2.5 standard
is not adequate (Cox, 2019b, p. 9 of
consensus responses). Furthermore, the
Administrator takes note of public
commenters who also noted that the
epidemiologic studies that restrict PM2.5
concentrations to below the current
standard provide support, along with
the other available information, for
lowering the level of the primary annual
PM2.5 standard. In considering the
studies that include restricted analyses,
along with the CASAC’s advice and
public comments on these types of
studies, the Administrator concludes
that, although there are inherent
uncertainties associated with this
limited body of evidence, these studies
that apply restricted analyses provide
support for serious effects (e.g.,
mortality) at concentrations below 10.0
mg/m3. Given this, the Administrator
concludes that it is appropriate to place
some weight on these studies, and in
doing so, notes that a standard level of
9.0 mg/m3 would be below the reported
mean PM2.5 concentrations of 9.6 mg/m3
in these studies and would, thus, be
expected to provide protection against
exposures related to these reported
mean concentrations.
The Administrator also takes into
consideration recent U.S. accountability
studies, which assess the health effects
associated with actions that improve air
quality (e.g., air quality policies or
implementation of an intervention).
These types of studies can also reduce
uncertainties related to residual
confounding of temporal and spatial
factors (U.S. EPA, 2022a, p. 3–25). The
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16282
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
Administrator notes that in the 2020
review, the available accountability
studies had ‘‘starting’’ annual average
PM2.5 concentrations (i.e., mean
concentration prior to reductions being
evaluated) from 13.2–31.5 mg/m3, and
the then-Administrator cited the lack of
accountability studies in areas where
the ‘‘starting’’ concentration met the
current primary PM2.5 standards as part
of his rationale for retaining the
standards. As at the time of proposal,
the current Administrator notes that in
three studies newly available in this
reconsideration and assessed in the ISA
Supplement, prior to implementation of
the policies, mean PM2.5 concentrations
in these studies were below the level of
the current annual standard level (12.0
mg/m3) and ranged from 10.0 mg/m3 to
11.1 mg/m3. These studies report
positive and significant associations
between mortality and cardiovascular
morbidity and reductions in ambient
PM2.5 following the implementation of a
policy (Henneman et al., 2019; Corrigan
et al., 2018; Sanders et al., 2020a; 88 FR
5627, January 27, 2023). These studies
suggest that public health improvements
may occur following the
implementation of a policy that reduces
annual average PM2.5 concentrations
below the level of the current standard
of 12.0 mg/m3. The Administrator
recognizes that in their review of the
2021 draft PA, the CASAC noted that
the availability of recent accountability
studies was one area where the evidence
had been strengthened and that the
studies assessed in the ISA Supplement
provide evidence of mortality effects at
annual average PM2.5 concentrations
below the current NAAQS (Sheppard,
2022a, p. 5 of consensus responses). The
Administrator recognizes that the
CASAC also concluded that, along with
other lines of evidence, the
accountability studies with starting
concentrations below the levels of the
current standards are appropriate to
consider for informing conclusions on
alternative standard levels (Sheppard,
2022a, p. 13 of consensus responses).
The Administrator also notes the advice
of the CASAC in their review of the
2019 draft ISA, where they suggested
that accountability studies be taken into
account and such studies provide
potentially crucial information about
whether and how much decreasing
PM2.5 causes decreases in future health
effects, which reflects the primary
purpose of the NAAQS (Cox, 2019b, p.
8 and 10 of consensus responses). The
Administrator also notes that in their
review of the 2019 draft ISA, some
members of the CASAC cautioned
against placing more weight on the data
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
from accountability studies based on the
methodological limitations of the
studies (Cox, 2019b, p. 8 of consensus
responses). The Administrator notes
that the CASAC did not explicitly cite
to accountability studies in their
reviews of the 2019 draft PA or 2021
draft PA as support for their
recommendations on the adequacy of
the primary annual PM2.5 standard or
potential alternative standard levels. A
number of public commenters who
support revising the level of the
standard to 8 mg/m3 cite these
accountability studies, along with the
broader evidence base, as support for a
more protective standard. The
Administrator, in considering the
evidence, the advice from the CASAC,
and public comment, first recognizes
that accountability studies are just one
line of evidence to be considered in the
broader evaluations of the information
available to inform conclusions on the
level of the standard. In so doing, he
notes that public health improvements
may occur following the
implementation of a policy that reduces
annual average PM2.5 concentrations
below the level of the current standard
of 12.0 mg/m3, and potentially below the
lowest ‘‘starting’’ concentrations in
these studies of 10.0 mg/m3. However,
the Administrator concludes that the
limited number of accountability
studies provide limited information for
informing decisions on the appropriate
level of the primary annual PM2.5
standard but recognizes that these
studies provide supplemental
information for consideration along
with the full body of evidence. Taken
together, the Administrator notes a
revised annual standard level of 9.0 mg/
m3 is at or below the lowest starting
concentration of these accountability
studies (i.e., 10.0 mg/m3), and judges
that it is appropriate to place some
weight on these studies, particularly for
informing his public policy judgments
regarding an adequate margin of safety.
In addition to his consideration of and
conclusions regarding the available
scientific evidence, the Administrator
also considers the results of the
quantitative risk assessment to inform
his conclusions regarding the
appropriate level for the primary annual
PM2.5 standard. The Administrator
recognizes that the risk estimates can
help to place the evidence for specific
health effects into a broader public
health context, but should be
considered along with the inherent
uncertainties and limitations of such
analyses when informing judgments
about the potential for additional public
health protection associated with PM2.5
PO 00000
Frm 00082
Fmt 4701
Sfmt 4700
exposure and related health effects. The
Administrator recognizes that the
overall risk assessment estimates
suggest that the current primary annual
PM2.5 standard could allow a substantial
number of PM2.5-associated deaths in
the U.S. The Administrator also
recognizes that the CASAC concurred
with the 2021 draft PA’s assessment that
meaningful risk reductions will result
from lowering the annual PM2.5
standard (Sheppard, 2022a, p. 16 of
consensus responses).
Additionally, with respect to the
results of the quantitative risk
assessment, the Administrator
recognizes that the 2022 PA also
provides information on the distribution
of concentrations associated with the
estimated mortality risk at each
alternative standard level assessed (U.S.
EPA, 2022b, sections 3.4.2.2 and 3.6.2.2,
Figure 3–18 and 3–19). When meeting
an annual standard of 9.0 mg/m3 at the
design value monitor, the exposure
concentrations within an area are
estimated to be below 9 mg/m3, with the
majority of those exposures being at
concentrations of below 8 mg/m3. The
Administrator notes that this range of
concentrations is below the lowest
means in the key long- and short-term
epidemiologic studies (concentrations at
which the evidence is the strongest in
supporting an association between
exposure to PM2.5 and adverse health
effects observed in the key
epidemiologic studies available in this
reconsideration). Thus, the
Administrator concludes that the results
of the quantitative risk assessment
suggest that a revised annual standard
level of 9.0 mg/m3 is estimated to reduce
PM2.5 exposures to fall within the range
of concentrations in which there is the
most confidence in the associations and
thus, confidence that estimated risk
reductions will actually occur.
The Administrator also notes the
information provided by the
quantitative risk assessment on the
distribution of concentrations associated
with the estimated mortality risk for a
higher annual standard level of 10.0 mg/
m3 and a lower standard level of 8.0 mg/
m3 (U.S. EPA, 2022b, sections 3.4.2.2
and 3.6.2.2, Figure 3–18 and 3–19). The
Administrator finds that, for an annual
standard level of 10.0 mg/m3, the
quantitative risk assessment estimates
that the standard would allow multiple
exposures at concentrations above the
lowest means in the key epidemiologic
studies, and therefore, calls into
question whether a standard level of
10.0 mg/m3 would provide enough
public health protection. Additionally,
the Administrator also finds that, for a
lower annual standard level of 8.0 mg/
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
m3, the quantitative risk assessment
estimates the exposure concentrations to
be below 8 mg/m3, with the majority of
those exposures being at concentrations
of below 7 mg/m3. The Administrator
observes that the majority of exposure
concentrations under this air quality
scenario are estimated to fall outside of
the range of concentrations in which he
has the most confidence in the
associations and that the additional risk
reductions will actually occur.
Recognizing and building upon the
above considerations and judgments,
and with consideration of advice from
the CASAC and public comment, the
Administrator concludes that the
current body of scientific evidence and
quantitative risk assessment support his
judgment that the level of the primary
annual PM2.5 standard should be revised
to a level of 9.0 mg/m3. Revising the
level of the primary annual PM2.5
standard will, in the Administrator’s
judgment, provide requisite public
health protection with an adequate
margin of safety.
The Administrator recognizes that
placing weight on the information from
the epidemiologic studies allows for
examination of the entire population,
including those that may be at
comparatively higher risk of
experiencing a PM2.5-related health
effects (e.g., children, older adults,
minority populations) (88 FR 5624,
January 27, 2023). In considering the
epidemiologic evidence, the
Administrator judges that, in reaching
his decision on an appropriate level for
the annual standard that will protect
public health with an adequate margin
of safety, in the absence of any
discernible population-level thresholds,
and in recognizing the need to weigh
uncertainties associated with the
epidemiologic evidence, it is most
appropriate to examine where the
evidence of associations observed in the
epidemiologic studies is strongest and,
conversely, to place less weight where
he has less confidence in the
associations observed in the
epidemiologic studies. The
Administrator notes that in previous
reviews, evidence-based approaches
noted that the evidence of an
association in any epidemiologic study
is ‘‘strongest at and around the longterm average where the data in the study
are most concentrated’’ (78 FR 3140,
January 15, 2013). These approaches
were supported by previous CASAC
advice as well as the CASAC’s advice in
their review of the 2021 draft PA as a
part of this reconsideration. Given this,
the Administrator notes that in revising
the annual PM2.5 standard to a level of
9.0 mg/m3, he is setting the standard at
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
a level below the long-term mean PM2.5
concentrations in the key long- and
short-term epidemiologic studies,
including the lowest study reported
mean of 9.3 mg/m3, following an
approach that is consistent with
previous PM NAAQS reviews. The
Administrator additionally notes that air
quality analyses in the 2022 PA
demonstrate that areas meeting a revised
annual standard of 9.0 mg/m3 would be
expected to shift the distribution of
PM2.5 exposure concentrations in an
area such that the area’s highest monitor
would generally be at or below 9.0 mg/
m3 annually, and most of the resulting
PM2.5 concentrations across the area
would be even further below the studyreported means.119 120 Thus, a standard
level of 9.0 mg/m3 is expected to provide
sufficient protection not only in areas
where the highest allowable
concentration would be located (i.e.,
near design value monitors) but also in
other parts of the area where PM2.5
concentrations would be expected to be
maintained even lower.
Furthermore, the Administrator
recognizes the CASAC’s advice in their
review of the 2021 draft PA, as well as
public comments, that weight should be
placed on study-reported PM2.5
concentrations that are somewhat below
the mean, particularly for some of the
newer epidemiologic studies with larger
cohort sizes. In weighing uncertainties
associated with using these data to
inform a revised annual standard level,
as well as noting the limited studies for
which this information is available, the
Administrator judges that some weight
should be placed on these data, but they
should not receive the same weight as
the study-reported mean concentrations.
Thus, the Administrator concludes that
it would be appropriate to set the
annual standard level near the 25th
percentile PM2.5 concentrations in the
two newer key epidemiologic studies for
which these values were reported. In
doing so, the Administrator notes that a
decision to revise the annual standard to
9.0 mg/m3 would set a level of the
standard near and somewhat below the
reported 25th percentile PM2.5
concentrations of 9.1 mg/m3 in these two
more recent hybrid model-based
studies.
119 Analyses in the 2022 PA suggest that the
highest monitored value would be expected to be
greater than the study-reported mean values by 10–
20% for monitor-based studies and 15–18% for
hybrid modeling studies that apply aspects of
population weighting (U.S. EPA, 2022b, section
2.3.3.2.4).
120 The risk assessment in the 2022 PA used air
quality adjustments to simulate just meeting the
current primary PM2.5 standards, as well as
alternative standard levels (U.S. EPA, 2022b,
section 3.4.1.4 and Appendix C, section C.1.4).
PO 00000
Frm 00083
Fmt 4701
Sfmt 4700
16283
The Administrator also takes note of
the study-reported long-term mean
PM2.5 concentrations in the key
Canadian epidemiologic studies. While
the Administrator notes that these
studies provide additional support for
associations between PM2.5
concentrations and health effects, he is
also mindful that there are important
differences between the exposure
environments in the U.S. and Canada
that affect interpretation of the data in
the context of informing decisions
regarding potential alternative levels of
the annual standard. The Administrator
also recognizes that the majority of the
CASAC in their review of the 2021 draft
PA, as well as a number of public
commenters, placed weight on the
Canadian epidemiologic studies in
recommending that the level of the
primary annual PM2.5 standard be
revised to 8–10 mg/m3. The
Administrator notes that a decision to
revise the annual standard to 9.0 mg/m3
would set the level of the standard
within the range of levels recommended
by the majority of CASAC in their
consideration of these studies.
Additionally, the Administrator also
considers the information provided by
epidemiologic studies that use restricted
analyses, as well as accountability
studies. With respect to the restricted
analyses, the Administrator, in
considering the CASAC’s advice in their
review of the 2021 draft PA and many
public comments on these types of
studies, concludes that, although there
are inherent uncertainties associated
with this limited body of evidence, the
studies that apply restricted analyses
provide support for serious effects (e.g.,
mortality) at concentrations below 10.0
mg/m3. Additionally, in considering
accountability studies, the
Administrator concludes that while the
small number of these studies provide
limited information for informing
decisions on the appropriate level of the
primary annual PM2.5 standard, these
studies provide supplemental
information for consideration along
with the full body of evidence. The
Administrator further notes that these
studies suggest that public health
improvements may occur following the
implementation of a policy that reduces
annual average PM2.5 concentrations
below the level of the current standard
of 12.0 mg/m3, and potentially below the
lowest ‘‘starting’’ concentrations in
these studies of 10.0 mg/m3. Taken
together, the Administrator judges that
it is appropriate to place some weight
on these types of studies, particularly
for informing his public policy
judgments regarding an adequate margin
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16284
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
of safety, and notes that a revised
annual standard level of 9.0 mg/m3 is
below the lowest starting concentration
of the accountability studies (i.e., 10.0
mg/m3), and below the concentration at
which studies that apply restricted
analyses provide support for serious
effects (i.e., 9.6 mg/m3).
The Administrator also judges that the
results of the quantitative risk
assessment provide support for a
primary annual PM2.5 standard with a
level of 9.0 mg/m3. The results of the risk
assessment suggest that when meeting
an annual standard of 9.0 mg/m3, PM2.5
exposures are maintained below 9 mg/
m3 at the design value monitor, with the
majority of those exposures being at
concentrations below 8 mg/m3. Thus, the
Administrator notes that an annual
standard level of 9.0 mg/m3 would be
expected to provide protection from
exposures where he has the greatest
confidence in the associations between
health effects and PM2.5 exposures (i.e.
the long-term mean PM2.5
concentrations in the key U.S.
epidemiologic studies, of which the
lowest is 9.3 mg/m3) and would provide
an adequate margin of safety by
maintaining most PM2.5 exposures even
further below 9.0 mg/m3.
When considering adequate margin of
safety, the Administrator notes that in
his decision to revise the annual
standard level to 9.0 mg/m3, he is
placing weight on the information from
the epidemiologic studies which allows
for examination of the entire
population, including those that may be
at comparatively higher risk of
experiencing a PM2.5-related health
effects (e.g., children, older adults,
minority populations). Additionally, as
discussed above, the Administrator also
recognizes that setting the annual
standard level at 9.0 mg/m3, which is
below concentrations at which the
evidence is the strongest in supporting
an association between exposure to
PM2.5 and adverse health effects
observed in the key epidemiologic
studies available in this reconsideration,
would be expected to shift the
distribution of PM2.5 exposure
concentrations in an area such that the
area’s highest monitor would generally
be at or below 9.0 mg/m3 annually, and
most of the resulting PM2.5
concentrations across the area would be
even lower. In considering these air
quality relationships, the Administrator
judges that a revised annual standard
level of 9.0 mg/m3 would provide
requisite protection with adequate
margin of safety, for all populations,
including those most at-risk.
In reaching this conclusion, the
Administrator recognizes that in
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
establishing primary standards under
the Act that are requisite to protect
public health with an adequate margin
of safety, he is seeking to establish
standards that are neither more nor less
stringent than necessary for this
purpose. The Act does not require that
primary standards be set at a zero-risk
level or to protect the most sensitive
individual, but rather at a level that
avoids unacceptable risks to public
health. In this context, the
Administrator’s conclusion is that
revised primary annual standard, in
conjunction with the 24-hour standard,
provides the appropriate degree of
protection, and that more or less
stringent standards would not be
requisite.
In considering the requirement for an
adequate margin of safety, the
Administrator notes that the
determination of what constitutes an
adequate margin of safety is expressly
left to the judgment of the EPA
Administrator. See Lead Industries
Association v. EPA, 647 F.2d at 1161–
62; Mississippi, 744 F.3d at 1353. He
further notes that in evaluating how
particular standards address the
requirement to provide an adequate
margin of safety, it is appropriate to
consider such factors as the nature and
severity of the health effects, the size of
sensitive population(s) at risk, and the
kind and degree of the uncertainties
present. Consistent with past practice
and long-standing judicial precedent,
and as described in this section, the
Administrator takes the need for an
adequate margin of safety into account
as an integral part of his decision
making on a standard. See, e.g., NRDC
v. EPA, 902 F. 2d 962, 973–74 (D.C. Cir.
1990).
Given all of the evidence and
information discussed above, the
Administrator judges that a standard
with a level of 9.0 mg/m3 is requisite to
protect public health with an adequate
margin of safety. In so doing, he first
recognizes that a less stringent standard
would allow the occurrence of higher
long- and short-term PM2.5
concentrations at a level at or above the
mean PM2.5 concentrations in key U.S.
epidemiologic studies. That is, a less
stringent standard would be expected to
allow more PM2.5 exposures at
concentrations at or above which the
key U.S. epidemiologic studies have
reported associations between mean
PM2.5 concentrations and serious health
effects and would deviate from some
past approaches for selecting the
appropriate level of the annual
standard. A less stringent standard
would also not provide requisite
protection with an adequate margin of
PO 00000
Frm 00084
Fmt 4701
Sfmt 4700
safety against PM2.5 exposures in the
lower percentiles of the air quality
distribution (i.e., 25th percentile) for
which associations with health effects
have been observed in a limited number
of epidemiologic studies. Furthermore,
the Administrator notes that the primary
annual and 24-hour PM2.5 standards,
together, are intended to provide public
health protection against the full
distribution of long- and short-term
PM2.5 exposures. As noted above, the
Administrator recognizes that the
changes in PM2.5 air quality designed to
meet a less stringent annual standard
would likely result in higher exposures
across the distribution of air quality,
including both higher average (or
typical) concentrations as well as higher
short-term peak PM2.5 concentrations.
Taking into consideration both the full
evidence base for associations of PM2.5
with mortality and other adverse health
effects, including the reported mean
PM2.5 concentrations from key long- and
short-term U.S. epidemiologic studies,
information from epidemiologic studies
that report 25th percentile PM2.5
concentrations, supplemental
information from other epidemiologic
studies (i.e., epidemiologic studies that
use restricted analyses, accountability
studies, and Canadian epidemiologic
studies), and the results of the risk
assessment, as well as the advice from
the CASAC and public comments, the
Administrator concludes that a less
stringent standard would allow risks of
mortality and other adverse health
effects that are too great, and thus would
not provide sufficient protection for
public health as required by the CAA.
Additionally, in considering a less
stringent standard, the Administrator
recognizes that through its control of
long- and short-term PM2.5
concentrations, the annual standard
provides a margin of safety for less wellstudied exposure levels and population
groups for which the evidence is limited
or lacking. In so doing, he recognizes
that our understanding of the
relationships between the presence of a
pollutant in ambient air and associated
health effects is based on a broad body
of information encompassing not only
more established aspects of the
evidence, such as the conclusion that
long- and short-term exposures to PM2.5
are causally related to mortality and
cardiovascular effects and likely to be
causally related to respiratory effects,
but also aspects with which there may
be substantial uncertainty. In particular,
the Administrator notes that there are
other categories of effects with causality
determinations that are suggestive of,
but not sufficient to infer, a causal
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
relationship between PM2.5 exposure
and health outcomes. These include, but
are not limited t,o short-term exposure
and nervous system effects, as well as
long- and short-term exposure and
pregnancy and birth outcomes, where
the evidence is less certain but which
represent potentially substantial
additional risk to public health from
exposure to PM2.5. He recognizes the
CAA requirement that requires primary
standards to provide an adequate
margin of safety was intended to
address uncertainties associated with
inconclusive scientific and technical
information as well as to provide a
reasonable degree of protection against
hazards that research has not yet
identified and in his judgment, the
primary NAAQS must be set at a level
that is adequately protective against
these and other effects which research
has not yet identified. Thus, even if the
Administrator had somewhat greater
concerns about the possibility of
confounding, error and bias in the
epidemiologic studies, which reduced
his confidence in finding that PM2.5 is
causally related to mortality and
cardiovascular effects, he would still
find it appropriate to set the primary
NAAQS below the means of key U.S.
epidemiologic studies given the strength
of the evidence providing support for
the association, as well as additional
evidence linking PM2.5 to other
endpoints of substantial public health
concern, and the need to protect public
health with an adequate margin of
safety. In considering the uncertainties
in both the epidemiologic evidence and
the controlled human exposures studies,
the Administrator recognizes that
collectively, the health effects evidence
generally reflects a continuum,
consisting of levels at which scientists
generally agree that health effects are
likely to occur, through lower levels at
which the likelihood and magnitude of
the response become increasingly
uncertain. In light of these uncertainties,
the Administrator recognizes that the
CAA requirement that primary
standards provide an adequate margin
of safety, as summarized in section I.A
above, is intended to address
uncertainties associated with
inconclusive scientific and technical
information, as well as to provide a
reasonable degree of protection against
hazards that research has not yet
identified. The Administrator has taken
the need to provide for an adequate
margin of safety into account as an
integral part of his decision-making on
the appropriate standards in setting the
standard at a level below the level
where available epidemiologic studies,
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
which include diverse populations that
are broadly representative of the U.S.
population including at-risk
populations, have provided the
strongest evidence supporting effects,
and in other ways as well. For example,
consideration of a margin of safety is
reflected in the approach of setting the
level of the annual standard near and
somewhat below the 25th percentile
PM2.5 concentrations from key U.S.
epidemiologic studies (i.e., 9.1 mg/m3),
as well as recognition that attaining a
design value will generally result in
significantly broader and greater
improvements of air quality across an
area (including but certainly not limited
to areas near the design value monitor)
(U.S. EPA, 2022a, sections 2.3.3.2.4 and
3.3.3.2.1, Table 3–5). Based on all of the
considerations noted here, and
considering the current body of
evidence, including the associated
limitations and uncertainties, in
combination with the exposure/risk
information, the Administrator
concludes that a less stringent standard
than the current standard would not
provide the requisite protection of
public health, including an adequate
margin of safety.
Having concluded that a less stringent
standard would not provide the
requisite protection of public health, the
Administrator next considers whether a
more stringent standard would be
appropriate. In so doing, he notes that
a decision to set the level of the annual
standard to below 9.0 mg/m3 would
place a large amount of the emphasis on
potential public health importance of
further reducing the occurrence of PM2.5
concentrations of concern, though the
exposures about which he is most
concerned are well controlled with an
annual standard level of 9.0 mg/m3, as
demonstrated by the quantitative risk
assessment. Such a decision would also
place greater weight on (1) further
reducing ambient PM2.5 concentrations
relative to those observed in long-and
short-term epidemiologic studies,
including those that he had judged to
have significant uncertainties, including
Canadian studies, studies using
restricted analyses, and accountability
studies; (2) shifting the air quality
distribution in areas such that the
highest exposure concentrations are
reduced to below PM2.5 concentrations
observed in epidemiologic studies to be
in the 25th or lower percentile, for
which the evidence is limited; and (3)
further shifting exposure concentrations
to those shown at the lower end of the
distribution in the quantitative risk
assessment, despite the important
uncertainties in the overall risk
PO 00000
Frm 00085
Fmt 4701
Sfmt 4700
16285
assessment. As discussed in this section
and in responses to significant
comments above and in the Response to
Comments document, the Administrator
has concluded that placing a large
emphasis on these factors and revising
the standard to a level below 9.0 mg/m3
would result in a standard that is more
stringent than the evidence indicates to
be sufficient to protect public health
with an adequate margin of safety.
Compared to a primary annual PM2.5
standard set at a level of 9.0 mg/m3, the
Administrator concludes that the extent
to which lower standard levels could
result in further public health
improvements becomes notably less
certain.
Thus, having carefully considered the
scientific evidence, quantitative
information, CASAC advice, and public
comments relevant to his decision on
the level of the primary annual PM2.5
standard, as discussed above and in the
Response to Comments document, the
Administrator is revising the level of the
primary annual PM2.5 standard to 9.0
mg/m3. In the Administrator’s judgment,
based on the currently available
evidence and information, an annual
standard set at this level and using the
specified indicator, averaging time, and
form, in conjunction with the other
primary PM standards, would be
requisite to protect public health with
an adequate margin of safety. The
Administrator judges that such a
standard would protect, with an
adequate margin of safety, the health of
at-risk populations, including children,
older adults, those with pre-existing
cardiovascular and respiratory diseases,
minority populations, and low SES
populations. The Administrator believes
that a standard set at 9.0 mg/m3 would
be sufficient to protect public health
with a margin of safety, and believes
that a lower standard would be more
than what is necessary to provide this
degree of protection. This judgment by
the Administrator appropriately
considers the degree of protection that
is neither more nor less stringent than
necessary for this purpose and
recognizes that the CAA does not
require that primary standards be set at
a zero-risk level, but rather at a level
that reduces risk sufficiently so as to
protect public health with an adequate
margin of safety.
In reaching his conclusions on
adequacy of the current suite of primary
PM2.5 standards, based on consideration
of the available scientific evidence and
quantitative information, the CASAC’s
advice and public comments, the
Administrator finds that the available
information is insufficient to call into
question the adequacy of the public
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16286
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
health protection afforded by the
current primary 24-hour PM2.5 standard.
As described earlier in this section, the
Administrator concludes that it is
appropriate to retain the current
indicator (PM2.5), averaging time (24hour), and form (98th percentile,
averaged over three years) for the
primary 24-hour PM2.5 standard and
below explains the basis for his final
decision that is also appropriate to
retain the current level of the primary
24-hour PM2.5 standard.
In reaching his conclusion to retain
the current primary 24-hour PM2.5
standard the Administrator does so in
light of the conclusion that the
epidemiologic evidence supports
associations between short- and longterm PM2.5 exposures and adverse
health effects, but that the
epidemiologic evidence does not
identify specific concentrations at
which those effects occur and the
Administrator has greatest confidence in
effects where the bulk of the data is
reported (i.e., the mean PM2.5
concentration, with some consideration
for the 25th percentile of the air quality
distribution). Thus, in considering the
epidemiologic evidence, the
Administrator concludes it is
appropriate to focus on setting a
generally controlling annual standard as
the most effective and efficient way to
reduce total population risk associated
with both long- and short-term PM2.5
exposures, and that it is appropriate to
revise the level of the annual standard
level to 9.0 mg/m3. In addition to the
epidemiologic evidence, the
Administrator also considers the
available controlled human exposure
studies, which provide evidence for
health effects following single, shortterm PM2.5 exposures to concentrations
that typically correspond to upper end
of the PM2.5 air quality distribution in
the U.S. (i.e., ‘‘peak’’ concentrations). In
so doing, the Administrator notes that
these studies report statistically
significant effects on one or more
indicators of cardiovascular function
following 2-hour exposures to PM2.5
concentrations at and above 120 mg/m3
and at and above 149 mg/m3 for vascular
impairment, the effect shown to be most
consistent across studies. In particular,
the Administrator notes that a single
study is assessed in the ISA Supplement
that reports effects following 4-hour
exposures at 37.8 mg/m3, although the
results of this study are inconsistent
with the results of the controlled human
exposure studies assessed in the 2019
ISA. Along with the inconsistent results
from the controlled human exposure
studies, the Administrator also
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
recognizes that effects observed in these
studies are intermediate effects which
are not typically considered adverse and
that the study participants were healthy
individuals. Taking into consideration
the available scientific evidence,
including the uncertainties and
limitations, along with the CASAC’s
advice, the Administrator concludes
that it is appropriate to maintain a
primary 24-hour PM2.5 standard to
protect against peak exposures.
Thus, the Administrator considers
what primary 24-hour PM2.5 standard is
requisite to provide supplemental
protection against peak exposures.
While having confidence that the
revised annual standard will result in
lowering risk associated with both longand short-term PM2.5 exposure by
lowering the overall air quality
distribution, as in the 2012 review, the
Administrator recognizes that an annual
standard alone would not be expected to
offer sufficient protection with an
adequate margin of safety against the
effects of short-term PM2.5 exposures in
all parts of the country. Therefore, he
continues to conclude that it is
appropriate to continue to provide
supplemental protection by means of a
24-hour standard, in conjunction with a
revised annual standard level of 9.0 mg/
m3.
In considering the available scientific
evidence assessed in the 2019 ISA and
ISA Supplement, the Administrator first
considers the controlled human
exposure studies for informing his
decisions on the primary 24-hour PM2.5
standard. In so doing, he notes that in
their review of the 2021 draft PA, the
majority of CASAC members expressed
the view that controlled human
exposure studies are not the best
evidence to use for justifying retaining
the 24-hour standard without revision,
in part because these studies
preferentially recruit less susceptible
individuals and have a typical exposure
duration much shorter than 24 hours.
Thus, in the view of the majority, ‘‘the
evidence of effects from controlled
human exposure studies with exposures
close to the current 24-hour standard
supports epidemiological evidence for
lowering the standard’’ (Sheppard,
2022a, p. 3–4 of consensus letter). In
reviewing the controlled human
exposure studies, the Administrator
agrees with the majority of CASAC that
these controlled human exposure
studies generally do not include
populations with substantially
increased risk from exposure to PM2.5,
such as children, older adults, or those
with more severe underlying illness.
However, he disagrees with any
conclusion that they should not be used
PO 00000
Frm 00086
Fmt 4701
Sfmt 4700
to inform a decision about the adequacy
of the current standard. The
Administrator finds the information
available from these studies to be useful,
noting that the recently available
controlled human exposure studies
provide evidence for health effects
following single, short-term exposures
to PM2.5 concentrations that are greater
than those allowed under the current
standard. The results of the controlled
human exposure studies are
inconsistent, particularly at lower PM2.5
concentrations, but some studies do
report statistically significant effects on
one or more indicators of cardiovascular
function following 2-hour exposures to
PM2.5 concentrations at and above 120
mg/m3 (and at and above 149 mg/m3 for
vascular impairment, the effect shown
to be most consistent across studies).
Additionally, one controlled human
exposure study assessed in the ISA
Supplement reports evidence of some
effects for cardiovascular markers
following 4-hour exposures to 37.8 mg/
m3 (Wyatt et al., 2020). However, there
is inconsistent evidence for
inflammation in other controlled human
exposure studies evaluated in the 2019
ISA. The Administrator finds these
studies are important in establishing
biological plausibility for PM2.5
exposures causing more serious health
effects, such as those seen in short-term
exposure epidemiologic studies, and
they provide support that more adverse
effects may be experienced following
longer exposure durations and/or
exposure to higher concentrations. As
described in more detail in responding
to public comments in section II.B.3
above, he notes that although the
controlled human exposure studies do
not provide a threshold below which no
effects occur, the observed effects in
these controlled human exposures
studies are ones that signal an
intermediate effect in the body, likely
due to short-term exposure to PM2.5, and
typically would not, by themselves, be
judged as adverse. As noted in sections
II.A.2 and II.B.3 above, associated
judgments regarding adversity or health
significance of measurable physiological
responses to air pollutants in previous
NAAQS reviews have been informed by
guidance, criteria or interpretative
statements developed within the public
health community. This type of
information on adversity of effects is
particularly informative to the
Administrator’s judgments regarding the
adversity of the effects observed in the
controlled human exposure studies
which are short-term in nature (i.e.,
generally ranging from 2- to 5-hours),
including those studies that are
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
conducted at near-ambient PM2.5
concentrations. Based on the
observation that the effects observed in
Wyatt et al. (2020) are not by themselves
adverse, and the fact that the findings of
this study are inconsistent with other
currently available evidence regarding
the level at which effects are observed,
the Administrator disagrees with the
view expressed by the majority of
CASAC that this study supports
epidemiologic evidence for lowering the
24-hour standard.
Consistent with his approach in
reaching his proposed decision and
taking into consideration these points as
well as balancing these limitations (i.e.,
that the health outcomes observed in
these controlled human exposure
studies are not clearly adverse and that
the studies generally do not include
those at increased risk from PM2.5
exposure), the Administrator still
considers it appropriate to ensure that
the 24-hour PM2.5 standard provides
protection against health effects
consistently observed in the controlled
human exposure studies. He next
examines the air quality analyses,
described in more detail in section
II.A.c.i above, to assess whether during
recent air quality conditions, areas
meeting the current standards would
experience PM2.5 concentrations
reported in these controlled human
exposure studies. He observes that air
quality analyses demonstrate that the
PM2.5 exposures shown to cause
consistent effects in the controlled
human exposure studies are well above
the ambient concentrations typically
measured in locations meeting the
current primary standards, and therefore
suggest that the current primary PM2.5
standards provide protection against
these ‘‘peak’’ concentrations. In fact, at
air quality monitoring sites meeting the
current primary PM2.5 standards (i.e.,
the 24-hour standard of 35 mg/m3 and
the annual standard of 12 mg/m3), the 2hour concentrations generally remain
below 10 mg/m3, and rarely exceed 30
mg/m3. Though two-hour concentrations
are higher at monitoring sites violating
the current standards, they generally
remain below 16 mg/m3 and rarely
exceed 80 mg/m3, still below
concentrations in CHE studies where
consistent effects are observed (e.g.,
greater than 120 mg/m3) (U.S. EPA,
2022b, section 2.3.2.2.3, Figure 2–19,
and section 3.3.3.1). Additionally, and
in response to public comments, the
Administrator notes additional air
quality analyses conducted by the
EPA,121 that provide a more refined
121 Jones et al. (2023). Comparison of Occurrence
of Scientifically Relevant Air Quality Observations
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
analysis of whether areas that meet the
current standards experience peak
concentrations reported in controlled
human exposure studies. He notes that
2-hour observations greater than 120 mg/
m3 and 4-hour observations greater than
38 mg/m3 rarely occur (e.g., 0.025% of
rolling 2-hour observations are greater
than 120 mg/m3 and 0.78% of rolling 4hour observations greater than 38 mg/
m3). Based on this information, the
Administrator finds that the current
suite of standards maintains subdaily
concentrations of PM2.5 in ambient air
far below the exposure concentrations
in controlled human exposure studies
where consistent effects have been
observed, and notes that while these
studies generally do not include the
most at-risk individuals, the exposure
concentrations in these studies also do
not elicit adverse effects.
Further, in light of the
Administrator’s emphasis on the annual
standard as the controlling standard,
with the 24-hour standard providing
supplemental protection against peak
concentrations, he next considers the
potential impact of a revised annual
standard of 9.0 mg/m3 on the occurrence
of peak sub-daily PM2.5 concentrations.
Specifically, the Administrator takes
note of the new air quality analyses 122
where he observes that lower
percentages of concentrations greater
than 120 mg/m3 and 38 mg/m3 occur in
areas meeting an annual standard of 9.0
mg/m3 and a 24-hour standard of 35 mg/
m3, versus an annual standard of 12.0
mg/m3 and a 24-hour standard of 35 mg/
m3. Thus, he concludes that an annual
standard that is controlling across most
areas of the country will continue to
effectively limit peak daily
concentrations in conjunction with the
existing 24-hour standard, with its level
of 35 mg/m3 and 98th percentile form,
which continues to provide
supplemental protection against peak
concentrations.
In addition, the Administrator also
notes that the majority of the CASAC in
their review of the 2021 draft PA, as
well as a number of public commenters,
support their recommendation to revise
the current 24-hour standard by
Between Design Value Groups. Memorandum to the
Rulemaking Docket for the Review of the National
Ambient Air Quality Standards for Particulate
Matter (EPA–HQ–OAR–2015–0072). Available at:
https://www.regulations.gov/docket/EPA-HQ-OAR2015-0072.
122 Jones et al. (2023). Comparison of Occurrence
of Scientifically Relevant Air Quality Observations
Between Design Value Groups. Memorandum to the
Rulemaking Docket for the Review of the National
Ambient Air Quality Standards for Particulate
Matter (EPA–HQ–OAR–2015–0072). Available at:
https://www.regulations.gov/docket/EPA-HQ-OAR2015-0072.
PO 00000
Frm 00087
Fmt 4701
Sfmt 4700
16287
pointing to ‘‘substantial epidemiologic
evidence from both morbidity and
mortality studies’’ which ‘‘includes
three U.S. air pollution studies with
analyses restricted to 24-hour
concentrations below 25 mg/m3’’
(Sheppard, 2022a, p. 17 consensus
responses). The Administrator notes
that the epidemiologic evidence
available in this reconsideration,
including the studies that restrict shortterm PM2.5 exposures (i.e., 24-hour
PM2.5 concentrations) to levels below 25
mg/m3, provides support for positive and
statistically significant associations
between short-term exposure to PM2.5
and all-cause mortality (Di et al., 2017a)
and CVD hospital admissions (deSouza
et al., 2021; Di et al., 2017a). He agrees
that these studies help to provide
additional support for reaching
conclusions on causality in the 2019
ISA. He further agrees that the available
epidemiologic studies provide
important information that it is
appropriate to consider in this
reconsideration, including information
on associations between health effects
and PM2.5 exposures in diverse
populations that are broadly
representative of the U.S. population,
and include populations identified as
at-risk (e.g., older adults, minority
populations), as well as evidence of
linear, no-threshold concentrationresponse relationships in those
associations, although with less
certainty in the shape of the curve at
long-term average concentrations below
about 8 mg/m3.
However, the Administrator also
notes significant limitations in the
currently available epidemiologic
information that limit his ability to draw
conclusions from the key short-term
studies, including those that employ
restricted analyses, to inform his
decision regarding the level of the 24hour PM2.5 standard. As a result of these
limitations, the Administrator does not
find that the short-term epidemiologic
studies, or the other evidence such as
the controlled human exposure studies
or the risk assessment, provide a
sufficient justification for revising the
24-hour standard.
First, he notes that short-term
epidemiologic studies examine
associations between day-to-day
variations in PM2.5 concentrations and
health outcomes, often over multi-year
study periods. As such, these studies
report long-term mean 24-hour PM2.5
concentrations (e.g., mean 24-hour
PM2.5 concentrations over multi-year
study periods), rather than at specific
points in the distribution (i.e., 90th or
98th percentile 24-hour concentrations)
at which effects occur. Further, he notes
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16288
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
that while there can be considerable
variability in daily exposures over a
multi-year study period, the bulk of the
observations reflect days with ambient
PM2.5 concentrations in the middle of
the air quality distribution (i.e.,
‘‘typical’’ days rather than days with
extremely low or extremely high
concentrations). As a result, the results
of these studies are more directly
applicable to decisions regarding the
annual standard (which is based on the
long-term mean of both short- and longterm epidemiologic studies), and the
fact that they do not report other air
quality statistics, such as the 98th
percentile concentrations which might
be more directly compared to the level
of the 24-hour standard, makes them
less useful for informing decisions on
the 24-hour standard. As discussed in
responding to comments above, the
form of the annual standard is based on
the annual mean PM2.5 concentration
averaged over three years,123 which
makes it better suited as a basis for
controlling air quality to avoid effects
observed in both long-term and shortterm epidemiologic studies. By contrast,
the form of the 24-hour standard is the
98th percentile averaged over three
years, which makes it appropriate for
controlling short-term peak
concentrations. However, based on the
available air quality information,
including distribution statistics of PM2.5
concentrations and health events
reported in the short-term
epidemiologic studies, these studies are
too limited in their ability to identify
health effects attributable to specific
short-term peak concentrations that are
necessary to evaluate whether the 24hour standard with its 98th percentile
form should be revised (e.g., restricted
epidemiologic studies do not report the
number or the percentile of health
events or the percentile of PM2.5
concentrations across the highest part of
the restricted air quality distribution,
including the 98th percentile). Thus, the
Administrator does not consider it
appropriate to use the reported means
from short-term studies to determine the
appropriate level for a 24-hour standard
with a 98th percentile form.
Similarly, the Administrator does not
consider the results of the restricted
analyses to be well suited to informing
the choice of level for a 24-hour
standard. Restricted analyses use a
subset of data from their main analyses
to evaluate health events that occur at
concentrations below a certain
concentration (e.g., 25 mg/m3). The
Administrator notes that the
associations between the health effects
(e.g., mortality and cardiovascular
morbidity) and PM2.5 concentrations
remain even after excluding higher
concentrations in the restricted
analyses, and he also recognizes that the
magnitude of the effect is generally
greater in the restricted analyses
compared to the associations reported in
the main analysis. He considers such
analyses to be informative in indicating
that the health effects association
reported in the main (unrestricted)
analysis are not driven only by the
upper peaks of the PM2.5 air quality
distribution, but rather persist at lower
portions of the distribution (consistent
with his emphasis on the annual
standard, which is focused on exposures
near the mean concentration, where the
bulk of the exposure distribution is
concentrated). Indeed, he notes that if
peak concentrations were the principal
driver of health effects associated with
PM2.5 exposure, one might expect the
associations to become weaker as the
upper portion of the data is excluded in
the restricted analyses, which is not
what is reported by the analyses (e.g.,
the restricted analyses generally report
associations that are greater in
magnitude compared to the main
analyses). However, he disagrees with
the assertion by the CASAC in their
review of the 2021 draft PA and some
public commenters that it would be
appropriate to focus on the specific
PM2.5 concentration (e.g., 25 or 30 mg/
m3) at which the analysis was restricted
as the basis for choosing a 24-hour
standard level. The Administrator
recognizes that in restricted analyses,
while an association continues to persist
across the full range of the air quality
distribution, and that the cutpoint
concentration at which the analysis was
restricted (e.g., 25 or 30 mg/m3) becomes
the maximum PM2.5 concentration in
the distribution, he also notes that these
studies do not provide information
related to the distribution of health
events and PM2.5 concentrations, and as
such, he is more uncertain where the
bulk of the data are and where he has
confidence in the reported
association.124 He notes that no
evidence exists to support a conclusion
that the PM2.5 concentration chosen as
the cutpoint in a restricted analysis has
any bearing on the concentration at
which effects are likely to occur (or not
occur). He notes that, as with long-term
123 The annual mean is calculated by averaging
daily values in a calendar quarter and then
averaging calendar quarters. See 40 CFR part 50
Appendix N, section 4.4.
124 These studies do not report information about
the distribution of the health events and PM2.5
concentrations (e.g., means, medians, other
percentiles) in the restricted analyses.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
PO 00000
Frm 00088
Fmt 4701
Sfmt 4700
studies, the evidence does not suggest
there is a specific point in the air quality
distribution of these short-term studies
that represents a ‘‘bright line’’ at and
above which effects have been observed
and below which effects have not been
observed. In order to identify a level of
the 24-hour standard based on
associations between the ‘‘upper end’’ of
exposures, either in the unrestricted or
the restricted analyses, and adverse
health effects, it would be necessary to
have a better understanding of how
specific 24-hour concentrations
correspond to the frequency and total
number of observed health events in the
study. Currently, such information,
including 98th percentile statistics, are
not reported in the key short-term
epidemiologic studies (and if they were
reported, the Administrator would have
to carefully consider how to weigh the
data). As such, in reaching his decision
on the primary 24-hour PM2.5 standard,
the Administrator judges that the
currently available information from
short-term epidemiologic studies,
including those that employ restricted
analyses, does not provide a sufficient
basis to revise the current 24-hour
standard, given that the 24-hour
standard focuses on reducing ‘‘peak’’
exposures (with its 98th percentile
form), but rather that such information
supports his judgment that it is
appropriate to focus on revising the
annual standard for purposes of
reducing all exposures, across the entire
distribution of air quality, to increase
public health protection.
In reaching final decisions regarding
the adequacy of the primary 24-hour
PM2.5 standard, the Administrator
continues to view an approach that
focuses on setting a generally
controlling annual standard as the most
effective and efficient way to reduce
total population risk associated with
both long- and short-term PM2.5
exposures. Additionally, he emphasizes
that improvements in air quality
associated with meeting an annual
standard level of 9.0 mg/m3 will result
in lowering risk associated with both
long- and short-term PM2.5 exposure by
lowering the overall air quality
distribution. The Administrator
concludes that reducing the annual
standard is the most efficient way to
reduce the risks from short-term
exposures identified in the
epidemiologic studies, as the available
evidence suggests the bulk of the risk
comes from the large number of days
across the bulk of the air quality
distribution, not the relatively small
number of days with peak
concentrations. However, as in the 2012
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
review, the Administrator recognizes
that an annual standard alone would not
be expected to offer sufficient protection
with an adequate margin of safety
against the effects of short-term PM2.5
exposures in all parts of the country and
concludes that, in conjunction with a
revised annual standard level of 9.0 mg/
m3, it is appropriate to continue to
provide supplemental protection by
means of a 24-hour standard,
particularly for areas with high peak-tomean ratios possibly associated with
strong local or seasonal sources.
In selecting the level of a 24-hour
standard designed to provide
supplemental protection against peak
exposures (in conjunction with a
revised annual standard of 9.0 mg/m3),
the Administrator considers the
information from the controlled human
exposure studies and the EPA’s analysis
of peak concentrations observed in areas
meeting the current standard of 35 mg/
m3 in conjunction with a revised
standard of 9.0 mg/m3 to be of particular
relevance. He notes the controlled
human exposure evidence includes
studies reporting effects on one or more
indicators of cardiovascular function
following 2-hour exposures at and above
120 mg/m3, including effects reported at
and above 149 mg/m3 for vascular
impairment, the effect shown to be most
consistent across studies, and less
consistent effects at lower
concentrations, including a single study
at near ambient concentrations (Wyatt et
al., 2020) reporting effects following 4hour exposures at 37.8 mg/m3. He
recognizes that the effects observed (in
those studies that observed effects) are
ones that signal an intermediate effect in
the body, likely due to short-term
exposure to PM2.5, and typically would
not, by themselves, be judged as
adverse, and the study participants were
healthy individuals.
He notes in particular that, in the
EPA’s analysis, in areas meeting the
current 24-hour standard and the
revised annual standard 0.029 percent
of 2-hour observations and 0.41 percent
of 4-hour observations reach PM2.5
concentrations higher than 120 mg/m3
and 37.8 mg/m3, respectively. He also
notes the lack of evidence of effects
from controlled human exposure studies
at levels below the current 24-hour
standard and the fact that the results of
Wyatt et al. (2020) are inconsistent with
other available studies, as well as the
intermediate nature of effects observed
in this study. In his judgment, the small
number of occurrences of peak
exposures indicate that, in conjunction
with a revised annual standard of 9.0
mg/m3, the current 24-hour standard of
35 mg/m3 remains requisite to protect
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
public health with an adequate margin
of safety, and that there is substantial
basis to doubt whether further
improvements in public health would
be achieved by further reducing these
exposures. Furthermore, the
Administrator concludes that due to the
limitations and uncertainties outlined
above, the information from recent
short-term epidemiologic studies,
including those that use restricted
analyses, is inadequate to inform
decisions regarding the adequacy of the
current 24-hour standard. Thus, in
reaching his decision on the primary 24hour PM2.5 standard, the Administrator
concludes that currently available
evidence does not call into question the
adequacy of the current standard.
In addition to the scientific evidence,
the Administrator also considers the
risk assessment in evaluating the
appropriate level of the 24-hour PM2.5
standard. The risk assessment indicates
that the annual standard is the
controlling standard across most of the
urban study areas evaluated (i.e., when
air quality related to the annual average
PM2.5 concentrations decrease, daily
average PM2.5 concentrations are also
expected to decrease). When air quality
is adjusted to just meet an alternative
24-hour standard level of 30 mg/m3 in
the areas where the 24-hour standard is
controlling, the risk assessment
estimates reductions in PM2.5-associated
risks across a more limited population
and number of areas compared to when
air quality is adjusted to simulate
alternative levels for the annual
standard (i.e., where the annual
standard is controlling), and these
predictions are largely confined to areas
located in the western U.S., several of
which are also likely to experience risk
reductions upon meeting a revised
annual standard. With respect to the
CASAC’s advice in their review of the
2021 draft PA, the Administrator notes
that the minority of CASAC advised that
these results suggest that the annual
standard can be used to limit both longand short-term PM2.5 concentrations and
views these risk assessment results as
supporting the conclusion that the
current 24-hour standard is adequate
(Sheppard, 2022a, p. 4 of consensus
letter). In contrast, the majority of
CASAC members in their review of the
2021 draft PA, as well as a number of
public commenters that support
revision of the 24-hour standard, placed
greater weight on the evidence-based
considerations (e.g. scientific evidence,
like the restricted analyses) than on the
values estimated by the risk assessment,
noting the potential for uncertainties in
how the risk assessment was able to
PO 00000
Frm 00089
Fmt 4701
Sfmt 4700
16289
‘‘capture areas with wintertime
stagnation and residential wood-burning
where the annual standard is less likely
to be protective’’ (Sheppard, 2022a, p. 4
of consensus letter).
In considering the application of the
risk assessment to judgments about the
adequacy of the current primary 24-hour
PM2.5 standard, the Administrator again
notes that the risk assessment analyses
of PM2.5-attributable mortality use input
data that include C–R functions from
epidemiologic studies that have no
threshold and a linear C–R relationship
down to zero, as well an air quality
adjustment approach that incorporates
proportional decreases in PM2.5
concentrations to meet lower standard
levels. As such, the Administrator notes
that this quantitative approach does not
incorporate any elements of uncertainty
in associations of health effects at lower
concentrations and that simulated air
quality improvements will always lead
to proportional decreases in risk (i.e.,
each additional mg/m3 reduction
produces additional benefits with no
clear stopping point at any PM2.5
concentration). Therefore, the
Administrator recognizes that while the
risk estimates can help to place the
evidence for specific health effects into
a broader public health context, the
results should be considered along with
the inherent uncertainties and
limitations of such analyses when
informing judgments about the potential
for additional public health protection
associated with PM2.5 exposure and
related health effects. Further, the
Administrator notes additionally that air
quality analyses have also been
considered in looking at the adequacy of
the 24-hour standard in controlling peak
PM2.5 concentrations of potential
concern,125 and that those analyses
included monitoring information from
across the entire U.S., specifically
highlighting areas with higher peak
concentrations and including areas
impacted by wintertime stagnation and
residential wood-burning. Thus, while
the risk assessment may have focused
on a subset of areas across the U.S.
based on the study area selection
criteria, the Administrator is
considering a broader set of information
in reaching his conclusions regarding
the appropriateness of the current 24125 Jones et al. (2023). Comparison of Occurrence
of Scientifically Relevant Air Quality Observations
Between Design Value Groups. Memorandum to the
Rulemaking Docket for the Review of the National
Ambient Air Quality Standards for Particulate
Matter (EPA–HQ–OAR–2015–0072). Available at:
https://www.regulations.gov/docket/EPA-HQ-OAR2015-0072.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16290
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
hour standard to control peak
concentrations.
The Administrator also considers the
advice from the CASAC in their reviews
of the 2019 draft PA and 2021 draft PA.
In their review of the 2019 draft PA, the
CASAC ‘‘agrees with the EPA and finds
that the available evidence does not call
into question the adequacy of public
health protection afforded by the
current 24-hour PM2.5 standard and
concurs that it be retained’’ (Cox, 2019b,
p. 3 of letter). He also notes that in their
review of the 2021 draft PA, the CASAC
did not reach consensus on whether the
current 24-hour standard is adequate,
with the majority of the CASAC
recommending that the 24-hour
standard be revised and the minority of
the CASAC recommending that the
standard be retained. The majority of
the CASAC members further stated that
‘‘[t]here is also less confidence that the
annual standard could adequately
protect against health effects of shortterm exposures. A range of 25–30 mg/m3
for the 24-hour PM2.5 standard would be
adequately protective’’ (Sheppard,
2022a, p. 4 of consensus letter). The
Administrator also acknowledges that
some public commenters agreed with
the majority of the CASAC in
supporting a revision to the level of the
24-hour standard to a range between 25–
30 mg/m3. These commenters cite a
number of reasons, including: (1)
Results from controlled human
exposure studies at near ambient
concentrations; (2) aspects of the
scientific evidence, including restricted
analyses that report positive and
significant associations below 35 mg/m3;
and (3) quantitative risk analyses that
show decreasing risk with decreasing
PM2.5 concentrations. In responding to
these comments, the Administrator
recognizes that some commenters have
different interpretations of the evidence,
air quality information, and quantitative
results from the risk assessment in this
review and would make different
judgments about the weight to place on
the relative strength and limitations of
the currently available scientific
evidence and information and how such
information could be used in making
public health policy decisions on the
24-hour standard. However, as outlined
above, the Administrator has carefully
considered the information available
from controlled human exposure studies
and short-term epidemiologic studies,
and weighed the strengths and
limitations of this evidence in
formulating his decisions. Furthermore,
as discussed above the Administrator
has noted significant uncertainties and
limitations inherent in the risk
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
estimates, as well as noting that very
few areas were included. In addition, he
has given careful consideration to the
majority of the CASAC’s advice in their
review of the 2021 draft PA, but has
drawn different conclusions with
respect to how currently available
evidence and air quality information
inform the selection of level for the 24hour primary PM2.5 standard.
In considering the advice of the
majority of CASAC, the Administrator
notes that a decision to set the level of
the 24-hour standard to below 35 mg/m3
would place a large amount of emphasis
on the potential public health
importance of further reducing the
occurrence of peak PM2.5
concentrations. However, the
Administrator concludes that there is
insufficient basis to conclude that a
more stringent standard to further
reduce peak concentrations is needed or
would benefit public health. As
discussed above, he judges that the
PM2.5 exposures in controlled human
exposure studies that correspond to
peak concentrations will already be well
controlled via the combination of the
revised annual standard, with a level of
9.0 mg/m3, and the 24-hour standard
with its level 35 mg/m3 and its 98th
percentile form. Taking into
consideration the inconsistent results
reported in controlled human exposure
studies, the intermediate nature of the
health effects observed in the controlled
human exposure studies that are not
typically considered adverse, the health
status of the study participants, and
how infrequently peak concentrations of
potential concern are anticipated to
occur in areas meeting the revised
primary annual PM2.5 standard, he
judges that the current 24-hour standard
is requisite to protect against the effects
reported in these studies with an
adequate margin of safety. Likewise, he
judges that neither the epidemiologic
studies (including the studies that use
restricted analyses) nor the risk
assessment provide a sufficient basis for
revising the 24-hour standard. As
discussed above, the epidemiologic
studies, including short-term studies
and those with restricted analyses, are
not well-suited for identifying a level for
a 24-hour standard to address health
effects associated with peak
concentrations. The restricted analyses
support the conclusion that the health
effects associated with PM2.5 is not
associated primarily with exposure to
higher concentrations of the main
analyses, but like other epidemiologic
studies they typically report only longterm mean 24-hour concentrations (e.g.,
restricted epidemiologic studies do not
PO 00000
Frm 00090
Fmt 4701
Sfmt 4700
report the number or the percentile of
health events or the percentile of PM2.5
concentrations across the highest part of
the restricted air quality distribution,
including the 98th percentile) and do
not identify any particular
concentration within the air quality
distribution above which effects have
been observed and below which effects
have not been observed. Similarly, the
risk assessment highlights that the
annual standard is controlling across
much of the U.S. and is generally more
effective at reducing risk than the 24hour standard and, taking into account
the limitations and assumptions of the
risk assessment discussed above, does
not provide a basis for revising the 24hour standard. For the reasons
discussed herein, the Administrator
judges that the uncertainties as to
whether there would be public health
benefits from a more stringent 24-hour
standard are too great to justify revising
the standard.
Thus, having carefully considered the
scientific evidence, quantitative
information, CASAC advice, and public
comments, the Administrator is
retaining the current primary 24-hour
PM2.5 standard, with its level of to 35
mg/m3 and its 98th percentile form. In
the Administrator’s judgment, based on
the currently available evidence and
information, a 24-hour standard set at
this level and using the specified
indicator, averaging time, and form
would be requisite to protect public
health with an adequate margin of
safety, in conjunction with the annual
standard. As noted, in evaluating the
adequacy of the current standards, the
Administrator focuses on evaluating the
public health protection afforded by the
annual and 24-hour standards, taken
together, against adverse health effects
associated with long- or short-term
PM2.5 exposures. A 24-hour standard set
at a level of 35 mg/m3, in conjunction
with a revised annual standard level of
9.0 mg/m3, in the judgment of the
Administrator, provides an appropriate
level of public health protection, for
both long- and short-term PM2.5
exposures. The Administrator believes
that a 24-hour standard set at 35 mg/m3
would continue to be sufficient to
protect public health with a margin of
safety, and believes that a lower
standard would be more than what is
necessary to provide this degree of
protection when considered in
conjunction with a revised annual
standard. The Administrator concludes
the current 24-hour standard at a level
of 35 mg/m3, in conjunction with a
revised annual standard level of 9.0 mg/
m3, will provide appropriate protection
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
in areas in which the long-term mean
concentrations are already relatively
low (i.e., below 9 mg/m3) but where
there may be elevated short-term peak
PM2.5 concentrations, often associated
with strong local or seasonal sources.
This judgment by the Administrator
appropriately considers the degree of
protection that is neither more nor less
stringent than necessary for this purpose
and recognizes that the CAA does not
require that primary standards be set at
a zero-risk level, but rather at a level
that reduces risk sufficiently so as to
protect public health with an adequate
margin of safety.
In making this decision to retain the
current level of the primary PM2.5 24hour standard at 35 mg/m3 in
conjunction with revising the annual
standard level from 12.0 mg/m3 to 9.0
mg/m3, given all of the evidence and
information discussed above, the
Administrator judges that the revised
suite of primary PM2.5 standards and the
rationale supporting these levels
appropriately reflects consideration of
the strength of the available evidence
and other information and its associated
uncertainties as well as the advice of
CASAC and consideration of public
comments. He additionally judges that
this suite of primary PM2.5 standards is
requisite to protect public health,
including at-risk populations, with an
adequate margin of safety from effects
associated with long and short-term
exposures to fine particles. This
judgment by the Administrator
appropriately considers the requirement
for standards that are requisite to protect
public health but are neither more nor
less stringent than necessary.
C. Decisions on the Primary PM2.5
Standards
For the reasons discussed above, and
taking into account the information and
assessments presented in the 2019 ISA
and ISA Supplement, the scientific and
quantitative risk information in the 2022
PA, the advice and recommendations of
the CASAC, and public comments, the
Administrator revises the current suite
of primary PM2.5 standards. Specifically,
the Administrator revises the level of
the primary annual PM2.5 standard to
9.0 mg/m3 while retaining its form,
indicator and averaging time. In
conjunction with revising the primary
annual PM2.5 standard level to provide
protection from effects associated with
long- and short-term PM2.5 exposures,
the Administrator retains the level of 35
mg/m3 and the 98th percentile form,
indicator and averaging time of the
primary 24-hour PM2.5 standard to
continue to provide supplemental
protection for areas with high peak
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
PM2.5 concentrations. The
Administrator concludes that this suite
of standards is requisite to protect
public health with an adequate margin
of safety against health effects
potentially associated with long- and
short-term PM2.5 exposures.
III. Rationale for Decisions on the
Primary PM10 Standard
This section presents the rationale for
the Administrator’s decision to retain
the existing primary PM10 standard.
This decision is based on a thorough
review of the latest scientific
information, published through January
2018 126 and evaluated in the 2019 ISA,
on human health effects associated with
PM10–2.5 in ambient air. As described in
section I above and in section 1.2 of the
ISA Supplement, the scope of the
updated scientific evaluation of the
health effects evidence is based on those
PM size fractions, exposure durations,
and health effects category
combinations where the 2019 ISA
concluded a causal relationship exists
(U.S. EPA, 2019a, U.S. EPA, 2022b).).
Therefore, because the 2019 ISA did not
conclude a causal relationship for
PM10–2.5 for any exposure durations or
health effect categories, the ISA
Supplement does not include an
evaluation of additional studies for
PM10–2.5. As a result, the 2019 ISA
continues to serve as the scientific
foundation for assessing the adequacy of
the primary PM10 standard in this
reconsideration of the 2020 final
decision (U.S. EPA, 2019a, section 1.7;
U.S. EPA, 2022a). The Administrator’s
decision also takes into account the
2022 PA evaluation of the policyrelevant information in the 2019 ISA,
CASAC advice and recommendations,
and public comments.
In presenting the rationale for the
Administrator’s final decision and its
foundations, Section III.A provides
background on the 2020 final decision
to retain the primary PM10 and a brief
summary of key aspects of the currently
available health effects information.
Section III.B summarizes the CASAC
advice and the Administrator’s
proposed conclusions to retain the
existing primary PM10 standard,
addresses public comments received on
126 In addition to the review’s opening ‘‘call for
information’’ (79 FR 71764, December 3, 2014), the
2019 ISA identified and evaluated studies and
reports that have undergone scientific peer review
and were published or accepted for publication
between January 1, 2009, through approximately
January 2018 (U.S. EPA, 2019a, p. ES–2). References
cited in the 2019 ISA, the references considered for
inclusion but not cited, and electronic links to
bibliographic information and abstracts can be
found at: https://hero.epa.gov/hero/particulatematter.
PO 00000
Frm 00091
Fmt 4701
Sfmt 4700
16291
the proposal, and presents the
Administrator’s conclusions on the
adequacy of the current standard,
drawing on consideration of information
in the 2019 ISA and the 2022 PA, advice
from the CASAC, and comments from
the public. Section III.C summarizes the
Administrator’s decision on the primary
PM10 standard.
A. Introduction
The general approach for this
reconsideration of the 2020 final
decision on the primary PM10 standard
relies on the scientific information
available for this review, as well as the
Administrator’s judgments regarding the
available public health effects evidence,
and the appropriate degree of public
health protection for the existing
standards. With the 2020 decision, the
then-Administrator retained the existing
primary 24-hour PM10 standard, with its
level of 150 mg/m3 and its one-expectedexceedance form on average over three
years, to continue to provide public
health protection against short-term
exposures to PM10–2.5 (85 FR 82725,
December 18, 2020).
1. Background on the Current Standard
Consistent with the 2009 ISA, the
2019 ISA concluded that the available
epidemiologic, controlled human
exposure, and animal toxicological
studies, including uncertainties,
provided support for the causality
determinations of ‘‘suggestive of, but not
sufficient to infer, a causal relationship’’
between short-term exposures to
PM10–2.5 and cardiovascular effects,
respiratory effects, and mortality (U.S.
EPA, 2019a, section 1.4.2). The 2019
ISA also reached the conclusion that the
evidence supports a ‘‘suggestive of, but
not sufficient to infer, a causal
relationship’’ between short-term
PM10–2.5 exposures and metabolic
effects, an endpoint that was not
evaluated in the 2009 ISA (U.S. EPA,
2019a, section 1.4.2).
Compared to the 2009 ISA, the 2019
ISA includes expanded evidence for the
relationships between long-term
exposures and cardiovascular effects,
metabolic effects, nervous system
effects, cancer, and mortality. The 2019
ISA concluded that the small number of
epidemiologic and experimental
studies, including uncertainties,
contribute to the determination that,
‘‘the evidence is suggestive of, but not
sufficient to infer, a causal relationship
between long-term PM10–2.5 exposure
and cardiovascular effects, metabolic
effects, nervous system effects, cancer,
and mortality and cancer (U.S. EPA,
2019a, p. 10–87). For long-term
exposures and cardiovascular effects,
E:\FR\FM\06MRR3.SGM
06MRR3
16292
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
cardiovascular effects, and cancer, this
is an upgrade from the ‘‘inadequate to
infer the presence or absence of a causal
relationship’’ conclusions in the 2009
ISA (U.S. EPA, 2019a, section 1.4.2).
This determination is also the first for
long-term exposures and metabolic
effects, as the 2009 ISA did not include
metabolic effects as an endpoint (U.S.
EPA, 2019a section 1.4.2).
In considering the available body of
evidence, it was noted in the 2020
review there were considerable
uncertainties and limitations associated
with the experimental evidence for
PM2.5 exposures and health effects, and
as such more weight was placed on the
available epidemiologic evidence.
Therefore, the primary focus in the 2020
review was on multi-city and single-city
epidemiologic studies that evaluated
associations between short-term
PM10–2.5 and mortality, cardiovascular
effects (hospital admissions and
emergency department visits, as well as
blood pressure and hypertension), and
respiratory effects. Despite differences
in the approaches 127 used to estimate
ambient PM10–2.5 concentrations, the
majority of the studies reported positive,
though often not statistically significant,
associations with short-term PM10–2.5
exposures. Most PM10–2.5 effect
estimates remained positive in
copollutant models that included either
gaseous pollutants or other particulate
matter size fractions (e.g., PM2.5). In U.S.
study locations likely to have met the
PM10 standard during the study period,
a few studies reported positive
associations between PM10–2.5 and
mortality that were statistically
significant and remained so in
copollutant models (U.S. EPA, 2019a).
In addition to the epidemiologic studies,
there were a small number of controlled
human exposure studies evaluated in
the 2019 ISA that reported alterations in
heart rate variability or increased
pulmonary inflammation following
short-term exposure to PM10–2.5,
providing some support for the
associations in the epidemiologic
studies. Animal toxicological studies
examined the effect of short-term
PM10–2.5 exposures using non-inhalation
(e.g., intratracheal instillation) route.128
127 As discussed further below, methods
employed by the epidemiologic studies to estimate
ambient PM10–2.5 concentrations include: (1)
Calculating the difference between PM10 and PM2.5
at co-located monitors, (2) calculating the difference
between county-wide averages of monitored PM10
and PM2.5 based on monitors that are not
necessarily co-located, and (3) direct measurement
of PM10–2.5 using a dichotomous sampler (U.S. EPA,
2019a, section 1.4.2).
128 Non-inhalation exposure experiments (i.e.,
intratracheal [IT] instillation) are informative for
size fractions (e.g., PM10–2.5) that cannot penetrate
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
Therefore, these studies provided
limited evidence for the biological
plausibility of PM10–2.5-induced effects
(U.S. EPA, 2019a). Although the
scientific evidence available in the 2019
ISA expanded the understanding of
health effects associated with PM10–2.5
exposures, a number of important
uncertainties remained. These
uncertainties, and their implications for
interpreting the scientific evidence,
include the following:
• The potential for confounding by
copollutants, notably PM2.5, was
addressed with copollutant models in a
relatively small number of PM10–2.5
epidemiologic studies (U.S. EPA,
2019a). This was particularly important
given the relatively small body of
experimental evidence (i.e., controlled
human exposure and animal
toxicological studies) available to
support the independent effect of
PM10–2.5 on human health. This
increases the uncertainty regarding the
extent to which PM10–2.5 itself, rather
than one or more copollutants, is
responsible for the mortality and
morbidity effects reported in
epidemiologic studies.
• There was greater spatial variability
in PM10–2.5 concentrations than PM2.5
concentrations, resulting in the
potential for increased exposure error
for PM10–2.5 (U.S. EPA, 2019a). Available
measurements did not provide sufficient
information to adequately characterize
the spatial distribution of PM10–2.5
concentrations (U.S. EPA, 2019a). The
limitations in estimates of ambient
PM10–2.5 concentrations ‘‘would tend to
increase uncertainty and make it more
difficult to detect effects of PM10–2.5 in
epidemiologic studies’’ (U.S. EPA,
2019a).
• Estimation of PM10–2.5
concentrations over which reported
health outcomes occur remain highly
uncertain. When compared with PM2.5,
there is uncertainty spanning all
epidemiologic studies examining
associations with PM10–2.5 including
deficiencies in the existing monitoring
networks, the lack of a systematic
evaluation of the various methods used
to estimate PM10–2.5 concentrations and
the resulting uncertainty in the spatial
as well as the temporal variability in
PM10–2.5 concentration (U.S. EPA,
2019a).). Given these limitations in
routine monitoring, epidemiologic
studies employed a number of different
approaches for estimating PM10–2.5
concentrations, including (1) calculating
the difference between PM10 and PM2.5
the airway of a study animal and may provide
information relevant to biological plausibility and
dosimetry (U.S. EPA, 2019a, section A–12).
PO 00000
Frm 00092
Fmt 4701
Sfmt 4700
at co-located monitors, (2) calculating
the difference between county-wide
averages of monitored PM10 and PM2.5
based on monitors that are not
necessarily co-located, and (3) direct
measurement of PM10–2.5 using a
dichotomous sampler (U.S. EPA, 2019a,
section 1.4.2). Given the relatively small
number of PM10–2.5 monitoring sites, the
relatively large spatial variability in
ambient PM10–2.5 concentrations, the use
of different approaches to estimating
ambient PM10–2.5 concentrations across
epidemiologic studies, and the
limitations inherent in such estimates,
the distributions of PM10–2.5
concentrations over which reported
health outcomes occur remain highly
uncertain (U.S. EPA, 2019a).
There was relatively little information
available to characterize potential
exposure differences that may inform
the apparent variability in associations
between short-term PM10–2.5 exposures
and health effects across study locations
(U.S. EPA, 2019a). Specifically, the
potential spatial and temporal
variability in PM10–2.5 exposures
complicates the interpretation of results
between study locations as well as the
relative lack of information on the
chemical and biological composition of
PM10–2.5 (U.S. EPA, 2009a U.S. EPA,
2019a).
In reaching his decision in 2020 to
retain the existing 24-hour primary
PM10 standard, the then-Administrator
specifically noted that, while the health
effects evidence was somewhat
expanded since the prior reviews, the
overall conclusions in the 2019 ISA,
including uncertainties and limitations,
were generally consistent with what was
considered in the 2012 review (85 FR
82725, December 18, 2020). In addition,
the then-Administrator recognized that
there were still a number of
uncertainties and limitations associated
with the available evidence.
With regard to the evidence on
PM10–2.5-related health effects, the thenAdministrator noted that epidemiologic
studies continued to report positive
associations with mortality and
morbidity in cities across North
America, Europe, and Asia, where
PM10–2.5 sources and composition were
expected to vary widely. While
significant uncertainties remained in the
2020 review, the then-Administrator
recognized that this expanded body of
evidence had broadened the range of
effects that have been linked with
PM10–2.5 exposures. The studies
evaluated in the 2019 ISA expanded the
scientific foundation presented in the
2009 ISA and led to revised causality
determinations (and new
determinations) for long-term PM10–2.5
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
exposures and mortality, cardiovascular
effects, metabolic effects, nervous
system effects, and cancer (85 FR 82726,
December 18, 2020). Drawing from his
consideration of this evidence, the thenAdministrator concluded that the
scientific information available since
the time of the last review supported a
decision to maintain a primary PM10
standard to provide public health
protection against PM10–2.5 exposures,
regardless of location, source of origin,
or particle composition (85 FR 82726,
December 18, 2020). With regard to
uncertainties in the available evidence,
the then-Administrator first noted that a
number of limitations were identified in
the 2012 review related to: (1) Estimates
of ambient PM10–2.5 concentrations used
in epidemiologic studies; (2) limited
evaluation of copollutant models to
address the potential for confounding;
and (3) limited experimental studies
supporting biological plausibility for
PM10–2.5-related effects. Despite the
expanded body of evidence for PM10–2.5
exposures and health effects, the thenAdministrator recognized that
uncertainties in the 2020 review
continued to include those associated
with the exposure estimates used in
epidemiologic studies, the
independence of the PM10–2.5 health
effect associations, and the biologically
plausible pathways for PM10–2.5 health
effects (85 FR 82726, December 18,
2020). These uncertainties contributed
to the 2019 ISA determinations that the
evidence is at most ‘‘suggestive of, but
not sufficient to infer’’ causal
relationships (85 FR 82726, December
18, 2020). In considering the available
evidence in his basis for the decision,
the then-Administrator emphasized
evidence supporting ‘‘causal’’ and
‘‘likely to be causal’’ relationships, and
therefore, judged that the PM10–2.5related health effects evidence provided
an uncertain scientific foundation for
making standard-setting decisions. He
further judged limitations in the
evidence raised questions as to whether
additional public health improvements
would be achieved by revising the
existing PM10 standard (85 FR 24126,
April 30, 2020). In the 2020 decision, for
all of the reasons discussed above and
recognizing the CASAC conclusion that
the evidence provided support for
retaining the current standard, the thenAdministrator concluded that it was
appropriate to retain the existing
primary PM10 standard, without
revision. His decision was consistent
with the CASAC advice related to the
primary PM10 standard. Specifically, the
CASAC agreed with the 2020 PA
conclusions that, while these effects are
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
important, the ‘‘evidence does not call
into question the adequacy of the public
health protection afforded by the
current primary PM10 standard’’ and
‘‘supports consideration of retaining the
current standard in this review’’ (Cox,
2019b, p. 3 of consensus letter). Thus,
the then-Administrator concluded that
the primary PM10 standard (in all of its
elements (i.e., indicator, averaging time,
form, and level)) was requisite to protect
public health with an adequate margin
of safety against effects that have been
associated with PM10–2.5. In light of this
conclusion, the EPA retained the
existing PM10 standard.
2. Overview of the Health Effects
Evidence
The information summarized here is
based on the scientific assessment of the
health effects evidence available in this
reconsideration; this evaluation is
documented in the 2019 ISA and its
policy implications are discussed
further in the 2022 PA. As noted above,
the ISA Supplement does not include an
evaluation of studies for PM10–2.5, and
the 2019 ISA continues to serve as the
scientific foundation for this
reconsideration.
a. Nature of Effects
For the health effect categories and
exposure duration combinations
evaluated, the 2019 ISA concludes that
the evidence supports causality
determinations for PM10–2.5 that are at
most ‘‘suggestive of, but not sufficient to
infer, a causal relationship’’. While the
evidence supporting the causal nature of
relationships between exposure to
PM10–2.5 has been strengthened for some
health effect categories since the
completion of the 2009 ISA, the 2019
ISA concludes that overall ‘‘the
uncertainties in the evidence identified
in the 2009 ISA have, to date, still not
been addressed’’ (U.S. EPA, 2019a,
section 1.4.2, p. 1–41; U.S. EPA, 2022b,
section 4.3.1). Specifically,
epidemiologic studies available in the
2012 review relied on various methods
to estimate PM10–2.5 concentrations, and
these methods had not been
systematically compared to evaluate
spatial and temporal correlations in
PM10–2.5 concentrations. Methods
included: (1) Calculating the difference
between PM10 and PM2.5 concentrations
at co-located monitors, (2) calculating
the difference between county-wide
averages of monitored PM10- and PM2.5based on monitors that are not
necessarily co-located, and (3) direct
measurement of PM10–2.5 using a
dichotomous sampler (U.S. EPA, 2019a,
section 1.4.2). As described in the 2019
ISA, there continues to be variability
PO 00000
Frm 00093
Fmt 4701
Sfmt 4700
16293
across epidemiologic studies in the
approaches used to estimate PM10–2.5
concentrations. Additionally, some
studies estimate long-term PM10–2.5
exposures as the difference between
PM10 and PM2.5 concentrations based on
information from spatiotemporal or land
use regression (LUR) models, in
addition to monitors. The various
methods used to estimate PM10–2.5
concentrations have not been
systematically evaluated (U.S. EPA,
2019a, section 3.3.1.1), contributing to
uncertainty regarding the spatial and
temporal correlations in PM10–2.5
concentrations across methods and in
the PM10–2.5 exposure estimates used in
epidemiologic studies (U.S. EPA, 2019a,
section 2.5.1.2.3). Given the greater
spatial and temporal variability of
PM10–2.5 and the lower number of
PM10–2.5 monitoring sites, compared to
PM2.5, this uncertainty is particularly
important for the coarse size fraction.
Beyond the uncertainty associated with
PM10–2.5 exposure estimates in
epidemiologic studies, the limited
information on the potential for
confounding by copollutants and the
limited support available for the
biological plausibility of health effects
following PM10–2.5 exposures also
continue to contribute to uncertainty in
the PM10–2.5 health evidence.
Uncertainty related to potential
confounding stems from the relatively
small number of epidemiologic studies
that have evaluated PM10–2.5 health
effect associations in copollutants
models with both gaseous pollutants
and other PM size fractions. On the
other hand, uncertainty related to the
biological plausibility of effects
attributed to PM10–2.5 exposures results
from the small number of controlled
human exposure and animal
toxicological studies that have evaluated
the health effects of experimental
PM10–2.5 inhalation exposures. The
evidence supporting the 2019 ISA’s
‘‘suggestive of, but not sufficient to
infer, a causal relationship’’ causality
determinations for PM10–2.5, including
uncertainties in this evidence, is
summarized below in sections III.B.1.a
through III.B.1.f.
i. Mortality
Due to the dearth of studies
examining the association between longterm PM10–2.5 exposure and mortality,
the 2009 ISA concluded that the
evidence was ‘‘inadequate to determine
if a causal relationship exists’’ (U.S.
EPA, 2009a). As reported in the 2019
ISA, some cohort studies conducted in
the U.S. and Europe report positive
associations between long-term PM10–2.5
exposure and total (nonaccidental)
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16294
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
mortality, though results are
inconsistent across studies (U.S. EPA,
2019a, Table 11–11). The examination
of copollutant models in these studies
remains limited and, when included,
PM10–2.5 effect estimates are often
attenuated after adjusting for PM2.5 (U.S.
EPA, 2019a, Table 11–11). Across
studies, PM10–2.5 exposure
concentrations are estimated using a
variety of approaches, including direct
measurements from dichotomous
samplers, calculating the difference
between PM10 and PM2.5 concentrations
measured at collocated monitors, and
calculating difference of area-wide
concentrations of PM10 and PM2.5. As
discussed above, temporal and spatial
correlations between these approaches
have not been evaluated, contributing to
uncertainty regarding the potential for
exposure measurement error (U.S. EPA,
2019a, section 3.3.1.1 and Table 11–11).
The 2019 ISA concludes that this
uncertainty ‘‘reduces the confidence in
the associations observed across
studies’’ (U.S. EPA, 2019a, p. 11–125).
The 2019 ISA additionally concludes
that the evidence for long-term PM10–2.5
exposures and cardiovascular effects,
respiratory morbidity, and metabolic
disease provide limited biological
plausibility for PM10–2.5-related
mortality (U.S. EPA, 2019a, sections
11.4.1 and 11.4). Taken together, the
2019 ISA concludes that, ‘‘this body of
evidence is suggestive, but not sufficient
to infer, that a causal relationship exists
between long-term PM10–2.5 exposure
and total mortality’’ (U.S. EPA, 2019a, p.
11–125).
With regard to short-term PM10–2.5
exposures and mortality, the 2009 ISA
concluded that the evidence is
‘‘suggestive of a causal relationship
between short-term exposure to PM10–2.5
and mortality’’ (U.S. EPA, 2009a). The
2019 ISA included multicity
epidemiologic studies conducted
primarily in Europe and Asia that
continue to provide consistent evidence
of positive associations between shortterm PM10–2.5 exposure and total
(nonaccidental) mortality (U.S. EPA,
2019a, Table 11–9). Although these
studies contribute to increasing
confidence in the PM10–2.5-mortality
relationship, the use of various
approaches to estimate PM10–2.5
exposures continues to contribute
uncertainty to the associations observed.
Recent studies expand the assessment of
potential copollutant confounding of the
PM10–2.5-mortality relationship and
provide evidence that PM10–2.5
associations generally remain positive
in copollutant models, though
associations are attenuated in some
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
instances (U.S. EPA, 2019a, section
11.3.4.1, Figure 11–28, Table 11–10).
The 2019 ISA concludes that, overall,
the assessment of potential copollutant
confounding is limited due to the lack
of information on the correlation
between PM10–2.5 and gaseous pollutants
and the small number of locations in
which copollutant analyses have been
conducted. Associations with causespecific mortality (i.e., cardiovascular
and respiratory mortality) provide some
support for associations with total
(nonaccidental) mortality, though
associations with respiratory mortality
are more uncertain (i.e., wider
confidence intervals) and less consistent
(U.S. EPA, 2019a, section 11.3.7). The
2019 ISA concludes that the evidence
for PM10–2.5-related cardiovascular
effects provides only limited support for
the biological plausibility of a
relationship between short-term
PM10–2.5 exposure and cardiovascular
mortality (U.S. EPA, 2019a, section
11.3.7). Based on the overall evidence,
the 2019 ISA concludes that, ‘‘this body
of evidence is suggestive, but not
sufficient to infer, that a causal
relationship exists between short-term
PM10–2.5 exposure and total mortality’’
(U.S. EPA, 2019a, p. 11–120).
ii. Cardiovascular Effects
In the 2009 ISA, the evidence
describing the relationship between
long-term exposure to PM10–2.5 and
cardiovascular effects was characterized
as ‘‘inadequate to infer the presence or
absence of a causal relationship.’’ The
limited number of epidemiologic
studies reported contradictory results
and experimental evidence
demonstrating an effect of PM10–2.5 on
the cardiovascular system was lacking
(U.S. EPA, 2019a, section 6.4).
The evidence relating long-term
PM10–2.5 exposures to cardiovascular
mortality remains limited, with no
consistent pattern of associations across
studies and, as discussed above,
uncertainty stemming from the use of
various approaches to estimate PM10–2.5
concentrations (U.S. EPA, 2019a, Table
6–70). The evidence for associations
with cardiovascular morbidity has
grown and, while results across studies
are not entirely consistent, some
epidemiologic studies report positive
associations with ischemic heart disease
(IHD) and MI (U.S. EPA, 2019a, Figure
6–34); stroke (U.S. EPA, 2019a, Figure
6–35); atherosclerosis (U.S. EPA, 2019a,
section 6.4.5); venous thromboembolism
(VTE) (U.S. EPA, 2019a, section 6.4.7);
and blood pressure and hypertension
(U.S. EPA, 2019a, Section 6.4.6).
PM10–2.5 cardiovascular mortality effect
estimates are often attenuated, but
PO 00000
Frm 00094
Fmt 4701
Sfmt 4700
remain positive, in copollutants models
that adjust for PM2.5. For morbidity
outcomes, associations are inconsistent
in copollutant models that adjust for
PM2.5, NO2, and chronic noise pollution
(U.S. EPA, 2019a, p. 6–276). The lack of
toxicological evidence for long-term
PM10–2.5 exposures represents a data gap
(U.S. EPA, 2019a, section 6.4.10),
resulting in the 2019 ISA conclusion
that ‘‘evidence from experimental
animal studies is of insufficient quantity
to establish biological plausibility’’ (U.S.
EPA, 2019a, p. 6–277). Based largely on
the observation of positive associations
in some epidemiologic studies, the 2019
ISA concludes that ‘‘evidence is
suggestive of, but not sufficient to infer,
a causal relationship between long-term
PM10–2.5 exposure and cardiovascular
effects’’ (U.S. EPA, 2019a, p. 6–277).
With regard to short-term PM10–2.5
exposures and cardiovascular effects,
the 2009 ISA found that the available
evidence for short-term PM10–2.5
exposure and cardiovascular effects was
‘‘suggestive of a causal relationship.’’
This conclusion was based on several
epidemiologic studies reporting
associations between short-term
PM10–2.5 exposure and cardiovascular
effects, including IHD hospitalizations,
supraventricular ectopy, and changes in
heart rate variability (HRV). In addition,
dust storm events resulting in high
concentrations of crustal material were
linked to increases in total
cardiovascular disease emergency
department visits and hospital
admissions. However, the 2009 ISA
noted the potential for exposure
measurement error primarily due to the
different methods used across studies to
estimate PM10–2.5 concentrations and
copollutant confounding in these
epidemiologic studies. In addition, there
was only limited evidence of
cardiovascular effects from a small
number of experimental studies (e.g.
animal toxicological studies and
controlled human exposure studies) that
examined short-term PM10–2.5 exposures
(U.S. EPA, 2009a, section 6.2.12.2). In
the 2019 ISA, key uncertainties
included the potential for exposure
measurement error, copollutant
confounding, and limited evidence of
biological plausibility for cardiovascular
effects following inhalation exposure
(U.S. EPA, 2019a, section 6.3.13).
The evidence for short-term PM10–2.5
exposure and cardiovascular outcomes
has expanded since the 2009 ISA,
though important uncertainties remain.
The 2019 ISA notes that there are a
small number of epidemiologic studies
reporting positive associations between
short-term exposure to PM10–2.5 and
cardiovascular-related morbidity
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
outcomes. However, the 2019 ISA notes
that there is limited evidence to support
that these associations are biologically
plausible, or independent of copollutant
confounding. The 2019 ISA also
concludes that it remains unclear how
the approaches used to estimate PM10–2.5
concentrations in epidemiologic studies
compare amongst one another and
subsequently how exposure
measurement error varies between each
method. Specifically, it is unclear how
well-correlated PM10–2.5 concentrations
are both temporally and spatially across
these methods and therefore whether
exposure measurement error varies
across these methods. Taken together,
the 2019 ISA concludes that ‘‘the
evidence is suggestive of, but not
sufficient to infer, a causal relationship
between short-term PM10–2.5 exposures
and cardiovascular effects’’ (U.S. EPA,
2019a, p. 6–254).
iii. Respiratory Effects
With regard to short-term PM10–2.5
exposures and respiratory effects, the
2009 ISA (U.S. EPA, 2009a) concluded
that the relationship between short-term
exposure to PM10–2.5 and respiratory
effects is ‘‘suggestive of a causal
relationship’’ based on a small number
of epidemiologic studies observing
associations with some respiratory
effects and limited evidence from
experimental studies to support
biological plausibility. Epidemiologic
findings were consistent for respiratory
infection and combined respiratoryrelated diseases, but not for COPD.
Studies were characterized by overall
uncertainty in the exposure assignment
approach and limited information
regarding potential copollutant
confounding. Controlled human
exposure studies of short-term PM10–2.5
exposures found no lung function
decrements and inconsistent evidence
for pulmonary inflammation. Animal
toxicological studies were limited to
those using non-inhalation (e.g., intratracheal instillation) routes of PM10–2.5
exposure.
Recent epidemiologic findings
consistently link PM10–2.5 exposure to
asthma exacerbation and respiratory
mortality, with some evidence that
associations remain positive (though
attenuated in some studies of mortality)
in copollutant models that include
PM2.5 or gaseous pollutants.
Epidemiologic studies provide limited
evidence for positive associations with
other respiratory outcomes, including
COPD exacerbation, respiratory
infection, and combined respiratoryrelated diseases (U.S. EPA, 2019a, Table
5–36). As noted above for other
endpoints, an uncertainty in these
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
epidemiologic studies is the lack of a
systematic evaluation of the various
methods used to estimate PM10–2.5
concentrations and the resulting
uncertainty in the spatial and temporal
variability in PM10–2.5 concentrations
compared to PM2.5 (U.S. EPA, 2019a,
sections 2.5.1.2.3 and 3.3.1.1).
Specifically, the existing monitoring
networks do not provide a good
characterization of how well correlated
concentrations are both spatially and
temporally across the PM10–2.5
estimation methods and overall spatial
and temporal patterns in PM10–2.5
concentrations. Taken together, the 2019
ISA concludes that ‘‘the collective
evidence is suggestive of, but not
sufficient to infer, a causal relationship
between short-term PM10–2.5 exposure
and respiratory effects’’ (U.S. EPA,
2019a, p. 5–270).
iv. Cancer
In the 2012 review, little information
was available from studies of cancer
following inhalation exposures to
PM10–2.5. Thus, the 2009 ISA determined
the evidence was ‘‘inadequate to
evaluate the relationship between longterm PM10–2.5 exposures and cancer’’
(U.S. EPA, 2009a). The scientific
information evaluated in the 2019 ISA
of long-term PM10–2.5 exposure and
cancer remains limited, with a few
recent epidemiologic studies reporting
positive, but imprecise, associations
with lung cancer incidence (U.S. EPA,
2019a). Moreover, uncertainty remains
in these studies with respect to
exposure measurement error due to the
use of PM10–2.5 predictions that have not
been validated by monitored PM10–2.5
concentrations (U.S. EPA, 2019a,
sections 3.3.2.3 and 10.3.4). Relatively
few experimental studies of PM10–2.5
have been conducted, though available
studies indicate that PM10–2.5 exhibits
two key characteristics of carcinogens:
genotoxicity and oxidative stress. While
limited, such experimental studies
provide some evidence of biological
plausibility for the findings in a small
number of epidemiologic studies (U.S.
EPA, 2019a, section 10.3.4).
Taken together, the small number of
epidemiologic and experimental
studies, along with uncertainty with
respect to exposure measurement error,
contribute to the determination in the
2019 ISA that, ‘‘the evidence is
suggestive of, but not sufficient to infer,
a causal relationship between long-term
PM10–2.5 exposure and cancer’’ (U.S.
EPA, 2019a, p. 10–87).
v. Metabolic Effects
The 2009 ISA did not make a
causality determination for PM10–2.5-
PO 00000
Frm 00095
Fmt 4701
Sfmt 4700
16295
related metabolic effects. One
epidemiologic study in the 2019 ISA
reports an association between longterm PM10–2.5 exposure and incident
diabetes, while additional crosssectional studies report associations
with effects on glucose or insulin
homeostasis (U.S. EPA, 2019a, section
7.4). As discussed above for other
outcomes, uncertainties with the
epidemiologic evidence include the
potential for copollutant confounding
and exposure measurement error due to
the different methods used across
studies to estimate PM10–2.5
concentrations (U.S. EPA, 2019a, Tables
7–14 and 7–15). The evidence base to
support the biological plausibility of
metabolic effects following PM10–2.5
exposures is limited, but a crosssectional study that investigated
biomarkers of insulin resistance and
systemic and peripheral inflammation
may support a pathway leading to type
2 diabetes (U.S. EPA, 2019a, sections
7.4.1 and 7.4.3). Based on the expanded,
though still limited evidence base, the
2019 ISA concludes that, ‘‘[o]verall, the
evidence is suggestive of, but not
sufficient to infer, a causal relationship
between [long]-term PM10–2.5 exposure
and metabolic effects’’ (U.S. EPA, 2019a,
p. 7–56).
vi. Nervous System Effects
The 2009 ISA did not make a
causality determination for PM10–2.5related nervous system effects. In the
2019 ISA, available epidemiologic
studies report associations between
PM10–2.5 and impaired cognition and
anxiety in adults in longitudinal
analyses (U.S. EPA, 2019a, Table 8–25,
section 8.4.5). Associations of long-term
exposure with neurodevelopmental
effects are not consistently reported in
children (U.S. EPA, 2019a, sections
8.4.4 and 8.4.5). Uncertainties in these
studies include the potential for
copollutant confounding, as no studies
examined copollutants models (U.S.
EPA, 2019a, section 8.4.5), and for
exposure measurement error, given the
use of various methods to estimate
PM10–2.5 concentrations (U.S. EPA,
2019a, Table 8–25). In addition, there is
limited animal toxicological evidence
supporting the biological plausibility of
nervous system effects (U.S. EPA,
2019a, sections 8.4.1 and 8.4.5). Overall,
the 2019 ISA concludes that, ‘‘the
evidence is suggestive of, but not
sufficient to infer, a causal relationship’’
between long-term PM10–2.5 exposure
and nervous system effects (U.S. EPA,
2019a, p. 8–75).
E:\FR\FM\06MRR3.SGM
06MRR3
16296
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
B. Conclusions on the Primary PM10
Standard
In drawing conclusions on the
adequacy of the current primary PM10
standard, in view of the advances in
scientific knowledge and additional
information now available, the
Administrator has considered the
evidence base, information, and policy
judgments that were the foundation of
the 2020 review and reflects upon the
body of information and evidence
available in this reconsideration. In so
doing, the Administrator has taken into
account both evidence-based and
quantitative information-based
considerations, as well as advice from
the CASAC and public comments.
Evidence-based considerations draw
upon the EPA’s integrated synthesis of
the scientific evidence from animal
toxicologic, controlled human exposure,
and epidemiologic studies evaluating
health effects related to exposures to
PM10–2.5 as presented in the 2019 ISA
and discussed in section III.A.2. In
addition to the evidence, the
Administrator has weighed a range of
policy-relevant considerations as
discussed in the 2022 PA and
summarized in sections III.B and III.C of
the proposal and summarized in section
III.B.2 below. These considerations,
along with the advice from the CASAC
(section III.B.1) and public comments
(section III.B.3), are discussed below. A
more detailed summary of all significant
comments, along with the EPA’s
responses in the Response to Comments
document, can be found in the docket
for this rulemaking (Docket No. EPA–
HQ–OAR–2015–00072). This document
is available for review in the docket for
this rulemaking and through EPA’s
NAAQS website (link). The
Administrator’s conclusions in this
reconsideration regarding the adequacy
of the current primary PM10 standard
and whether any revisions are
appropriate are described in section
III.B.4.
1. CASAC Advice
As described in section I.X, the EPA
decided to prepare a revised PA for the
reconsideration of the 2020 final
decision. The CASAC’s advice on the
2019 draft PA and the 2021 draft PA
was documented in letters to the prior
and current Administrators (Cox, 2019b;
Sheppard, 2022a) and is summarized
below. In reviewing both the 2019 draft
PA and the 2021 draft PA, the CASAC
agreed with the EPA’s preliminary
conclusion that the available scientific
evidence, including its uncertainties
and limitations, does not call into
question the adequacy of the current
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
primary PM10 standard and that the
standard should be retained, without
revision.
In its review of the 2019 draft PA, the
CASAC concurred with the overall
preliminary conclusion that it is
appropriate to consider retaining the
current primary PM10 standard, without
revision. In their agreement with the
conclusions in the 2019 draft PA, the
CASAC stated that ‘‘that key
uncertainties identified in the last
review remain’’ (Cox, 2019b) and that
‘‘none of the identified health outcomes
linked to PM10–2.5’’ were judged to be
causal or likely to be causal (Cox, 2019b,
p. 12 of consensus responses).
Moreover, to reduce these uncertainties
in future reviews, the CASAC
recommended improvements to PM10–2.5
exposure assessment, including a more
extensive network for direct monitoring
of the PM10–2.5 fraction (Cox, 2019b, p.
13 of consensus responses). The CASAC
also recommended additional controlled
human exposure and animal
toxicological studies of the PM10–2.5
fraction to improve the understanding of
biological mechanisms and pathways
(Cox, 2019b, p. 13 of consensus
responses). Overall, the CASAC agreed
with the EPA’s preliminary conclusion
in the 2019 draft PA that ‘‘. . . the
available evidence does not call into
question the adequacy of the public
health protection afforded by the
current primary PM10 standard and that
evidence supports consideration of
retaining the current standard in this
review’’ (Cox, 2019b, p. 3 of letter).
In its review of the 2021 draft PA, the
CASAC provided advice on the
adequacy of the current primary PM10
standard in the context of its review of
the revised PA for this reconsideration
(Sheppard, 2022a) 129.) 130. In this
context, the CASAC supported the
preliminary conclusion in the 2021 draft
PA that the evidence reviewed in the
2019 ISA does not call into question the
public health protection provided by the
current primary PM10 standard against
PM10–2.5 exposures and concurs with the
2021 draft PA’s overall preliminary
conclusion that it is appropriate to
consider retaining the current primary
PM10 standard (Sheppard, 2022a, p. 4 of
consensus letter). Additionally, the
129 As described in section I.C.5.b above, the
scope of the ISA Supplement did not include
consideration of studies of health effects associated
with exposure to PM10–2.5. Therefore, the
information and conclusions presented in the 2022
PA are very similar to those in the 2020 PA.
130 As described in section I.C.5.b above, the
scope of the ISA Supplement did not include
consideration of studies of health effects associated
with exposure to PM10–2.5. Therefore, the
information and conclusions presented in the 2022
PA are very similar to those in the 2020 PA.
PO 00000
Frm 00096
Fmt 4701
Sfmt 4700
CASAC concurred that ‘‘. . . at this
time, PM10 is an appropriate choice as
the indicator for PM10–2.5’’ and ‘‘that it
is important to retain the level of
protection afforded by the current PM10
standard’’ (Sheppard, 2022a, p. 4 of
consensus letter). The CASAC also
recognized uncertainties associated with
the scientific evidence, including
‘‘compared to PM2.5 studies, the more
limited number of epidemiology studies
with positive statistically significant
findings, and the difficulty in extracting
the sole contribution of coarse PM to
observed adverse health effects’’
(Sheppard, 2022a, p. 19 of consensus
responses).
The CASAC recommended several
areas for additional research to reduce
uncertainties in the PM10–2.5 exposure
estimates used in the epidemiologic
studies, to evaluate the independence of
PM10–2.5 health effect associations, to
evaluate the biological plausibility of
PM10–2.5-related effects, and to increase
the number of studies examining
PM10–2.5-related health effects in at-risk
populations (Sheppard, 2022a, p. 20 of
consensus responses). Furthermore, the
CASAC ‘‘recognizes a need for, and
supports investment in research and
deployment of measurement systems to
better characterize PM10–2.5’’ and to
‘‘provide information that can improve
public health’’ (Sheppard, 2022a, p. 20
of consensus responses).
2. Basis for the Proposed Decision
At the time of the proposal, the
Administrator carefully considered the
assessment of the current evidence and
conclusions reached in the 2019 ISA,
considerations and staff conclusions
and associated rationales presented in
the 2020 PA and 2022 PA, and advice
and recommendations of the CASAC (88
FR 5634, January 27, 2023). Consistent
with previous reviews, the
Administrator first considered the
available scientific evidence for
PM10–2.5-related exposures and health
effects, as evaluated in the 2019 ISA. As
an initial matter, the Administrator
recognized that the scientific evidence
for PM10–2.5-related effects available in
this reconsideration is the same body of
evidence that was available at the time
of the 2020 review, as evaluated in the
2019 ISA and summarized in section
III.A.2 above. The 2019 ISA concludes
that the evidence supports ‘‘suggestive
of, but not sufficient to infer’’ causal
relationships between short- and longterm exposures to PM10–2.5 and
cardiovascular effects, cancer, and
mortality and long-term PM10–2.5
exposures and metabolic effects and
nervous system effects (U.S. EPA,
2019a). The Administrator noted that
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
the evidence for several PM10–2.5-related
health effects has expanded since the
completion of the 2009 ISA, but
important uncertainties remain. The
uncertainties in the epidemiologic
studies contribute to the determinations
in the 2019 ISA that the evidence for
short and long-term PM10–2.5 exposures
and mortality, cardiovascular effects,
metabolic effects, nervous system
effects, and cancer is ‘‘suggestive of, but
not sufficient to infer’’ causal
relationships (U.S. EPA, 2019a; U.S.
EPA, 2022b, section 4.3.1). Drawing
from the evidence evaluated in the 2019
ISA and consideration of the scientific
evidence in the 2022 PA, the
Administrator noted that, consistent
with previous reviews, the 2019 ISA
and the 2022 PA highlight a number of
uncertainties associated with the
evidence, including: (1) PM10–2.5
exposure estimates used in
epidemiologic studies, (2) independence
of PM10–2.5 health effect associations,
and (3) biological plausibility of the
PM10–2.5-related effects. These
uncertainties contribute to the
determinations in the 2019 ISA that the
evidence for short-term PM10–2.5
exposures and key health effects is
‘‘suggestive of, but not sufficient to
infer’’ causal relationships. In
considering the available scientific
evidence, consistent with approaches
employed in past NAAQS reviews, the
Administrator placed the most weight
on evidence supporting ‘‘causal’’ and
‘‘likely to be causal’’ relationships. In so
doing, he noted that the available
evidence for short- and long-term
PM10–2.5 exposures and health effects
does not support causality
determinations of a ‘‘causal
relationship’’ or ‘‘likely to be causal
relationship.’’ Furthermore, the
Administrator recognized that, because
of the uncertainties and limitations in
the evidence base, the 2022 PA does not
include a quantitative assessment of
PM10–2.5 exposures and risk that might
further inform decisions regarding the
adequacy of the current 24-hour primary
PM10 standard. Therefore, in light of the
2019 ISA conclusions that the evidence
supports ‘‘suggestive of, but not
sufficient to infer’’ causal relationships.
The Administrator judged that there are
substantial uncertainties that raise
questions regarding the degree to which
additional public health improvements
would be achieved by revising the
existing PM10 standard. In considering
the available evidence for long-term
PM10–2.5 exposures, the Administrator
noted that there is limited evidence that
would support consideration of an
annual standard to provide protection
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
against such effects, in conjunction with
the current primary 24-hour PM10
standard. He preliminarily concluded
that the current primary 24-hour PM2.5
standard that reduces 24-hour exposures
also likely reduces long-term average
exposures, and therefore provides some
margin of safety against the health
effects associated with long-term
PM10–2.5 exposures.
In reaching his proposed decision on
the adequacy of the current primary 24hour PM10 standard, the Administrator
also considered advice from the CASAC.
As noted above in section III.B.1, the
CASAC recognized uncertainties
associated with the scientific evidence
and agreed with the 2019 draft PA and
2021 draft PA conclusions that the
scientific evidence does not call into
question the adequacy of the primary
PM10 standard and supports
consideration of retaining the current
standard.
When considering the above
information together, the Administrator
proposed to conclude that the available
scientific evidence continues to support
a PM10 standard to provide some
measure of protection against PM10–2.5
exposures. Additionally, he recognized
that there are important uncertainties
and limitations associated with the
available evidence for PM10–2.5-related
health effects, for both short and longterm exposure, as evaluated in the 2019
ISA. Consistent with the decisions in
the previous reviews, the Administrator
proposed to conclude that these
limitations lead to considerable
uncertainty regarding the potential
public health implications of revising
the level of the current primary 24-hour
PM10 standard. Thus, based on his
consideration of the evidence and
associated uncertainties and limitations
for PM10–2.5-related health effects and
his consideration of CASAC advice on
the primary PM10 standard, the
Administrator proposed to retain the
current primary PM10 standard, without
revision.
3. Comments on the Proposed Decision
Of the public comments received on
the proposal, very few commenters
provided comments on the primary
PM10 standard. Of those commenters
who did provide comments on the
primary PM10 standard, the majority
agree with the EPA’s proposed decision
to retain the primary PM10 standard. In
so doing, these commenters agree with
the EPA’s rationale regarding the
available scientific information,
including uncertainties and limitations,
for informing decisions on the standard.
These commenters state that no new
scientific evidence or quantitative
PO 00000
Frm 00097
Fmt 4701
Sfmt 4700
16297
information has emerged since the 2020
decision to retain the current standard.
Furthermore, these commenters note
that the EPA did not evaluate any new
scientific evidence related to PM10–2.5
exposures and health effects as a part of
the 2022 ISA Supplement developed for
this reconsideration, nor did the revised
2022 PA consider any new or different
information from the 2020 PA, and
therefore, the EPA reached the same
conclusion as is the 2020 PA that the
current standard is adequate and should
be retained. This group includes
industries and industry groups, as well
as some State and local governments.
All of these commenters generally note
their agreements with the rationale
provided in the proposal and the
CASAC concurrence with the 2021 draft
PA conclusion that the available
information does not call into question
the adequacy of the current standard,
and therefore, does not support revision
and that the current standard should be
retained.
Some commenters, including those
from environmental and public health
organizations and groups, some states,
and individuals, disagreed with the
Administrator’s proposed decision to
retain the current primary PM10
standard. These commenters
recommend that the EPA revise the
primary PM10 standard to a lower level
to provide increased public health
protection, citing to the available
scientific evidence, as well as the
proposed revision to the primary PM2.5
standard.
Commenters who disagreed with the
proposal to retain the current standard
state that revision to the primary PM10
standard is necessary to protect public
health with an adequate margin of
safety. In their recommendations for
revising the standard, some commenters
contend that the current standard, with
its indicator of PM10 to target exposures
to PM10–2.5, has become less protective
as ambient concentrations of PM2.5 have
been reduced with revisions to that
standard. These commenters assert that
the current primary PM10 standard
allows increased exposure to PM10–2.5 in
ambient air because retaining the
primary PM10 would allow
proportionately more PM10–2.5 mass as
the PM2.5 standard has been revised
downward. Moreover, in support of
their recommendations, the commenters
note that the available evidence of
PM10–2.5-related health effects has been
expanded and strengthened since the
time of the last review. Taken together,
the commenters contend that the
primary PM10 standard should be
revised and failure to do so would be
arbitrary and capricious. Some of these
E:\FR\FM\06MRR3.SGM
06MRR3
16298
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
commenters assert that the level of the
primary PM10 standard should be
revised to 140 or 145 mg/m3, concurrent
with a strengthened primary 24-hour
PM2.5 standard, while other commenters
recommend revising the level of the
standard to within the range of 65–75
mg/m3, to provide increased public
health protection.
We disagree with the commenters that
the primary PM10 standard should be
revised because of reductions in
ambient concentrations of PM2.5. As an
initial matter, we note that overall,
ambient concentrations of both PM10
and PM2.5 have declined significantly
over time. Ambient concentrations of
PM10 have declined by 46% across the
U.S. from 2000 to 2019,131 while PM2.5
concentrations in ambient air have
declined by 43% during this same time
period.132 As noted in the 2022 PA (p.
2–41), the majority of PM10–2.5 sites have
generally remained steady and do not
exhibit a trend of increasing or
decreasing concentrations during this
time period, reflecting the relatively
consistent level of dust emission across
the U.S. from 2000 to 2019 (U.S. EPA,
2022b).
The 2019 ISA provides a comparison
of the relative contribution of PM2.5 and
PM10–2.5 to PM10 concentrations by
region and season using the more
comprehensive monitoring data from
the NCore network available in this
reconsideration (U.S. EPA, 2019, section
2.5.1.1.4). The data indicate that, for
urban areas, there are roughly
equivalent amounts of PM2.5 and
PM10–2.5 contributing to PM10 in ambient
air, while rural locations have a slightly
higher contribution of PM10–2.5
contributing to PM10 concentrations
than PM2.5 (U.S. EPA, 2019, section
2.5.1.1.4, Table 2–7). There is generally
a greater contribution from the PM2.5
fraction in the East and a greater
contribution from the PM10–2.5 fraction
in the West and Midwest.
The EPA recognizes that when the
primary annual PM2.5 standard was
revised from 15.0 mg/m3 to 12.0 mg/m3
while leaving the 24-hour PM2.5
standards unchanged at 35 mg/m3 and
the 24-hour PM10 standard unchanged at
150 mg/m3, the PM10–2.5 fraction of PM10
could increase in some areas as the
PM2.5 fraction decreases (78 FR 3085,
131 PM
10 concentrations presented as the annual
second maximum 24-hour concentration (in mg/m3)
at 262 sites in the U.S. For more information, see:
https://www.epa.gov/air-trends/particulate-matterpm10-trends
132 PM
2.5 concentrations presented as the
seasonally-weighted annual average concentration
(in mg/m3) at 406 sites in the U.S. For more
information, see: https://www.epa.gov/air-trends/
particulate-matter-pm25-trends
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
March 03, 2013). As described in the
2019 ISA, PM10 has become
considerably coarser across the U.S.
compared to similar observations in the
2009 ISA such that, in urban areas, the
mass of the coarse fraction of PM is
similar to or greater than the mass of the
fine fraction of PM (U.S. EPA, 2019,
section 2.5.1.1.4; U.S. EPA, 2009c).
However, in considering recent air
quality data, the EPA notes that in most
areas of the country PM2.5 and PM10
concentrations have declined and are
well below their respective 24-hour
standards. While the contribution of
fine and coarse PM to PM10 mass
concentrations may vary spatially and
temporally, based on the trends in
recent air quality data, the
Administrator concludes that the
current primary 24-hour PM10 standard
is maintaining air quality at level that
provides requisite protection against
PM10–2.5. That is, recent air quality data
does not suggest that PM10–2.5
concentrations have been increasing as
PM2.5 concentrations have been
decreasing. In considering the available
PM10–2.5 health effects evidence in this
reconsideration, there continue to be
significant uncertainties and limitations,
specifically with respect to the exposure
assessment methods used to estimate
PM10–2.5 concentrations, that make it
difficult to fully assess the public health
implications of revising the primary
PM10 standard even considering the
possibility for additional variability in
the relative ratio of PM2.5 to PM10–2.5 in
current PM10 air quality across the U.S.
As described in detail above in section
III.A.2 and in the proposal (85 FR 5558,
January 27, 2023), the uncertainties and
limitations in the health effects
evidence for PM10–2.5 contributed to the
determinations in the 2019 ISA that the
evidence for key PM10–2.5 health effects
is ‘‘suggestive of, but not sufficient to
infer, a causal relationship’’ or
‘‘inadequate to infer the presence, or
absence of a causal relationship’’ (U.S.
EPA, 2019a). While the evidence base
for PM10–2.5-related health effects has
somewhat expanded since the 2009 ISA,
the Administrator concludes that the
evidence remains too limited to inform
judgments regarding whether a more
protective primary PM10 standard is
warranted at this time.
Beyond the uncertainties and
limitations associated with the available
scientific evidence, the EPA also notes
that, while the NCore monitoring
network has been expanded since the
time of the last review, epidemiologic
studies available in this review do not
use PM10–2.5 NCore data in evaluating
associations between PM10–2.5 in
PO 00000
Frm 00098
Fmt 4701
Sfmt 4700
ambient air and long- or short-term
exposures. In the absence of such
evidence, the public health implications
of changes in ambient PM10–2.5
concentrations as PM2.5 concentrations
decrease remain unclear. Therefore, the
EPA continues to recognize this as an
area for future research, to address the
existing uncertainties (U.S. EPA, 2022b,
section 4.6), and inform future reviews
of the PM NAAQS. Taken together, as at
the time of proposal, the Administrator
concludes that these and other
limitations in the PM10–2.5 evidence
raised questions as to whether
additional public health improvements
would be achieved by revising the
existing PM10 standard, particularly
when considering such judgments along
with his decision to retain the current
primary 24-hour PM2.5 standard.
Therefore, the EPA does not agree with
the commenters that the currently
available air quality information or
scientific evidence support revisions to
the primary PM10 standard in this
reconsideration.
Consistent with their comments on
the 2020 proposal, some commenters
disagreed with the Administrator’s
proposed conclusion to retain the
current primary PM10 standard,
primarily focusing their comments on
the need for revisions to the form of the
standard or the level of the standard.
With regard to comments on the form of
the standard, some commenters assert
that the EPA should revise the standard
by adopting a separate form (or a
‘‘compliance threshold’’ in their
words)—the 99th percentile, averaged
over three years—for the primary PM10
standard for continuous monitors,
which provide data every day, while
maintaining the current form of the
standard (one exceedance, averaged
over three years) for 1-in-6 samplers,
given the increased use of continuous
monitoring and to ease the burden of
demonstrating exceptional events.
These commenters, in support of their
comment, contend that the 99th
percentile would effectively change the
form from the 2nd highest to the 4th
highest and would allow no more than
three exceedances per year, averaged
over three years. These commenters
additionally highlight the EPA’s
decision in the 1997 review to adopt a
99th percentile form, averaged over
three years, citing to advantages of a
percentile-based form in the
Administrator’s rationale in that review.
The comments further assert that a 99th
percentile form for the primary PM10
standard is still more conservative than
the form for other short-term NAAQS
(e.g., PM2.5 and NO2).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
First, the EPA has long recognized
that the form is an integral part of the
NAAQS and must be selected together
with the other elements (i.e., indicator,
averaging time, level) of the NAAQS to
ensure the appropriate stringency and
requisite degree of public health
protection. Thus, if the EPA were to
change the form according to the
monitoring method it would be
establishing two different NAAQS,
varying based on the monitoring
method. The EPA has not done this to
date, did not propose such an approach,
and declines to adopt it for the final
rule, as we believe such a decision in
this final rule is beyond the scope of the
proposal, and that each PM standard
should have a single form, indicator,
level and averaging time, chosen by the
Administrator as necessary and
appropriate. While certain continuous
monitors may be established and
approved as a Federal Equivalent
Method (FEM) for PM10, as an
alternative to a Federal Reference
Method (FRM), the use of an FEM is
intended as an alternative means of
determining compliance with the
NAAQS, not as authorizing a different
NAAQS.
Even if the commenters had asked
that the change in form be made without
regard to monitoring method, the EPA
does not believe such a change would
be warranted. The change in form for
continuous monitors suggested by the
commenters, without also lowering the
level of such a standard, would allow
more exceedances and thereby reduce
the public health protection provided
against exposures to PM10–2.5 in ambient
air, resulting in a less stringent primary
PM10 standard than the current
standard. These commenters have not
provided new evidence or analyses to
support their conclusion that an
appropriate degree of public health
protection could be achieved by
allowing the use of an alternative form
(i.e., 99th percentile), while retaining
the other elements of the standard.
With regard to the commenters’
assertion that an alternate form of the
standard would ease the burden of
demonstrating exceptional events, the
EPA recognizes, consistent with the
CAA, that it may be appropriate to
exclude monitoring data influenced by
‘‘exceptional’’ events when making
certain regulatory determinations.
However, the EPA notes that the cost of
implementation of the standards may
not be considered by the EPA in
reviewing the standards. The EPA
continues to update and develop
documentation and tools to facilitate the
implementation of the 2016 Exceptional
Events Rule, including new PM2.5
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
implementation focused products under
development that are intended to assist
air agencies with the development of
demonstrations for specific types of
exceptional events. With regard to the
commenters’ specific concerns for
wildfires or high winds, the EPA
released updated guidance documents
on the preparation of exceptional event
demonstrations related to wildfires in
September 2016, high wind dust events
in April 2019, and prescribed fires in
August 2019. These guidance
documents outline the regulatory
requirements and provide examples for
air agencies preparing demonstrations
for wildfires, high wind dust, and
prescribed fire events. For all of the
reasons discussed above, the EPA does
not agree with the commenters that the
form of the primary PM10 standard
should be revised to a 99th percentile
for continuous monitors.
4. Administrator’s Conclusions
This section summarizes the
Administrator’s considerations and
conclusions related to the current
primary PM10 standard. In establishing
primary standards under the Act that
are ‘‘requisite’’ to protect the public
health with an adequate margin of
safety, the Administrator is seeking to
establish standards that are neither more
nor less stringent than necessary for this
purpose. In so doing, the Administrator
notes that his final decision in this
reconsideration is a public health policy
judgment that draws upon scientific
information, as well as judgments about
how to consider the range and
magnitude of uncertainties that are
inherent in the information.
Accordingly, he recognizes that his
decision requires judgments based on
the interpretation of the evidence that
neither overstates nor understates the
strength or limitations of the evidence
nor the appropriate inferences to be
drawn. He recognizes, as described in
section I.A above, that the Act does not
require that primary standards be set at
a zero-risk level; rather, the NAAQS
must be sufficient but not more
stringent than necessary to protect
public health, including the health of
sensitive groups with an adequate
margin of safety.
Given these requirements, and
consistent with the primary PM2.5
standards discussed above (section
II.C.3), the Administrator’s final
decision in this reconsideration of the
current primary PM10 standard will be
a public health policy judgment that
draws upon the scientific information
examining the health effects of PM10–2.5
exposures, including how to consider
the range and magnitude of
PO 00000
Frm 00099
Fmt 4701
Sfmt 4700
16299
uncertainties inherent in that
information. The Administrator’s final
decision is based on an interpretation of
the scientific evidence that neither
overstates nor understates its strengths
and limitations, nor the appropriate
inferences to be drawn.
Having carefully considered advice
from the CASAC and public comments,
as discussed above, the Administrator
notes that the fundamental scientific
conclusions on health effects of PM10–2.5
in ambient air that were reached in the
2019 ISA and summarized in the 2020
PA and 2022 PA remain valid.
Additionally, the Administrator believes
the judgments he proposed (85 FR 5558,
January 27, 2023) with regard to the
evidence remain appropriate. Further,
in considering the adequacy of the
current primary PM10 standard in this
reconsideration, the Administrator has
carefully considered the policy-relevant
evidence and conclusions contained in
the 2019 ISA; the rationale and
conclusions presented in the 2020 PA
and 2022 PA; the advice and
recommendations from the CASAC in
their reviews of the 2019 draft PA and
2021 draft PA; and public comments, as
addressed in section III.B.3 above and in
the RTC document. In the discussion
below, the Administrator gives weight
to the conclusions in the 2020 PA and
2022 PA, with which the CASAC has
concurred, as summarized in section
III.C of the proposal and takes note of
the key aspects of the rationale for those
conclusions that contribute to his
decision in this review. In considering
this information, the Administrator
concludes that the preliminary
conclusions and policy judgments
supporting his proposed decision
remain valid, and that the current
primary PM10 standard provides
requisite protection of public health
with an adequate margin of safety and
should be retained. In considering the
2020 PA and 2022 PA evaluations and
conclusions, the Administrator notes
that, while the health effects evidence is
somewhat expanded since the 2009 ISA
as described in section III.A.2 above, the
overall conclusions are generally
consistent with those reached in the
2009 ISA (U.S. EPA, 2020b, section 4.4).
In so doing, he additionally notes that
the CASAC supported the preliminary
conclusion in the 2019 draft PA and
2021 draft PA that the evidence
reviewed in the 2019 ISA does not call
into question the public health
protection provided by the current
primary PM10 standard against PM10–2.5
exposures and concurs that it is
appropriate to consider retaining the
current primary PM10 standard (Cox,
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16300
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
2019b, p. 13 of consensus responses;
Sheppard, 2022a, p. 4 of consensus
letter).
As noted below, the scientific
evidence for PM10–2.5-related health
effects has expanded somewhat since
the 2012 review, in particular for longterm exposures. The Administrator
recognizes, however, that there are a
number of uncertainties and limitations
associated with the available
information, as described in the
proposal (85 FR 5558, January 27, 2023)
and below. With regard to the current
evidence on PM10–2.5-related health
effects, the Administrator takes note of
recent epidemiologic studies that
continue to report positive associations
with mortality and morbidity in cities
across North America, Europe, and Asia,
where PM10–2.5 sources and composition
are expected to vary widely. While
significant uncertainties remain, as
described below, the Administrator
recognizes that this expanded body of
evidence has broadened the range of
effects that have been linked with
PM10–2.5 exposures. These studies
provide an important part of the
scientific foundation supporting the
2019 ISA’s revised causality
determinations (and new
determinations) for long-term PM10–2.5
exposures and mortality, cardiovascular
effects, metabolic effects, nervous
system effects, and cancer (U.S. EPA,
2019a; U.S. EPA, 2022b, section 4.2).
Drawing from his consideration of this
evidence, the Administrator concludes
that the available scientific information
supports a decision to maintain a
primary PM10 standard to provide
public health protection against PM10–2.5
exposures, regardless of location, source
of origin, or particle composition. With
regard to uncertainties in the evidence,
the Administrator first notes that a
number of limitations were identified in
the 2012 review related to: (1) Estimates
of ambient PM10–2.5 concentrations used
in epidemiologic studies; (2) limited
evaluation of copollutant models to
address the potential for confounding;
and (3) limited experimental studies
supporting biological plausibility for
PM10–2.5-related effects. Despite the
expanded body of evidence for PM10–2.5
exposures and health effects assessed in
the 2019 ISA, the Administrator
recognizes that uncertainties remain,
similar to those in the 2012 review. As
summarized in section III.A.2 above and
in responding to public comments,
uncertainties in the available scientific
evidence continue to include those
associated with the exposure estimates
used in epidemiologic studies, the
independence of the PM10–2.5 health
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
effect associations, and the biologically
plausible pathways for PM10–2.5 health
effects (U.S. EPA, 2022b, section 4.3).
These uncertainties contribute to the
2019 ISA determinations that the
evidence is ‘‘suggestive of, but not
sufficient to infer’’ causal relationships
(U.S. EPA, 2019a). The Administrator
recognizes that the NAAQS must allow
for a margin of safety but also places
emphasis on evidence supporting
‘‘causal’’ or ‘‘likely to be causal’’
relationships (as described in sections
II.A.2 and III.A.2 above). Finding that
there is too much uncertainty that a
more stringent standard would improve
public health, the Administrator judges
that the available evidence provides
support for his conclusion that the
current standard provides the requisite
level of protection from the effects of
PM10–2.5. In making this judgment, the
Administrator considers whether this
level of protection is more than what is
requisite and whether a less stringent
standard would be appropriate to
consider. He notes that there continues
to be uncertainty associated with the
evidence, as reflected by the ‘‘suggestive
of, but not sufficient to infer’’ causal
determinations. The Administrator
recognizes that the CAA requirement
that primary standards provide an
adequate margin of safety, as
summarized in section I.A above, is
intended to address uncertainties
associated with inconclusive scientific
evidence and technical information, as
well as to provide a reasonable degree
of protection against hazards that
research has not yet identified. In light
of these considerations and the current
body of evidence, including
uncertainties and limitations, the
Administrator concludes that a less
stringent standard would not provide
the requisite protection of public health,
including an adequate margin of safety.
The Administrator also considers
whether the level of protection
associated with the current standard is
less than what is requisite and whether
a more stringent standard would be
appropriate to consider. In so doing, the
Administrator considers, as discussed
above, the level of protection offered
from exposures for which public health
implications are less clear. In so doing,
he again notes the significant
uncertainties and limitations that persist
in the scientific evidence. In particular,
he notes limitations in the approaches
used to estimate ambient PM10–2.5
concentrations in epidemiologic studies,
limited examination of the potential for
confounding by co-occurring pollutants,
and limited support for the biological
plausibility of the serious effects
PO 00000
Frm 00100
Fmt 4701
Sfmt 4700
reported in many epidemiologic studies
that are reflected by the ‘‘suggestive of,
but not sufficient to infer’’ causal
determinations. Thus, in light of the
currently available information,
including the uncertainties and
limitations of the evidence base
available to inform his judgments
regarding protection against PM10–2.5related effects, the Administrator does
not find it appropriate to increase the
stringency of the standard in order to
provide the requisite public health
protection. Rather, he judges it
appropriate to maintain the level of
protection provided by the current
primary PM10 standard for PM10–2.5
exposures and he does not judge that
the available information and the
associated uncertainties indicate the
need for a greater level of public health
protection.
In reaching his conclusions on the
primary PM10 standard, the
Administrator also considers advice
from the CASAC. In their comments, the
CASAC noted that uncertainties that
were identified in the 2012 review
persist in the evidence for PM10–2.5related health effects (Cox, 2019b, p. 13
of consensus responses; Sheppard,
2022a, p. 4 of consensus letter) In
considering these comments, the
Administrator takes note of the CASAC
consideration of the evidence, and
associated uncertainties, and its
conclusion that the evidence reviewed
in the 2019 ISA does not call into
question the adequacy of the public
health protection afforded by the
current primary PM10 standard (Cox,
2019b, p. 3 of letter; Sheppard, 2022a,
p. 4 of consensus letter). The
Administrator further notes the
unanimous conclusions of the CASAC
that evidence supports consideration of
retaining the current primary PM10
standard (Cox, 2019b, p. 3 of consensus
letter; Sheppard, 2022a, p. 4 of
consensus letter). In addition to the
CASAC’s advice, the Administrator also
considers public comments, the
majority of which supported retaining
the primary PM10 standard, citing to and
agreeing with the Administrator’s
rationale for his proposed decision. The
Administrator also recognizes that a few
public commenters supported revising
the primary PM10 standard in order to
provide increased protection against
PM10–2.5-related health effects.
The Administrator also notes that the
scientific record for his decision on the
primary PM10 standard is the same as
the record before the then-Administrator
in 2020, as the scope of the ISA
Supplement focused on health effect
categories where the 2019 ISA
concluded a causal relationship (i.e.,
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
short- and long-term PM2.5 exposure and
cardiovascular effects and mortality).
Therefore, because no health outcome
categories for short- or long-term
PM10–2.5 exposure in the 2019 ISA were
greater than ‘‘suggestive of, but not
sufficient to infer, a causal
relationship’’, the ISA Supplement did
not evaluate studies published after the
literature cutoff date of the 2019 ISA
related to PM10–2.5 exposures and health
effects. The Administrator further notes
his decision is consistent with the
decision of the prior Administrator in
2020 to retain the primary PM10
standard.
With regard to the indicator for the
primary PM10 standard, the
Administrator recognizes that the 2022
PA notes that the evidence continues to
support retaining the PM10 indicator to
provide public health protection against
PM10–2.5-related effects. He notes that,
consistent with the approaches in
previous reviews, a standard with a
PM10 mass-based indicator, in
conjunction with a PM2.5 mass-based
standard, will result in controlling
allowable concentrations of PM10–2.5.
The Administrator also takes note of the
2019 ISA comparison that showed that
the relative contribution of PM2.5 and
PM10–2.5 to PM10 concentrations can vary
across the U.S. by region and season,
with urban locations having a somewhat
higher contribution of PM2.5
contributing to PM10 concentrations
than PM10–2.5 (U.S. EPA, 2019a, section
2.5.1.1.4, Table 2–7). In these urban
locations, where PM2.5 concentrations
are somewhat higher than in rural
locations, the toxicity of the PM10 may
be higher due to contaminating PM2.5.
Further, although uncertainties with the
evidence persist, the strongest health
effects evidence associated with PM10–2.5
comes from epidemiologic studies
conducted in urban areas. He also notes
that the CASAC agreed with the EPA’s
conclusions that a PM10 indicator
remained appropriate (Cox, 2019b, p. 13
of consensus responses; Sheppard,
2022a, p. 4 of letter). In light of this
information, the Administrator
concludes that the PM10 indicator
remains appropriate and provides
protection from exposure to all coarse
PM, regardless of location, source of
origin, or particle composition.
Similarly, with regard to averaging
time, form, and level of the standard,
the Administrator takes note of
uncertainties in the available evidence
and information and continues to find
that the current standard, as defined by
in all of its elements, is requisite. As an
initial matter, the Administrator notes
that the current primary PM10 standard,
with its level of 150 mg/m3, 24-hour
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
averaging time, not to be exceeded more
than once per year on average over three
years, is intended to protect against
short-term peak PM10–2.5 exposures. In
so doing, while the Administrator notes
that changes in PM2.5 concentrations in
ambient air can influence the
contribution of the fine and coarse
fractions to PM10 mass, such that
reductions in PM2.5 concentrations can
lead to more allowable PM10–2.5 under
the current primary PM10 standard, he
recognizes that there is no new
information available in this
reconsideration to suggest that the
public health protection provided by the
current standard is not requisite or that
a more stringent standard is warranted
at this time. The Administrator
concludes that, particularly in light of
his decision to retain the primary 24hour PM2.5 standard with its level of 35
mg/m3 as described in section II.B.4
above, the primary PM10 standard
would be expected to maintain PM10–2.5
concentrations in ambient air below
those that have been considered to be
associated with serious health effects in
past NAAQS reviews. The
Administrator also notes that while the
scientific evidence available in the 2019
ISA has expanded since the completion
of the 2009 ISA, he concludes that this
information does not provide support
for the causal or likely to be causal
relationships upon which he places the
greatest weight in considering the
adequacy of the current standards. He
further concludes that the uncertainties
and limitations of the scientific
evidence, along with the absence of
information to inform a quantitative
exposure or risk assessment, make it
difficult to reach decisions regarding
whether a more protective standard is
warranted at this time. He has
additionally considered the public
comments regarding revisions to these
elements of the standard and continues
to judge that the existing level and the
existing form, in all its aspects, together
with the other elements of the existing
standard provide an appropriate level of
public health protection. For all of the
reasons discussed above and
recognizing the CASAC’s conclusion
that the current evidence provides
support for retaining the current
standard, the Administrator concludes
that the current primary PM10 standard
(in all of its elements) is requisite to
protect public health with an adequate
margin of safety from effects of PM10–2.5
in ambient air and should be retained
without revision.
PO 00000
Frm 00101
Fmt 4701
Sfmt 4700
16301
C. Decision on the Primary PM10
Standard
For the reasons discussed above and
considering information and
assessments presented in the 2019 ISA
and the 2022 PA, the advice from the
CASAC, and public comments, the
Administrator concludes that the
current primary PM10 standard is
requisite to protect public health with
an adequate margin of safety, including
the health of at-risk populations, and is
retaining the current standard without
revision.
IV. Communication of Public Health
A. Air Quality Index Overview
Information about the public health
implications of ambient concentrations
of criteria pollutants is communicated
to the public using the Air Quality
Index (AQI) reported on the EPA’s
AirNow website.133 The current AQI has
been in use since its inception in
1999.134 It provides useful, timely, and
easily understandable information about
the daily degree of pollution. The goal
of the AQI is to establish a nationally
uniform system of indexing pollution
concentrations for ozone, carbon
monoxide, nitrogen dioxide, PM, and
sulfur dioxide. The AQI is recognized
internationally as a proven tool to
effectively communicate air quality
information to the public as
demonstrated by the fact that many
countries have created similar indices
based on the AQI.
The AQI converts an individual
pollutant concentration in a
community’s air to a number on a scale
from 0 to 500. Reported AQI values for
specific pollutants enable the public to
know whether air pollution levels in a
particular location are characterized as
good (0–50), moderate (51–100),
unhealthy for sensitive groups (101–
150), unhealthy (151–200), very
unhealthy (201–300), or hazardous
(301+). Across criteria pollutants, the
AQI value of 100 typically corresponds
to the level of the short-term (e.g., 24hour, 8-hour, or 1-hour standard)
NAAQS for each pollutant. Below an
index value of 100, an intermediate
value of 50 is defined either as the level
of the annual standard if an annual
standard has been established (e.g.,
PM2.5, nitrogen dioxide), a
133 See
https://www.airnow.gov/.
1976, the EPA established a nationally
uniform air quality index, then called the Pollutant
Standard Index (PSI), for use by State and local
agencies on a voluntary basis (41 FR 37660,
September 7, 1976; 52 FR 24634, July 1,1987). In
August 1999, the EPA adopted revisions to this air
quality index (64 FR 42530, August 4, 1999) and
renamed the index the AQI.
134 In
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16302
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
concentration equal to one-half the
value of the 24-hour standard used to
define an index value of 100 (e.g.,
carbon monoxide), or a concentration
based directly on health effects evidence
(e.g., ozone). An AQI value greater than
100 means that a pollutant is in one of
the unhealthy categories (i.e., unhealthy
for sensitive groups, unhealthy, very
unhealthy, or hazardous). An AQI value
at or below 100 means that a pollutant
concentration is in one of the
satisfactory categories (i.e., moderate or
good). The scientific evidence on
pollutant-related health effects for each
NAAQS review support decisions
related to pollutant concentrations at
which to set the various AQI
breakpoints, which delineate the AQI
categories for each individual pollutant
(i.e., the pollutant concentrations
corresponding to index values of 150,
200, 300, and 500). The AQI is reported
three ways by the EPA and State, local
and Tribal agencies, all of which are
useful and complementary. The daily
AQI is reported for the previous day and
used to observe trends in community air
quality, the AQI forecast helps people
plan their outdoor activities for the next
day, and the near-real-time AQI, or
NowCast AQI, tells people whether it is
a good time for outdoor activity.
Historically, State and local agencies
have primarily used the AQI to provide
general information to the public about
air quality and its relationship to public
health. For more than two decades,
many State and local agencies, as well
as the EPA and other Federal agencies,
have been developing new and
innovative programs and initiatives to
provide more information related to air
quality and health messaging to the
public in a more timely way. These
initiatives, including air quality
forecasting, near real-time data reporting
through the AirNow website, use of data
from air quality sensors on the EPA and
U.S. Forest Service’s (USFS) Fire and
Smoke Map, and air quality action day
programs, provide useful, up-to-date,
and timely information to the public
about air pollution and its health effects.
Such information can help the public
learn when their well-being may be
compromised, so they can take actions
to avoid or to reduce exposures to
ambient air pollution at concentrations
of concern. This information can also
encourage the public to take actions that
will reduce air pollution on days when
concentrations are projected to be of
concern to local communities (e.g., air
quality action day programs can
encourage individuals to drive less or
carpool).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
B. Air Quality Index Category
Breakpoints for PM2.5
Recognizing the scientific information
available and current AQI reporting
practices, the EPA proposed several
revisions to the AQI PM2.5 breakpoints.
EPA solicited and received comments
on these proposed revisions. Upon
reviewing the information in the
proposal and considering the comments
received EPA is making final revisions
to the AQI category breakpoints for
PM2.5. This section summarizes the
proposed revisions, which can be read
in full in the proposal (88 FR 5638,
January 27, 2023), significant comments,
and final revisions.
1. Summary of Proposed Revisions
One purpose of the AQI is to
communicate to the public when air
quality is poor and thus when they
should consider taking actions to reduce
their exposures. The higher the AQI
value, the higher the level of air
pollution and the greater the health
concern. In recognition of the scientific
information available that is informing
the reconsideration of the 2020 final
decision on the primary PM2.5
standards, including a number of new
controlled human exposure and
epidemiologic studies published since
the completion of the 2009 ISA, as well
as additional epidemiologic studies
from other peer reviewed documents
that evaluate the health effects of
wildfire smoke exposure and that can
inform the selection of AQI breakpoints
at higher PM2.5 concentrations,135 the
EPA proposed to make two sets of
changes to the PM2.5 sub-index of the
AQI. First, the EPA proposed to
continue to use the approach used in
the revisions to the AQI in 2012 (77 FR
38890, June 29, 2012) of setting the
lower breakpoints (50, 100 and 150) to
be based on the levels of the primary
PM2.5 annual and 24-hour standards and
proposed to revise the lower
breakpoints to be consistent with
changes to the primary PM2.5 standards
that are part of this reconsideration.
Second, the EPA proposed to revise the
135 In evaluating the scientific evidence available
to inform decisions regarding the AQI breakpoints,
the EPA considered studies that were included as
a part of the 2019 ISA and ISA Supplement, but
also considered other studies that were not
included as a part of the review of the air quality
criteria. The ISAs have specific criteria for study
inclusion and consideration in reaching
conclusions regarding causal relationships, and
some studies that may not have met those criteria
(e.g., epidemiologic studies that evaluate the health
effects of wildfire smoke exposure that would have
higher PM2.5 concentrations, which are outside of
the scope of the ISA) were identified as studies that
could be used to inform decisions on the AQI,
particularly for the upper breakpoints.
PO 00000
Frm 00102
Fmt 4701
Sfmt 4700
upper AQI breakpoints (200 and above)
and to replace the linear-relationship
approach used in 1999 to set these
breakpoints, with an approach that more
fully considers the PM2.5 health effects
evidence from controlled human
exposure and epidemiologic studies that
have become available in the last 20
years (64 FR 42530, August 4, 1999).
a. Air Quality Index Values of 50, 100
and 150
With respect to the lower AQI
breakpoints in the proposal (88 FR 5638,
January 27, 2023), the EPA proposed to
conclude that it is appropriate to
continue setting these breakpoints to be
consistent with the primary annual and
24-hour PM2.5 standard levels. The
lowest AQI value of 50 provides the
breakpoint between the ‘‘good’’ and
‘‘moderate’’ categories. At and below
this concentration, air quality is
considered ‘‘good’’ for everyone. Above
this concentration, in the ‘‘moderate’’
category, the AQI contains advisories for
unusually sensitive individuals. The
EPA has historically set this breakpoint
at the level of the primary annual PM2.5
standard. In doing so, the EPA has
recognized that: (1) The annual standard
is set to provide protection to the
public, including at-risk populations,
from PM2.5 concentrations, which, when
experienced on average for a year, have
the potential to result in adverse health
effects; and (2) the AQI exposure period
represents a shorter exposure period
(e.g., 24-hour (or less)) while focusing
on the most sensitive individuals. The
EPA saw no basis for deviating from this
approach in this reconsideration. Thus,
the EPA proposed to set the AQI value
of 50 at a daily (i.e., 24-hour) average
concentration equal to the level of the
primary annual PM2.5 standard that is
promulgated.
The historical approach to setting an
AQI value of 100, which is the
breakpoint between the ‘‘moderate’’ and
‘‘unhealthy for sensitive groups’’
categories, and above which advisories
are generated for sensitive groups, is to
set it at the same level as the primary
24-hour PM2.5 standard. In so doing, the
EPA has recognized that the primary 24hour PM2.5 standard is set to provide
protection to the public, including atrisk populations, from short-term
exposures to PM2.5 concentrations that
have the potential to result in adverse
health effects. Given this, it is
appropriate to generate advisories for
sensitive groups at concentrations above
this level. In the past, State, local, and
Tribal air quality agencies have
expressed strong support for this
approach (78 FR 3086, January 15,
2013). The EPA saw no basis to deviate
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
from this approach in this
reconsideration. In the proposal (88 FR
5638, January 27, 2023), the EPA
proposed to retain the current primary
24-hour PM2.5 standard with its level of
35 mg/m3 but took comment on revising
the level of that standard to 25 mg/m3
(section II.D.3.b). Thus, the EPA
proposed to retain the AQI value of 100
set at the level of the current primary
24-hour PM2.5 standard concentration of
35 mg/m3 (i.e., 24-hour average).
With respect to an AQI value of 150,
which is the breakpoint between the
‘‘unhealthy for sensitive groups’’ and
‘‘unhealthy categories,’’ this breakpoint
concentration in this reconsideration is
based upon the considering the same
health effects information, as assessed
in the 2019 ISA and ISA Supplement
and described in section II above, that
informs the proposed decisions on the
level of the 24-hour standard and the
AQI value of 100. Previously, the
Agency has used a proportional
adjustment in which the AQI value of
150 was set proportionally to the AQI
value of 100. This proportional
adjustment inherently recognizes that
the available epidemiologic studies
provide no evidence of discernible
thresholds, below which effects do not
occur in either sensitive groups or in the
general population, that could inform
conclusions regarding concentrations at
which to set this breakpoint. Given that
the epidemiologic evidence continues to
be the most relevant health effects
evidence for informing this range of AQI
values, the EPA saw no basis to deviate
from this approach in this
reconsideration. Therefore, the EPA
proposed to set an AQI value of 150
proportionally, depending on the
breakpoint concentration of the AQI
value of 100 (i.e., 55.4 for a 24-hour
standard of 35 mg/m3).
b. Air Quality Index Values of 200 and
Above
In the proposal (88 FR 5639, January
27, 2023), the EPA summarized the
history of setting the AQI values of 300
and above in the 1999 rule (64 FR
42530, August 4, 1999) and established
breakpoints for PM2.5 in that range. In
general, the AQI values between 100
and 500 were based on PM2.5
concentrations that generally reflected a
linear relationship between increasing
index values and increasing PM2.5
concentrations.136 It was found that this
linear relationship was generally
consistent with the health effects
136 The AQI breakpoint at 150 was originally set
in 1999 to be linearly related to the concentrations
at the 100 and 500 breakpoints but then revised in
2012 to be proportional to the AQI breakpoint
concentration at 100 (78 FR 3181, January 15, 2013).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
evidence, which suggested that as PM2.5
concentrations increase, increasingly
larger numbers of people are likely to
experience serious health effects in this
range of PM2.5 concentrations (64 FR
42536, August 4, 1999). For the AQI
breakpoint of 500, the concentration
was based on the method used to
establish a previously existing PM10
breakpoint that was informed by studies
conducted in London using the British
Smoke method, which uses a different
particle size cutpoint as noted in the
proposal (88 FR 5639, January 27, 2023).
Due to limited ambient PM2.5
monitoring data available at that time,
the decision on the 500 value
concentration for PM2.5 was based on
the stated assumption that PM
concentrations measured by the British
Smoke method were approximately
equivalent to PM2.5 concentrations (64
FR 42530, August 4, 1999). Given that
the British Smoke method has a larger
particle size cutpoint than the current
PM2.5 monitoring method, which has a
cutpoint of 2.5 microns, a concentration
of 500 mg/m3 based on the British
Smoke method would be equivalent to
a lower PM2.5 concentration. With
respect to the upper breakpoints of the
AQI, the EPA has historically been
concerned about establishing these
upper breakpoints using evidence based
on larger size fractions of PM, given that
PM2.5 is the indicator for the AQI. While
monitoring data for higher PM2.5
concentrations in ambient air has been
available for many years, the health
effects evidence has only recently
become available for consideration in
informing decisions on the upper
breakpoints of the AQI.
As part of this reconsideration, the
EPA recognized that the health effects
evidence associated with PM2.5
exposure has greatly expanded in recent
years. Multiple controlled human
exposure studies have become available
that provide information about health
effects across a range of concentrations.
While many of the new studies
evaluated in the 2019 ISA focused on
examining health effects associated with
exposure to lower PM2.5 concentrations,
there are also several new controlled
human exposure studies that provide
information about the health effects
observed in study participants at
concentrations well above the standard
levels. Additionally, there are also
epidemiologic studies now available
and evaluated in other Agency peerreviewed documents that can inform
health effects associated with higher
PM2.5 concentrations (U.S. EPA,
PO 00000
Frm 00103
Fmt 4701
Sfmt 4700
16303
2021b).137 Thus, the EPA concluded
that it is appropriate to reevaluate the
upper AQI breakpoints, taking into
account the expanded body of scientific
evidence, particularly given several new
epidemiologic studies conducted during
high pollution events like wildfires and
multiple controlled human exposure
studies. While it remains unclear the
exact PM2.5 concentrations at which
specific health effects occur, the more
recent studies do provide more refined
information about the concentration
range in which these effects might occur
in some populations. These studies
provide support for coherence of effects
across scientific disciplines and
potentially biologically plausible
pathways for the overt population-level
health effects observed in epidemiologic
studies. Therefore, taking into account
the short exposure time period in these
studies (e.g., 1–6 hours) and that the
studies generally do not include at-risk
(or sensitive) populations, but rather
young, healthy adults, these studies, in
conjunction with information from
epidemiologic studies, the EPA
preliminarily concluded it would be
appropriate to be more cautionary and
offer advisories to the public for
reducing exposures at lower
concentrations than recommended with
the current AQI breakpoints.
The AQI value of 200 is the
breakpoint between the ‘‘unhealthy’’
and ‘‘very unhealthy’’ categories. At
AQI values above 200, the AQI would
be providing a health warning that the
risk of anyone experiencing a health
effect following short-term exposures to
these PM2.5 concentrations has
increased. To inform proposed
decisions on this breakpoint, the EPA
takes note of studies indicating the
potential for respiratory or
cardiovascular effects that are on their
own representative of or are on the
biologically plausible pathway to more
serious health outcomes (e.g.,
emergency department visits, hospital
admissions). The controlled human
exposure studies evaluated in the 2009
and 2019 ISAs provide evidence of
inflammation as well as cardiovascular
effects in healthy subjects at and above
120 mg/m3. For example, Ramanathan et
al. (2016) observed a transient reduction
in antioxidant/anti-inflammatory
function after exposing healthy young
subjects to a mean concentration of 150
mg/m3 of PM2.5 for 2 hours. Urch et al.
137 In this reconsideration, the controlled human
exposure studies were evaluated in the 2019 ISA,
whereas the epidemiologic studies of wildfire
smoke exposures were included in the EPA
Comparative Assessment of the Impacts of
Prescribed Fire Versus Wildfire (CAIF): A Case
Study in the Western U.S. (U.S. EPA 2021b).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16304
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
(2010) also reported increased markers
of inflammation when exposing both
asthmatic and non-asthmatic subjects to
a mean concentration of 140 mg/m3 of
PM2.5 for 3 hours. In studies specifically
examining cardiovascular effects, Ghio
et al. (2000) and Ghio et al. (2003)
exposed healthy subjects to a mean
concentration of 120 mg/m3 for 2 hours
and reported significantly increased
levels of fibrinogen, a marker of
coagulation that increases during
inflammation. Sivagangabalan et al.
(2011) exposed healthy subjects to a
mean concentration of 150 mg/m3 of
PM2.5 for 2 hours and noted an
increased QT interval (3.4 ± 1.4)
indicating some evidence for
conduction abnormalities, an indicator
of possible arrhythmias. Lastly, Brook et
al. (2009) reported a transient increase
of 2.9 mm Hg in diastolic blood pressure
in healthy subjects during the 2-hour
exposure to a mean concentration of 148
mg/m3 of PM2.5.
In addition to epidemiologic studies
evaluated in the 2019 ISA that analyzed
exposures at ambient PM2.5
concentrations, there are a number of
recent epidemiologic studies focusing
on wildfire smoke that have become
available that were evaluated in the
EPA’s recently released peer-reviewed
assessment on wildland fire (U.S. EPA,
2021b). One of these studies,
Hutchinson et al. (2018), conducted a
bidirectional case-crossover analysis to
examine associations between wildfirespecific PM2.5 exposure and respiratoryrelated healthcare encounters (i.e., ED
visits, inpatient hospital admissions,
and outpatient visits) prior and during
the 2007 San Diego wildfires. This study
found positive and significant
associations to PM2.5 exposures and
respiratory-related healthcare
encounters. Further, during the initial 5day period of the wildfire event, the
study observed that there was evidence
of increases in a number of respiratoryrelated outcomes particularly ED visits
for asthma, upper respiratory infection,
respiratory symptoms, acute bronchitis,
and all respiratory-related visits
(Hutchinson et al., 2018). When
examining the air quality during the
wildfire event, PM2.5 concentrations
were highest during the initial five days
of the wildfire, with 24-hour average
PM2.5 concentrations of 89.1 mg/m3
across all zip codes and with the highest
24-hour average of 160 mg/m3 on the
first day (Hutchinson et al., 2018).
When considering this collective body
of evidence from controlled human
exposure and epidemiologic studies, the
Agency proposed to set an AQI value of
200 at a daily (i.e., 24-hour average)
concentration of PM2.5 of 125 mg/m3. As
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
discussed above and in the proposal (88
FR 5640, January 27, 2023), this
concentration is at the lower end of the
concentrations consistently shown to be
associated with respiratory and
cardiovascular effects in controlled
human exposure studies following
short-term exposures (e.g., 2–3 hours)
and in young, healthy adults (Ghio et
al., 2000; Ghio et al., 2003; Urch et al.,
2010; Ramanathan et al., 2016;
Sivagangabalan et al., 2011; and Brook
et al., 2009) and also within the range
of 5-day average and maximum
concentrations observed to be associated
with respiratory-related outcomes
following exposure to wildfire smoke
(Hutchinson et al., 2018).
The AQI value of 300 denotes the
breakpoint between the ‘‘very
unhealthy’’ and ‘‘hazardous’’ categories,
and thus marks the beginning of the
‘‘hazardous’’ AQI category. At AQI
values above 300, the AQI provides a
health warning that everyone is likely to
experience effects following short-term
exposures to these PM2.5 concentrations.
To inform decisions on this AQI
breakpoint, the EPA takes note of
controlled human exposure studies that
consistently show subclinical effects
which are often associated with more
severe cardiovascular outcomes. As
discussed above, Brook et al. (2009)
reported a transient increase of 2.9 mm
Hg in diastolic blood pressure in
healthy subjects during the 2-hour
exposure to a mean concentration of 148
mg/m3 of PM2.5. Bellavia et al. (2013)
exposed healthy subjects to an average
PM2.5 concentration of 242 mg/m3 for 2
hours and reported increased systolic
blood pressure (2.53 mm Hg). Tong et al.
(2015) exposed healthy subjects to an
average PM2.5 concentration of 253 mg/
m3 for 2 hours and observed a
significant increase in diastolic blood
pressure (2.1 mm Hg) and a
nonsignificant increase in systolic blood
pressure (2.5 mm Hg). Lucking et al.
(2011) reported impaired vascular
function and increased potential for
coagulation when exposing healthy
subjects to diesel exhaust (DE) with an
average PM2.5 concentration of 320 mg/
m3 for a duration of 1 hour.138 These
studies all provided evidence of
impaired vascular function, including
vasodilatation impairment and
increased thrombus formation, with
Tong et al. (2015), Bellavia et al. (2013),
Brook et al. (2009) all reporting
138 Although participants in Lucking et al. (2011)
were exposed to diesel exhaust (DE), the authors
also conducted analyses using a particle trap, and
as noted in the 2019 ISA, this type of study design
allows for the assessment of the role of PM2.5 on the
health effects observed by removing PM from the
DE mixture.
PO 00000
Frm 00104
Fmt 4701
Sfmt 4700
increases in blood pressure.
Additionally, Behbod et al. (2013)
reported increased inflammatory
markers following a 2-hour exposure to
an average PM2.5 concentration of 250
mg/m3 in healthy subjects.
In addition to the controlled human
exposure studies discussed above, the
epidemiologic study conducted by
DeFlorio-Barker et al. (2019) examined
the relationship between wildfire smoke
and cardiopulmonary hospitalizations
among adults 65 years of age and older
from 2008–2010 in 692 U.S. counties.
The authors reported a 2.22% increase
in all-cause respiratory hospitalizations
on wildfire smoke days for a 10 mg/m3
increase in 24-hour average PM2.5
concentrations (DeFlorio-Barker et al.,
2019). The maximum 24-hour average
concentration in this study on wildfire
smoke days was 212.5 mg/m3 (DeFlorioBarker et al., 2019). In considering this
study, the EPA notes the increased
probability that even healthy adults
experience effects at this maximum
exposure concentration, particularly
given that this maximum concentration
is near the exposure concentrations in
controlled human exposure studies that
consistently reported evidence of
impaired vascular function and several
that reported increases in blood
pressure in healthy adults following 2hour exposures.
Based on the information discussed
above and in the proposal (88 FR 5640,
January 27, 2023), the EPA proposed to
revise the 300 level of the AQI, which
marks the beginning of the ‘‘hazardous’’
AQI category, to a concentration that is
consistent with the PM2.5 concentrations
associated with health effects as
reported in the controlled human
exposure (Brook et al., 2009; Bellavia et
al., 2013; Tong et al., 2015; Behbod et
al., 2013) and epidemiologic studies
(DeFlorio-Barker et al. (2019).
Specifically, the Agency proposed to set
an AQI value of 300 at a daily (i.e., 24hour average) PM2.5 concentration of
225 mg/m3. This concentration falls
between the 2-hour average
concentrations reported in controlled
human exposure studies found to be
consistently associated, in healthy
adults, with impaired vascular function
and/or increases in blood pressure,
which could both be a precursor to more
severe cardiovascular effects following
short-term (1- to 2-hour) exposures, and
the maximum 24-hour average PM2.5
concentrations on wildfire smoke days
reported in the epidemiologic study
conducted by DeFlorio-Barker et al.
(2019).
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
c. Air Quality Index Value of 500
Lastly, the EPA also proposed
revisions to the 500 value of the AQI.
The 500 value of the AQI is within the
‘‘hazardous’’ category but is specified
and used to calculate the slope of the
AQI values in the ‘‘hazardous category’’
above and below AQI values of 500. In
the past, this breakpoint had a very
prominent role in determining the
current upper AQI values given that it
was used as part of the linear
relationship with the concentration at
the AQI value of 100 to determine the
AQI values of 200 and 300 in 1999 (64
FR 42530, August 4, 1999).
As discussed above and in the
proposal (88 FR 5641, January 27, 2023),
the current breakpoint concentration for
the 500 value of the AQI was set in 1999
at a 24-hour average PM2.5 concentration
of 500 mg/m3 and was based on studies
conducted in London using the British
Smoke method, which used a different
particle size cutpoint and likely
overestimated the PM2.5 concentration.
In looking to improve upon that
approach, the EPA considered several
recent controlled human exposure
studies that observe health effects that
are on the biologically plausible
pathway to more severe cardiovascular
outcomes and note that these seem to
follow exposures to high PM2.5
concentrations that are well above those
typically observed in ambient air. More
specifically, in controlled human
exposure studies, Vieira et al. (2016a)
and Vieira et al. (2016b) exposed
healthy subjects and subjects with heart
failure to diesel exhaust (DE) with a
mean PM2.5 concentration of 325 mg/m3
for 21 minutes and reported decreased
stroke volume, and increased arterial
stiffness (an indicator of endothelial
dysfunction) in both healthy and heart
failure subjects.139 Also as summarized
above and discussed in the proposal (88
FR 5641, January 27, 2023), Lucking et
al. (2011) exposed healthy subjects to
DE with a mean PM2.5 concentration of
320 mg/m3 for 1 hour.140 Epidemiologic
studies have linked the types of
cardiovascular effects observed in these
controlled human exposure studies with
the exacerbation of ischemic heart
disease (IHD) and heart failure as well
as myocardial infarction (MI) and
stroke.
In addition to the controlled human
exposure studies discussed in the
proposal (88 FR 5641, January 27, 2023)
and summarized above, recent
epidemiologic studies examining the
relationship between concentrations of
PM2.5 during wildfires and respiratory
health also informed the proposed
decisions on the concentration for the
AQI value of 500. As discussed in the
proposal (88 FR 5641, January 27, 2023)
and summarized earlier in this section,
Hutchinson et al. (2018) reported
increases in a number of respiratoryrelated ED visits for asthma, upper
respiratory infection, respiratory
symptoms, acute bronchitis, and all
combined respiratory-related visits
based on data from Medi-Cal claims for
emergency department presentations,
inpatient hospitalizations, and
outpatient visits during the initial 5-day
period of the 2007 San Diego fire.
During the initial 5-day window, PM2.5
concentrations were found to be at their
highest with the 95th percentile of 24hour average concentrations of 333 mg/
m3.
Although studies of short-term (i.e.,
daily) exposures to wildfire smoke are
more informative in considering
alternative level for the AQI value of
500 since they mirror the 24-hour
exposure timeframe, additional
information from epidemiologic studies
of longer-term exposures (i.e., over
many weeks) during wildfire events can
provide supporting information. As
discussed in the proposal (88 FR 5641,
January 27, 2023) and summarized here,
Orr et al. (2020) conducted a
16305
longitudinal study that reported
exposure to wildfire smoke from a
multi-month fire resulted in reduced
lung function in subsequent years and
concluded that exposure to high PM2.5
concentrations during a multi-week fire
event may lead to health consequences,
such as declines in lung function.
During the 2017 wildfire event (August
1 to September 19, 2017), Orr et al.
(2020) reported that many days during
the multi-month fire had PM2.5
concentrations above 300 mg/m3,
resulting in a daily average PM2.5
concentration of 220.9 mg/m3 with a
maximum PM2.5 concentration of 638
mg/m3.
The controlled human exposure
studies provide biological plausibility
for results of epidemiologic studies that
document increases in respiratoryrelated health care events during the
wildfires. The collective evidence from
controlled human exposure and
epidemiologic studies, which includes
decreases in stroke volume, increased
arterial stiffness, impaired vascular
function and respiratory-related
healthcare encounters provide healthbased evidence that informed the
proposed decisions on the level of the
AQI value of 500. Given the
concentrations observed in these
studies, the Agency proposed to revise
the AQI value of 500 to a level set at a
daily (i.e., 24-hour average) PM2.5
concentration of 325 mg/m3. This
concentration is at or below the lowest
concentrations observed in the
controlled human exposure studies
associated with more severe effects
discussed above and also at the low end
of the daily concentrations observed in
the epidemiologic studies conducted by
Hutchinson et al. (2018) and Orr et al.
(2020).
Table 1 below summarizes the
proposed breakpoints for the PM2.5 subindex.
TABLE 1—PROPOSED BREAKPOINTS FOR PM2.5 SUB-INDEX
ddrumheller on DSK120RN23PROD with RULES3
AQI category
Index values
Good ................................................................................................................................................
Moderate ..........................................................................................................................................
Unhealthy for Sensitive Groups .......................................................................................................
Unhealthy .........................................................................................................................................
Very Unhealthy ................................................................................................................................
139 These effects were attenuated when the DE
was filtered, to reduce PM2.5 concentrations,
indicating the effects were likely associated with
PM2.5 exposure.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
140 When applying a particle trap, PM
2.5
concentrations were reduced, and effects associated
with cardiovascular function including impaired
vascular function, as measured by vasodilatation
PO 00000
Frm 00105
Fmt 4701
Sfmt 4700
0–50
51–100
101–150
151–200
201–300
Proposed
breakpoints
(μg/m3, 24-hour
average)
0.0–(9.0–10.0)
(9.1–10.1)–35.4
35.5–55.4
55.5–125.4
125.5–225.4
and thrombus formation were attenuated indicating
associations with PM2.5.
E:\FR\FM\06MRR3.SGM
06MRR3
16306
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
TABLE 1—PROPOSED BREAKPOINTS FOR PM2.5 SUB-INDEX—Continued
AQI category
Index values
Hazardous 1 .....................................................................................................................................
301+
Proposed
breakpoints
(μg/m3, 24-hour
average)
225.5
1 AQI
values between breakpoints are calculated using equation 1 in appendix G. For AQI values in the hazardous category, AQI values greater than 500 should be calculated using equation 1 and the PM2.5 concentration specified for the AQI value of 500.
ddrumheller on DSK120RN23PROD with RULES3
2. Summary of Significant Comments on
Proposed Revisions
The EPA received many comments on
the proposed changes to the PM2.5 AQI
breakpoints. Many commenters
generally supported all the proposed
revisions to the AQI breakpoints based
on the revisions to the primary annual
and daily PM2.5 standards and recent
scientific evidence discussed in the
proposal (88 FR 5558, January 27, 2023).
However, we received specific
comments on proposed revisions to the
breakpoints in the lower end of the AQI,
related to their linkage to the annual
and daily PM2.5 standards, and proposed
revisions to the breakpoints at the upper
end of the AQI, based on EPA’s
interpretation of available health effects
evidence.
a. Air Quality Index Values of 50, 100,
and 150
Some commenters agreed with using
the historical approach of setting the 50,
100 and 150 breakpoints of the AQI to
be consistent with the primary PM2.5
standards. Some cited the reason that
this approach creates consistent
communication with respect to air
quality and the standards, and this is
how the other AQI sub-indices are set.
A few commenters disagreed with the
historical approach and suggested
instead that the 50 breakpoint of the
AQI should not be revised at all, or that
the 50 and 100 breakpoints of the AQI
should be supported directly by health
data similar to the basis for the
proposed 200, 300 and 500 breakpoints.
The few commenters that disagreed
with the historical approach of the 50
breakpoint of the AQI noted that setting
a short-term breakpoint to annual
standard was not logical since it is a
long-term standard and not meant to be
interpreted for short-term messaging
with the AQI, in particular when
reported hourly via the NowCast. These
commenters also noted that additional
studies are needed to identify the health
impacts of short-term exposures at low
concentrations. They also noted that
lowering the 50 breakpoint of the AQI
in conjunction with the annual standard
may cause confusion with the public
because some State programs and policy
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
decisions are connected to the AQI
while others are based on PM
concentrations, which could lead to
inconsistent messaging reducing the
public’s trust. These comments were
supported by noting that revised
breakpoints could lead to more
moderate days than in the past, but the
monitor values would be the same as
before when the commenters considered
it ‘‘healthy,’’ possibly eroding trust in
air agencies’ messaging. Commenters
also noted if the breakpoints are revised,
the public will not visually be able to
detect the difference between what was
considered a good AQI day versus a
now moderate AQI day.
The EPA disagrees with these
commenters. With respect to setting a
short-term breakpoint to the level of a
much longer-term (annual) standard,
setting the lower AQI breakpoints at the
level of the annual and daily PM2.5
standards for communication purposes
was discussed in the proposed
reconsideration (88 FR 5558, January 27,
2023) and previously supported by State
organizations in the 2012 PM Final Rule
(77 FR 38890, June 29, 2012). Both the
AQI and the Pollutant Standards Index,
which came before it, have historically
been normalized across pollutants by
defining an index value of 50 and 100
as the numerical level of the annual
(when defined) and short-term (i.e.,
averaging time of 24-hours or less)
primary NAAQS for each pollutant.
This approach clearly communicates the
air quality to the public. The EPA
considers this approach to be
appropriate given the available evidence
and structure of the standard. As
discussed in section II.B above and in
the notice of final rulemaking for the
2012 review (77 FR 38890, June 29,
2012), the primary annual and 24-hour
PM2.5 standards work together in
concert to provide public health
protection. The annual PM2.5 standard is
generally viewed as the principal means
of providing public health protection
against ‘‘typical’’ daily and annual PM2.5
exposures, while the 24-hour PM2.5
standard is generally viewed as a means
of providing protection against shortterm exposures to ‘‘peak’’ PM2.5
concentrations, such as can occur in
PO 00000
Frm 00106
Fmt 4701
Sfmt 4700
areas with strong contributions from
local or seasonal sources, even when
annual average PM2.5 concentrations
remain relatively low. Because the
annual standard provides public health
protection for typical daily PM2.5
exposures, the EPA thinks it is
appropriate to use that level for the 50
breakpoint of the AQI and describe
daily air quality at and below the level
of the annual standard ‘‘Good.’’ Since an
annual standard allows for days with air
quality above that level, it is appropriate
to call days just above it ‘‘Moderate.’’ If
the 50 breakpoint of the AQI was set at
a level above the annual standard, it
would be possible for the majority of
days to be called ‘‘good’’ in a year when
an area exceeds the annual standard.
This could cause confusion with the
public about air quality if the general
perception is that local air quality is
‘‘good,’’ but the area fails to meet the
annual standard. In addition, the EPA
continues to find it appropriate to use
the NowCast with the PM2.5 AQI index
to provide more real-time information to
the public. As discussed in the AQI
Technical Assistance Document, while
the NowCast algorithm is approximating
a 24-hour average exposure, it can
reflect concentrations observed over
shorter averaging times when air quality
is changing rapidly (U.S. EPA, 2018a).
The EPA continues to consider the use
of the primary annual standard level
suitable in the NowCast given the health
evidence supporting the standard and
given that the reported concentrations
are an approximation of ‘‘typical’’ daily
exposure. Additionally, the EPA reflects
the nature of the NowCast in the
associated health messaging.
With regard to the commenter stating
the public may not be able to visually
detect a difference in the air quality, the
EPA notes that the AQI is intended to
be a communication tool for public
awareness precisely because it is
generally difficult for the public to
visually judge air quality risks when air
pollution is ‘‘moderate.’’ Moreover,
since the establishment of the AQI, the
EPA and State and local air agencies
and organizations have developed
experience in educating the public
about changes in the standards and,
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
concurrently, related changes to AQI
breakpoints and advisories. When the
standards change, the EPA and State
and local agencies have sought to help
the public understand that air quality is
not getting worse, it’s that the health
evidence underlying the standards and
the AQI has changed. The EPA’s Air
Quality System (AQS), the primary
repository for air quality monitoring
data, is also adjusted to reflect the
revised breakpoints. Specifically, all
historical AQI values in AQS are
recomputed with the revised
breakpoints, so that all data queries and
reports downstream of AQS will show
appropriate trends in AQI values over
time. If any State, local or Tribal air
agency is concerned that people are or
will be confused on a moderate AQI
day, then they could use the
communication information that has
been developed with this rulemaking.
Some commenters stated that the AQI
should not necessarily be linked to the
primary PM2.5 standards. One example
is the comment that if the annual
standard is not lowered to 8 mg/m3, the
EPA should lower the 50 breakpoint of
the AQI to that level to better inform the
public of the need for behavioral
modifications to reduce the harm to
health from PM2.5 exposure. Similar to
the reasons discussed above, the EPA
concludes that setting the 50 breakpoint
of the AQI at the level of the annual
PM2.5 standard is appropriate from a
health perspective and for
communication purposes. The
Administrator has judged the primary
annual standard (in conjunction with
the other primary standards) as revised
in this final action to be requisite to
protect public health with an adequate
margin of safety, based on the health
evidence discussed in section II.A.2.
Setting the 50 breakpoint lower than the
annual standard also has the potential to
cause confusion with the public since it
does not reflect the standards and the
Administrator’s judgments about the
standards as well.
With regard to the 100 breakpoint of
the AQI, several commenters expressed
the view that the level of the 24-hour
PM2.5 standard and an AQI value of 100
should be set at 25 mg/m3 based on the
body of evidence and lower end of the
range recommended by CASAC. These
commenters noted that if the current 24hour standard and AQI value of 100 is
retained at 35 mg/m3 then the public
will not be able to make informed
decisions about actions to take to
protect their health. Many of these
commenters further recommended that
the AQI value of 100 should be lowered
to 25 mg/m3 even if the standard is
retained. Commenters expressed the
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
view that this would more adequately
allow the public to take healthprotective actions.
The EPA disagrees with these
commenters and notes that many State,
Tribal and local air agencies have
expressed strong support for aligning
the 100 breakpoint of the AQI with the
short-term 24-hour primary PM2.5
standards as discussed in the proposal
(88 FR 5558, January 27, 2023). The EPA
agrees with the view, expressed by
State, local and Tribal entities, that
aligning the lower breakpoints with the
standards enables clear communication
of the standards. This alignment
approach is also utilized in the other
AQI sub-indices lower breakpoints and
taking a different approach with the
PM2.5 AQI could cause confusion.
Additionally, the Administrator has
judged that it is appropriate to retain the
24-hour standard at a level of 35 mg/m3
(in conjunction with the other primary
standards) to protect public health with
an adequate margin of safety, based on
the health evidence discussed in section
II.A.2. Thus, EPA disagrees that it is
necessary or appropriate to set the 100
breakpoint at a lower concentration to
provide further information to the
public. The 50 breakpoint, which is set
at a level below 25 mg/m3, will continue
to provide information to members of
the public particularly concerned about
exposures to PM2.5. As with the 50
breakpoint, aligning the breakpoint with
the standard both reflects the
Administrator’s judgment about the
health risks and eliminates the potential
to cause confusion in the public about
those risks.
b. Air Quality Index Values of 200 and
Above
Some commenters supported the
proposed revisions to the 200, 300 and
500 breakpoints that recognize the
expanded body of scientific evidence,
particularly several new epidemiologic
studies conducted during high pollution
events such as wildfires and multiple
controlled human exposure studies. A
few commenters agreed with
incorporating the expanded body of
scientific evidence into the 200, 300 and
500 breakpoints, but suggested a
modified linear approach between 200
(115 mg/m3) and 500 (312 mg/m3, setting
the 300 breakpoint to 187 mg/m3) based
on recent epidemiologic wildfire smoke
studies.
Other commenters disagreed with the
proposed revisions and suggested the
EPA should continue using the previous
breakpoints that follow the 1999 linear
approach (64 FR 42530, August 4, 1999),
because not changing the breakpoints
would simplify communications. A few
PO 00000
Frm 00107
Fmt 4701
Sfmt 4700
16307
commenters stated the proposed
revisions to the AQI upper breakpoints
are not justified because the scientific
evidence supporting the revisions is
inadequate. To support this view, the
commenters suggest that only three
epidemiologic studies were used in
determining the upper breakpoints and
none of them were representative of
potential effects in the general public; of
the 13 studies cited only three were near
the proposed revised breakpoints; four
of the studies involved exposure to PM
from diesel and traffic pollution, which
is different than PM from wildfire
smoke; and the data supporting the
revisions only indicated ‘‘mild’’ health
effects that were mostly in sensitive
populations.
The EPA agrees with the majority of
commenters that supported utilizing the
expanded body of scientific evidence to
revise the 200, 300 and 500 breakpoints
of the AQI. The EPA appreciates the
suggestion of using a revised linear
approach from 200 to 500. But rather
than using the available evidence to
only set the breakpoint of 500, the EPA
finds it appropriate to set the
breakpoints for 200, 300 and 500 using
an evidence-based approach, by relying
on information presented in both
controlled human exposure studies and
epidemiologic studies that examine
relationships between high PM2.5
exposure episodes (i.e., periods of
wildfire smoke) and various health
outcomes. Setting these breakpoints
based directly on health effects
evidence, which can be communicated,
is more useful and appropriate than
using a linear approach, because it can
better describe the potential health
effects and symptoms which also helps
the public better understand why more
health protective actions are needed. By
its nature, a linear approach does not
evaluate and identify associated health
effects and risk factors.
The EPA disagrees with the
commenters that expressed the view
that these upper breakpoints should not
be revised based largely on the
numerous peer-reviewed studies
published since the 200, 300 and 500
breakpoints were originally established
in 1999 (64 FR 42530, August 4, 1999).
As discussed in the proposal (88 FR
5641, January 27, 2023), the rationale
behind the proposed revisions is rooted
in the fact the upper AQI breakpoints
are based on outdated scientific
evidence. Specifically, the traditional
linear approach was predicated on the
500 value of the AQI, which was
estimated using health studies that used
the British Smoke Method. The British
Smoke Method is based on a particle
size fraction (4.5 microns) that is larger
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16308
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
than PM2.5. Given that the British Smoke
method has a larger particle size
cutpoint than the current PM2.5
monitoring method, which has a
cutpoint of 2.5 microns, a concentration
of 500 mg/m3 based on the British
Smoke method would be equivalent to
a lower PM2.5 concentration (88 FR
5641, January 27, 2023). The
combination of a larger particle size
fraction informing previous decisions
around upper AQI breakpoints and
more recent scientific evidence than the
London Fog Episode, on the potential
health consequences of what we
currently consider to be high PM2.5
exposures, provides the underlying
basis for revising the upper breakpoints
to better inform the public about air
quality to allow the public to take health
protective actions as appropriate.
Moreover, as discussed above, until
recently there was limited information
upon which to base the breakpoints
between 150 and 500, so the linear
approach was a reasonable substitute.
While not changing the breakpoints may
be easier because there is no change to
communicate, using a health-based
approach is more appropriate, because it
helps the public better understand that
more health protective actions are
needed.
The Agency disagrees that the
scientific evidence discussed in the
proposal is inadequate to revise the 200,
300 and 500 breakpoints of the AQI (88
FR 5640, January 27, 2023). The EPA
disagrees that these studies should not
be considered because they ‘‘indicated
mild health effects in sensitive
populations.’’ The EPA notes that many
of the subclinical effects discussed in
the proposal (88 FR 5640, January 27,
2023) that informed the breakpoints are
on the biologically plausible pathway
(see 2019 ISA, section 6.1.1 and Figure
6–1) to more severe cardiovascular
outcomes, such as ED visits, hospital
admissions, and death as depicted in
the large number of epidemiologic
studies evaluated in the 2019 ISA and
ISA Supplement. From a public health
perspective, the purpose of the AQI is
to inform the public when air quality
could adversely affect their health. The
scientific evidence informed revisions
to the breakpoints at the upper end of
the AQI allow it to better reflect the risk
of experiencing health effects at higher
PM2.5 concentrations. In addition, the
EPA disagrees with the commenter that
the effects reported at these higher
concentrations were observed only in
sensitive populations as these effects
were also reported in healthy
populations (Ghio et al., 2000; Ghio et
al., 2003; Urch et al., 2010; Ramanathan
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
take actions to reduce PM2.5 exposures.
Furthermore, with regard to an
additional warning system, the EPA is
concerned that having two air quality
communication systems operating at the
same time would likely be confusing to
c. Other Comments
the public and reduce the effectiveness
The EPA received a few additional
of the systems.
comments on elements of the PM2.5 AQI,
At the same time, the EPA recognizes
including the averaging time. Some
that when air quality is rapidly
commenters expressed the view that the changing, such as during wildfire smoke
24-hour averaging time was not useful
events, reporting information based on a
when informing the public how to
24-hour metric may not be as useful for
protect their health, particularly during
the public as reporting more frequently
rapidly changing conditions such as
would be. The EPA has balanced
wildfire smoke events. Instead, they
concerns about being able to provide
suggested a subdaily averaging time of
timely communication of air quality
1–3 hours would be more effective
hazards when conditions are changing
because it more closely aligns with how quickly with the goal of limiting the
people breathe.
number of air quality communications
A few of these commenters suggested
systems and its judgment that the
that instead of changing the AQI
evidence supports a 24-hour-based
averaging time, which aligns with the
metric linked to the daily standard by
short-term standard, the EPA could
establishing the NowCast, which takes
create a public health warning system
into consideration subdaily PM2.5
for unhealthy PM2.5 levels. The
concentrations and provides a near realcommenters noted that aligning the AQI time AQI value based on the AQI colors
averaging time with the short-term
and scale. Specifically, the NowCast
standard could be useful for consistent
shows air quality conditions for the
communication with the standards and
most current hour of PM2.5 data
attainment but suggested that a subdaily available by using a calculation that
warning system could better allow the
involves multiple hours of past data. As
public to take health protective actions.
noted in the AQI Technical Assistance
The EPA disagrees that a shorter
Document, the NowCast currently uses
averaging period for the PM2.5 AQI sublonger averages during periods of stable
index would be better. The health
air quality and shorter averages (down
effects evidence supporting a subdaily
to a 3-hour average) when air quality is
metric is limited and inconsistent. As
changing rapidly, such as during a
part of its review of the health effects
wildfire (U.S. EPA, 2018a). As discussed
evidence, the 2019 ISA evaluated
further in section IV.D.2 of this notice,
whether a subdaily metric would be
the EPA uses the NowCast to
more closely related to health effects.
approximate the complete daily AQI
Most epidemiologic studies that
(24-hour average) during any given
examined the relationship between
hour. This means the subdaily NowCast
short-term PM2.5 exposures and health
is approximating a 24-hour average
effects evaluated an exposure metric
exposure, which aligns with the health
averaged over 24-hours. Some recent
evidence and the existing AQI
studies, focusing on respiratory and
communications network, while also
cardiovascular effects and mortality,
being capable of communicating rapidly
have examined whether there is
changing conditions to the public.
evidence that subdaily exposure metrics
are more closely related to health effects 3. Summary of Final Revisions
than a traditional 24-hour average
Upon reviewing and considering the
metric. After evaluating this limited
comments on the proposed revisions
newer evidence, the 2019 ISA
(summarized above in Section IV.C)
concluded that ‘‘collectively, the
along with the scientific evidence
available evidence does not indicate
outlined in the proposal (88 FR 5639,
that subdaily averaging periods for
January 27, 2023) and summarized
PM2.5 are more closely associated with
above in section IV.A, the EPA is
finalizing the proposed changes to the
health effects than the 24-hour avg
AQI.
exposure metric,’’ (2019 ISA, chapter 1,
Thus, as discussed in section IV of the
section 1.5.2.1, pp. 146–147; U.S. EPA,
preamble (88 FR 5639, January 27, 2023)
2022a).
to the proposed rule, the EPA is taking
In addition, there are communication
benefits to aligning the averaging time of final action to revise the AQI value of
50 to 9.0 mg/m3, 24-hour average,
the AQI with the daily standard, as
some of these commenters note, such as consistent with the final decision on the
providing consistent messages about
primary annual PM2.5 standard level as
when it may be beneficial for people to
summarized in section II.C of the
et al., 2016; Sivagangabalan et al., 2011;
Brook et al., 2009; Bellavia et al. (2013);
Tong et al. (2015); Behbod et al. (2013);
Vieira et al. (2016a) Vieira et al. (2016b);
and Lucking et al. (2011)).
PO 00000
Frm 00108
Fmt 4701
Sfmt 4700
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
preamble to the final rule; retain the
AQI value of 100 at 35 mg/m3, 24-hour
average, consistent with the final
decision on the primary 24-hour PM2.5
standard level as summarized in section
II.C of the preamble to the final rule;
and retain the AQI value of 150 at 55 mg/
m3, 24-hour average. The EPA is also
taking action to revise the AQI value of
200 to 125 mg/m3, 24-hour average; 300
to 225 mg/m3, 24-hour average; and 500
to 325 mg/m3, 24-hour average,
consistent with the rationale discussed
above and the health evidence
discussed in section IV of the preamble
(88 FR 5639, January 27, 2023) to the
proposed rule. The EPA has prepared
communications materials to assist
States with adjusting to the revised AQI
and looks forward to working with, and
learning from the experiences of, State,
local, and Tribal governments in
implementing these changes.
C. Air Quality Index Category
Breakpoints for PM10
The EPA proposed to retain the PM10
sub-index of the AQI consistent with the
proposed decision to retain the primary
PM10 standard, and consistent with the
health effects information that supports
this proposed decision, as discussed in
section III.D of the proposal (88 FR
5632, January 27, 2023). EPA did not
receive comments on this and is taking
final action to retain the PM10 sub-index
of the AQI for the reasons stated in the
preamble to the proposed rule (88 FR
5642, January 27, 2023).
ddrumheller on DSK120RN23PROD with RULES3
D. Air Quality Index Reporting
With respect to the reporting
requirements for the AQI and as noted
in the proposal (88 FR 5642, January 27,
2023) there have been many
technological advances in air quality
monitoring and data reporting since the
appendix G to 40 CFR part 58 was last
revised in 1999. Federal, State, local,
and Tribal agencies have used these
changes to make health information and
air quality data more readily available
and easier to access. Given this, it is
useful to update the reporting
requirements and recommendations to
match current practices and ensure the
public has the most useful and timely
information to take health-protective
behaviors.
1. Summary of Proposed Revisions
Currently, appendix G defines daily
reporting as five days per week. When
this reporting requirement was
originated in 1999 the technology
available at that time was not sufficient
to calculate and report the AQI more
than five days per week without
requiring additional staffing on the
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
weekends. Since that time, advances in
technology have allowed for reporting
seven days per week automatically
without expending additional resources
on weekends. As a result, most State,
local, and Tribal air agencies now report
the AQI seven days per a week. Given
these technological advances and noting
that reporting agencies currently report
the AQI seven days per week, the EPA
proposed that State, local, and Tribal
agencies that report the AQI be required
to report it seven days a week, ensuring
that the members of the public continue
to have access to daily air quality and
health information that they can use to
take steps to protect their health.
Improvements in monitoring
networks and modeling capabilities
have also enabled the ability to report
the AQI in near real-time. This allows
State, local, and Tribal air agencies to
provide timely air quality information to
the public for making health-protective
decisions and to help satisfy AQI
reporting requirements. The availability
of near real-time AQI data also allows
for more timely responses by the public
when air quality conditions are
changing rapidly, such as during
wildfire smoke events. Subdaily
reporting of the AQI can be critical
when there are rapidly change
conditions and/or high pollution events
so that the public is able to make
informed decisions to protect their
health. Many State, local, and Tribal air
agencies currently report the AQI hourly
to ensure that the public has access to
accurate and timely information. In
recognition of these advances, and to
continue to provide for near-real time
AQI reporting that the public has come
to rely on, the EPA proposed to
recommend that State, local, and Tribal
agencies report the AQI in near-real
time.
In lieu of or along with reporting the
near-real-time AQI directly to the
public, most State/local and Tribal
agencies submit hourly air quality data
to the EPA. The EPA and some State,
local and Tribal air quality agencies use
this near-real-time data to create
products for use by the public, weather
service providers and the media as
discussed in the proposal (88 FR 5643,
January 27, 2023). To continue to ensure
the availability of the products that the
public and many stakeholders rely
upon, the EPA proposed to recommend
that State, local, and Tribal air quality
agencies submit hourly data to the
EPA’s air quality database. Submitting
hourly data to the EPA for use on the
AirNow website and in other products
also enables State, local, and Tribal air
quality agencies to meet the
PO 00000
Frm 00109
Fmt 4701
Sfmt 4700
16309
recommendation to report the AQI in
near-real-time.
In addition to the proposed updates to
the reporting requirements and
recommendations for near-real-time
reporting and data submission
recommendations, the Agency also
proposed reformatting the question-andanswer format used in appendix G to
align with the current standard
formatting used in the Code of Federal
Regulations. In proposing to update the
format, the EPA did not reopen the
language that has merely been moved or
rearranged as there are no substantive
changes.
Another change the EPA proposed to
make to appendix G is with regard to
Table 2—Breakpoints for the AQI for
purposes of clarity. As discussed in the
proposal (88 FR 5642, January 27, 2023)
and summarized here, the EPA
proposed to collapse the two rows
presented for the Hazardous Category
into one. The two rows in the current
table specify pollutant concentrations
for two AQI ranges within the
Hazardous category (301–400 and 401–
500), with an intermediate break at 400.
The 400 breakpoint for all criteria
pollutants in the current Table 2 is set
at the proportional pollutant
concentration approximately halfway
between the Index values of 300 and
500. In proposing updated AQI
breakpoints for PM2.5, the EPA
considered adjusting the 400 breakpoint
similarly. However, the EPA concluded
that collapsing the two rows into a
single range (301–500) would provide a
more transparent and easy-to-follow
presentation of the pollutant
concentrations corresponding to the
AQI range for the Hazardous category.
Moreover, collapsing the Hazardous
category into a single row in Table 2 has
no substantive effect on the Emergency
Episode program in 40 CFR part 51,
appendix L. Thus, the EPA proposed to
remove the breakpoint of 400 from the
table in appendix G but this change
would not substantively affect the
derivation of the AQI for any pollutant.
In addition, the EPA proposed to
move some information currently in
appendix G into the Technical
Assistance Document for the Reporting
of Daily Air Quality, or TAD (U.S. EPA,
2018a), so that it can be updated in a
more timely manner to reflect current
scientific and health effects evidence
and current communication methods,
thereby assisting State, local, and Tribal
agencies in providing accurate and
timely information to the public.
Information that was proposed to be
moved from appendix G to the TAD
included the definitions of the sensitive
(at-risk) populations for each pollutant.
E:\FR\FM\06MRR3.SGM
06MRR3
16310
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
This definition is typically evaluated
and updated, as warranted, in most
NAAQS reviews, even if the standard is
not revised. Generally, if the standard is
not revised in a review of the NAAQS,
then appendix G is also not revised.
Moving the definitions of sensitive
groups to the TAD allows them to be
updated even when a NAAQS is not
revised to be consistent with the
definitions of the sensitive (at-risk)
populations identified in the ISA for
that NAAQS review. Also, the proposal
(88 FR 5642, January 27, 2023)
recognized that the ways that air quality
and health information is supplied to
the news media and public changes
regularly and thus proposed that
information about suggested approaches
for public communication be taken out
of appendix G and discussed in the
TAD.
2. Summary of Significant Comments on
the Proposed Revisions
The EPA received many comments on
the proposed changes to AQI reporting,
many of which supported the proposed
revisions. EPA discusses several of the
topics that received the most attention
from commenters below. Discussion of
other comments received on the
proposed changes to the AQI can be
found in section IV of the Responses to
Significant Comments on the 2023
Proposed Reconsideration of the
National Ambient Air Quality Standards
for Particulate Matter.
Most commenters expressed support
for revising the definition of ‘‘daily
reporting’’ from five days a week to
seven days a week. A commenter did
not support this change and
recommended the EPA maintain the
definition of daily as five days per week,
noting that State and local air agencies
do not routinely work seven days per
week and would not be available to
perform quality control of this data and
report it reliably on weekends.
The EPA appreciates the support for
this proposed revision and disagrees
that the proposed change would require
personnel to perform quality control of
AQI data on weekends. 40 CFR part 58
Appendix D defines continuous
monitoring requirements for agencies
participating in the State/Local Air
Monitoring Stations (SLAMS) network,
and Appendix G states that agencies ‘‘
. . . must use concentration data from
State/Local Air Monitoring Stations
(SLAMS) required by 40 CFR 58.10’’
when reporting the AQI. Therefore, as
noted in Appendix D and G, Agencies
are required to report the AQI using
monitors within SLAMS, which are not
subject to daily quality control/
validation.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
A few commenters noted that the
proposal preamble language mentioned
AQI is reported three ways (88 FR 5637,
5638, January 27, 2023): ‘‘The AQI is
reported three ways all of which are
useful and complementary. The daily
AQI is reported for the previous day and
used to observe trends in community air
quality, the AQI forecast helps people
plan their outdoor activities for the next
day, and the near-real-time AQI, or
NowCast AQI, tells people whether it is
a good time for outdoor activity.’’ These
commenters suggested that the NowCast
is being codified in 40 CFR part 58
Appendix G as a method of calculating
the AQI, which they oppose, saying that
codifying its use is inappropriate given
the shortest averaging period of the
PM2.5 NAAQS remains at 24-hours.
Some stated that NowCast values have
no direct correlation to the AQI
calculation methodology codified in 40
CFR part 58 Appendix G. These
commenters say that codifying the
NowCast would impose a significant
burden on States’ forecasting staff.
However, some other commenters
noted they appreciate the publicfriendly format and near real-time data
the NowCast provides and use it in their
clinical encounters with patients. One
air agency recognized the importance of
the NowCast near real-time AQI during
high pollution events and suggested the
EPA should provide more ‘‘concrete’’
health messaging for these short-term
spikes.
The EPA disagrees that the preamble
language proposed to codify the
NowCast or to impose a burden on
reporting agencies. The preamble to the
proposed rule references the AQI being
reported in three ways and it does so
because the EPA and many State, local
and Tribal air quality agencies already
report it these three ways. However, text
included in the preamble is generally
explanatory and does not alter
regulatory provisions. Comments that
State that EPA is codifying the NowCast
into Appendix G are incorrect. Further,
in proposed revisions to 40 CFR part 58
Appendix G, the EPA recommended,
but did not propose to require, the use
of air quality forecasts and a subdaily
AQI. Consistent with the proposal, the
EPA is therefore not finalizing any
additional requirement or burden on
States’ forecasting staff relative to
forecasts or a subdaily AQI.
The EPA disagrees with the comment
that the NowCast values have no direct
correlation to the AQI calculation
methodology codified in 40 CFR part 58
Appendix G. As noted in the AQI
Technical Assistance Document
(Technical Assistance Document for the
Reporting of Daily Air Quality—the Air
PO 00000
Frm 00110
Fmt 4701
Sfmt 4700
Quality Index (AQI)), the NowCast
algorithm is based on the AQI
methodology but provides more realtime information to the public (U.S.
EPA, 2018a). While the NowCast
algorithm is approximating a 24-hour
average exposure, it can reflect
concentrations observed over shorter
averaging times when air quality is
changing rapidly (U.S. EPA, 2018a). The
EPA reflects the nature of the NowCast
in the health messaging provided there.
As noted in the above discussion of
the AQI, air quality can change quickly
during the day. A central purpose of the
AQI is to help the public know when it
is prudent to take action to reduce their
exposure to pollution. Accordingly, the
EPA developed the NowCast to estimate
the 24-hour AQI for the current hour to
give people information and tools to
reduce their exposures to protect their
health, particularly when air quality
may be changing. The NowCast gives
people the knowledge and ability to take
timely action. They can use this
information to reduce their exposure—
reducing exposures if PM2.5 is high only
during a few hours a day will help
reduce a person’s 24-hour exposure—or
be active when air quality is better.
The first NowCast method was
developed in 2003 and was designed so
‘‘current conditions’’ represent the 24hour PM2.5 standard as closely as
possible. This method proved to be slow
to respond during rapid air quality
changes. In 2013, the EPA developed an
updated NowCast method for PM2.5 141
that responds more quickly to rapidly
changing air quality conditions, such as
those we see during wildfires, to make
air quality alerts more timely. We
analyzed millions of data points in
developing this NowCast method and
presented this information to State,
local and Tribal air agencies. The
updated NowCast, which is still in use,
was launched August 1, 2013, on
AirNow.gov. It was designed to
represent a shorter average (target 3hour) when air quality is changing
rapidly, in part because 3-hour averages
from some continuous monitors are
more stable than 1-hour averages. The
NowCast reflects a longer-term (12-hour)
average when air quality is stable.
After evaluating the 2013 NowCast
method, the EPA concluded that it
matched the desired characteristics. The
NowCast method responds to rapid
changes in air quality yet still reflects a
141 U.S. EPA. (2013). Transitioning to a New
NowCast Method. Presentation available in the
Rulemaking Docket for the Review of the National
Ambient Air Quality Standards for Particulate
Matter (EPA–HQ–OAR–2015–0072), at: https://
www.regulations.gov/docket/EPA-HQ-OAR-20150072.
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
longer-term average when air quality is
stable; will work in any location with
adequate air quality data and for any air
quality situation; gives people the best
possible estimate of a 24-hour exposure;
allows the EPA to caution people in
time for them to take protective action
and reduce their 24-hour exposure; and
ensures that AQI maps on AirNow more
closely match what people see.
The AQI is designed to allow people
to reduce their exposure when pollution
levels are higher and be active outdoors
when pollution levels are lower. Since
air quality almost always changes
during the day, that level of granularity
is not possible with a 24-hour forecast.
If the public has only the 24-hour
forecast, they may miss the times to be
active outdoors when air quality is
better and may be active outdoors when
air quality is worse.
Also as noted above, many entities
appreciate the near real-time reporting
of the AQI that the NowCast provides
and suggested more specific messaging
is needed. The EPA appreciates this
insight and will continue to consider
ways to communicate air quality
information most effectively to the
public. For example, in light of recent
wildfire events, the EPA worked with
the USFS to pilot the AirNow Fire and
Smoke Map.
3. Summary of Final Revisions
Upon reviewing and considering the
comments on the proposed revisions
(summarized above in Section IV.C)
along with the rationale outlined in the
proposal (88 FR 5638, January 27, 2023)
and summarized above in section IV.C,
the EPA is finalizing the proposed
changes to the AQI reporting
requirements. Thus, as discussed in
section IV of the preamble to the
proposed rule, the EPA is taking final
action to require the AQI be reported
16311
seven days a week; recommend that
State, local, and Tribal agencies report
the AQI in near-real time; recommend
that State, local, and Tribal air quality
agencies submit hourly data to the
EPA’s air quality database; reformat
appendix G to align with the current
standard formatting used in the Code of
Federal Regulations; collapse the two
rows in Table 2 presented for the
Hazardous Category into one by
removing the 400 breakpoint; and move
some information currently in appendix
G into the Technical Assistance
Document for the Reporting of Daily Air
Quality, or TAD (U.S. EPA, 2018a) such
as including the definitions of the
sensitive (at-risk) populations for each
pollutant and suggested approaches for
public communication as stated in the
revised Appendix G.
Table 2 below summarizes the
breakpoints for the PM2.5 sub-index.
TABLE 2—BREAKPOINTS FOR PM2.5 SUB-INDEX
AQI category
Index values
Good ................................................................................................................................................
Moderate ..........................................................................................................................................
Unhealthy for Sensitive Groups .......................................................................................................
Unhealthy .........................................................................................................................................
Very Unhealthy ................................................................................................................................
Hazardous 1 .....................................................................................................................................
0–50
51–100
101–150
151–200
201–300
301+
Breakpoints
(μg/m3, 24-hour average)
0.0–9.0
9.1–35.4
35.5–55.4
55.5–125.4
125.5–225.4
225.5
1 AQI values between breakpoints are calculated using equation 1 in appendix G. For AQI values in the hazardous category, AQI values greater than 500 should be calculated using equation 1 and the PM2.5 concentration specified for the AQI value of 500.
V. Rationale for Decisions on the
Secondary PM Standards
ddrumheller on DSK120RN23PROD with RULES3
This section presents the rationale for
the Administrator’s decision that no
change to the current secondary PM
standards is required at this time to
provide requisite protection against the
public welfare effects of PM within the
scope of this reconsideration (i.e.,
visibility, climate, and materials
effects).142 This decision is based on a
thorough review of the scientific
evidence generally published through
142 Consistent with the 2016 Integrated Review
Plan (U.S. EPA, 2016), other welfare effects of PM,
including ecological effects, are being considered in
the separate, on-going review of the secondary
NAAQS for oxides of nitrogen, oxides of sulfur and
PM. Accordingly, the public welfare protection
provided by the secondary PM standards against
ecological effects such as those related to deposition
of nitrogen- and sulfur-containing compounds in
vulnerable ecosystems is being considered in that
separate review. Thus, the Administrator’s decision
in this reconsideration will be focused only and
specifically on the adequacy of public welfare
protection provided by the secondary PM standards
from effects related to visibility, climate, and
materials and hereafter ‘‘welfare effects’’ refers to
non-ecological welfare effects (i.e., visibility,
climate, and materials effects).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
December 2017,143 as presented in the
2019 ISA (U.S. EPA, 2019a), on the nonecological public welfare effects of PM
pertaining to the presence of PM in
ambient air, specifically visibility,
climate, and materials effects.
Additionally, this decision is based on
a thorough evaluation of some studies
that became available after the literature
cutoff date of the 2019 ISA that could
either further inform the adequacy of
the current PM NAAQS or address key
scientific topics that have evolved since
the literature cutoff date for the 2019
ISA, generally through March 2021, as
presented in the ISA Supplement 144
143 In addition to the 2020 review’s opening ‘‘call
for information’’ (79 FR 71764, December 3, 2014),
the 2019 ISA identified and evaluated studies and
reports that have undergone scientific peer review
and were published or accepted for publication
between January 1, 2009 through approximately
January 2018 (U.S. EPA, 2019a, p. ES–2). References
that are cited in the 2019 ISA, the references that
were considered for inclusion but not cited, and
electronic links to bibliographic information and
abstracts can be found at: https://hero.epa.gov/hero/
particulate-matter.
144 As described in more detail in the ISA
Supplement, ‘‘the scope of this Supplement
provides specific criteria for the types of studies
considered for inclusion within the Supplement.
PO 00000
Frm 00111
Fmt 4701
Sfmt 4700
(U.S. EPA, 2022a). The selection of
welfare effects evaluated within the ISA
Supplement was based on the causality
determinations reported in the 2019 ISA
and the subsequent use of scientific
evidence in the 2020 PA.145
Specifically, studies must be peer reviewed and
published between approximately January 2018 and
March 2021’’ (U.S. EPA, 2022a, section 1.2.2).
145 As described in section 1.2.1 of the ISA
Supplement, ‘‘the selection of welfare effects to
evaluate within this Supplement is based on the
causality determinations reported in the 2019 PM
ISA and the subsequent use of scientific evidence
in the 2020 PM PA. The 2019 PM ISA concluded
a causal relationship for each of the welfare effects
categories evaluated (i.e., visibility, climate effects,
and materials effects). While the 2020 PM PA
considered the broader set of evidence for these
effects, for climate effects and material effects, it
concluded that there remained ‘substantial
uncertainties with regard to the quantitative
relationships with PM concentrations and
concentration patterns that limit[ed] [the] ability to
quantitatively assess the public welfare protection
provided by the standards from these effects (U.S.
EPA, 2020b). Given these uncertainties and
limitations, the basis of the discussion on
conclusions regarding the secondary standards in
the 2020 PM PA primarily focused on visibility
effects. Therefore, this Supplement focuses only on
visibility effects in evaluating newly available
scientific information and is limited to studies
E:\FR\FM\06MRR3.SGM
Continued
06MRR3
16312
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
Specifically, for welfare effects, the
focus within the ISA Supplement is on
visibility effects. The ISA Supplement
does not include an evaluation of
studies on climate or materials effects.
The Administrator’s decision also takes
into account the 2022 PA evaluation of
the policy-relevant information in the
2019 ISA and ISA Supplement and
presentation of quantitative analysis of
air quality related to visibility
impairment; CASAC advice and
recommendations, as reflected in
discussions of the drafts of the ISA
Supplement and 2022 PA at public
meetings and in the CASAC’s letters to
the Administrator; and public
comments received on the proposal.
In presenting the rationale for the
Administrator’s final decision and its
foundations, section V.A provides
background on the 2020 final decision
to retain the secondary PM standards
(section V.A.1), and also provides brief
summaries of key aspects of the
currently available welfare effects
evidence (section V.A.2) and
quantitative information (section V.A.3).
Section V.B summarizes the CASAC’s
advice (section V.B.1) and the proposed
conclusions (section V.B.2), addresses
public comments received on the
proposal (section V.B.3), and presents
the Administrator’s conclusions on the
adequacy of the current standards
(section V.B.4), drawing on
consideration of the available scientific
and quantitative information, advice
from the CASAC, and comments from
the public. Section V.C summarizes the
Administrator’s decision on the
secondary PM standards.
A. Introduction
The general approach for this
reconsideration of the 2020 final
decision on the secondary PM standards
relies on the EPA’s assessments of the
current scientific evidence and
associated quantitative analyses to
inform the Administrator’s judgments
regarding secondary standards that are
requisite to protect the public welfare
from known or anticipated adverse
effects associated with the pollutant’s
presence in the ambient air. The EPA’s
assessments are primarily documented
in the 2019 ISA, ISA Supplement, and
2022 PA, which builds on the 2020 PA,
all of which have received CASAC
review and public comment (83 FR
53471, October 23, 2018; 83 FR 55529,
November 6, 2018; 85 FR 4655, January
27, 2020; 86 FR 52673, September 22,
2021; 86 FR 54186, September 30, 2021;
86 FR 56263, October 8, 2021; 87 FR
conducted in the U.S. and Canada’’ (U.S. EPA,
2022a, section 1.2.1).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
958, January 7, 2022; 87 FR 22207, April
14, 2022; 87 FR 31965, May 26, 2022).
In bridging the gap between the
scientific assessments of the 2019 ISA
and ISA Supplement and the judgments
required of the Administrator in
determining whether the current
standards provide the requisite public
welfare protection, the 2022 PA
evaluates policy implications of the
evaluation of the current evidence in the
2019 ISA and ISA Supplement, and the
quantitative information documented in
the 2022 PA. In evaluating the public
welfare protection afforded by the
current standards against PM-related
effects within the scope of this
reconsideration, the four basic elements
of the NAAQS (indicator, averaging
time, level, and form) are considered
collectively.
The final decision on the adequacy of
the current secondary standards is a
public welfare policy judgment to be
made by the Administrator. In reaching
conclusions with regard to the standard,
the decision draws on the scientific
information and analyses about welfare
effects, and associated public welfare
significance, as well as judgments about
how to consider the range and
magnitude of uncertainties that are
inherent in the scientific evidence and
analyses. This approach is based on the
recognition that the available evidence
generally reflects a continuum that
includes ambient air exposures at which
scientists agree that effects are likely to
occur through lower levels at which the
likelihood and magnitude of responses
become increasingly uncertain. This
approach is consistent with the
requirements of the provisions of the
Clean Air Act related to the review of
NAAQS and with how the EPA and the
courts have historically interpreted the
Act. These provisions require the
Administrator to establish secondary
standards that, in the judgment of the
Administrator, are requisite to protect
public welfare from known or
anticipated adverse effects associated
with the presence of the pollutant in the
ambient air. In so doing, the
Administrator seeks to establish
standards that are neither more nor less
stringent than necessary for this
purpose. The Act does not require that
standards be set at a zero-risk level, but
rather at a level that reduces risk
sufficiently so as to protect the public
welfare from known or anticipated
adverse effects.
1. Background on the Current Standards
The current secondary PM standards
were retained in 2020 based on the
scientific and technical information
available at that time, as well as the
PO 00000
Frm 00112
Fmt 4701
Sfmt 4700
then-Administrator’s judgments
regarding the available welfare effects
evidence, the appropriate degree of
public welfare protection for the
existing standards, and available air
quality information on visibility
impairment that may be allowed by
such a standard (85 FR 82684, December
18, 2020). With the 2020 decision, the
then-Administrator retained the
secondary 24-hour PM2.5 standard, with
its level of 35 mg/m3, the annual PM2.5
standard, with its level of 15.0 mg/m3,
and the 24-hour PM10 standard, with its
level of 150 mg/m3. The subsections
below focus on the key considerations
and the then-Administrator’s
conclusions in the 2020 final decision
for climate and materials effects (section
V.A.1.a) and visibility effects (section
V.A.2.b).
a. Non-Visibility Effects
In light of the robust evidence base,
the 2019 ISA concluded there to be
causal relationships between PM and
climate effects and materials effects
(U.S. EPA, 2019a, sections 13.3.9 and
13.4.2). The 2020 final decision was
based on a thorough review in the 2019
ISA of the scientific information on PMinduced climate and materials effects.
The decision also took into account: (1)
Assessments in the 2020 PA of the most
policy-relevant information in the 2019
ISA regarding evidence of adverse
effects of PM to climate and materials,
(2) uncertainties in the available
evidence to inform a quantitative
assessment of PM-related climate and
materials effects, (3) CASAC advice and
recommendations, and (4) public
comments received during the
development of these documents and on
the proposal document.
In considering non-visibility welfare
effects in the 2020 decision, the thenAdministrator concluded that, while it
is important to maintain an appropriate
degree of control of fine and coarse
particles to address non-visibility
welfare effects, ‘‘it is generally
appropriate to retain the existing
standards and that there is insufficient
information to establish any distinct
secondary PM standards to address
climate and materials effects of PM’’ (85
FR 82744, December 18, 2020).
With regard to climate, the thenAdministrator recognized that there
were a number of improvements and
refinements to climate models since the
2012 review. However, while the
evidence continued to support a causal
relationship between PM and climate
effects, the then-Administrator noted
that significant limitations continued to
exist related to quantifying the
contributions of direct and indirect
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
effects of PM and PM components on
climate forcing (U.S. EPA, 2020b,
sections 5.2.2.1.1 and 5.4). He also
recognized that the models continued to
exhibit considerable variability in
estimates of PM-related climate impacts
at regional scales (e.g., ∼100 km) as
compared to simulations at global
scales. Therefore, the resulting
uncertainty led the then-Administrator
to conclude in the 2020 decision that
the available scientific information
remained insufficient to quantify
climate impacts associated with
particular concentrations of PM in
ambient air (U.S. EPA, 2020b, section
5.2.2.2.1) or to evaluate or consider a
level of PM air quality in the U.S. to
protect against climate effects and that
there was insufficient information
available to base a national ambient
standard on climate impacts (85 FR
82744, December 18, 2020).
With regard to materials effects, the
then-Administrator noted that the
evidence available in the 2019 ISA
continued to support a causal
relationship between materials effects
and PM deposition (U.S. EPA, 2019a,
section 13.4). He recognized that the
deposition of fine and coarse particles to
materials can lead to physical damage
and/or impaired aesthetic qualities.
Particles can contribute to materials
damage by adding to the natural
weathering processes and by promoting
the corrosion of metals, the degradation
of building materials, and the
weakening of material components.
While some new information was
available in the 2019 ISA, the
information was from studies primarily
conducted outside of the U.S. in areas
where PM concentrations in ambient air
are higher than those observed in the
U.S. (U.S. EPA, 2020b, section 13.4).
Additionally, the information assessed
in the 2019 ISA did not support
quantitative analyses of PM-related
materials effects in the 2020 PA (U.S.
EPA, 2020b, section 5.2.2.2.2). Given the
limited amount of information available
and its inherent uncertainties and
limitations, the Administrator
concluded that he was unable to relate
soiling or damage to specific levels of
PM in ambient air or to evaluate or
consider a level of air quality to protect
against such materials effects, and that
there was insufficient information
available to support a distinct national
ambient standard based on materials
effects (85 FR 82744, December 18,
2020).
In reviewing the 2019 draft PA, the
CASAC agreed with staff conclusions
that, while these effects are important,
‘‘the available evidence does not call
into question the protection afforded by
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
the current secondary PM standards’’
and recommended that the secondary
standards ‘‘should be retained’’ (Cox,
2019b, p. 3 of letter). In reaching a final
decision in 2020, for all of the reasons
discussed above and recognizing the
CASAC conclusion that the evidence
provided support for retaining the
current secondary PM standards, the
then-Administrator concluded that it
was appropriate to retain the existing
secondary PM standards, without
revision. For climate and materials
effects, this conclusion reflected his
judgment that, although it remains
important to maintain secondary PM2.5
and PM10 standards to provide some
degree of control over long- and shortterm concentrations of both fine and
coarse particles, there was insufficient
information to establish distinct
secondary PM standards to address nonvisibility PM-related welfare effects (85
FR 82744, December 18, 2020).
b. Visibility Effects
The 2019 ISA concluded that, ‘‘the
evidence is sufficient to conclude that a
causal relationship exists between PM
and visibility impairment’’ (U.S. EPA,
2019a, section 13.2.6). The 2020
decision on the adequacy of the
secondary standards with regard to
visibility effects was a public welfare
policy judgment made by the thenAdministrator, which drew upon the
available scientific evidence for PMrelated visibility effects and on analyses
of visibility impairment, as well as
judgments about the appropriate weight
to place on the range of uncertainties
inherent in the evidence and analyses.
The 2020 final decision was based on a
thorough review in the 2019 ISA of the
scientific information on PM-related
visibility effects. The decision also took
into account: (1) Assessments in the
2020 PA of the most policy-relevant
information in the 2019 ISA regarding
evidence of adverse effects of PM on
visibility; (2) air quality analyses of the
PM2.5 visibility index and design values
based on the form and averaging time of
the existing secondary 24-hour PM2.5
standard; (3) CASAC advice and
recommendations; and (4) public
comments received during the
development of these documents and on
the 2020 proposal document.
In considering the visibility effects in
the 2020 review, the then-Administrator
noted the long-standing body of
evidence for PM-related visibility
impairment. This evidence, which is
based on the fundamental relationship
between light extinction and PM mass,
demonstrated that ambient PM can
impair visibility in both urban and
remote areas, and had changed very
PO 00000
Frm 00113
Fmt 4701
Sfmt 4700
16313
little since the 2012 review (U.S. EPA,
2019a, section 13.1; U.S. EPA, 2009a,
section 9.2.5). The evidence related to
public perception of visibility
impairment was from studies from four
areas in North America.146 These
studies provided information to inform
our understanding of levels of visibility
impairment that the public judged to be
‘‘acceptable’’ (U.S. EPA, 2010b; 85 FR
24131, April 30, 2020). In considering
these public preference studies, the
then-Administrator noted that no new
visibility studies conducted in the U.S.
were discussed in the 2019 ISA, and
there was little newly available
information with regard to acceptable
levels of visibility impairment in the
U.S. The Administrator recognized that
visibility impairment can have
implications for people’s enjoyment of
daily activities and their overall wellbeing, and therefore, considered the
degree to which the current secondary
standards protect against PM-related
visibility impairment.
Consistent with the 2012 review, in
the 2020 review, the then-Administrator
first concluded that a target level of
protection for a secondary PM standard
is most appropriately defined in terms
of a visibility index that directly takes
into account the factors (i.e., species
composition and relative humidity) that
influence the relationship between
PM2.5 in ambient air and PM-related
visibility impairment. In defining a
target level of protection, the thenAdministrator considered the specific
aspects of such an index, including the
appropriate indicator, averaging time,
form and level (78 FR 82742–82744,
December 18, 2020).
First, with regard to indicator, the
then-Administrator noted that in the
2012 review, the EPA used an index
based on estimates of light extinction by
PM2.5 components calculated using an
adjusted version of the IMPROVE
algorithm, which allows the estimation
of the light extinction using routinely
monitored components of PM2.5,
PM10–2.5 mass, and estimates of relative
humidity. The then-Administrator
recognized that, while there have been
some revisions to the IMPROVE
algorithm since the time of the 2012
146 Preference studies were available in four
urban areas. Three western preference studies were
available, including one in Denver, Colorado (Ely et
al., 1991), one in the lower Fraser River valley near
Vancouver, British Columbia, Canada (Pryor, 1996),
and one in Phoenix, Arizona (BBC Research &
Consulting, 2003). A pilot focus group study was
also conducted for Washington, DC (Abt Associates,
2001), and a replicate study with 26 participants
was also conducted for Washington, DC (Smith and
Howell, 2009). More details about these studies are
available in Appendix D of the 2022 PA (U.S. EPA,
2022b).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16314
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
review, our fundamental understanding
of the relationship between PM in
ambient air and light extinction had
changed little and the various IMPROVE
algorithms appropriately reflected this
relationship across the U.S. In the
absence of a monitoring network for
direct measurement of light extinction,
he concluded that a calculated light
extinction indicator that utilizes the
IMPROVE algorithms continued to
provide a reasonable basis for defining
a target level of protection against PMrelated visibility impairment (78 FR
82742–82744, December 18, 2020).
In further defining the characteristics
of a visibility index, the thenAdministrator next considered the
appropriate averaging time, form, and
level of the index. Given the available
scientific information the review, and in
considering the CASAC’s advice and
public comments, the thenAdministrator concluded that,
consistent with the decision in the 2012
review, a visibility index with a 24-hour
averaging time and a form based on the
3-year average of annual 90th percentile
values remained reasonable. With
regard to the averaging time and form of
such an index, the Administrator noted
analyses conducted in the last review
that demonstrated relatively strong
correlations between 24-hour and
subdaily (i.e., 4-hour average) PM2.5
light extinction (78 FR 3226, January 15,
2013), indicating that a 24-hour
averaging time is an appropriate
surrogate for the subdaily time periods
of the perception of PM-related
visibility impairment and the relevant
exposure periods for segments of the
viewing public. This decision in the
2020 review also recognized that a 24hour averaging time may be less
influenced by atypical conditions and/
or atypical instrument performance (78
FR 3226, January 15, 2013). The thenAdministrator recognized that there was
no new information to support updated
analyses of this nature, and therefore, he
believed these analyses continued to
provide support for consideration of a
24-hour averaging time for a visibility
index in this review. With regard to the
statistical form of the index, the
Administrator noted that, consistent
with the 2012 review: (1) A multi-year
percentile form offers greater stability
from the occasional effect of interannual
meteorological variability (78 FR 3198,
January 15, 2013; U.S. EPA, 2011, p. 4–
58); (2) a 90th percentile represents the
median of the distribution of the 20
percent worst visibility days, which are
targeted in Federal Class I areas by the
Regional Haze Program; and (3) public
preference studies did not provide
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
information to identify a different target
than that identified for Federal Class I
areas (U.S. EPA, 2011, p. 4–59).
Therefore, the then-Administrator
judged that a visibility index based on
estimates of light extinction, with a 24hour averaging time and a 90th
percentile form, averaged over three
years, remained appropriate (78 FR
82742–82744, December 18, 2020).
With regard to the level of a visibility
index, consistent with the 2012 review,
the then-Administrator judged that it
was appropriate to establish a target
level of protection of 30 deciviews
(dv),147 148 reflecting the upper end of
the range of visibility impairment
judged to be acceptable by at least 50%
of study participants in the available
public preference studies (78 FR 3226,
January 15, 2013). The 2011 PA
identified a range of levels from 20 to
30 dv based on the responses in the
public preference studies available at
that time (U.S. EPA, 2011, section 4.3.4).
At the time of the 2012 review, the thenAdministrator noted a number of
uncertainties and limitations in public
preference studies, including the small
number of stated preference studies
available, the relatively small number of
study participants, the extent to which
the study participants may not be
representative of the broader study area
population in some of the studies, and
the variations in the specific materials
and methods used in each study. In
considering the available preference
studies in 2012, with their inherent
uncertainties and limitations, the thenAdministrator concluded that the
substantial degree of variability and
uncertainty in the public preference
studies should be reflected in a target
level of protection based on the upper
end of the range of candidate protection
levels (CPLs).
Given that there were no new
preference studies in the 2019 ISA, the
then-Administrator’s judgments in 2020
were based on the same studies, with
the same range of levels, available in the
2012 review. The 2020 PA (U.S. EPA,
2020b, section 5.5), discussed a number
of limitations and uncertainties
associated with these studies. In
considering the scientific information,
with its uncertainties and limitations, as
well as public comments on the level of
the target level of protection against
visibility impairment, the then147 Deciview (dv) refers to a scale for
characterizing visibility that is defined directly in
terms of light extinction. The deciview scale is
frequently used in the scientific and regulatory
literature on visibility.
148 For comparison, 20 dv, 25 dv, and 30 dv are
equivalent to 64, 112, and 191 megameters (Mm¥1),
respectively.
PO 00000
Frm 00114
Fmt 4701
Sfmt 4700
Administrator concluded that it was
appropriate to again use a level of 30 dv
for the visibility index (78 FR 82742–
82744, December 18, 2020).
Having concluded that the protection
provided by a standard defined in terms
of a PM2.5 visibility index, with a 24hour averaging time, and a 90th
percentile form, averaged over 3 years,
set at a level of 30 dv, was requisite to
protect public welfare with regard to
visual air quality, the Administrator
next considered the degree of protection
from visibility impairment afforded by
the existing suite of secondary PM
standards.
In this context, the thenAdministrator considered the updated
analyses of visibility impairment
presented in the 2020 PA (U.S. EPA,
2020b, section 5.2.1.2), which reflected
a number of improvements since the
2012 review. Specifically, the updated
analyses examined multiple versions of
the IMPROVE equation, including the
version incorporating revisions since
the time of the 2012 review. These
updated analyses provided a further
understanding of how variation in the
inputs to the algorithms affect the
estimates of light extinction (U.S. EPA,
2020b, Appendix D). Additionally, for a
subset of monitoring sites with available
PM10–2.5 data, the updated analyses
better characterized the influence of
coarse PM on light extinction than in
the 2012 review (U.S. EPA, 2020b,
section 5.2.1.2).
The results of the updated analyses in
the 2020 PA were consistent with those
from the 2012 review. Regardless of
which version of the IMPROVE equation
was used, the analyses demonstrated
that, based on 2015–2017 data, the 3year visibility metric was at or below
about 30 dv in all areas meeting the
current 24-hour PM2.5 standard, and
below 25 dv in most of those areas. In
locations with available PM10–2.5
monitoring, which met both the current
24-hour secondary PM2.5 and PM10
standards, 3-year visibility index
metrics were at or below 30 dv
regardless of whether the coarse fraction
was included as an input to the
algorithm for estimating light extinction
(U.S. EPA, 2020b, section 5.2.1.2).
While the inclusion of the coarse
fraction had a relatively modest impact
on the estimates of light extinction, the
then-Administrator recognized the
continued importance of the PM10
standard given the potential for larger
impacts on light extinction in areas with
higher coarse particle concentrations,
which were not included in the analyses
in the 2020 PA due to a lack of available
data (U.S. EPA, 2019a, section 13.2.4.1;
U.S. EPA, 2020b, section 5.2.1.2). He
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
noted that the air quality analyses
showed that all areas meeting the
existing 24-hour PM2.5 standard, with its
level of 35 mg/m3, had visual air quality
at least as good as 30 dv, based on the
visibility index. Thus, the secondary 24hour PM2.5 standard would likely be
controlling relative to a 24-hour
visibility index set at a level of 30 dv.
Additionally, areas would be unlikely to
exceed the target level of protection for
visibility of 30 dv without also
exceeding the existing secondary 24hour PM2.5 standard. Thus, the thenAdministrator judged that the 24-hour
PM2.5 standard provided sufficient
protection in all areas against the effects
of visibility impairment, i.e., that the
existing 24-hour PM2.5 standard would
provide at least the target level of
protection for visual air quality of 30 dv
which he judged appropriate (78 FR
82742–82744, December 18, 2020).
2. Overview of Welfare Effects Evidence
The information summarized here is
based on the scientific assessment of the
welfare effects evidence available in this
reconsideration; this assessment is
documented in the 2019 ISA and ISA
Supplement and its policy implications
are further discussed in the 2022 PA.
While the 2019 ISA provides the broad
scientific foundation for this
reconsideration, additional literature
has become available since the cutoff
date of the 2019 ISA that expands the
body of evidence related to visibility
effects that can inform the
Administrator’s judgment on the
adequacy of the current secondary PM
standards. As such, the ISA Supplement
builds on the information in the 2019
ISA with a targeted identification and
evaluation of new scientific information
regarding visibility effects. As described
in the ISA Supplement and the 2022
PA, the selection of welfare effects to
evaluate within the ISA Supplement
were based on the causality
determinations reported in the 2019 ISA
and the subsequent use of scientific
evidence in the 2020 PA (U.S. EPA,
2019a, section 1.2; U.S. EPA, 2022a,
section 1.4.2). The ISA Supplement
focuses on U.S. and Canadian studies
that provide new information on public
preferences for visibility impairment
and/or developed new methodologies or
conducted quantitative analyses of light
extinction (U.S. EPA, 2022a, section
1.2). Such studies of visibility effects
and quantitative relationships between
visibility impairment and PM in
ambient air were considered to be of
greatest utility in informing the
Administrator’s conclusions on the
adequacy of the current secondary PM
standards. The visibility effects
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
evidence presented within the 2019
ISA, along with the targeted
identification and evaluation of new
scientific information in the ISA
Supplement, provides the scientific
basis for the reconsideration of the 2020
final decision on the secondary PM
standards for visibility effects. For
climate and materials effects, the 2020
PA concluded that there were
substantial uncertainties associated with
the quantitative relationships with PM
concentrations and the concentration
patterns that limited the ability to
quantitatively assess the public welfare
protection provided by the standards
from these effects. Therefore, the
evaluation of the information related to
these effects draws heavily from the
2019 ISA and 2020 PA. The subsections
below briefly summarize the nature of
PM-related visibility (section V.B.1.a),
climate (section V.B.1.b), and materials
(section V.B.1.c) effects.
a. Nature of Effects
Visibility impairment can have
implications for people’s enjoyment of
daily activities and for their overall
sense of well-being (U.S. EPA, 2009a,
section 9.2). The strongest evidence for
PM-related visibility impairment comes
from the fundamental relationship
between light extinction and PM mass
(U.S. EPA, 2009a), which confirms a
well-established ‘‘causal relationship
exists between PM and visibility
impairment’’ (U.S. EPA, 2009a, p. 2–28).
Beyond its effects on visibility, the 2009
ISA also identified a causal relationship
‘‘between PM and climate effects,
including both direct effects of radiative
forcing and indirect effects that involve
cloud and feedbacks that influence
precipitation formation and cloud
lifetimes’’ (U.S. EPA, 2009a, p. 2–29).
The evidence also supports a causal
relationship between PM and effects on
materials, including soiling effects and
materials damage (U.S. EPA, 2009a, p.
2–31).
The evidence available in this
reconsideration is consistent with the
evidence available at the time of the
2012 and 2020 reviews and supports the
conclusions of causal relationships
between PM and visibility, climate, and
materials effects (U.S. EPA, 2019a,
chapter 13). Evidence newly available in
this reconsideration augments the
previously available evidence of the
relationship between PM and visibility
impairment (U.S. EPA, 2019a, section
13.2; U.S. EPA, 2022a, section 4),
climate effects (U.S. EPA, 2019a, section
13.3), and materials effects (U.S. EPA,
2019a, section 13.4).
PO 00000
Frm 00115
Fmt 4701
Sfmt 4700
16315
i. Visibility
The fundamental relationship
between light extinction and PM mass,
and the EPA’s understanding of this
relationship, has changed little since the
2009 ISA (U.S. EPA, 2009a). The
combined effect of light scattering and
absorption by particles and gases is
characterized as light extinction, i.e., the
fraction of light that is scattered or
absorbed per unit of distance in the
atmosphere.149 Light extinction is
measured in units of 1/distance, which
is often expressed in the technical
literature as visibility per megameter
(abbreviated Mm¥1). Higher values of
light extinction (usually given in units
of Mm¥1 or dv) correspond to lower
visibility. When PM is present in the air,
its contribution to light extinction is
typically much greater than that of gases
(U.S. EPA, 2019a, section 13.2.1). The
impact of PM on light scattering
depends on particle size and
composition, as well as relative
humidity. All particles scatter light, as
described by the Mie theory, which
relates light scattering to particle size,
shape, and index of refraction (U.S.
EPA, 2019a, section 13.2.3; Mie, 1908,
Van de Hulst, 1981). Fine particles
scatter more light than coarse particles
on a per unit mass basis and include
sulfates, nitrates, organics, lightabsorbing carbon, and soil (Malm et al.,
1994). Hygroscopic particles like
ammonium sulfate, ammonium nitrate,
and sea salt increase in size as relative
humidity increases, leading to increased
light scattering (U.S. EPA, 2019a,
section 13.2.3).
As at the time of the 2012 and 2020
reviews, direct measurements of PM
light extinction, scattering, and
absorption continue to be considered
more accurate for quantifying visibility
than PM mass-based estimates because
measurements do not depend on
assumptions about particle
characteristics (e.g., size, shape, density,
component mixture, etc.) (U.S. EPA,
2019a, section 13.2.2.2). Measurements
of light extinction can be made with
high time resolution, allowing for
characterization of subdaily temporal
patterns of visibility impairment. A
number of measurement methods have
been used for visibility impairment (e.g.,
149 All particles scatter light and, although a
larger particle scatters more light than a similarly
shaped smaller particle of the same composition,
the light scattered per unit of mass is greatest for
particles with diameters from ∼0.3–1.0 mm (U.S.
EPA, 2009a, section 2.5.1; U.S. EPA, 2019a, section
13.2.1). Particles with hygroscopic components
(e.g., particulate sulfate and nitrate) contribute more
to light extinction at higher relative humidity than
at lower relative humidity because they change size
in the atmosphere in response to relative humidity.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16316
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
transmissometers, integrating
nephelometers, teleradiometers,
telephotometers, and photography and
photographic modeling), although each
of these methods has its own strengths
and limitations (U.S. EPA, 2019a, Table
13–1). While some recent research
confirms and adds to the body of
knowledge regarding direct
measurements as is described in the
2019 ISA and ISA Supplement, no
major new developments have been
made with these measurement methods
since prior reviews (U.S. EPA, 2019a,
section 13.2.2.2; U.S. EPA, 2022a,
section 4.2).
In the absence of a robust monitoring
network for the routine measurement of
light extinction across the U.S.,
estimation of light extinction based on
existing PM monitoring can be used.
The theoretical relationship between
light extinction and PM characteristics,
as derived from Mie theory (U.S. EPA,
2019a, Equation 13.5), can be used to
estimate light extinction by combining
mass scattering efficiencies of particles
with particle concentrations (U.S. EPA,
2019a, section 13.2.3; U.S. EPA, 2009a,
sections 9.2.2.2 and 9.2.3.1). This
estimation of light extinction is
consistent with the method used in
previous reviews. The algorithm used to
estimate light extinction, known as the
IMPROVE algorithm,150 provides for the
estimation of light extinction (bext), in
units of Mm 1, using routinely
monitored components of fine (PM2.5)
and coarse (PM10–2.5) PM. Relative
humidity data are also needed to
estimate the contribution by liquid
water that is in solution with the
hygroscopic components of PM. To
estimate each component’s contribution
to light extinction, their concentrations
are multiplied by extinction coefficients
and are additionally multiplied by a
water growth factor that accounts for
their expansion with moisture. Both the
extinction efficiency coefficients and
water growth factors of the IMPROVE
algorithm have been developed by a
combination of empirical assessment
and theoretical calculation using
particle size distributions associated
with each of the major aerosol
components (U.S. EPA, 2019a, sections
13.2.3.1 and 13.2.3.3).
At the time of the 2012 review, two
versions of the IMPROVE algorithm
were available in the literature—the
150 The algorithm is referred to as the IMPROVE
algorithm as it was developed specifically to use
monitoring data generated at IMPROVE network
sites and with equipment specifically designed to
support the IMPROVE program and was evaluated
using IMPROVE optical measurements at the subset
of monitoring sites that make those measurements
(Malm et al., 1994).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
original IMPROVE algorithm
(Lowenthal and Kumar, 2004, Malm and
Hand, 2007, Ryan et al., 2005) and the
revised IMPROVE algorithm (Pitchford
et al., 2007). As described in detail in
the 2022 PA (U.S. EPA, 2022b, section
5.3.1.1) and the 2019 ISA (U.S. EPA,
2019a, section 13.2.3), the algorithm has
been further evaluated and refined since
the time of the 2012 review (Lowenthal
and Kumar, 2016), particularly for PM
characteristics and relative humidity in
remote areas. All three versions of the
IMPROVE algorithm were considered in
evaluating visibility impairment in this
reconsideration.
Consistent with the evidence
available at the time of the 2012 and
2020 reviews, our understanding of
public perception of visibility
impairment comes from visibility
preference studies conducted in four
areas in North America.151 The detailed
methodology for these studies are
described in the 2022 PA (U.S. EPA,
2022b, section 5.3.1.1), the 2019 ISA
(U.S. EPA, 2019a), and the 2009 ISA
(U.S. EPA, 2019a). In summary, the
study participants were queried
regarding multiple images that were
either photographs of the same location
and scenery that had been taken on
different days on which measured
extinction data were available or
digitized photographs onto which a
uniform ‘‘haze’’ had been
superimposed. Results of the studies
indicated a wide range of judgments on
what study participants considered to
be acceptable visibility across the
different study areas, depending on the
setting depicted in each photograph.
Based on the results of the four cities,
a range encompassing the PM2.5
visibility index values from images that
were judged to be acceptable by at least
50 percent of study participants across
all four of the urban preference studies
was identified (U.S. EPA, 2010b, p. 4–
24; U.S. EPA, 2020b, Figure 5–2). Much
lower visibility (considerably more haze
resulting in higher values of light
extinction) was considered acceptable
in Washington, DC, than was in Denver,
and 30 dv reflected the level of
impairment that was determined to be
‘‘acceptable’’ by at least 50 percent of
study participants (78 FR 3226–3227,
January 15, 2013).
Since the completion of the 2009 and
2019 ISAs, there has been only one
public preference study that has become
available in the U.S. This study uses
151 Preference studies were available in four
urban areas in the last review: Denver, Colorado
(Ely et al., 1991), Vancouver, British Columbia,
Canada (Pryor, 1996), Phoenix, Arizona (BBC
Research & Consulting, 2003), and Washington, DC
(Abt Associates, 2001; Smith and Howell, 2009).
PO 00000
Frm 00116
Fmt 4701
Sfmt 4700
images of the Grand Canyon, AZ,
described in the ISA Supplement (U.S.
EPA, 2022a). The Grand Canyon study,
conducted by Malm et al. (2019), has a
similar study design to that used in the
public preference studies discussed
above; however, there are several
important differences that make it
difficult to directly compare the results
of the Malm et al. (2019) study with
other public preference studies. As an
initial matter, the Grand Canyon study
was conducted in a Federal Class I area,
as opposed to in an urban area, with a
scene depicted in the photographs that
did not include urban features.152 We
recognize that public preferences with
respect to visibility in Federal Class 1
areas may well differ from visibility
preferences in urban areas and other
contexts, although there is currently a
lack of information to on such
questions. Further, the Malm et al.
(2019) study also used a much lower
range of superimposed ‘‘haze’’ than the
preference studies discussed above.153 It
is unclear whether the participant
preferences are a function in part of the
range of potential values presented,
such that the participant preferences for
the Grand Canyon were generally
lower 154 than the other preference
studies in part because of the lower
range of superimposed ‘‘haze’’ for the
images in that study, or if their
preferences would vary if presented
with images with a range of
superimposed ‘‘haze’’ more comparable
to the levels used in the other studies
(i.e., more ‘‘haze’’ superimposed on the
images).
The Malm et al. (2019) study also
explored alternate methods for
evaluating ‘‘acceptable’’ levels of visual
air quality from the preference studies,
including the use of scene-specific
visibility indices as potential indicators
of visibility levels as perceived by the
observer (Malm et al., 2019). In addition
to measures of atmospheric haze, such
152 The Grand Canyon study used a single scene
looking west down the canyon with a small
landscape feature of a 100-km-distant mountain
(Mount Trumbull), along with other closer
landscape features. The scenes presented in the
previously available visibility preference studies are
presented in more detail in Table D–9 in the 2022
PA (U.S. EPA, 2022b, Appendix D).
153 The Grand Canyon study superimposed light
extinction ranging from 3 dv to 20 dv on the image
slides shown to participants compared to the
previously available preference studies. In those
studies, the visibility ranges presented were as low
as 9 dv and as high as 45 dv. The visibility ranges
presented in the previously available visibility
preference studies are described in more detail in
Table D–9 in the 2022 PA (U.S. EPA, 2022b,
Appendix D).
154 In the Grand Canyon study, the level of
impairment that was determined to be ‘‘acceptable’’
by at least 50 percent of study participants was 7
dv (Malm et al., 2019).
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
as atmospheric extinction, used in
previously available preference studies,
other indices for visual air quality
include color and achromatic contrast of
single landscape figures, average and
equivalent contrast of an entire scene,
edge detection algorithms such as the
Sobel index, and just-noticeable
difference or change indexes. The
results reported by Malm et al. (2019)
suggest that scene-dependent metrics,
such as contrast, may be useful alternate
predictors of preference levels
compared to universal metrics like light
extinction (U.S. EPA, 2022a, section
4.2.1). This is because extinction alone
is not a measure of ‘‘haze,’’ but of light
attenuation per unit distance, and
visible ‘‘haze’’ is dependent on both
light extinction and distance to a
landscape feature (U.S. EPA, 2022a,
section 4.2.1). However, there are very
few studies available that use scenedependent metrics (i.e., contrast) to
evaluate public preference information,
which makes it difficult to evaluate
them as an alternative to the light
extinction approach.
ii. Climate
ddrumheller on DSK120RN23PROD with RULES3
The available evidence continues to
support the conclusion of a causal
relationship between PM and climate
effects (U.S. EPA, 2019a, section 13.3.9).
Since the 2012 review, climate impacts
have been extensively studied and
recent research reinforces and
strengthens the evidence evaluated in
the 2009 ISA. Recent evidence provides
greater specificity about the details of
radiative forcing effects 155 and
increases the understanding of
additional climate impacts driven by
PM radiative effects. The
Intergovernmental Panel on Climate
Change (IPCC) assesses the role of
anthropogenic activity in past and
future climate change, and since the
completion of the 2009 ISA, has issued
the Fifth IPCC Assessment Report (AR5;
IPCC, 2013), which summarizes any key
scientific advances in understanding the
climate effects of PM since the previous
report. As in the 2009 ISA, the 2019 ISA
draws substantially on the IPCC report
to summarize climate effects. As
155 Radiative forcing (RF) for a given atmospheric
constituent is defined as the perturbation in net
radiative flux, at the tropopause (or the top of the
atmosphere) caused by that constituent, in watts per
square meter (Wm 2), after allowing for
temperatures in the stratosphere to adjust to the
perturbation but holding all other climate responses
constant, including surface and tropospheric
temperatures (Fiore et al., 2015; Myhre et al., 2013).
A positive forcing indicates net energy trapped in
the Earth system and suggests warming of the
Earth’s surface, whereas a negative forcing indicates
net loss of energy and suggests cooling (U.S. EPA,
2019a, section 13.3.2.2).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
discussed in more detail in the 2022 PA
(U.S. EPA, 2022b, section 5.3.2.1.1), the
general conclusions are similar between
the IPCC AR4 and AR5 reports with
regard to effects of PM on global
climate. Consistent with the evidence
available in the 2012 review, the key
components, including sulfate, nitrate,
organic carbon (OC), black carbon (BC),
and dust, that contribute to climate
processes vary in their reflectivity,
forcing efficiencies, and direction of
forcing. Since the completion of the
2009 ISA, the evidence base has
expanded with respect to the
mechanisms of climate responses and
feedbacks to PM radiative forcing;
however, the recently published
literature assessed in the 2019 ISA does
not reduce the considerable
uncertainties that continue to exist
related these mechanisms.
As described in the proposal (88 FR
5650, January 27, 2023), PM has a very
heterogeneous distribution globally and
patterns of forcing tend to correlate with
PM loading, with the greatest forcings
centralized over continental regions.
The climate response to this PM forcing,
however, is more complicated since the
perturbation to one climate variable
(e.g., temperature, cloud cover,
precipitation) can lead to a cascade of
effects on other variables. While the
initial PM radiative forcing may be
concentrated regionally, the eventual
climate response can be much broader
spatially or be concentrated in remote
regions, and may be quite complex,
affecting multiple climate variables with
possible differences in the direction of
the forcing in different regions or for
different variables (U.S. EPA, 2019a,
section 13.3.6). The complex climate
system interactions lead to variation
among climate models, which have
suggested a range of factors that can
influence large-scale meteorological
processes and may affect temperature,
including local feedback effects
involving soil moisture and cloud cover,
changes in the hygroscopicity of the PM,
and interactions with clouds (U.S. EPA,
2019a, section 13.3.7). As a result, there
remains insufficient evidence to related
climate effects to specific PM levels in
ambient air or to establish a quantitative
relationship between PM and climate
effects, particularly at a regional scale.
Further research is needed to better
characterize the effects of PM on
regional climate in the U.S. before PM
climate effects can be quantified.
iii. Materials
Consistent with the evidence assessed
in the 2009 ISA, the available evidence
continues to support the conclusion that
there is a causal relationship between
PO 00000
Frm 00117
Fmt 4701
Sfmt 4700
16317
PM deposition and materials effects.
Effects of deposited PM, particularly
sulfates and nitrates, to materials
include both physical damage and
impaired aesthetic qualities, generally
involving soiling and/or corrosion (U.S.
EPA, 2019a, section 13.4.2). Because of
their electrolytic, hygroscopic, and
acidic properties and their ability to
sorb corrosive gases, particles contribute
to materials damage by adding to the
effects of natural weathering processes,
by potentially promoting or accelerating
the corrosion of metals, degradation of
painted surfaces, deterioration of
building materials, and weakening of
material components.156 There is a
limited amount of recently available
data for consideration in this review
from studies primarily conducted
outside of the U.S. on buildings and
other items of cultural heritage.
However, these studies involved
concentrations of PM in ambient air
greater than those typically observed in
the U.S. (U.S. EPA, 2019a, section 13.4).
Building on the evidence available in
the 2009 ISA, and as described in detail
in the proposal (88 FR 5650, January 27,
2023) and in the 2019 ISA (U.S. EPA,
2019a, section 13.4), research has
progressed on (1) the theoretical
understanding of soiling of items of
cultural heritage; (2) the quantification
of degradation rates and further
characterization of factors that influence
damage of stone materials; (3) materials
damage from PM components besides
sulfate and black carbon and
atmospheric gases besides SO2; (4)
methods for evaluating soiling of
materials by PM mixtures; (5) PMattributable damage to other materials,
including glass and photovoltaic panels;
(6) development of dose-response
relationships for soiling of building
materials; and (7) damage functions to
quantify material decay as a function of
pollutant type and load. While the
evidence of PM-related materials effects
has expanded somewhat since the
completion of the 2009 ISA, there
remains insufficient evidence to relate
soiling or damage to specific PM levels
in ambient air or to establish a
quantitative relationship between PM
and materials degradation. The recent
evidence assessed in the 2019 ISA is
generally similar to the evidence
available in the 2009 ISA, including
156 As discussed in the 2019 ISA (U.S. EPA,
2019a, section 13.4.1), corrosion typically involves
reactions of acidic PM (i.e., acidic sulfate or nitrate)
with material surfaces, but gases like SO2 and nitric
acid (HNO3) also contribute. Because ‘‘the impacts
of gaseous and particulate N and S wet deposition
cannot be clearly distinguished’’ (U.S. EPA, 2019a,
p. 13–1), the assessment of the evidence in the 2019
ISA considers the combined impacts.
E:\FR\FM\06MRR3.SGM
06MRR3
16318
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
associated limitations and uncertainties
and a lack of evidence to inform
quantitative relationships between PM
and materials effects, therefore leading
to similar conclusions about the PMrelated effects on materials.
ddrumheller on DSK120RN23PROD with RULES3
3. Summary of Air Quality and
Quantitative Information
Beyond the consideration of the
scientific evidence, as discussed in
section V.A.2 above, quantitative
analyses of PM air quality, when
available, can also inform conclusions
on the adequacy of the public welfare
protection provided by the current
secondary PM standards.
a. Visibility Effects
In the 2012 and 2020 reviews,
quantitative analyses for PM-related
visibility effects focused on daily
visibility impairment, given the shortterm nature of PM-related visibility
effects. The evidence and information
available in this reconsideration
continues to provide support for the
short-term (i.e., hourly or daily) nature
of PM-related visibility impairment. As
such, the quantitative analyses
presented in the 2022 PA continue to
focus on daily visibility impairment and
utilize a two-phase assessment approach
for visibility impairment, consistent
with the approaches taken in past
reviews. First, the 2022 PA considers
the appropriateness of the elements
(indicator, averaging time, form, and
level) of the visibility index for
providing protection against PM-related
visibility effects. Second, recent air
quality was used to evaluate the
relationship between the current
secondary 24-hour PM2.5 standard and
the visibility index. The information
available since the 2012 review includes
an updated equation for estimating light
extinction, summarized in the 2022 PA
(U.S. EPA, 2022b, section 5.3.1.1) and
described in the 2019 ISA (U.S. EPA,
2019a, section 13.2.3.3), as well as more
recent air monitoring data, that together
allow for development of an updated
assessment of PM-related visibility
impairment in study locations in the
U.S.
i. Target Level of Protection in Terms
of a PM2.5 Visibility Index
In evaluating the adequacy of the
current secondary PM standards, the
2022 PA first evaluates the
appropriateness of the elements
(indicator, averaging time, form, and
level) identified for a visibility index to
protect against visibility effects. In
previous reviews, the visibility index as
set at a level of 30 dv, with estimated
light extinction as the indicator, a 24hour averaging time, and a 90th
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
percentile form, averaged over three
years.
With regard to an indicator for the
visibility index, the 2022 PA recognizes
the lack of availability of methods and
an established network for directly
measuring light extinction (U.S. EPA,
2022b, section 5.3.1.1). Therefore,
consistent with previous reviews, the
2022 PA concludes that a visibility
index based on estimates of light
extinction by PM2.5 components derived
from an adjusted version of the original
IMPROVE algorithm to be the most
appropriate indicator for the visibility
index in this reconsideration. As
described in section 5.3.1.1 of the 2022
PA, the IMPROVE algorithm estimates
light extinction using routinely
monitored components of PM2.5 and
PM10–2.5, along with estimates of relative
humidity (U.S. EPA, 2022b, section
5.3.1.1).
With regard to averaging time, the
2022 PA notes that the evidence
continues to provide support for the
short-term nature of PM-related
visibility effects. Given that there is no
new information available regarding the
time periods during which visibility
impairment occurs or public preferences
related to specific time periods for
visibility impairment, the 2022 PA
concludes that it is appropriate to
continue to focus on daily visibility
impairment. In so doing, the 2022 PA
relies on analyses that were conducted
in the 2012 review that showed
relatively strong correlations between
24-hour and subdaily (i.e., 4-hour
average) PM2.5 light extinction that
indicated that a 24-hour averaging time
is an appropriate surrogate for the
subdaily time periods relevant for visual
perception (U.S. EPA, 2011, Figures G–
4 and G–5; Frank, 2012). These analyses
continue to provide support for a 24hour averaging time for the visibility
index in this reconsideration. Consistent
with previous reviews, the 2022 PA also
notes that the 24-hour averaging time
may be less influenced by atypical
conditions and/or atypical instrument
performance than a subdaily averaging
time (85 FR 82740, December 18, 2020;
78 FR 3226, January 15, 2013).
With regard to the form for the
visibility index, the available
information continues to provide
support for a 3-year average of annual
90th percentile values. Given that there
is no new information to inform
selection of an alternate form, as in
previous reviews, the 2022 PA notes
that the 3-year average form provides
stability from the occasional effect of
inter-annual meteorological variability
that can result in unusually high
pollution levels for a particular year (85
PO 00000
Frm 00118
Fmt 4701
Sfmt 4700
FR 82741, December 18, 2020; 78 FR
3198, January 15, 2013; U.S. EPA, 2011,
p. 4–58). In so doing, the 2022 PA
considers the evaluation in the 2010
Urban-Focused Visibility Assessment
(UFVA) of three different statistical
forms: 90th, 95th, and 98th percentiles
(U.S. EPA, 2010b, Chapter 4).). In
considering this evaluation of statistical
forms from the 2010 UFVA, consistent
with the 2011 PA, the 2022 PA notes
that the Regional Haze Program targets
the 20 percent most impaired days for
visibility improvements in visual air
quality in Federal Class I areas and that
the median of the distribution of these
20 percent most impaired days would
be the 90th percentile. The 2011 PA also
noted that strategies that are
implemented so that 90 percent of days
would have visual air quality that is at
or below the level of the visibility index
would reasonably be expected to lead to
improvements in visual air quality for
the 20 percent most impaired days.
Additionally, as in the 2011 PA, the
2022 PA recognizes that the available
public preference studies do not address
frequency of occurrence of different
levels of visibility (U.S. EPA, 2022b,
section 5.3.1.2). Therefore, the analyses
and consideration for the form of a
visibility index from the 2011 PA
continue to provide support for a 90th
percentile form, averaged across three
years, in defining the characteristics of
a visibility index in this
reconsideration.
With regard to the level for the
visibility index, the 2022 PA recognizes
that there is an additional public
preference study (Malm et al., 2019)
available in this reconsideration. As
noted above, however, this study differs
from the previously available public
preference studies in several ways,
which makes it difficult to integrate this
newly available study with the
previously available studies. Most
significantly, this study was evaluated
public preferences for visibility in the
Grand Canyon, perhaps the most
notable Class I area in the country for
visibility purposes. Therefore, the 2022
PA concludes that the Grand Canyon
study is not directly comparable to the
other available preferences studies and
public preferences of visibility
impairment in the Malm et al. (2019)
study are not appropriate to consider in
identifying a range of levels for the
target level of protection against
visibility impairment for this
reconsideration of the secondary PM
NAAQS.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
Therefore, the 2022 PA continues to
rely on the same studies 157 and the
range of 20 to 30 dv identified from
those studies in previous reviews. With
regard to selecting the appropriate target
level of protection for visibility
impairment within this range, the 2022
PA notes that in previous reviews, a
level at the upper end of the range (i.e.,
30 dv) was selected given the
uncertainties and limitations associated
with the public preference studies (U.S.
EPA, 2022b, section 5.3.1.1). However,
the 2022 PA also recognizes that (1) the
degree of protection provided by a
secondary PM NAAQS is not
determined solely by any one element of
the standard but by all elements (i.e.,
indicator, averaging time, form, and
level) being considered together, and (2)
decisions regarding the adequacy of the
current secondary standards is a public
welfare policy judgment to be made by
the Administrator. As such, the
Administrator may judge that a target
level of protection below the upper end
of the range (i.e., less than 30 dv) is
appropriate, depending on his public
welfare policy judgments, which draw
upon the available scientific evidence
for PM-related visibility effects and on
analyses of visibility impairment, as
well as judgments about the appropriate
weight to place on the range of
uncertainties inherent in the evidence
and analyses.
In considering the available public
preference studies, consistent with past
reviews, the 2022 PA concludes that it
is reasonable to consider a range of 20
to 30 dv for selecting a target level of
protection, including a high value of 30
dv, a midpoint value of 25 dv, and a low
value of 20 dv. A target level of
protection at or in the upper end of the
range would focus on the Washington,
DC, preference study results (Abt
Associates, 2001; Smith and Howell,
2009), which identified 30 dv as the
level of impairment that was
determined to be ‘‘acceptable’’ by at
least 50 percent of study participants.
The public preferences of visibility
impairment in the Washington, DC,
study are likely to be generally
representative of urban areas that do not
have valued scenic elements (e.g.,
mountains) in the distant background.
This would be more representative of
areas in the middle of the country and
many areas in the eastern U.S., as well
157 As noted above, the available public
preference studies include those conducted in
Denver, Colorado (Ely et al., 1991), Vancouver,
British Columbia, Canada (Pryor, 1996), Phoenix,
Arizona (BBC Research & Consulting, 2003), and
Washington, DC (Abt Associates, 2001; Smith and
Howell, 2009).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
as possibly some areas in the western
U.S.
A target level of protection in the
middle of the range would be most
closely associated with the level of
impairment that was determined to be
‘‘acceptable’’ by at least 50 percent of
study participants in the Phoenix, AZ,
study (BBC Research & Consulting,
2003), which was 24 dv. This study,
while methodologically similar to the
other public preference studies,
included participants that were selected
as a representative sample of the
Phoenix area population 158 and used
computer-generated images to depict
specific uniform visibility impairment
conditions. This study yielded the best
results of the four public preference
studies in terms of the least noisy
preference results and the most
representative selection of participants.
Therefore, based on this study, the use
of 25 dv to represent a midpoint within
the range of target levels protection is
well supported.
A target level of protection at or just
above the lower end of the range would
focus on the Denver, CO, study, but may
not be as strongly supported as higher
levels within the range (Ely et al., 1991).
Older studies, such as those conducted
in Denver, CO (Ely et al., 1991), and
British Columbia, Canada (Pryor, 1996),
used photographs that were taken at
different times of the day and on
different days to capture a range of light
extinction levels needed for the
preference studies. Compared to studies
that used computer-generated images
(i.e., those in Phoenix, AZ, and
Washington, DC) there was more
variability in scene appearance in these
older studies that could affect
preference rating and includes
uncertainties associated with using
ambient measurements to represent
sight path-averaged light extinction
values rather than superimposing a
computer-generated amount of haze
onto the images. When using
photographs, the intrinsic appearance of
the scene can change due to
meteorological conditions (i.e., shadow
patterns and cloud conditions) and
spatial variations in ambient air quality
that can result in ambient light
158 The other preference studies did not include
populations that were necessarily representative of
the population in the area for which the images
being judged. For example, in the Denver, CO,
study, participants were from intact groups (i.e.,
those who were meeting for other reasons) and were
asked to provide a period of time during a regularly
scheduled meeting to participate in the study (Ely
et al., 1991). As another example, in the British
Columbia, Canada, study, participants were
recruited from undergraduate and graduate students
enrolled in classes at the University of British
Columbia’s Department of Geography (Pryor, 1996).
PO 00000
Frm 00119
Fmt 4701
Sfmt 4700
16319
extinction measurement not being
representative of the sight-path-averaged
light extinction. Computer-generated
images, such as those generated with
WinHaze, do not introduce such
uncertainties, as the same base
photograph is used (i.e., there is no
intrinsic change in scene appearance)
and the modeled haze that is
superimposed on the photograph is
determined based on uniform light
extinction throughout the scene.
In addition to differences in
preferences that may arise from
photographs versus computer-generated
images, urban visibility preference may
differ by location, and such differences
may arise from differences in the
cityscape scene that is depicted in the
images. These differences are related to
the perceived value of objects and
scenes that are included in the image, as
objects at a greater distance have a
greater sensitivity to perceived visibility
changes as light extinction is changed
compared to similar scenes with objects
at shorter distances. For example, a
person (regardless of their location)
evaluating visibility in an image with
more scenic elements such as
mountains or natural views may value
better visibility conditions in these
images compared to the same level of
visibility impairment in an image that
only depicts urban features such as
buildings and roads. That is, if a person
was shown the same level of visibility
impairment in two images depicting
different scenes—one with mountains in
the background and urban features in
the foreground and one with no
mountains in the background and
nearby buildings in the image without
mountains in the distance—may find
the amount of haze to be unacceptable
in the image with the mountains in the
distance because of a greater perceived
value of viewing the mountains, while
finding the amount of haze to be
acceptable in the image with the
buildings because of a lesser value of
viewing the cityscape or an expectation
that such urban areas may generally
have higher levels of haze in general.
This is consistent when comparing the
differences between the Denver, CO,
study results (which found the 50%
acceptance criteria occurred at the best
visual air quality levels among the four
cities) and the Washington, DC, results
(which found the 50% acceptability
criteria occurred at the worst visual air
quality levels among the four cities).
These results may occur because the
most prominent and picturesque feature
of the cityscape of Denver is the visible
snow-covered mountains in the
distance, while the prominent and
E:\FR\FM\06MRR3.SGM
06MRR3
16320
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
picturesque features of the Washington,
DC, cityscape are buildings relatively
nearby without prominent and/or
valued scenic features that are more
distant. Given these variabilities in
preferences it is unclear to what extent,
the available evidence provides strong
support for a target level of protection
at the lower end of the range. Future
studies that reduce sources of noisiness
and uncertainty in the results could
provide more information that would
support selection of a target level of
protection at or just above the lower end
of the range.
Taken together, the 2022 PA
concludes that available information
continues to support a visibility index
with estimated light extinction as the
indicator, a 24-hour averaging time, and
a 90th percentile form, averaged over
three years, with a level within the
range of 20 to 30 dv.
ddrumheller on DSK120RN23PROD with RULES3
ii. Relationship Between the PM2.5
Visibility Index and the Current
Secondary 24-Hour PM2.5 Standard
The 2022 PA presents quantitative
analyses based on recent air quality that
evaluate the relationship between recent
air quality and calculated light
extinction. As in previous reviews,
these analyses explored this
relationship as an estimate of visibility
impairment in terms of the 24-hour
PM2.5 standard and the visibility index.
Generally, the results of the updated
analyses are similar to those based on
the data available at the time of the 2012
and 2020 reviews (U.S. EPA, 2022b,
section 5.3.1.2). As discussed in section
V.C.1.a above, the 2022 PA concludes
that the available evidence continues to
support a visibility index with
estimated light extinction as the
indicator, a 24-hour averaging time, and
a 90th percentile form, averaged over
three years, with a level within the
range of 20 to 30 dv. These analyses
evaluate visibility impairment in the
U.S. under recent air quality conditions,
particularly those conditions that meet
the current standards, and the relative
influence of various factors on light
extinction. Given the relationship of
visibility with short-term PM, we focus
particularly on the short-term PM
standards.159 Compared to the 2012
159 The analyses presented in the 2022 PA focus
on the visibility index and the current secondary
24-hour PM2.5 standard with a level of 35 mg/m3.
However, we recognize that all three secondary PM
standards influence the PM concentrations
associated with the air quality distribution. As
noted in section V.A.1 above, the current secondary
PM standards include the 24-hour PM2.5 standard,
with its level of 35 mg/m3, the annual PM2.5
standard, with its level of 15.0 mg/m3, and the 24hour PM10 standard, with its level of 150 mg/m3.
With regard to the annual PM2.5 standard, we note
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
review, updated analyses incorporate
several refinements, including (1) the
evaluation of three versions of the
IMPROVE equation to calculate light
extinction (U.S. EPA, 2022b, Appendix
D, Equations D–1 through D–3) in order
to better understand the influence of
variability in equation inputs; 160 (2) the
use of 24-hour relative humidity data,
rather than monthly average relative
humidity as was used in the 2012
review (U.S. EPA, 2022b, section
5.3.1.2, Appendix D); and (3) the
inclusion of the coarse fraction in the
estimation of light extinction (U.S. EPA,
2022b, section 5.3.1.2, Appendix D).
The analyses in the reconsideration are
updated from the 2012 and 2020
reviews and include 60 monitoring sites
that measure PM2.5 and PM10 and are
geographically distributed across the
U.S. in both urban and rural areas (U.S.
EPA, 2022b, Appendix D, Figure D–1).
When light extinction was calculated
using the revised IMPROVE equation, in
areas that meet the current 24-hour
PM2.5 standard for the 2017–2019 time
period, all sites have light extinction
estimates at or below 26 dv (U.S. EPA,
2022b, Figure 5–3). For the four
locations that exceed the current 24hour PM2.5 standard, light extinction
estimates range from 22 dv to 27 dv
(U.S. EPA, 2022b, Figure 5–3). These
findings are consistent with the findings
of the analyses using the same
IMPROVE equation in the 2012 review
with data from 102 sites with data from
2008–2010 and in the 2020 review with
data from 67 sites with data from 2015–
2017. The analyses presented in the
2022 PA indicate similar findings to
those from the analyses in the 2012 and
2020 reviews, i.e., the updated
quantitative analysis shows that the 3year visibility metric was no higher than
30 dv 161 at sites meeting the current
that all 60 areas included in the analyses meet the
current secondary annual PM standard (U.S. EPA,
2022b, Table D–7).
160 While the PM
2.5 monitoring network has an
increasing number of continuous FEM monitors
reporting hourly PM2.5 mass concentrations, there
continue to be data quality uncertainties associated
with providing hourly PM2.5 mass and component
measurements that could be input into IMPROVE
equation calculations for subdaily visibility
impairment estimates. As detailed in the 2022 PA,
there are uncertainties associated with the precision
and bias of 24-hour PM2.5 measurements (U.S. EPA,
2022b, p. 2–18), as well as to the fractional
uncertainty associated with 24-hour PM component
measurements (U.S. EPA, 2022b, p. 2–21). Given
the uncertainties present when evaluating data
quality on a 24-hour basis, the uncertainty
associated with subdaily measurements may be
even greater. Therefore, the inputs to these light
extinction calculations are based on 24-hour
average measurements of PM2.5 mass and
components, rather than subdaily information.
161 A 3-year visibility metric with a level of 30 dv
would be at the upper end of the range of levels
identified from the public preference studies.
PO 00000
Frm 00120
Fmt 4701
Sfmt 4700
secondary PM standards, and at most
such sites the 3-year visibility index
values are much lower (e.g., an average
of 20 dv across the 60 sites).162
When light extinction was calculated
using the revised IMPROVE equation,163
the resulting 3-year visibility metrics are
nearly identical to light extinction
estimates calculated using the original
IMPROVE equation (U.S. EPA, 2022b,
Figure 5–4), but some sites are just
slightly higher. Using the revised
IMPROVE equation, for those sites that
meet the current 24-hour PM2.5
standard, the 3-year visibility metric is
at or below 26 dv. For the four locations
that exceed the current 24-hour PM2.5
standard, light extinction estimates
range from 22 dv to 29 dv (U.S. EPA,
2022b, Figure 5–4). These results are
similar to those for light extinction
calculated using the original IMPROVE
equation,164 and those from previous
reviews.
When light extinction was calculated
using the refined equation from
Lowenthal and Kumar (2016), the
resulting 3-year visibility metrics are
slightly higher at all sites compared to
light extinction estimates calculated
using the original IMPROVE equation
(U.S. EPA, 2022b, Figure 5–5).165 These
higher estimates are to be expected,
given the higher OC multiplier included
in the IMPROVE equation from
Lowenthal and Kumar (2016), which
reflects the use of data from remote
areas with higher concentrations of
organic PM when validating the
equation. As such, it is important to
note that the Lowenthal and Kumar
(2016) version of the equation may
overestimate light extinction in nonremote areas, including the urban areas
in the updated analyses in this
reconsideration.
Nevertheless, when light extinction is
calculated using the Lowenthal and
162 When light extinction is calculated using the
original IMPROVE equation, all 60 sites have 3-year
visibility metrics below 30 dv, 58 sites are at or
below 25 dv, and 26 sites are at or below 20 dv (see
U.S. EPA, 2022b, Appendix D, Table D–3).
163 As described in more detail in the 2022 PA,
the revised IMPROVE equation divides PM
components into smaller and larger sizes of
particles in PM2.5, with separate mass scattering
efficiencies and hygroscopic growth functions for
each size category (U.S. EPA, 2022b, section
5.3.1.1).
164 When light extinction is calculated using the
revised IMPROVE equation, all 60 sites have 3-year
visibility metrics below 30 dv, 56 sites are at or
below 25 dv, and 26 sites are at or below 20 dv (see
U.S. EPA, 2022b, Appendix D, Table D–3).
165 When light extinction is calculated using the
Lowenthal and Kumar IMPROVE equation, 59 sites
have 3-year visibility metrics below 30 dv, 45 sites
are at or below 25 dv, and 15 sites are at or below
20 dv. The one site with a 3-year visibility metric
of 32 dv exceeds the secondary 24-hour PM2.5
standard, with a design value of 56 mg/m3 (see U.S.
EPA, 2022b, Appendix D, Table D–3).
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
Kumar (2016) equation for those sites
that meet the current 24-hour PM2.5
standard, the 3-year visibility metric is
generally at or below 28 dv. For those
sites that exceed the current 24-hour
PM2.5 standard, three of these sites have
a 3-year visibility metric ranging
between 26 dv and 30 dv, while one site
in Fresno, California that exceeds the
current 24-hour PM2.5 standard and has
a 3-year visibility index value of 32 dv
(compared to 29 dv when light
extinction is calculated with the original
IMPROVE equation) (see U.S. EPA,
2022b, Appendix D, Table D–3). At this
site, it is likely that the 3-year visibility
metric using the Lowenthal and Kumar
(2016) equation would be below 30 dv
if PM2.5 concentrations were reduced
such that the 24-hour PM2.5 level of 35
mg/m3 was attained.
In considering visibility impairment
under recent air quality conditions, the
2022 PA recognizes that the differences
in the inputs to equations estimating
light extinction can influence the
resulting values. For example, given the
varying chemical composition of
emissions from different sources, the 2.1
multiplier for converting OC to organic
matter (OM) in the Lowenthal and
Kumar (2016) equation may not be
appropriate for all source types. At the
time of the 2012 review, the EPA judged
that a 1.6 multiplier was more
appropriate, for the purposes of
estimating visibility index at sites across
the U.S., than the 1.4 or 1.8 multipliers
used in the original and revised
IMPROVE equations, respectively. A
multiplier of 1.8 or 2.1 would account
for the more aged and oxygenated
organic PM that tends to be found in
more remote regions than in urban
regions, whereas a multiplier of 1.4 may
underestimate the contribution of
organic PM found in remote regions
when estimating light extinction (78 FR
3206, January 15, 2013; U.S. EPA, 2012,
p. IV–5). The available scientific
information and results of the air quality
analyses indicate that it may be
appropriate to select inputs to the
IMPROVE equation (e.g., the multiplier
for OC to OM) on a regional basis rather
than a national basis when calculating
light extinction. This is especially true
when comparing sites with localized
PM sources (such as sites in urban or
industrial areas) to sites with PM
derived largely from biogenic precursor
emissions (that contribute to
widespread secondary organic aerosol
formation), such as those in the
southeastern U.S. The 2022 PA notes,
however, that conditions involving PM
from such different sources have not
been well studied in the context of
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
applying a multiplier to estimate light
extinction, contributing uncertainty to
estimates of light extinction for such
conditions.
At the time of the 2012 review, the
EPA noted that PM2.5 is the size fraction
of PM responsible for most of the
visibility impairment in urban areas (77
FR 38980, June 29, 2012). Data available
at the time of the 2012 review suggested
that, generally, PM10–2.5 was a minor
contributor to visibility impairment
most of the time (U.S. EPA, 2010b)
although the coarse fraction may be a
major contributor in some areas in the
desert southwestern region of the U.S.
Moreover, at the time of the 2012
review, there were few data available
from PM10–2.5 monitors to quantify the
contribution of coarse PM to calculated
light extinction. Since that time, an
expansion in PM10–2.5 monitoring efforts
has increased the availability of data for
use in estimating light extinction with
both PM2.5 and PM10–2.5 concentrations
included as inputs in the equations. The
analysis in the 2020 PA addressed light
extinction at 20 of the 67 PM2.5 sites
where collocated PM10–2.5 monitoring
data were available. Since that time,
PM10–2.5 monitoring data are available at
more locations and the analyses
presented in the 2022 PA include those
for light extinction estimated with
coarse and fine PM at all 60 sites.
Generally, the contribution of the coarse
fraction to light extinction at these sites
is minimal, contributing less than 1 dv
to the 3-year visibility metric (U.S. EPA,
2020b, section 5.2.1.2). However, the
2022 PA notes that in the updated
quantitative analyses, only a few sites
were in locations that would be
expected to have high concentrations of
coarse PM, such as the Southwest.
These results are consistent with those
in the analyses in the 2019 ISA, which
found that mass scattering from
PM10¥2.5 was relatively small (less than
10%) in the eastern and northwestern
U.S., whereas mass scattering was much
larger in the Southwest (more than 20%)
particularly in southern Arizona and
New Mexico (U.S. EPA, 2019a, section
13.2.4.1, p. 13–36).
Overall, the findings of these updated
quantitative analyses are generally
consistent with those in the 2012 and
2020 reviews. The 3-year visibility
metric was generally below 26 dv in
most areas that meet the current 24-hour
PM2.5 standard. Small differences in the
3-year visibility metric were observed
between the variations of the IMPROVE
equation, which may suggest that it may
be more appropriate to use one version
over another in different regions of the
U.S. based on PM characteristics such as
PO 00000
Frm 00121
Fmt 4701
Sfmt 4700
16321
particle size and composition to more
accurately estimate light extinction.
b. Non-Visibility Effects
Consistent with the evidence
available at the time of the 2012 and
2020 reviews, and as described in detail
in the 2022 PA (U.S. EPA, 2022b,
section 5.3.2.2), the data remain
insufficient to conduct quantitative
analyses for PM effects on climate and
materials. For PM-related climate
effects, as explained in more detail in
the proposal (88 FR 5654, January 27,
2023), our understanding of PM-related
climate effects is still limited by
significant key uncertainties. The
recently available evidence does not
appreciably improve our understanding
of the spatial and temporal
heterogeneity of PM components that
contribute to climate forcing (U.S. EPA,
2022b, sections 5.3.2.1.1 and 5.5).
Significant uncertainties also persist
related to quantifying the contributions
of PM and PM components to the direct
and indirect effects on climate forcing,
such as changes to the pattern of
rainfall, changes to wind patterns, and
effects on vertical mixing in the
atmosphere (U.S. EPA, 2022b, sections
5.3.2.1.1 and 5.5). Additionally, while
improvements have been made to
climate models since the completion of
the 2009 ISA, the models continue to
exhibit variability in estimates of the
PM-related climate effects on regional
scales (e.g., ∼100 km) compared to
simulations at the global scale (U.S.
EPA, 2022b, sections 5.3.2.1.1 and 5.5).
While our understanding of climate
forcing on a global scale is somewhat
expanded since the 2012 review,
significant limitations remain to
quantifying potential adverse PMrelated climate effects in the U.S. and
how they would vary in response to
incremental changes in PM
concentrations across the U.S. As such,
while recent research is available on
climate forcing on a global scale, the
remaining limitations and uncertainties
are significant, and the recent global
scale research does not translate directly
for use at regional spatial scales.
Therefore, the evidence does not
provide a clear understanding at the
necessary spatial scales for quantifying
the relationship between PM mass in
ambient air and the associated climaterelated effects in the U.S. that would be
necessary to evaluate or consider a level
of air quality to protect against such
effects and for informing consideration
of a national PM standard on climate in
this reconsideration (U.S. EPA, 2022b,
section 5.3.2.2.1; U.S. EPA, 2019a,
section 13.3).
E:\FR\FM\06MRR3.SGM
06MRR3
16322
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
For PM-related materials effects, as
explained in more detail in the 2022 PA
(U.S. EPA, 2022b, section 5.3.2.2), the
available evidence has been somewhat
expanded to include additional
information about the soiling process
and the types of materials impacted by
PM. This evidence provides some
limited information to inform doseresponse relationships and damage
functions associated with PM, although
most of these studies were conducted
outside of the U.S. where PM
concentrations in ambient air are
typically above those observed in the
U.S. (U.S. EPA, 2022b, section 5.3.2.1.2;
U.S. EPA, 2019a, section 13.4). The
evidence on materials effects
characterized in the 2019 ISA also
includes studies examining effects of
PM on the energy efficiency of solar
panels and passive cooling building
materials, although the evidence
remains insufficient to establish
quantitative relationships between PM
in ambient air and these or other
materials effects (U.S. EPA, 2022b,
section 5.3.2.1.2). While the available
evidence assessed in the 2019 ISA is
somewhat expanded since the time of
the 2012 review, quantitative
relationships have not been established
for PM-related soiling and corrosion and
frequency of cleaning or repair that
further the understanding of the public
welfare implications of materials effects
(U.S. EPA, 2022b, section 5.3.2.2.2; U.S.
EPA, 2019a, section 13.4). Therefore,
there is insufficient information to
inform quantitative analyses assessing
materials effects to inform consideration
of a national PM standard on materials
in this reconsideration (U.S. EPA,
2022b, section 5.3.2.2.2; U.S. EPA,
2019a, section 13.4).
B. Conclusions on the Secondary PM
Standards
In drawing conclusions on the
adequacy of the current secondary PM
standards, in view of the advances in
scientific knowledge and additional
information now available, the
Administrator has considered the
evidence base, information, and policy
judgments that were the foundation of
the 2020 decision and reflects upon the
body of information and evidence
available in this reconsideration. In so
doing, the Administrator has taken into
account both evidence-based and
quantitative information-based
considerations, as well as advice from
the CASAC and public comments.
Evidence-based considerations draw
upon the EPA’s assessment and
integrated synthesis of the scientific
evidence from studies evaluating
welfare effects related to visibility,
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
climate, and materials associated with
PM in ambient air as discussed in the
2022 PA (summarized in sections V.B
and V.D.2 of the proposal, section V.A.2
above). The quantitative informationbased considerations draw from the
results of the quantitative analyses of
visibility impairment presented in the
2022 PA (as summarized in section V.C
of the proposal and V.A.3 above) and
consideration of these results in the
2022 PA.
Consideration of the scientific
evidence and quantitative information
in the 2022 PA and by the
Administrator is framed by
consideration of a series of policyrelevant questions. Section V.B.2 below
summarizes the rationale for the
Administrators proposed decision,
drawing from section V.D.3 of the
proposal. The advice and
recommendations of the CASAC and
public comments on the proposed
decision are addressed below in
sections V.B.1 and V.B.3, respectively.
The Administrator’s conclusions in this
reconsideration regarding the adequacy
of the secondary PM standards and
whether any revisions are appropriate
are described in section V.D.4.
1. CASAC Advice
In comments on the 2019 draft PA,
the CASAC concurred with the staff’s
overall preliminary conclusions that it
is appropriate to consider retaining the
current secondary standards without
revision (Cox, 2019b). The CASAC
‘‘finds much of the information . . . on
visibility and materials effects of PM2.5
to be useful, while recognizing that
uncertainties and controversies remain
about the best ways to evaluate these
effects’’ (Cox, 2019b, p. 13 of consensus
responses). Regarding climate, while the
CASAC agreed that research on PMrelated effects has expanded since the
2012 review, it also concluded that
‘‘there are still significant uncertainties
associated with the accurate
measurement of PM to the direct and
indirect effects of PM on climate’’ (Cox,
2019b, pp. 13–14 of consensus
responses). The committee
recommended that the EPA summarize
the ‘‘current scientific knowledge and
quantitative modeling results for effects
of reducing PM2.5’’ on several climaterelated outcomes (Cox, 2019b, p. 14 of
consensus responses), while also
recognizing that ‘‘it is appropriate to
acknowledge uncertainties in climate
change impacts and resulting welfare
impacts in the United States of
reductions in PM2.5 levels’’ (Cox, 2019b,
p. 14 of consensus responses). When
considering the overall body of
scientific evidence and technical
PO 00000
Frm 00122
Fmt 4701
Sfmt 4700
information for PM-related effects on
visibility, climate, and materials, the
CASAC agreed with the EPA’s
preliminary conclusions in the 2019
draft PA, stating that ‘‘the available
evidence does not call into question the
protection afforded by the current
secondary PM standards and concurs
that they should be retained’’ (Cox,
2019b, p. 3 of letter).
In this reconsideration, the CASAC
provided its advice regarding the
current secondary PM standards in the
context of its review of the 2021 draft
PA (Sheppard, 2022a). In its comments
on the 2021 draft PA, the CASAC first
recognized that the scientific evidence
is sufficient to support a causal
relationship between PM and visibility
effects, climate effects and materials
effects.
With regard to visibility effects, the
CASAC recognized that the
identification of a target level of
protection for the visibility index is
based on a limited number of studies
and suggested that ‘‘additional regionand view-specific visibility preference
studies and data analyses are needed to
support a more refined visibility target’’
(Sheppard, 2022a, p. 21 of consensus
responses). While the CASAC did not
recommend revising either the target
level of protection for the visibility
index or the level of the current 24-hour
PM2.5 standard, they did state that a
visibility index of 30 deciviews ‘‘needs
to be justified’’ and ‘‘[i]f a value of 20–
25 deciviews is deemed to be an
appropriate visibility target level of
protection, then a secondary 24-hour
PM2.5 standard in the range of 25–35
mg/m3 should be considered’’
(Sheppard, 2022a, p. 21 of consensus
responses).
The CASAC also recognized the
limited availability of monitoring
methods and networks for directly
measuring light extinction. As such,
they suggest that ‘‘[a] more extensive
technical evaluation of the alternatives
for visibility indicators and practical
measurement methods (including the
necessity for a visibility FRM) is need
for future reviews’’ (Sheppard, 2022a, p.
22 of consensus letter). The majority of
the CASAC ‘‘recommend[ed] that an
FRM for a directly measured PM2.5 light
extinction indicator be developed’’ to
inform the consideration of the
protection afforded by the secondary
PM standards against visibility
impairment, the minority of the CASAC
‘‘believe that a light extinction FRM is
not necessary to set a secondary
standard protective of visibility’’
(Sheppard, 2022a, p. 22 of consensus
responses).
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
With regard to climate, the CASAC
noted that ‘‘there is a causal relationship
between PM and climate change, but
large uncertainties remain’’ and
recommended additional research
(Sheppard, 2022a, p. 22 of consensus
responses). With respect to materials
damage, the CASAC noted that
‘‘[q]uantitative information on the
relationship between PM and material
damage is lacking’’ and suggested some
additional studies and research
approaches that could provide
additional information on the effects of
PM on materials and the quantitative
assessment of the relationship between
materials effects and PM in ambient air
(Sheppard, 2022a, p. 23 of consensus
responses).
2. Basis for the Proposed Decision
In reaching his proposed conclusions,
the Administrator first recognized that,
consistent with the scope of this
reconsideration, his decision in this
reconsideration will be focused only
and specifically on the adequacy of
public welfare protection provided by
the secondary PM standards from effects
related to visibility, climate, and
materials. He then considered the
assessment of the current evidence and
conclusions reached in the 2019 ISA
and ISA Supplement; the currently
available quantitative information,
including associated limitations and
uncertainties, described in detail and
characterized in the 2022 PA;
considerations and staff conclusions
and associated rationales presented in
the 2022 PA; and the advice and
recommendations from the CASAC (88
FR 5655, January 27, 2023).
With respect to visibility, the
Administrator noted the longstanding
body of evidence that demonstrates a
causal relationship between ambient PM
and effects on visibility (U.S. EPA,
2019a, section 13.2). and that visibility
impairment can have implications for
people’s enjoyment of daily activities
and for their overall sense of well-being.
Therefore, as in previous reviews, he
considered the degree to which the
current secondary standards protect
against PM-related visibility
impairment. In so doing, and consistent
with previous reviews, the
Administrator considered the protection
provided by the current secondary
standards against PM-related visibility
impairment in conjunction with the
Regional Haze Program 166 for protecting
166 The Regional Haze Program was established
by Congress specifically to achieve ‘‘the prevention
of any future, and the remedying of existing,
impairment of visibility in mandatory Class I areas,
which impairment results from man-made air
pollution,’’ and that Congress established a long-
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
visibility in Class I areas,167 which
together would be expected to achieve
appropriate visual air quality across all
areas (88 FR 5658, January 27, 2023).
The Administrator proposed to
conclude that addressing visibility
impairment in Class I areas is beyond
the scope of the secondary PM NAAQS
and that setting the secondary PM
NAAQS at a level that would remedy
visibility impairment in Class I areas
would result in standards that are more
stringent than is requisite.
In further considering what standards
are requisite to protect against adverse
public welfare effects from visibility
impairment, the Administrator adopted
an approach consistent with the
approach used in previous reviews (88
FR 5645, January 27, 2023). That is, he
first identified an appropriate target
level of protection in terms of a PM
visibility index that accounts for the
factors that influence the relationship
between particles in the ambient air and
visibility (i.e., size fraction, species
composition, and relative humidity). He
then considered air quality analyses
examining the relationship between this
PM visibility index and the current
secondary 24-hour PM2.5 standard in
locations meeting the current 24-hour
PM2.5 and PM10 standards (U.S. EPA,
2022b, section 5.3.1.2; 88 FR 5650,
January 27, 2023).
To identify a target level of protection,
the Administrator first considered the
characteristics of the visibility index
and defines its elements (indicator,
averaging time, form, and level). With
regard to the indicator for the visibility
index, the Administrator recognized
that there is a lack of availability of
methods and an established network for
directly measuring light extinction,
consistent with the conclusions reached
in the 2022 PA (U.S. EPA, 2022b,
section 5.3.1.1) and with the CASAC’s
recommendation for additional research
on direct measurement methods for
light extinction in their review of the
2021 draft PA (Sheppard, 2022a, p. 22
of consensus responses). Consistent
with the approaches used in reaching
decisions in 2012 and 2020, given the
lack of such monitoring data, the
Administrator preliminarily judged that
estimated light extinction, as calculated
using one or more versions of the
term program to achieve that goal (CAA section
169A).
167 In adopting section 169A, Congress set a goal
of eliminating anthropogenic visibility impairment
at Class I areas, as well as a framework for achieving
that goal which extends well beyond the planning
process and timeframe for attaining secondary
NAAQS. Thus, the Regional Haze Program will
continue to contribute to reductions in visibility
impairment in Class I areas.
PO 00000
Frm 00123
Fmt 4701
Sfmt 4700
16323
IMPROVE algorithms, continues to be
the most appropriate indicator for the
visibility index in this reconsideration
(88 FR 5659, January 27, 2023).
In further defining the characteristics
of a visibility index based on estimates
of light extinction, the Administrator
considered the appropriate averaging
time, form, and level of the index. With
regard to the averaging time and form,
the Administrator noted that in previous
reviews, a 24-hour averaging time was
selected and the form was defined as the
3-year average of annual 90th percentile
values. The Administrator recognized
that the evidence available in this
reconsideration and described in the
2022 PA continue to provide support for
the short-term nature of PM-related
visibility effects. Considering the
available analyses of 24-hour and
subdaily PM2.5 light extinction, and
noting that the CASAC did not provide
advice or recommendations with regard
to the averaging time of the visibility
index, the Administrator preliminarily
judged that the 24-hour averaging time
continues to be appropriate for the
visibility index (88 FR 5659, January 27,
2023).
With regard to the form of the
visibility index, the Administrator noted
that, consistent with the approach taken
in other NAAQS, including the current
secondary 24-hour PM2.5 NAAQS, a
multi-year percentile form offers greater
stability to the air quality management
process by reducing the possibility that
statistically unusual indicator values
will lead to transient violations of the
standard. Using a 3-year average
provides stability from the occasional
effects of inter-annual meteorological
variability that can result in unusually
high pollution levels for a particular
year (88 FR 5659, January 27, 2023). In
considering the percentile that would be
appropriate with the 3-year average, the
Administrator first noted that the
Regional Haze Program targets the 20%
most impaired days for improvements
in visual air quality in Class I areas.168
Based on analyses examining 90th, 95th,
and 98th percentile forms, the
Administrator preliminarily judged that
a focus similar to the Regional Haze
Program focused on improving the 20%
most impaired days suggest that the
90th percentile, which represents the
median of the 20% most impaired days,
such that 90% of days have visual air
quality that is at or below the target
level of protection of the visibility
168 As noted above, the Administrator viewed the
Regional Haze Program as a complement to the
secondary PM NAAQS, and thus took into
consideration its approach to improving visibility
in considering how to address visibility outside of
Class I areas.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16324
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
index, would be reasonably expected to
lead to improvements in visual air
quality for the 20% most impaired days
(88 FR 5659, January 27, 2023). In the
analyses of percentiles, the results
suggest that a higher percentile value
could have the effect of limiting the
occurrence of days with peak PMrelated light extinction in areas outside
of Federal Class I areas to a greater
degree. However, the Administrator
preliminarily concluded that it is
appropriate to balance concerns about
focusing on the group of most impaired
days with concerns about focusing on
the days with peak visibility
impairment. Additionally, the
Administrator noted that the CASAC
did not provide advice or
recommendations related to the form of
the visibility index. Therefore, the
Administrator preliminarily judged that
it remains appropriate to define a
visibility index in terms of a 24-hour
averaging time and a form based on the
3-year average of annual 90th percentile
values (88 FR 5659, January 27, 2023).
With regard to the level of the
visibility index, the Administrator first
noted that the scientific evidence that is
available to inform the level of the
visibility index is largely the same as in
previous reviews, and continues to
provide support for a level within the
range of 20 to 30 dv (88 FR 5659–5660,
January 27, 2023). The Administrator
recognized that significant uncertainties
and limitations remained, in particular
those related to the public preference
studies, including methodological
differences between the studies, and
that the available studies may not
capture the full range of visibility
preferences in the U.S. population (88
FR 5659–5660, January 27, 2023). The
Administrator also noted that, in their
review of the 2021 draft PA, the CASAC
recognized that a judgment regarding
the appropriate target level of protection
for the visibility index is based on a
limited number of visibility preference
studies, with studies conducted in the
western U.S. reporting public
preferences for visibility impairment
associated with the lower end of the
range of levels, while studies conducted
in the eastern U.S. reporting public
preferences associated with the upper
end of the range (Sheppard, 2022a, p. 21
of consensus responses). The
Administrator noted that there have
long been significant questions about
how to set a national standard for
visibility that is not overprotective for
some areas of the U.S. In establishing
the Regional Haze Program to improve
visibility in Class I areas, Congress
noted that ‘‘as a matter of equity, the
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
national ambient air quality standards
cannot be revised to adequately protect
visibility in all areas of the country.’’
H.R. Rep. 95–294 at 205. Thus, in
reaching his proposed conclusion, the
Administrator recognized that there are
substantial uncertainties and limitations
in the public preference studies that
should be considered when selecting a
target level of protection for the
visibility index and took the
uncertainties and variability inherent in
the public preference studies into
account. In so doing, the Administrator
first preliminarily judged that,
consistent with similar judgments in
past reviews, it is appropriate to
recognize that the secondary 24-hour
PM2.5 standard is intended to address
visibility impairment across a wide
range of regions and circumstances, and
that the current standard works in
conjunction with the Regional Haze
Program to improve visibility, and
therefore, it is appropriate to establish a
target level of protection based on the
upper end of the range of levels. In
considering the information available in
this reconsideration and the CASAC’s
advice, the Administrator proposed to
conclude that the protection provided
by a visibility index based on estimated
light extinction, a 24-hour averaging
time, and a 90th percentile form,
averaged over 3 years, set at a level of
30 dv (the upper end of the range of
levels) would be requisite to protect
public welfare with regard to visibility
impairment (88 FR 5660, January 27,
2023).
In preliminarily concluding that it
remains appropriate in this
reconsideration to define the target level
of protection in terms of a visibility
index based on estimated light
extinction as described above (i.e., with
a 24-hour averaging time; a 3-year, 90th
percentile form; and a level of 30 dv),
the Administrator next considered the
degree of protection from visibility
impairment afforded by the existing
secondary standards. He considered the
updated analyses of PM-related
visibility impairment presented in the
2022 PA (U.S. EPA, 2022b, section
5.3.1.2), which reflect several
improvements over the analyses
conducted in the 2012 review.
Specifically, the updated analyses
examine multiple versions of the
IMPROVE algorithm, including the
version incorporating revisions since
the 2012 review (section V.B.1.a), which
provides an improved understanding of
how variation in equation inputs
impacts calculated light extinction (U.S.
EPA, 2022b, Appendix D). In addition,
unlike the analyses in the 2012 review
PO 00000
Frm 00124
Fmt 4701
Sfmt 4700
and the 2020 PA, all of the sites
included in the analyses had PM10–2.5
data available, which allows for better
characterization of the influence of the
coarse fraction on light extinction (U.S.
EPA, 2022b, section 5.3.1.2).
The Administrator noted that the
results of these updated analyses are
consistent with the results from the
2012 and 2020 reviews (88 FR 5660,
January 27, 2023). Regardless of the
IMPROVE equation used, these analyses
demonstrate that the 3-year visibility
metric is at or below 28 dv in all areas
meeting the current 24-hour PM2.5
standard (section V.C.1.b). Given the
results of these analyses, the
Administrator preliminarily concluded
that the updated scientific evidence and
technical information support the
adequacy of the current secondary PM2.5
and PM10 standards to protect against
PM-related visibility impairment. While
the inclusion of the coarse fraction had
a relatively modest impact on calculated
light extinction in the analyses
presented in the 2022 PA, he
nevertheless recognized the continued
importance of the PM10 standard given
the potential for larger impacts in
locations with higher coarse particle
concentrations, such as in the
southwestern U.S., for which only a few
sites met the criteria for inclusion in the
analyses in the 2022 PA (U.S. EPA,
2019a, section 13.2.4.1; U.S. EPA,
2022b, section 5.3.1.2).
With regard to the adequacy of the
secondary 24-hour PM2.5 standard, the
Administrator noted that the CASAC
stated that ‘‘[i]f a value of 20–25
deciviews is deemed to be an
appropriate visibility target level of
protection, then a secondary 24-hour
PM2.5 standard in the range of 25–35 mg/
m3 should be considered’’ (Sheppard,
2022a, p. 21 of consensus responses).
The Administrator recognized that the
CASAC recommended that the
Administrator provide additional
justification for a visibility index target
of 30 dv but did not specifically
recommend that he choose an
alternative level for the visibility index.
The Administrator considered the
CASAC’s advice, together with the
available scientific evidence and
quantitative information, in reaching his
proposed conclusions. He recognized
conclusions regarding the appropriate
weight to place on the scientific and
technical information examining PMrelated visibility impairment including
how to consider the range and
magnitude of uncertainties inherent in
that information is a public welfare
policy judgment left to the
Administrator. As such, the
Administrator noted his conclusion on
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
the appropriate visibility index (i.e.,
with a 24-hour averaging time; a 3-year,
90th percentile form; and a level of 30
dv) and his conclusions regarding the
quantitative analyses of the relationship
between the visibility index and the
current secondary 24-hour PM2.5
standard. In so doing, he proposed to
conclude that the current secondary
standards provide requisite protection
against PM-related visibility effects (88
FR 5661, January 27, 2023).
In reaching his proposed conclusions,
the Administrator also recognized that
the available evidence on visibility
impairment generally reflects a
continuum and that the public
preference studies did not identify a
specific level of visibility impairment
that would be perceived as ‘‘acceptable’’
or ‘‘unacceptable’’ across the whole U.S.
population. However, he noted that a
judgment regarding the appropriate
target level of protection would take
into consideration the appropriate
weight to place on the individual public
preference studies. In so doing, he noted
that placing more weight on the public
preference study from Washington, DC,
could provide support for a target level
of protection at or near 30 dv, whereas
placing more weight on the public
preference study performed in the
Phoenix, AZ, study could provide
support for a target level of protection
below 30 dv and down to 25 dv. While
the Administrator noted that, in their
review of the 2021 draft PA, the CASAC
did not recommend revising the level of
the current 24-hour PM2.5 standard, the
Administrator recognized that they did
recommend greater justification for a
target level of protection of 30 dv, and
noted that if a target level of protection
of 20–25 dv was identified, then a
secondary 24-hour PM2.5 standard in the
range of 25–35 mg/m3 should be
considered (Sheppard, 2022a, p. 21 of
consensus responses). For these reasons,
the Administrator solicited comment on
his proposed decision to retain the
current secondary 24-hour PM2.5
standard, as well as the appropriateness
of a target level of protection for
visibility below 30 dv and as low as 25
dv, and on revising the level of the
current secondary 24-hour PM2.5
standard to a level as low as 25 mg/m3.
With respect to climate effects, the
Administrator recognized that a number
of improvements and refinements have
been made to climate models since the
time of the 2012 review. However,
despite continuing research and the
strong evidence supporting a causal
relationship with climate effects (U.S.
EPA, 2019a, section 13.3.9), the
Administrator noted that there are still
significant limitations in quantifying the
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
contributions of the direct and indirect
effects of PM and PM components on
climate forcing (U.S. EPA, 2022b,
sections 5.3.2.1.1 and 5.5). He also
recognized that models continue to
exhibit considerable variability in
estimates of PM-related climate impacts
at regional scales (e.g., ∼100 km),
compared to simulations at the global
scale (U.S. EPA, 2022b, sections
5.3.2.1.1 and 5.5). As noted above, the
CASAC recognized a causal relationship
between PM and climate effects but also
the large uncertainties associated with
quantitatively assessing such effects,
particularly on a national level in the
context of a U.S.-based standard. These
uncertainties led the Administrator to
preliminarily conclude that the
scientific information available in this
reconsideration remains insufficient to
quantify, with confidence, the impacts
of ambient PM on climate in the U.S.
(U.S. EPA, 2022b, section 5.3.2.2.1) and
that there is insufficient information at
this time to revise the current secondary
PM standards or to promulgate a
distinct secondary standard to address
PM-related climate effects (88 FR 5661,
January 27, 2023).
With respect to materials effects, the
Administrator noted that the available
evidence continues to support the
conclusion that there is a causal
relationship with PM deposition (U.S.
EPA, 2019a, section 13.4). He
recognized that deposition of particles
in the fine or coarse fractions can result
in physical damage and/or impaired
aesthetic qualities. Particles can
contribute to materials damage by
adding to the effects of natural
weathering processes and by promoting
the corrosion of metals, the degradation
of painted surfaces, the deterioration of
building materials, and the weakening
of material components. While some
recent evidence on materials effects of
PM is available in the 2019 ISA, the
Administrator noted that this evidence
is primarily from studies conducted
outside of the U.S. in areas where PM
concentrations in ambient air are higher
than those observed in the U.S. (U.S.
EPA, 2019a, section 13.4). The CASAC
also noted the lack of quantitative
information relating PM and material
effects. Given the limited amount of
information on the quantitative
relationships between PM and materials
effects in the U.S., and uncertainties in
the degree to which those effects could
be adverse to the public welfare, the
Administrator preliminarily judged that
the scientific information available in
this reconsideration remains insufficient
to quantify, with confidence, the public
welfare impacts of ambient PM on
PO 00000
Frm 00125
Fmt 4701
Sfmt 4700
16325
materials and that there is insufficient
information at this time to revise the
current secondary PM standards or to
promulgate a distinct secondary
standard to address PM-related
materials effects (88 FR 5661, January
27, 2023).
Taken together, the Administrator
proposed to conclude that the scientific
and technical information for PMrelated visibility impairment, climate
impacts, and materials effects, with its
attendant uncertainties and limitations,
supports the current level of protection
provided by the secondary PM
standards as being requisite to protect
against known and anticipated adverse
effects on public welfare. For visibility
impairment, this proposed conclusion
reflected his consideration of the
evidence for PM-related light extinction,
together with his consideration of
updated analyses of the protection
provided by the current secondary PM2.5
and PM10 standards. For climate and
materials effects, this conclusion
reflected his preliminary judgment that,
although it remains important to
maintain secondary PM2.5 and PM10
standards to provide some degree of
control over long- and short-term
concentrations of both fine and coarse
particles, it is generally appropriate not
to change the existing secondary
standards at this time and that it is not
appropriate to establish any distinct
secondary PM standards to address PMrelated climate and materials effects at
this time. As such, the Administrator
recognized that current suite of
secondary standards (i.e., the 24-hour
PM2.5, 24-hour PM10, and annual PM2.5
standards) together provide such control
for both fine and coarse particles and
long- and short-term visibility and nonvisibility (e.g., climate and materials) 169
effects related to PM in ambient air. His
proposed conclusions on the secondary
standards were consistent with advice
from the CASAC, which noted
substantial uncertainties remain in the
scientific evidence for climate and
materials effects. Thus, based on his
consideration of the evidence and
analyses for PM-related welfare effects,
as described above, and his
consideration of CASAC advice on the
secondary standards, the Administrator
proposed not to change those standards
(i.e., the current 24-hour and annual
PM2.5 standards, 24-hour PM10 standard)
at this time (88 FR 5662, January 27,
2023).
169 As noted earlier, other welfare effects of PM,
such as ecological effects, are being considered in
the separate, on-going review of the secondary
NAAQS for oxides of nitrogen, oxides of sulfur and
PM.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16326
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
3. Comments on the Proposed Decision
Of the public comments received on
the proposal, very few were specific to
the secondary PM standards. Of those
commenters who did provide comments
on the secondary PM standards, the
majority support the Administrator’s
proposed decision to retain the current
standards. Some commenters disagree
with the Administrator’s proposed
conclusion to retain the current
secondary standards, primarily focusing
their comments on the need for a
revised standard to protect against
visibility impairment. In addition to the
comments addressed in this notice, the
EPA has prepared a Response to
Comments document that addresses
other specific comments related to
setting the secondary PM standards.
This document is available for review in
the docket for this rulemaking and
through the EPA’s NAAQS website
(https://www.epa.gov/naaqs/particulatematter-pm-air-quality-standards).
We first note that some commenters
raise questions about the protection
provided by the secondary PM
standards for ecological effects (e.g.,
effects on ecosystems, ecosystem
services, or species). However,
consistent with the 2016 IRP and as
described in the proposal (88 FR 5643,
January 27, 2023), other welfare effects
of PM, such as the ecological effects
identified by commenters, are being
considered as part of the separate,
ongoing review of the secondary
standards for oxides of sulfur, oxides of
nitrogen and PM, and thus, those
comments are beyond the scope of this
action.
Of the comments addressing the
proposed decision for the secondary PM
standards, many of the commenters
support the Administrator’s proposed
decision to retain the current secondary
PM standards, without revision. This
group includes industries and industry
groups and State and local governments
and organizations. All of these
commenters generally note their
agreement with the rationale provided
in the proposal, with a focus on the
strength of the available scientific
evidence for PM-related welfare effects.
Most also recognize that the scientific
evidence and quantitative information
available in this reconsideration have
not substantially altered our previous
understanding of PM-related effects on
non-ecological welfare effects (i.e.,
visibility, climate, and materials) and do
not call into question the adequacy of
the current secondary standards. They
find the proposed decision not to
change the standards at this time to be
well supported and a reasonable
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
exercise of the Administrator’s public
welfare policy judgment under the CAA.
The EPA agrees with these comments
regarding the adequacy of the current
secondary PM standards and the lack of
support for revision of these standards
at this time.
The EPA received relatively few
comments on the proposed decision that
it is not appropriate to establish any
distinct secondary PM standards to
address PM-related climate effects.
Several commenters agree that the
available scientific evidence provides
support for the 2019 conclusion that
there is a causal relationship between
PM and climate effects, and the
commenters also agree with the EPA
that the currently available information
is not sufficient for supporting
quantitative analyses for the climate
effects of PM in ambient air. These
commenters support the Administrator’s
proposed decision not to set a distinct
standard for climate.
There were also very few commenters
who commented on the proposed
decision that it is not appropriate to
establish any distinct secondary PM
standards to address PM-related
materials effects. As with comments on
climate effects, commenters generally
agree with the EPA that the evidence is
not sufficient to support quantitative
analyses for PM-related materials
effects. However, some commenters
contend that EPA failed to explain in
the proposal how the current standard
is appropriate to protect materials from
the effects of PM. These commenters
disagree with the EPA’s conclusion that
quantitative relationships have not been
established for PM-related soiling and
corrosion and frequency of cleaning or
repair of materials, and cite to several
studies conducted outside the U.S. that
they contend that the EPA should
consider since the same materials are
present in the U.S. They further contend
that, in discussing the available
scientific evidence in the 2019 ISA for
studies conducted outside of the U.S.,
the EPA did not provide references to
these studies and, therefore, the public
is unable to comment on these studies.
They further State that EPA failed to
consider the following information: (1)
Recent work related to soiling of
photovoltaic modules and other
surfaces, and; (2) damage and
degradation resulting from oxidant
concentrations and solar radiation for a
number of materials, including
polymeric materials, plastic, paint, and
rubber. These commenters further assert
that the EPA failed to propose a
standard that provides requisite
protection against materials effects
attributable to PM.
PO 00000
Frm 00126
Fmt 4701
Sfmt 4700
As an initial matter, we note that the
commenters submitted the same
comments related to materials effects
during the 2020 review. Consistent with
our response in the 2020 notice of final
rulemaking (85 FR 82737, December 18,
2020), we disagree with the commenters
that the EPA failed to consider the
relevant scientific information about
materials effects available in this
reconsideration. The 2019 ISA
considered and included studies related
to materials effects of PM, including
studies conducted in and outside of the
U.S., on newly studied materials
including photovoltaic modules that
were published prior to the cutoff date
for the literature search.170 These
include the Besson et al. (2017) study
referenced by the commenters (U.S.
EPA, 2019a, section 13.4.2). The
Gr2014
19:03 Mar 05, 2024
Jkt 262001
components, and specific measures of
materials damage, such as frequency of
painting or repair of materials, the EPA
finds the evidence is insufficient to
support a secondary NAAQS to protect
against materials effects.
With regard to studies conducted
outside of the U.S., including those
referenced by the commenters, as
described in the proposal, in reaching
his proposed conclusion, the
Administrator recognized that while
there was some newly available
information related to materials effects
of PM included in the 2019 ISA, ‘‘this
evidence is primarily from studies
conducted outside of the U.S. in areas
where PM concentrations in ambient air
are higher than those observed in the
U.S. (U.S. EPA, 2019a, section 13.4)’’
(88 FR 5661, January 27, 2023). We
disagree with the commenters that EPA
did not provide references for these
studies, nor that the lack of references
inhibited the public’s ability to provide
comment on this proposed conclusion.
First, the reference to section 13.4 in the
2019 ISA is a direct citation to the
evaluation of newly available studies on
PM-related materials effects, which
includes citations for all materials
effects evidence considered in the 2020
review and in this reconsideration.
Second, section 5.3.2.1.2 of the 2022 PA
considers the available scientific
evidence for PM-related materials
effects—including citations to the
studies newly available in the 2019
ISA—and how that evidence informs
conclusions regarding the adequacy of
the standard (U.S. EPA, 2022b, section
5.3.2.1.2). Therefore, the EPA disagrees
that the proposal failed to provide the
proper references to the studies
conducted outside of the U.S., and that
the public was not provided the
opportunity to provide comment on
these studies.
Moreover, we disagree with the
commenters that the EPA failed to
consider quantitative information from
studies available in this reconsideration.
As detailed in sections 5.3.2.1.2 and
5.3.2.2 of the 2022 PA, and consistent
with the information available in the
2020 review, a number of new studies
are available that apply new methods to
characterize PM-related effects on
previously studied materials; however,
the evidence remains insufficient to
relate soiling or damage to specific
levels of PM in ambient air or to
establish quantitative relationships
between PM and materials degradation.
The uncertainties in the evidence
identified in the 2012 review persist in
the evidence in the 2020 review and in
this reconsideration, with significant
uncertainties and limitations to
PO 00000
Frm 00127
Fmt 4701
Sfmt 4700
16327
establishing quantitative relationships
between particle size, concentration,
chemical components, and frequency of
painting or repair of materials. While
some new evidence is available in the
2019 ISA, overall, the data are
insufficient to conduct quantitative
analyses for PM-related materials
effects. Quantitative relationships have
not been established between
characteristics of PM and frequency of
repainting or cleaning of materials,
including photovoltaic panels and other
energy-efficient materials, that would
help inform our understanding of the
public welfare implications of soiling in
the U.S. (U.S. EPA, 2022b, section
5.3.2.2.2; U.S. EPA, 2019a, section 13.4).
Similarly, the information does not
support quantitative analyses between
microbial deterioration of surfaces and
the contribution of carbonaceous PM to
the formation of black crusts that
contribute to soiling (U.S. EPA, 2022b,
section 5.3.2.2.2; U.S. EPA, 2019a,
section 13.4). We also note that
quantitative relationships are difficult to
assess, in particular those characterized
using damage functions as these
approaches depend on human
perception of the level of soiling
deemed to be acceptable and evidence
in this area remains limited in this
reconsideration (U.S. EPA, 2022b,
section 5.3.2.1.2). Additionally, we note
the CASAC’s concurrence with
conclusions in the 2020 PA (Cox, 2019b,
p. 13 of consensus responses) and the
2022 PA (Sheppard, 2022a, p. 23 of
consensus responses) that uncertainties
remain about the best way to evaluate
materials effects of PM in ambient air.
Further, no new studies are available in
this reconsideration to link human
perception of reduced aesthetic appeal
of buildings and other objects to
materials effects and PM in ambient air.
Finally, uncertainties remain about
deposition rates of PM in ambient air to
surfaces and the interaction of PM with
copollutants on these surfaces (U.S.
EPA, 2022b, section 5.6).
With respect to the commenters’
assertion that the EPA failed to consider
information related to materials damage
and degradation from oxidant
concentrations and solar radiation for a
variety of materials, we first note that,
even assuming these sources of
materials damage are within the scope
of this review of the PM NAAQS, the
commenter did not provide any
references to the scientific studies that
they suggest that the EPA did not
consider. Despite the lack of a list of
specific references from the commenter,
we note that the 2019 ISA considered a
number of studies that examined the
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16328
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
relationships between PM and several of
the materials listed by the commenters
(e.g., paint, plastic, rubber). However, as
described in the 2022 PA, these studies
did not provide additional information
regarding quantitative relationships
between PM and materials that could
inform quantitative analyses (U.S. EPA,
2022b, sections 5.3.2.1.2 and 5.3.2.2.2),
nor did they alter conclusions regarding
the adequacy of the current standard
(U.S. EPA, 2022b, section 5.5).
As summarized above and in the
proposal, the evidence in the 2020
review and in this reconsideration for
PM-related effects on materials is not
substantively changed from that in the
2012 review. There continues to be a
lack of evidence related to materials
effects that establishes quantitative
relationships and supports quantitative
analyses of PM-related materials soiling
or damage. While the information
available in the 2020 review and in this
reconsideration continues to support a
causal relationship between PM in
ambient air and materials effects (U.S.
EPA, 2019a, section 13.4), the EPA is
unable to relate soiling or damage to
specific levels of PM in ambient air and
is unable to evaluate or consider a level
of air quality to protect against such
materials effects. Although the EPA did
not propose a distinct level of air quality
or a national standard based on air
quality impacts (88 FR 5662, January 27,
2023), we did identify data gaps that
prevented us from doing so. The EPA
identified a number of key uncertainties
and areas of future research (U.S. EPA,
2022b, section 5.6) that may inform
consideration of the materials effects of
PM in ambient air in future reviews of
the PM NAAQS. The EPA notes that one
commenter objected to the
Administrator’s proposed conclusion in
the proposal (88 FR 5661, January 27,
2023) that in light of the available
evidence for PM-related impacts on
climate and on materials that it is
appropriate not to change the existing
secondary standards at this time. The
EPA has explained, in both the proposal
and this final action, the basis for its
conclusion that there is insufficient
evidence to identify any particular
secondary standard or standards that
would provide requisite protection
against climate effects or materials
damage. The EPA acknowledges that, as
a result, the adoption of any distinct
secondary PM standards for those
effects would be inconsistent with the
requirements of the CAA. The EPA is
clarifying that it is not basing its
decisions on secondary standards in this
reconsideration to address these welfare
effects because it has concluded that the
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
available scientific evidence is
insufficient to allow the Administrator
to make a reasoned judgment about
what specific standard(s) would be
requisite to protect against known or
anticipated adverse effects to public
welfare from PM-related materials
damage or climate effects.
Some commenters agree with the
Administrator’s proposed conclusion
that a target level of protection for
visibility of 30 dv and the level of the
secondary 24-hour PM2.5 standard of 35
mg/m3 continues to be adequate to
protect visibility, highlighting
improvements in visibility in the U.S.
Other commenters who disagree with
the proposed decision indicated support
for a more stringent standard for
visibility impairment, although some of
these commenters did not necessarily
specify the alternative standard that
would, in their judgment, address their
concerns related to various aspects of
the EPA’s proposal, including the
available public preference studies,
specific aspects of the visibility index,
and the target level of protection
identified by the Administrator. Rather,
most commenters focused on particular
aspects of the visibility metric
underlying the current secondary 24hour PM2.5 standard, including the form,
averaging time, and target level of
protection necessary to protect against
visibility impairment.
With regard to the commenters’
assertion that the current secondary
standards are inadequate to protect the
public welfare from PM-related
visibility impairment, the EPA disagrees
that the currently available information
is sufficient to suggest that a more
stringent standard is warranted. The
EPA identified and addressed in great
detail the limitations and uncertainties
associated with the public preference
studies as a part of the 2012 review (78
FR 3210, January 15, 2013). Given that
the evidence related to public
preferences has not substantially
changed since the 2012 review, the EPA
reiterated the limitations and
uncertainties inherent in the evidence
as a part of the 2020 PA (U.S. EPA,
2020b, section 5.5), as well as in the
2022 PA for this reconsideration (U.S.
EPA, 2022b, section 5.6). The 2022 PA
highlights key uncertainties associated
with public perception of visibility
impairment and identifies areas for
future research to inform future PM
NAAQS reviews, including those raised
by the commenters (U.S. EPA, 2022b,
section 5.6). Specifically, the EPA agrees
with commenters that there are several
areas where additional information
would reduce uncertainty in our
interpretation of the available
PO 00000
Frm 00128
Fmt 4701
Sfmt 4700
information for purposes of
characterizing visibility impairment. As
described in more detail in the 2020 PA
(U.S. EPA, 2020b, p. 5–41) and the 2022
PA (U.S. EPA, 2022b, p. 5–53), briefly,
these areas include: (1) Expanding the
number and geographic coverage of
preference studies in urban, rural, and
Class I areas; (2) evaluating visibility
preferences of the U.S. population
today, given that the preference studies
were conducted more than 15 years ago,
during which time air quality in the
U.S. has improved; (3) accounting for
the influence of varying study methods
may have on an individual’s response as
to what level of visibility impairment is
acceptable, and; (4) information on
people’s judgments on acceptable
visibility based on factors that can
influence their perception of visibility
(e.g., duration of impairment
experiences, time of day, frequency of
impairment).
However, the EPA disagrees with the
commenters that the current secondary
PM standards are inadequate and
should be made more stringent because
of the limitations and uncertainties
associated with the available public
preference studies. The EPA does not
view the limitations of the preference
studies and other available evidence as
so significant as to render the EPA
unable to identify a secondary standard
to protect against the adverse effects of
PM on visibility, but the EPA also does
not believe that the limitations
themselves mean that the standards are
inadequate. In fact, there is a limited
amount of recently available scientific
evidence to further inform our
understanding of public preferences and
visibility impairment is recognized by
the Administrator in reaching his
proposed decision not to change the
current secondary PM standards at this
time, given that the evidence base is
largely the same as at the time of the
2012 and 2020 reviews.
These same commenters further
contend that the EPA failed to use the
latest science to develop a visibility
index, stating that the EPA failed to
consider the contrast of distance
methodology employed in a recent
meta-analysis of available preference
studies (Malm et al., 2019). Commenters
claim that the EPA draws conclusions
from the Malm et al. (2019) study about
how to relate contrast to acceptable
visibility preferences in the 2022 ISA
Supplement, yet ignores the findings of
the study and fails to consider the
‘‘contrast of distance’’ methodology in
the 2022 PA and the proposal, thereby,
in their view, departing from the
CASAC’s advice to consider this
evidence in setting the secondary
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
standard. Finally, the commenters assert
that the EPA did not explain why the
available public preference studies are
adequate for analysis using a light
extinction approach but not using the
contrast of distance approach, and that
such differential treatment is arbitrary.
We disagree with the commenters that
the EPA did not use the latest science
in evaluating the visibility index, and
that the EPA failed to consider the
contrast of distance methodology used
in Malm et al. (2019). As the
commenters state, the Malm et al. (2019)
study was included in the ISA
Supplement (U.S. EPA, 2022a, section
4.2.1). However, the EPA disagrees with
the assertion that the ISA Supplement
reached conclusions about how to relate
contrast to acceptable visibility
preferences. The ISA Supplement
provided an overview of the Malm et al.
(2019) study, stating that ‘‘[t]he main
conclusion of this study was that the
level of acceptable visual air quality is
more consistent across studies using
metrics that evaluate the distinction of
an object from a background than using
metrics that evaluate the greatest
distance at which an object can be
observed.’’ Furthermore, the statements
that the commenters are referencing in
support of this statement (i.e., U.S. EPA,
2022b, pp. 4–5–4–6) are in fact the
conclusions of the study itself, rather
than conclusions of the EPA. For
example, the ISA Supplement notes that
‘‘Malm et al. (2019) suggested that
scene-dependent metrics like contrast,
which integrate the effects of bext along
the sight paths between observers and
landscape features, are better predictors
of preference levels than universal
metrics like light extinction.’’ The
suggestion that the contrast of distance
methodology is a better predictor than
light extinction is one of the study
authors, not the EPA. The EPA has not
reached a conclusion on whether
contrast of distance methodology would
be a more appropriate indicator for a
visibility index than estimated light
extinction because the EPA finds that
there is insufficient information in the
record at this time to support that it is
practical to evaluate, much less adopt,
the contrast of distance methodology on
a national basis. Specifically, the Malm
et al. (2019) study does not provide as
a part of their publication the specific
input values to the equation to calculate
the contrast of distance associated with
the available public preference studies
(e.g., sight paths from the images), nor
do the preference studies present or
make publicly available these data in
their publications. In the absence of
additional studies or publicly available
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
data to further evaluate the contrast of
distance methodology, the EPA is
unable to consider contrast of distance
as an alternative to estimated light
extinction in this reconsideration,
although we note that it may be
appropriate to evaluate it more closely
in future reviews.
In reaching conclusions regarding the
appropriate indicator for the visibility
index, the 2022 PA specifically notes
‘‘that limited new research is available
on methods of characterizing visibility
or on how visibility is valued by the
public, such as visibility preference
studies. Thus, while limited new
research has further informed our
understanding of the influence of
atmospheric components of PM2.5 on
light extinction, the available evidence
to inform consideration of the public
welfare implications of PM-related
visibility impairment remains relatively
unchanged’’ (U.S. EPA, 2022b, p. 5–50).
The EPA again notes in the proposal
that ‘‘there are very few studies
available that use scene-dependent
metrics (i.e., contrast) to evaluate public
preference information, which makes it
difficult to evaluate them as an
alternative to the light extinction
approach’’ (88 FR 5649–5650, January
27, 2023). To further expand on this
statement, the Malm et al. (2019) study
does not provide enough information to
replicate the results of their contrast of
distance approach to allow for a
comprehensive evaluation of the
potential use of this methodology in
considering the results of the public
preference studies for determining the
target level of protection for visibility.
Some commenters suggests that the
methodology could be approximated by
simply ensuring that people could
always see distant scenic elements, and
that characterizing typical average and/
or maximal viewing distances cross
different geographical areas and regions
would be a straightforward
Geographical Information Systems (GIS)
exercise. The EPA disagrees that this
assessment would be straightforward,
given the lack of data establishing
viewing distances in the available
scientific record and the diversity of
distance to scenic elements across
different areas and regions of the U.S.,
and finds that this approach is also not
practical to adopt in this
reconsideration. Finally, while the
Malm et al. (2019) study is using an
alternative approach for evaluating
public preferences and acceptability, we
note that this study is evaluating the
same public preference studies that
have been available for the past several
decades. For these reasons, the EPA
disagrees with the commenters’
PO 00000
Frm 00129
Fmt 4701
Sfmt 4700
16329
allegation that the EPA ignored the
findings of the Malm et al. (2019) study
and failed to consider the contrast of
distance methodology in the 2022 PA
and the proposal, and ignored the
CASAC’s advice to consider this study.
The ISA Supplement and the 2022 PA
considered the Malm et al. (2019) study,
along with the full body of available
scientific evidence, and took into
account the uncertainties and
limitations associated with the evidence
for visibility preferences, in reaching
conclusions regarding the adequacy of
the secondary 24-hour PM2.5 standard
(U.S. EPA, 2022b, pp. 5–24–5.25, 5–50).
Several comments in support of
revising the secondary 24-hour PM2.5
standard to protect against visibility
generally recommend revisions to the
elements of the standard and visibility
index (indicator, averaging time, form,
and level) consistent with those
supported by the CASAC and public
comments in previous PM NAAQS
reviews. Some commenters assert that
the EPA’s approach in the 2022 PA and
in the proposal for this reconsideration
did not evaluate options for alternative
secondary PM standards and thereby is
flawed. We address comments on the
elements of a visibility index and a
revised standard for visibility effects
below.
As an initial matter, the EPA disagrees
to the extent commenters are suggesting
that the PA is legally required to analyze
options for alternative standards. The
PA is a document developed by the EPA
in order to assist the Administrator and
the CASAC in reaching conclusions
regarding the adequacy of the current
standards, and its scope is determined
by the EPA. Moreover, the 2022 PA did
assess a wide range of information
relevant to the Administrator’s decision
and considered a range of potential
standards.
First, in developing the 2022 PA and
in responding to CASAC’s advice and
recommendations during its review of
the 2021 draft PA, the EPA expanded
upon its discussion of determining the
target level of protection for the
visibility index and considered the
extent to which the available scientific
information would alter regarding the
visibility index and the appropriate
target level of protection against PMrelated visibility effects (U.S. EPA,
2022b, pp. 5–27–5–29). This detailed
discussion expands the consideration of
the target level of protection for the
visibility index presented in the 2020
PA (U.S. EPA, 2020b) and the 2021 draft
PA (U.S. EPA, 2021c), neither of which
specifically considered the elements of
the visibility index in determining the
appropriate target level of protection. In
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16330
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
considering the available information in
the 2022 PA, the EPA concluded that
the available information continued to
provide support for a visibility index
with a level of 30 dv, with estimated
light extinction as the indicator, a 24hour averaging time, and a 90th
percentile form, averaged over three
years.
Additionally, in summarizing the air
quality and quantitative information in
the proposal for this reconsideration,
the EPA further expands upon the
discussion added to the 2022 PA related
to the target level of protection in terms
of a PM2.5 visibility index. In so doing,
the EPA considers even more
extensively the available public
preference studies and quantitative
analyses (88 FR 5651–5652, January 27,
2023). In particular, there is a more
detailed discussion of the public
preference studies, including the levels
of impairment determined to be
‘‘acceptable’’ by at least 50 percent of
study participants and the
methodologies used in the studies,
including uncertainties and limitations
associated with the methodologies (88
FR 5652, January 27, 2023). In reaching
a proposed decision regarding the
adequacy of the secondary PM
standards, as well as the appropriate
target level of protection for the
visibility index, the Administrator
considered the available scientific
evidence and quantitative analyses, as
well as judgments about how to
consider the range and magnitude of
uncertainties that are inherent in the
scientific evidence and analyses. In so
doing, the Administrator proposed to
conclude that the protection provided
by a visibility index based on estimated
light extinction, a 24-hour averaging
time, and a 90th percentile form,
averaged over 3 years, set at a level of
30 dv would be requisite to protect
public welfare with regard to visibility
impairment (88 FR 5660, January 27,
2023).
Having provisionally concluded that
it was appropriate to define the target
level of protection in terms of a
visibility index based on estimated light
extinction as described above (i.e., with
a 24-hour averaging time; a 3-year, 90th
percentile form; and a level of 30 dv),
the Administrator next considered the
degree of protection from visibility
afforded by the current secondary PM
standards. In so doing, he considered
the updated analyses of PM-related
visibility impairment presented in the
2022 PA (U.S. EPA, 2022b, section
5.3.1.2) and described in more detail in
the proposal (88 FR 5656, January 27,
2023), which included estimating light
extinction using multiple versions of the
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
IMPROVE algorithm and inclusion of
PM10-2.5 data at all sites to allow for
better characterization of the influence
of the coarse fraction of PM on light
extinction. The Administrator noted
that the results of the analyses in the
2022 PA were consistent with those
from the 2012 and 2020 reviews. He also
recognized that, regardless of the
IMPROVE equation that was used, the
analyses demonstrated that the 3-year
visibility metric is at or below 28 dv in
all areas meting the current 24-hour
PM2.5 standard (88 FR 5657, January 27,
2023). The Administrator also noted
that, in their review of the 2021 draft
PA, the CASAC stated that ‘‘[i]f a value
of 20–25 deciviews is deemed to be an
appropriate visibility target level of
protection, then a secondary 24-hour
standard in the range of 25–35 mg/m3
should be considered (Sheppard, 2022a,
p. 21 of consensus responses). The
Administrator recognized that while the
CASAC recommended that additional
justification be provided for a visibility
index target level of protection of 30 dv,
they did not specifically recommend
that he choose an alternative level for
the visibility index. Therefore, the
Administrator considered the available
scientific evidence, quantitative
information, and the CASAC’s advice in
reaching his proposed conclusions. The
Administrator recognized conclusions
regarding the appropriate weight to
place on the scientific and technical
information, including how to consider
the range and magnitude of
uncertainties inherent in that
information, is a public welfare policy
judgment left to the Administrator. As
such, the Administrator noted his
preliminary conclusion on the
appropriate visibility index (i.e., with a
24-hour averaging time; a 3-year, 90th
percentile form; and a level of 30 dv)
and his preliminary conclusions
regarding the quantitative analyses of
the relationship between the visibility
index and the current secondary 24hour PM2.5 standard. In so doing, he
proposed to conclude that the current
secondary standards provide requisite
protection against PM-related visibility
effects (88 FR 5661, January 27, 2023).
However, the Administrator
additionally recognized that the
available evidence on visibility
impairment generally reflects a
continuum and that the public
preference studies did not identify a
specific level of visibility impairment
that would be perceived as ‘‘acceptable’’
or ‘‘unacceptable’’ across the whole U.S.
population. He noted a judgment of a
target level of protection, below 30 dv
and down to 25 dv, could be supported
PO 00000
Frm 00130
Fmt 4701
Sfmt 4700
if more weight was put on the public
preference study performed in the
Phoenix, AZ, study (BBC Research &
Consulting, 2003). As described above,
while the Administrator noted that the
CASAC did not recommend revising the
level of the current 24-hour PM2.5
standard in their review of the 2021
draft PA, they did state that, should an
alternative level be considered for the
visibility index, revisions to the
secondary 24-hour PM2.5 standard
should also be considered (Sheppard,
2022a, p. 21 of consensus responses).
Thus, the Administrator solicited
comment on the appropriateness of a
target level of protection for visibility
below 30 dv and down as low as 25 dv,
and of revising the level of the current
secondary 24-hour PM2.5 standard to a
level as low as 25 mg/m3 (88 FR 5662,
January 27, 2023), and the
Administrator considered these public
comments in reaching his final decision
on the secondary standards. Thus, the
EPA disagrees that the 2022 PA and the
proposal did not adequately consider
options for revising the secondary PM
NAAQS.
With regard to the elements of the
visibility index, in considering the
adequacy of the current secondary 24hour PM2.5 standard to protect against
visibility impairment, as described in
the proposal (88 FR 5658–5660, January
27, 2023), the Administrator first
defined an appropriate target level of
protection in terms of a PM visibility
index. In considering the information
available in this reconsideration and the
CASAC’s advice, the Administrator
proposed to conclude that the
protection provided by a visibility index
based on estimated light extinction, a
24-hour averaging time, and 90th
percentile form, averaged over 3 years,
set at a level of 30 dv, would be
requisite to protect public welfare with
regard to visibility impairment (88 FR
5660, January 27, 2023).
In defining this target level of
protection, the Administrator first
considered the indicator of such an
index. He noted that, given the lack of
availability of methods and an
established network for directly
measuring light extinctions, a visibility
index based on estimates of light
extinction by PM2.5 components derived
from an adjusted version of the original
IMPROVE algorithm would be most
appropriate, consistent with the 2012
and 2020 reviews. As described in the
proposal (88 FR 5649, January 27, 2023)
and above (section V.A.2), the
IMPROVE algorithm estimates light
extinction using routinely monitored
components of PM2.5 and PM10–2.5, along
with estimates of relative humidity. The
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
Administrator, while recognizing that
some revisions to the IMRPOVE
algorithm were newly available in the
2020 review, noted that the fundamental
relationship between ambient PM and
light extinction has changed very little
and the different versions of the
IMPROVE algorithms can appropriately
reflect this relationship across the U.S.
(88 FR 5658–5659, January 27, 2023). As
such, he judged that defining a target
level of protection in terms of estimated
light extinction continues to be a
reasonable approach in this
reconsideration.
Some commenters who criticized the
EPA’s interpretation and application of
the Malm et al. (2019) study also
contend that an indicator based on the
contrast of distance would be a
significant improvement over the
current indicator for the visibility index
and would more accurately evaluate
public preferences. However, as
described in the 2022 PA (U.S. EPA,
2022b, section 5.3.1.1), while scenedependent metrics, such as contrast,
may be useful alternative predictors of
preferences compared to universal
metrics like light extinction, there are a
very limited number of studies that use
such metrics to evaluate public
preferences of visibility impairment and
there is a lack of scientific evidence that
supports one metric over another.
Moreover, the EPA finds that even if the
Administrator agreed that the contrast of
distance methodology was an
improvement over light extinction, there
is insufficient information available to
evaluate and adopt contrast of distance
as an indicator for a national visibility
target at this time. While, in its review
of the 2021 draft PA the CASAC
suggested that the EPA consider this
method in developing the secondary PM
standards, the CASAC also noted that
‘‘more extensive technical evaluation of
the alternatives for visibility indicators
and practical measurement methods’’ is
needed to inform future reviews of the
secondary PM standards (Sheppard,
2022a, p. 22 of consensus responses).
The CASAC did not recommend using
a different indicator for this
reconsideration, with the majority of
CASAC members reiterated past advice
recommending development of a
visibility FRM for a directly measured
PM2.5 light extinction indicator
(Sheppard, 2022a, p. 22 of consensus
responses), a recommendation that was
supported by other public commenters
as well, and the minority of the CASAC
suggested that such an FRM is not
necessary. For these reasons, the EPA
does not consider it feasible or
appropriate to define the visibility index
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
in terms of a contrast of distance
indicator at this time.
With regard to averaging time, some
commenters suggested to the EPA that a
secondary standard with a different
form than the primary standard may be
a more relevant for welfare effects.
While they do not recommend a specific
alternative form, the commenters point
to CASAC advice in past reviews where
the CASAC stated that a subdaily
standard based on daylight hours better
reflects visibility impairment.
In defining the characteristics of a
visibility index, the EPA continues to
believe that a 24-hour averaging time is
reasonable. This is in part based on
analyses conducted in the 2012 review
that showed relatively strong
correlations between 24-hour and
subdaily (i.e., 4-hour average) PM2.5
light extinction (88 FR 5659, January 27,
2023; 85 FR 82740, December 18, 2020;
78 FR 3226, January 15, 2013),
indicating that a 24-hour averaging time
is an appropriate surrogate for the
subdaily time periods relevant for visual
perception. The EPA believes that these
analyses continue to provide support for
a 24-hour averaging time for the
visibility index in this reconsideration.
The EPA also recognizes that the longer
averaging time may be less influenced
by atypical conditions and/or atypical
instrument performance (88 FR 5659,
January 27, 2023; 85 FR 82740,
December 18, 2020; 78 FR 3226, January
15, 2013). When taken together, the
available scientific information and
updated analyses of calculated light
extinction available in this
reconsideration continue to support that
a 24-hour averaging time is appropriate
when defining a target level of
protection against visibility impairment
in terms of a visibility index.
Moreover, the EPA disagrees with
commenters that a secondary PM2.5
standard with a 24-hour averaging time
does not provide requisite protection
against the public welfare impacts of
visibility impairment. At the time of the
2012 review, the EPA recognized that
hourly or subdaily (i.e., 4- to 6-hour)
averaging times, within daylight hours
and excluding hours with high relative
humidity, are more directly related to
the short-term nature of visibility
impairment and the relevant viewing
periods for segments of the viewing
public than a 24-hour averaging time. At
the time of the 2012 review, the EPA
agreed that a subdaily averaging time
would generally be preferable. However,
the Agency noted significant data
quality uncertainties associated with the
instruments that would provide hourly
PM2.5 mass concentrations necessary to
inform a subdaily averaging time. These
PO 00000
Frm 00131
Fmt 4701
Sfmt 4700
16331
uncertainties, as described in the 2012
review, included short-term variability
in hourly data from available
continuous monitoring methods, which
would prohibit establishing a subdaily
averaging time (78 FR 3209, January 15,
2013). For all of these reasons, and
consistent with the 2020 review, the
EPA continues to believe that a subdaily
averaging time is not supported by the
information available in this
reconsideration.
With regard to the form of the
visibility index, some commenters
contend that the form used in evaluating
visibility impairment is not appropriate.
First, commenters contend that the EPA
incorrectly stated that the CASAC did
not provide advice on the 3-year, 90th
percentile form of the visibility index
and that the CASAC specifically
recommended that the EPA further
justify the metric and form, and by not
doing so, the proposal arbitrarily
departs from the CASAC’s
recommendations. The commenters also
contend that the EPA fails to explain
how averaging the form over three years
is protective given that the public does
not perceive visibility in three-year
averages.
We disagree with the commenters that
the EPA departed from the CASAC’s
recommendations that ‘‘[t]he final PA
should provide a robust justification for
the daily light extinction percentile
used in the analysis’’ (Sheppard, 2022a,
p. 22 of consensus responses). In this
statement, the CASAC did not make
explicit recommendations for revisions
to the form of the visibility index, as the
commenters assert, but rather requested
additional justification for the percentile
selected for the visibility index in the
2022 PA. In response to the CASAC’s
recommendation after reviewing the
2021 draft PA, the EPA included a new
section in the 2022 PA that explicitly
discusses the elements (i.e., indicator,
averaging time, form, and level) of the
visibility index, including additional
justification for the conclusions
regarding the appropriate elements for
the index (U.S. EPA, 2022b, pp. 5–27–
5–29). In so doing, the 2022 PA
recognizes that there is no new
information available in this
reconsideration to inform selection of an
alternative form of the visibility index,
and therefore, relied on the analyses
presented in the 2010 UFVA that
evaluated the different statistical forms
of the visibility index. The 2022 PA also
discusses the approach to improving
visual air quality in Federal Class I areas
as a part of the Regional Haze Program
(U.S. EPA, 2022b, p. 5–28).
Furthermore, as reflected in responding
to public comments below, and in
E:\FR\FM\06MRR3.SGM
06MRR3
16332
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
reaching his final conclusions in section
V.B.4 below, the Administrator further
considers the available scientific and
quantitative information, the CASAC’s
advice, and public comments in
informing his final conclusions
regarding the appropriate target level of
protection for the visibility index. With
regard to the commenters’ assertion that
the EPA did not justify why averaging
the form over three years is protective,
we agree with the commenters that
people do not perceive visibility
impairment in three year averages. As
described in the 2022 PA, visibilityrelated effects and perceived
impairment are often associated with
short-term PM concentrations, and
therefore, the focus of the visibility
analyses is centered on the adequacy of
the 24-hour PM2.5 standard (U.S. EPA,
2022b, p. 5–29). However, as described
in the 2022 PA, the 3-year average form
provides stability from the occasional
effect of inter-annual meteorological
variability that can result in unusually
high pollution levels for a particular
year (U.S. EPA, 2022b, p. 5–28).
Occasional meteorological variability is
of particular concern for the visibility
index, which can be impacted by not
only PM concentrations in ambient air
but also relative humidity. The D.C.
Circuit has previously recognized that it
is legitimate for the EPA to consider
overall stability of the standard and its
resulting promotion of overall
effectiveness of NAAQS control
programs in setting a standard. See
American Trucking Ass’ns v. Whitman,
283 F.3d 355, 375–76 (D.C. Cir. 2002).
The 2022 PA concluded that the
available information continues to
provide support for a 90th percentile
form, averaged over three years, and the
inclusion of additional justification for
the elements of the visibility index
responds to the CASAC’s
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
recommendation (U.S. EPA, 2022b,
section 5.3.1.2).
Some commenters suggest that the
90th percentile form is too low and
would result in 36 days being excluded
annually, presuming that the public
only finds it objectionable when
visibility is worse than the standard on
37 or more days per year. The
commenters also contend that the EPA’s
approach of using a 90th percentile form
for the visibility index is inconsistent
with the goals of the Regional Haze
Program. In so doing, the commenters
note that the Regional Haze Rule focuses
on improving conditions on the worst
days, while they argue that a 90th
percentile form for the visibility index
would ignore the 36 worst visibility
days, rather than identifying them and
reducing pollution on those days.
In reaching conclusions regarding the
appropriate form of the visibility index,
the EPA is following the same approach
employed in past reviews of the
secondary PM NAAQS, including those
in the 2012 and 2020 rulemakings. In
reaching conclusions regarding the
appropriate form of the visibility index
in the 2011 PA, the EPA considered the
percentile forms of the visibility index
assessed in the 2010 PA (i.e., 90th, 95th,
98th) along with the approach for
improving visual air quality under the
Regional Haze Program. In so doing, the
2011 PA notes that the Regional Haze
Program targets the 20% most impaired
days for improvements in visual air
quality in Federal Class I areas (i.e., the
days more impaired than the 80th
percentile). The 2011 PA recognized
that to increase the likelihood of
improving visual air quality on the
worst days, the form of the visibility
index should be set well above the 80th
percentile. The 2011 PA further
concluded that a 90th percentile form
would represent the median of the
distribution of the 20% most impaired
PO 00000
Frm 00132
Fmt 4701
Sfmt 4700
days, and meeting a visibility index
with a 90th percentile form would mean
that 90% of the days have visual air
quality that is at or below the level of
the visibility index and would
reasonably expected to lead to
improvements in visual air quality for
the 20% most impaired days (U.S. EPA,
2011, p. 4–59). The 2022 PA noted that
there is no new information from public
preference studies that would inform
the Administrator’s consideration of the
appropriate form for the visibility target
index, and reached conclusions
consistent with those of 2011 PA.
However, as discussed below, the EPA
disagrees that a focus on the 90th
percentile ‘‘ignores’’ any days with
worse visibility. It is possible to
examine past patterns of air quality to
judge the relationship between the 90th
percentile and higher percentiles, and to
assess whether achieving a 90th
percentile visibility target will also
result in air quality improvements,
where necessary, at higher percentiles.
Based on its assessment of past air
quality and potential alternative
percentiles for the form, the EPA judged
that a 90th percentile would
appropriately achieve improved air
quality both above and below that
percentile.
Some commenters suggest that the
analyses conducted in the 2010 UFVA
are based on a different metric than the
24-hour average being considered in the
reconsideration, that the analyses are
outdated and irrelevant. Therefore, the
commenters assert that relying on the
analyses in the 2010 UFVA is not a
rational justification for the use of a
90th percentile for the visibility index
in this reconsideration. Moreover, these
commenters state that, in past reviews,
both the EPA and the CASAC have
considered and recommended a 98th
percentile form, but the proposal does
not consider the 98th percentile.
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
These commenters assert that the
2010 UFVA was not considering the
same metric under consideration here.
However, the EPA was citing to the
2010 UFVA for the conclusion that there
are correlations between different
statistical forms of the visibility index.
To confirm whether these correlations
occur under recent air quality, we
conducted additional air quality
analyses evaluating the visibility index
using the current percentile form (i.e.,
90th) and two alternative forms (i.e.,
95th and 98th).172 While a higher
percentile form would further limit the
number of days with peak PM-related
light extinction, the analyses confirm
that a 90th percentile form is effective
in limiting visibility impairment at
higher percentiles. Based on these
analyses, depending on which version
of the IMPROVE equation is used to
estimate light extinction, the differences
in the 3-year averages of estimated light
extinction for the 90th, 95th, and 98th
percentile forms are small. For example,
in areas that meet the current 24-hour
PM2.5 standard, for light extinction
estimated using the original IMPROVE
equation, all sites have light extinction
estimates for a 90th percentile form at
or below 26 dv, for a 95th or 98th
percentile form at or below 29 dv.173 In
most locations, when estimating light
extinction based on the original
IMPROVE equation, the difference
between a 95th or 98th percentile form
and a 90th percentile form is generally
less than 3 dv.174 As noted in previous
reviews, a change of 1 to 2 dv in light
extinction under many viewing
conditions will be perceived as a small,
but noticeable, change in the
appearance of a scene, regardless of the
initial amount of visibility impairment
(88 FR 5657, January 27, 2023; U.S.
EPA, 2004b; U.S. EPA, 2010b). Thus,
differences between a 90th percentile
172 Gantt, B., and Hagan, N. (2023). Analysis of
Percentile Forms of the Visibility Index.
Memorandum to the Rulemaking Docket for the
Review of the National Ambient Air Quality
Standards for Particulate Matter (EPA–HQ–OAR–
2015–0072). Available at: https://
www.regulations.gov/docket/EPA-HQ-OAR-20150072.
173 Gantt, B., and Hagan, N. (2023). Analysis of
Percentile Forms of the Visibility Index.
Memorandum to the Rulemaking Docket for the
Review of the National Ambient Air Quality
Standards for Particulate Matter (EPA–HQ–OAR–
2015–0072). Available at: https://
www.regulations.gov/docket/EPA-HQ-OAR-20150072.
174 Gantt, B., and Hagan, N. (2023). Analysis of
Percentile Forms of the Visibility Index.
Memorandum to the Rulemaking Docket for the
Review of the National Ambient Air Quality
Standards for Particulate Matter (EPA–HQ–OAR–
2015–0072). Available at: https://
www.regulations.gov/docket/EPA-HQ-OAR-20150072.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
form and a 95th or 98th percentile form
remain small, and for any of these forms
of the visibility index, the estimated
light extinction based on the original
IMPROVE equation in areas meeting the
current secondary 24-hour PM2.5
standard is below the upper end of the
range of the levels considered for the
visibility index (i.e., below 30 dv).
Some commenters disagree with the
EPA’s proposed conclusion that a level
of 30 dv is appropriate for the visibility
index and support a lower level in order
to provide increased protection against
visibility impairment. Commenters who
support a revised level for the visibility
index state that a target level of
protection of 30 dv would mean that
less than 10% of participants in the
public preference studies, other than the
Washington, DC, study, would accept
visibility conditions above 29 dv. These
commenters further suggest that a 75%
acceptability, rather than 50%
acceptability, is requisite to protect
visibility sources, which would be on
average a level of 21 dv when using the
light extinction method or 18 dv when
using the contrast of distance method.
These commenters argue that, based on
the available information, a target level
of protection for the visibility index of
approximately 20 dv would be more
appropriate, and therefore, the level of
the secondary 24-hour PM2.5 standard
should be strengthened to 25 mg/m3.
Other commenters who support a
revised level for the visibility index
suggest that public preference studies
with longer sight paths to distant
landscape features or with lower target
levels than those in the Washington, DC
study, such as the Phoenix study, would
support a lower level. These
commenters support revising the target
level of protection for the visibility
index to a 25 dv, and revising the level
of the secondary 24-hour PM2.5 standard
to a level as low as 25 mg/m3, suggesting
that in low relative humidity
environments, 25 dv is consistent with
PM2.5 concentrations of less than 25 mg/
m3.
Some commenters state that EPA’s
justification for setting a target level of
protection at the upper end of the 20 to
30 dv range is arbitrary. These
commenters state that the EPA’s
reliance on the standard operating in
many regions and circumstances as
support for the upper end of the range
is irrational and illegal. Moreover, these
commenters contend that EPA provided
no rational connection between the
Regional Haze Program and the
proposed decision to set the target level
of protection at the upper end of the
range. They suggest that the EPA
proposed to rely exclusively on the
PO 00000
Frm 00133
Fmt 4701
Sfmt 4700
16333
Regional Haze Program to protect
visibility in Class I areas and to give
visibility in these areas no weight in
considering the secondary PM standard
and that it is not rational to entirely
ignore visibility in Class I areas when
setting the secondary standard. These
commenters assert that the Regional
Haze Program provides no rational basis
for a target level of protection at the
upper end of the range, nor does the
EPA identify one.
Some commenters contend that the
EPA failed to justify the adequacy of the
current secondary annual PM2.5
standard, noting that the secondary 24hour and annual PM2.5 standards work
together to provide protection against
short- and long-term effects of PM2.5.
These commenters point to CASAC
comments on the 2021 draft PA and the
comments of an individual CASAC
member’s support for strengthening the
secondary annual PM2.5 standard to
provide increased protection against
climate and materials effects over time.
They contend that EPA arbitrarily failed
to discuss the secondary annual PM2.5
standard not only in the proposal, but
also in the 2022 PA and in the 2020
final decision.
The EPA recognizes that the selection
of the target level of protection for the
visibility index is fundamentally a
public welfare policy judgment for the
Administrator. The Administrator is
tasked by the CAA to judge when
visibility impairment becomes an
adverse effect on public welfare. It is
clear that visibility impairment can
become adverse to public welfare, but
the Administrator does not consider that
every deciview of impairment is adverse
to public welfare. In considering the
point at which visibility impairment
becomes adverse to public welfare, such
that the attainment of the secondary PM
NAAQS would prevent the adverse
effect, the Administrator gives weight to
the public preference studies as to when
visibility impairment is unacceptable.
At the same time, the Administrator
recognizes the limitations of these
studies, which have been detailed in the
proposal and the 2022 PA. Similarly,
the EPA discussed the Regional Haze
program in the proposal to highlight
that there is a distinct program to
protect against visibility impairment in
Class I areas, and the existence of that
program is relevant to the
Administrator’s judgment about the
level of visibility impairment that is
adverse to public welfare under CAA
109(d), because in determining what is
requisite the Administrator is primarily
considering visibility impairment
outside of Class I areas.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16334
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
In considering how to use the results
of the public preference studies, the
Administrator concludes that a 50th
acceptability criterion is an appropriate
tool. The Administrator’s task is to set
standards that are neither more stringent
nor less stringent than necessary, and a
50% acceptability criterion seems most
appropriate to use in judging when
visibility impairments become adverse,
because it should more closely represent
when the median person would find the
impairment to be adverse. The
Administrator notes this conclusion is
consistent with the approach adopted in
the Denver study by Ely et al. (1991)
where the 50% acceptability criterion
for urban visibility was first presented.
This study discussed the use of the 50%
acceptability criteria as a reasonable
basis for setting a standard to protect
visibility in urban areas. In doing so, Ely
et al. (1991) noted that the 50%
acceptability criterion divided the slides
into two groups—those judged
acceptable and those judged
unacceptable by a majority of people in
the study—and therefore, was
reasonable since it defines the point
where the majority of the study
participants began to judge levels of
visibility impairment as unacceptable
(Ely et al., 1991).
In considering the appropriate target
level of protection, we next look to the
available public preference studies,
noting that the selecting of the range of
20 to 30 dv for the target level of
protection for the visibility index is
informed by the 50% acceptability
values from these studies. The Denver,
CO, (Ely et al., 1991) and British
Columbia, Canada, (Pryor, 1996) studies
met the 50% acceptability criteria at 20
dv and 19–23 dv, respectively (U.S.
EPA, 2022b, Table D–8). As described in
the proposal, these studies used
photographs that were taken at different
times of the day and on different days
to capture a range of light extinction
levels needed for the preference studies
(88 FR 5652, January 27, 2023).
Compared to studies that used
computer-generated images (i.e., those
in Phoenix, AZ, and Washington, DC)
there was more variability in scene
appearance in these older studies that
could affect preference rating and
includes uncertainties associated with
using ambient measurements to
represent sight path-averaged light
extinction values rather than
superimposing a computer-generated
amount of haze onto the images. When
using photographs, the intrinsic
appearance of the scene can change due
to meteorological conditions (i.e.,
shadow patterns and cloud conditions)
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
and spatial variations in ambient air
quality that can result in ambient light
extinction measurement not being
representative of the sight-path-averaged
light extinction. Computer-generated
images, such as those generated with
WinHaze, do not introduce such
uncertainties, as the same base
photograph is used (i.e., there is no
intrinsic change in scene appearance)
and the modeled haze that is
superimposed on the photograph is
determined based on uniform light
extinction throughout the scene.
Because of the uncertainties and
limitations associated with the Denver,
CO, and British Columbia, Canada, the
EPA concludes that it is appropriate to
place less weight on these studies, and
to instead focus on the public
preference studies that were designed to
reduce these uncertainties and
limitations.
In so doing, we focus on the public
preference studies that use computergenerated images (i.e., those in the
Phoenix, AZ, and Washington, DC)
studies. As described in the proposal,
the use of computer-generated images
have less variability in scene appears
than in those studies that use
photographs taken on different days and
at different times of the days (i.e., those
in the Denver, CO, study) that would be
likely to influence preference rating and
introduces uncertainties associated with
using ambient measurements to present
sight path-averaged light extinction
values rather than superimposing a
computer-generated amount of haze
onto the images (88 FR 5652, January
27, 2023).
The Phoenix, AZ, public preference
study (BBC Research & Consulting,
2003) had several strengths compared to
some of the other public preference
studies. The Phoenix, AZ, study had the
largest number of participants (385 in 27
separate focus group sessions) of all of
the public preference studies, with a
sample group designed to be
demographically representative of the
Phoenix population at that time. The
age range in the Phoenix study was also
more inclusive (18–65+), with the
distribution of the study participants
corresponding reasonably well to the
overall age distribution in the 2000 U.S.
Census for the Phoenix area (BBC
Research & Consulting, 2003).
Furthermore, the 21 images used in the
Phoenix, AZ, study were developed
using the WinHaze software with visual
air quality ranging from 15 to 35 dv, and
the view was toward the southwest,
including downtown Phoenix, with the
Sierra Estrella Mountains in the
background at a distance of 25 miles.
This study had the least noisy
PO 00000
Frm 00134
Fmt 4701
Sfmt 4700
preference results, perhaps because a
larger, more representative group of
participants combined with the use of
computer-generated images resulted in
the smoother distribution of responses
of ‘‘acceptable’’ visual air quality. Based
on the EPA’s evaluation of the public
preference studies in the 2012 review,
the 50% ‘‘acceptable’’ criteria was met
at approximately 24 dv (U.S. EPA, 2010,
Table 2–3).
We also consider the public
preferences for the Washington, DC,
studies (Abt Associates, 2001; Smith
and Howell, 2009). The 2001
Washington, DC study included nine
participants, and the 2009 Washington,
DC, study replicated the 2001 study
with 26 additional participants. Similar
to the Phoenix study, the Washington,
DC, studies also had the strength of
having the 20 images included in the
study generated using WinHaze with
visual air quality ranging from 9 to 45
dv. The study depicted a scene of a
panoramic view of the Potomac River,
the National Mall, and downtown
Washington, DC. All of the distinct
buildings in the scene were within four
miles and the higher elevations in the
background were less than 10 miles
from where the image was taken from
the Arlington National Cemetary in
Virginia. The 50% ‘‘acceptable’’ criteria
was met at approximately 29 dv (U.S.
EPA, 2010, Table 2–3).
As described in more detail in the
proposal, visibility preferences can vary
by location, and such differences may
arise based on the differences in the
cityscape scene that is depicted in the
images (88 FR 5652, January 27, 2023).
In considering the geographical
differences between the public
preference studies, we recognize that
the methodological differences between
the studies may influence the resulting
‘‘acceptable’’ level of visibility
impairment. In the Phoenix, AZ, study,
the image depicted mountains in the
background and urban features in the
foreground, whereas the Washington,
DC, study depicted nearby buildings in
the image without mountains in the
distance. As an initial matter, we note
that the object of interest to the study
participant could differ across the
studies based on the scenes included in
the images being evaluated—with the
mountains being of greater interest in
the images in the Phoenix, AZ, study,
despite also depicting buildings that are
similar to those shown and presumed to
be of interest in the images in the
Washington, DC, study (88 FR 5652,
January 27, 2023). We also agree with
the commenters that the distance
between the object of interest and the
camera is an important consideration in
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
evaluating the public preference studies.
Objects at greater distances from the
camera location (such as those in the
Phoenix, AZ, study which had a
maximum distance of 42 km (U.S. EPA,
2022b, Table D–8)) have a greater
sensitivity to light extinction, which
alone could explain differences in
preferences but coupled with an object
of greater interest results in lower
acceptable levels of visibility
impairment. Conversely, objects at
closer distances from the camera
location (such as those in the
Washington, DC, study which had a
maximum distance of 8 km (U.S. EPA,
2022b, Table D–8)) have less sensitivity
to light extinction, which coupled with
objects of interest (compared to the
mountainous views in the Phoenix, AZ,
study) result in higher acceptable levels
of visibility impairment. These studies
clearly demonstrate that there are
differences in the public preferences
across the studies depending on the
images that are used, in particular the
object of interest to the study participant
depicted in the image and the distance
of the sight path to the object, and that
such differences can influence
preference results.
However, we note that these
uncertainties and limitations have
persisted from past reviews, and there is
very little new information to inform
conclusions regarding the interpretation
of these results with regard to the target
level of protection. In selecting a target
level of protection, and in considering
the CASAC’s advice in their review of
the 2021 draft PA and public comments,
we conclude that it is appropriate to
consider the information from the
public preference studies in
Washington, DC, and Phoenix, AZ, and
in so doing, that it is appropriate to
place weight on both of these studies in
reaching conclusions on the appropriate
target level of protection. The EPA
recognizes that the scenes depicted in
these two studies are different and may
influence public preferences of visibility
impairment, but notes these studies can
be considered together as providing
information about different areas across
the U.S. with variations in the scenes
that people are likely to most commonly
encounter. The scene depicted in the
images used in the Washington, DC,
study have a mix of buildings,
landmarks, and open space. On the
other hand, the scene depicted in the
Phoenix, AZ, study included a mix of
buildings in the foreground and with
more distant mountains in the
background. The Administrator
considers it appropriate to consider
these studies together because in
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
combination, they provide a greater
diversity of scenes, which is more likely
to be representative of scenes people
typically experience around the country
(e.g., not only in eastern metropolitan
statistical areas, but also in western
areas with different vistas). In
considering these two studies together,
the EPA recognizes that, first, the
‘‘object of interest’’ is a subjective
judgment left to the participants of the
public preference studies, and second,
the images in these two studies may
differ in terms of sensitivity to changes
in light extinction because of the
distance between the object of interest
in the scene and the camera. As noted
by the public commenters, the sight
path for the images in the public
preference studies is an important
consideration in reaching conclusions
regarding the appropriate target level of
protection for the visibility index. In
addition, the Administrator judges that
giving weight to multiple studies is a
more appropriate approach than
focusing on a single study, particularly
where the study design (including the
representativeness of the participants
and the scenes depicted in the images)
may be important for interpreting the
results of the public preference studies
for informing conclusions regarding the
visibility index. Given these
considerations and taking into
consideration public comments on the
target level of protection for the
visibility index, the Administrator
recognizes that it is more appropriate to
consider a broader range of public
preferences, reflecting a broader range of
scenes, by putting significant weight on
both the Washington, DC, and Phoenix,
AZ, studies. In so doing, he reaches the
conclusion that it would be appropriate
to identify secondary PM standards that
generally limit visibility impairment to
a level between the two studies.
The Administrator next considers
what target level of protection would be
appropriate based on the available
information from these public
preference studies. He first recognizes
that, in the 2012 and 2020 final
decisions, the then-Administrators
selected a target level of protection of 30
dv, based on the upper end of the range.
In so doing, the then-Administrators
judged that it was appropriate to place
more weight on the uncertainties
associated with the public preference
studies in reaching their conclusions.
However, in this reconsideration, the
current Administrator, while continuing
to recognize that substantial
uncertainties remain and that there is
relatively limited new information
regarding public preferences of visibility
PO 00000
Frm 00135
Fmt 4701
Sfmt 4700
16335
impairment, judges that it is important
to balance the weight placed on
uncertainties with the strength of the
scientific evidence. As such, the
Administrator concludes that it is
appropriate to consider a target level of
protection within the range of 20 to 30
dv. He further concludes that in
selecting a target level within that range
it is appropriate to place weight on both
the mid-point of the range, as supported
by the study in Phoenix, AZ, as well as
the upper end, as supported by the
Washington, DC, study. The
Administrator notes that these two
studies both employ similar
methodologies that are subject to fewer
uncertainties than older public
preference studies (including their use
of WinHaze to reduce uncertainties in
the preference solicitations) although he
notes that the Phoenix, AZ, study
yielded the best results of the four
public preference studies in terms of the
least noisy preference results and the
most representative selection of
participants. Furthermore, he notes the
differences between the scenes used for
each study and finds that consideration
of these studies together is more
appropriate in selecting a national target
for visibility protection than considering
either study alone. Thus, in considering
this information, along with the
uncertainties and limitations of the
public preference studies, the
Administrator judges that it would be
appropriate to select a target level of
protection based on placing equal
weight on the upper end of the range
(i.e., 30 dv) and the middle of the range
(i.e., 24 dv based on the Phoenix, AZ,
study) in order to identify a nationwide
target for protection against visibility
impairment. In so doing, the
Administrator concludes that a visibility
index with a target level of protection of
27, defined in terms of estimated light
extinction, with a 24-hour averaging
time and a 3-year, 90th percentile form,
would provide adequate protection
against PM-related visibility effects on
public welfare. Such a target level of
protection balances the information
from two key studies reflecting different
participant preferences for different
vistas in different parts of the country,
appropriately weighting both near-field
and more distant landscape features that
may be of importance to public
perceptions of visibility.
The Administrator notes that the
available evidence indicates that the
relationship between PM and light
extinction is complex, depending on
factors such as PM composition, size
fraction, and age of the particles in
ambient air, as well as relative
E:\FR\FM\06MRR3.SGM
06MRR3
16336
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
humidity. These factors can vary across
the country based on differences in
regional influences, as well as
meteorological conditions that can vary
spatially and temporally in different
areas. The Administrator also recognizes
that this variability, coupled with the
age of the PM depending on the distance
from the source to the monitor location,
also complicates the selection of which
IMPROVE equation is most appropriate
in different areas, although he notes that
different IMPROVE equations will yield
similar, but not identical, results. In so
doing, the Administrator takes note of
the figures presented in the 2022 PA,
which depict the comparisons using the
original IMPROVE equation (Figure 5–
3), the revised IMPROVE equation
(Figure 5–4), and the Lowenthal &
Kumar equation (Figure 5–6), as well as
the estimated light extinction values for
the three different equations presented
in Table D–7.
The Administrator notes that when
light extinction is calculated using the
original IMPROVE equation, all 60 sites
have 3-year visibility metrics below 28
dv, 58 sites are at or below 25 dv, 26
sites are at or below 20 dv, and of the
two sites above 25 dv one is at 26 dv
and the other has a 24-hour PM2.5 design
value of 56 mg/m3 (i.e., well above the
current 24-hour standard). Results are
similar for other IMPROVE equations.175
Based on these analyses, and consistent
with the results of similar analyses in
the 2012 review and the 2020 PA, the
Administrator concludes that the
current secondary 24-hour PM2.5
standard, with its level of 35 mg/m3,
maintains the visibility index below 27
dv, and in fact, the current standard
maintains air quality such that many
areas have visibility index values that
range between 15 and 25 dv for all three
IMPROVE equations. In the areas that
meet the secondary 24-hour PM2.5
standard, all locations were below 27 dv
when using the original and revised
IMPROVE equation and all but three
locations were at or below 27 dv when
using the Lowenthal & Kumar IMPROVE
equation. Three locations (two in
California and one in Utah) had air
quality that was at 28 dv when the
Lowenthal & Kumar IMPROVE equation
was used. As described in more detail
175 When light extinction is calculated using the
revised IMPROVE equation, all 60 sites have 3-year
visibility metrics below 28 dv, 56 sites are at or
below 25 dv, and 26 sites are at or below 20 dv.
When light extinction is calculated using the
Lowenthal and Kumar IMPROVE equation, 59 sites
have 3-year visibility metrics below 28 dv, 45 sites
are at or below 25 dv, and 15 sites are at or below
20 dv. The one site with a 3-year visibility metric
of 32 dv exceeds the secondary 24-hour PM2.5
standard, with a design value of 56 mg/m3 (see U.S.
EPA, 2022b, Appendix D, Table D–3).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
in section V.A.1.3, we recognize that
there are differences in the inputs for
the three IMPROVE equations that can
influence the resulting estimated light
extinction values. The higher multiplier
for converting OC to OM in the
Lowenthal & Kumar IMPROVE equation
(i.e., a multiplier of 2.1) may be more
appropriate in more remote locations
where there is more aged and
oxygenated organic PM than in urban
locations. The three locations with air
quality at 28 dv are all in urban areas
(downtown Los Angeles, CA; Rubidoux,
CA; Salt Lake City, UT) and tend to have
higher levels of nitrate and OC,
especially during the wintertime when
peak PM2.5 concentrations typically
occur. In these locations, it may be more
appropriate to use either the original or
revised IMPROVE equation, which have
multipliers of 1.4 and 1.8, respectively,
in order to refine the inputs such that
estimated light extinction in these
locations is more accurately
characterized based on site-specific
characteristics.
We also note that the four areas that
exceed the secondary 24-hour PM2.5
standard also generally had air quality
that was below 27 dv in terms of the
visibility index, with only two locations
experiencing a visibility index above 27
dv. One location that exceeds the
secondary 24-hour PM2.5 standard had a
visibility index of 29 dv using the
original IMPROVE equation, while two
locations were 30 and 32 dv using the
Lowenthal & Kumar IMPROVE
equation. We believe attainment and
maintenance of the secondary 24-hour
PM2.5 standard will result in improved
air quality in these areas, such that the
visibility index values for these areas
will decrease even further.
The Administrator recognizes that in
concluding that it is appropriate to
identify secondary PM standards that
generally limit visibility impairment to
as low as 27 dv in terms of the visibility
index, the current secondary PM
standards continue to provide
protection against visibility impairment
associated with a visibility index as low
as, or even lower than, 27 dv. In so
doing, he notes that when meeting the
current 24-hour PM2.5 standard, all sites
have a visibility index at or below 27 dv
with the original and revised IMPROVE
equations, and all but three sites at or
below 27 dv with the Lowenthal and
Kumar IMPROVE equation.
Furthermore, the Administrator notes
that this conclusion is consistent with
the CASAC’s advice who, in their
review the 2021 draft PA, stated that
‘‘[i]f a value of 20–25 deciviews is
deemed to be an appropriate visibility
target level of protection, then a
PO 00000
Frm 00136
Fmt 4701
Sfmt 4700
secondary 24-hour PM2.5 standard in the
range of 25–35 mg/m3 should be
considered’’ (Sheppard, 2022a, p. 21 of
consensus responses).
Thus, the Administrator concludes
that weight on both the upper end of the
range of target levels of protection for
the visibility index identified in
previous reviews and the mid-point of
the range, as presented by the Phoenix,
AZ, public preference study, and
focusing on a target level of protection
of 27 dv, he still judges the current
secondary 24-hour PM2.5 standard
requisite to achieve that target because
the standard generally maintains the
visibility index at or below 27 dv such
that more stringent standards are not
warranted.
The EPA agrees with the commenters
that the secondary PM standards work
together to provide protection against
short- and long-term effects of both fine
and coarse particles (U.S. EPA, 2022b,
section 5.5; 88 FR 5661, January 27,
2023). However, the EPA disagrees with
commenters that we failed to discuss
the secondary annual PM2.5 standard in
the proposal, 2022 PA, and the 2020
final notice and that we failed to justify
the adequacy of the secondary annual
PM2.5 standard. As described in the
2022 PA and the proposal, we recognize
that PM2.5 is the size fraction of PM
responsible for most of the visibility
impairment in urban areas (U.S. EPA,
2022b, section 5.3.1.2; 88 FR 5654,
January 27, 2023). Analyses in the 2019
ISA found that mass scattering from
PM10–2.5 was relatively small (less than
10%) in the eastern and northwestern
U.S., whereas mass scattering was much
larger in the Southwest (more than
20%), particularly in southern Arizona
and New Mexico (U.S. EPA, 2019,
section 13.2.4.1, p. 13–36). Given the
relationship between visibility and
PM2.5 along with the short-term nature
of visibility effects, we focus more on
the adequacy of the secondary 24-hour
PM2.5 standard for providing protection
against visibility impairment (U.S. EPA,
2022b, section 5.3.1.2; 88 FR 5653,
January 27, 2023). In reaching his
proposed conclusions, the
Administrator clearly states that he
‘‘recognizes that the current suite of
secondary standards (i.e., the 24-hour
PM2.5, 24-hour PM10, and annual PM2.5
standards) together provide . . . control
for both fine and coarse particulates and
long- and short-term visibility and nonvisibility (e.g., climate and materials)
effects related to PM in ambient air’’ (88
FR 5661, January 27, 2023). Thus, by
explaining how the secondary standards
work together to provide protection
from adverse effects, why we focus on
the secondary 24-hour PM2.5 standard as
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
most relevant to visibility impairment,
and how the Administrator selected the
target level of protection for the
visibility index, we have addressed the
CASAC’s request to support the
proposed decision to revise the
secondary 24-hour PM2.5 standard while
retaining the secondary annual PM2.5
standard. The commenters also cite to
an individual CASAC member’s
comments for the review of the 2021
draft PA who stated ‘‘[f]or the limited
scope of this reconsideration review, I
see no reason to not simply set the
Secondary equal to the Primary PM
Standards, whatever they may be’’
(Sheppard, 2022a, p. A–3). This CASAC
member did not provide a supporting
rationale for revising the secondary
standards to levels equal to the primary
standards. Although areas across the
country are required to attain both the
primary and secondary PM2.5 standards
so air quality is unaffected by the
Administrator’s decision not to revise
the secondary standards to be equal to
the primary standards, as described in
responding to comments above, the
CAA provisions require the
Administrator to establish secondary
standards that, in the judgment of the
Administrator, are requisite to protect
public welfare from known or
anticipated adverse effects associated
with the presence of the pollutant in
ambient air. In so doing, the
Administrator seeks to establish
standards that are neither more nor less
stringent than necessary for this
purpose. The Act does not require that
standards be set at a zero-risk level, but
rather at a level that reduces risk
sufficiently so as to protect the public
welfare from known or anticipated
adverse effects. The final decision on
the adequacy of the current secondary
standards is a public welfare policy
judgment to be made by the
Administrator. In reaching his proposed
and final decisions regarding the
adequacy of the current secondary PM
standards, the Administrator considered
the available scientific information and
analyses about welfare effects, and
associated public welfare significance,
as well as judgments about how to
consider the range and magnitude of
uncertainties that are inherent in the
scientific evidence and analyses. In so
doing, the Administrator concluded that
the currently available scientific
evidence and quantitative analyses,
including uncertainties and limitations,
do not call into question the adequacy
of the current secondary PM standards
and that the current secondary PM
standards should be retained, without
revision. The Administrator’s judgments
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
and decisions on the primary and
secondary standards are independent
and consider different aspects of the
available scientific evidence and
information in reaching conclusions
regarding the adequacy of the standards
in protecting against PM-related health
and welfare effects.
4. Administrator’s Conclusions
This section summarizes the
Administrator’s considerations and
conclusions related to the current
secondary PM2.5 and PM10 standards
and presents the rationale for his
decision that no change is required for
those standards at this time. The CAA
provisions require the Administrator to
establish secondary standards that, in
the judgment of the Administrator, are
requisite to protect public welfare from
known or anticipated adverse effects
associated with the presence of the
pollutant in the ambient air. In so doing,
the Administrator seeks to establish
standards that are neither more nor less
stringent than necessary for this
purpose. The Act does not require that
standards be set at a zero-risk level, but
rather at a level that reduces risk
sufficiently so as to protect the public
welfare from known or anticipated
adverse effects. The final decision on
the adequacy of the current secondary
standards is a public welfare policy
judgment to be made by the
Administrator. The decision should
draw on the scientific information and
analyses about welfare effects, and
associated public welfare significance,
as well as judgments about how to
consider the range and magnitude of
uncertainties that are inherent in the
scientific evidence and analyses. This
approach is based on the recognition
that the available evidence generally
reflects a continuum that includes
ambient air exposures at which
scientists agree that effects are likely to
occur through lower levels at which the
likelihood and magnitude of responses
become increasingly uncertain. This
approach is consistent with the
requirements of the provisions of the
Clean Air Act related to the review of
NAAQS and with how the EPA and the
courts have historically interpreted the
Act.
Given these requirements, the
Administrator’s final decision in this
reconsideration is a public welfare
policy judgment that draws upon the
scientific and technical information
examining PM-related visibility
impairment, climate effects and
materials effects, including how to
consider the range and magnitude of
uncertainties inherent in that
information. The Administrator
PO 00000
Frm 00137
Fmt 4701
Sfmt 4700
16337
recognizes that his final decision is
based on an interpretation of the
scientific evidence and technical
analyses that neither overstates nor
understates their strengths and
limitations, or the appropriate
inferences to be drawn. In particular,
the Administrator notes that the
assessment of when visibility
impairment is adverse to public welfare
requires a public welfare policy
judgment informed by available
scientific and quantitative information.
In considering the adequacy of the
current secondary PM standards in this
reconsideration, the Administrator has
carefully considered the: (1) Policyrelevant evidence and conclusions
contained in the 2019 ISA and 2022 ISA
Supplement; (2) the quantitative
information presented and assessed in
the 2022 PA; (3) the evaluation of this
evidence, the quantitative information,
and the rationale and conclusions
presented in the 2022 PA; (4) the advice
and recommendations from the CASAC;
and (5) public comments. In the
discussion below, the Administrator
gives weight to the 2022 PA
conclusions, with which the CASAC
generally concurred during their review
of the 2019 draft PA and 2021 draft PA,
as summarized in section IV.B.1 of the
2020 final notice and section V.D.1 of
the 2022 proposal, and takes note of key
aspects of the rationale for those
conclusions that contribute to his
decision in this reconsideration. After
giving careful consideration to all of this
information, the Administrator judges
that no change is required for the
secondary PM standards at this time.
In considering the 2022 PA
evaluations and conclusions, the
Administrator takes note of the overall
conclusions that the non-ecological
welfare effects evidence and
quantitative information are generally
consistent with what was considered in
the 2020 final decision and in the 2012
review (U.S. EPA, 2022b, section 5.5).
The scientific evidence for nonecological welfare effects in this
reconsideration is largely the same as
that available in the 2019 ISA and 2020
PA. As described in section I.C.5.b
above, the 2022 ISA Supplement
included a limited number of newly
available studies on PM-related
visibility effects. This newly available
evidence on visibility effects, along with
the full body of non-ecological welfare
effects evidence assessed in the 2019
ISA, reaffirms conclusions on the
visibility, climate, and materials effects
recognized in the 2020 final decision
and in the 2012 review, including key
conclusions on which the standards are
based. Further, as discussed in more
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16338
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
detail above, the updated quantitative
analyses of visibility impairment for
areas meeting the current standards in
the 2022 PA support the adequacy of the
current secondary PM standards to
protect against PM-related visibility
impairment. The Administrator also
recognizes that uncertainties and
limitations continue to be associated
with the available scientific evidence
and quantitative information.
With regard to the current evidence
on visibility effects, as summarized in
the 2022 PA and discussed in detail in
the 2019 ISA and ISA Supplement, the
Administrator notes the long-standing
body of evidence for PM-related
visibility impairment. As in previous
reviews, this evidence continues to
demonstrate a causal relationship
between PM in ambient air and effects
on visibility (U.S. EPA, 2019a, section
13.2). The Administrator recognizes that
visibility impairment can have
implications for people’s enjoyment of
daily activities and for their overall
sense of well-being. Therefore, as in
previous reviews, he considers the
degree to which the current secondary
standards protect against PM-related
visibility impairment and the degree to
which PM-related visibility impairment
is adverse to public welfare. In
particular, in recognizing the short-term
nature of visibility impairment along
with the fact that PM2.5 is the size
fraction that contributes most to light
extinction, the Administrator especially
focuses on the adequacy of the current
secondary 24-hour PM2.5 standard in
providing protection against PM-related
visibility effects judged to be adverse.
The Administrator also considers the
protection provided by the current
secondary 24-hour PM2.5 standard
against PM-related visibility impairment
in conjunction with the Regional Haze
Program as a means of achieving
appropriate levels of protection against
PM-related visibility impairment in
urban, suburban, rural, and Federal
Class I areas across the U.S. Programs
implemented to meet the secondary PM
standards, along with the requirements
of the Regional Haze Program
established for protecting against
visibility impairment in Class I areas,
would be expected to improve visual air
quality across all areas of the country.
As described in the proposal (88 FR
5658, January 27, 2023), the
Administrator recognizes that the
Regional Haze Program was established
by Congress specifically to achieve ‘‘the
prevention of any future, and the
remedying of existing, impairment of
visibility in mandatory Class I areas,
which impairment results from manmade air pollution,’’ and that Congress
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
established a long-term program to
achieve that goal (CAA section 169A). In
adopting section 169, Congress set a
goal of eliminating anthropogenic
visibility impairment at Class I areas, as
well as a framework for achieving that
goal which extends well beyond the
planning process and timeframe for
attaining the secondary PM NAAQS.
Recognizing that the Regional Haze
Program will continue to contribute to
reductions in visibility impairment in
Class I areas, consistent with his
proposed conclusions, the
Administrator concludes that
addressing visibility impairment in
Class I areas is largely beyond the scope
of the secondary PM standards and that
setting the secondary 24-hour PM2.5
standard at a level that would remedy
visibility impairment in Class I areas
would result in standards that are more
stringent than is requisite.
In further considering what standards
are requisite to protect against adverse
public welfare effects from visibility
impairment, the Administrator
concludes that it is appropriate to use
an approach consistent with the
approach used past reviews (88 FR
5650, January 27, 2023). He first
identifies an appropriate target level of
protection in terms of a PM visibility
index that takes into account the factors
that influence the relationship between
PM in ambient air and visibility (i.e.,
size fraction, species composition, and
relative humidity). He then considers
the air quality analyses conducted in the
2022 PA that examine the relationship
between the PM visibility index and the
current secondary 24-hour PM2.5
standard in locations that meet the
current 24-hour PM2.5 and PM10
standards (U.S. EPA, 2022b, section
5.3.1.2).
In reaching conclusions regarding the
target level of protection, the
Administrator first considers the
characteristics of the visibility index
and defines its elements (indicator,
averaging time, form, and level). With
regard to the indicator for the visibility
index, the Administrator continues to
recognize that, consistent with the
conclusions of the 2022 PA and the
CASAC’s advice in their review of the
2021 draft PA, there is a lack of
availability of methods and an
established network for directly
measuring light extinction. Therefore,
the Administrator concludes that it
continues to be appropriate to using an
index based on estimates of light
extinction by PM2.5 components based
on the IMPROVE algorithm. In so doing,
the Administrator recognizes that the
fundamental understanding of the
relationship between ambient PM and
PO 00000
Frm 00138
Fmt 4701
Sfmt 4700
light extinction has generally changed
very little over time; however, several
versions of the IMPROVE equation have
been developed and evaluated that
could be used to estimate light
extinction. As at the time of the
proposal, the Administrator recognizes
that the results of the quantitative
analyses in the 2022 PA that examined
three versions of the IMPROVE equation
indicate that there are very small
differences in estimates of light
extinction between the equations, and
that it is not always clear that one
version of the IMPROVE equation is
more appropriate for estimating light
extinction across the U.S. than other
versions of the IMPROVE algorithm (88
FR 5659, January 27, 2023). He also
recognizes that the selection of inputs to
the IMPROVE equation (e.g., the
multiplier for OC to OM) may be more
appropriate on a regional basis rather
than a national basis when calculating
light extinction, and notes the CASAC’s
advice that PM-visibility relationships
are region specific (Sheppard, 2022a, p.
21 of consensus responses). The
Administrator further notes that neither
the CASAC nor public commenters
recommended a specific IMPROVE
equation or an approach for using
different IMPROVE equations across the
U.S. Therefore, given the absence of a
robust monitoring network to directly
measure light extinction, the
Administrator concludes that light
estimated light extinction, as calculated
using one or more versions of the
IMPROVE algorithms, continues to be
the most appropriate indicator for the
visibility index.
Having reached the conclusion that
estimated light extinction is the
appropriate indicator for the visibility
index, the Administrator next considers
the appropriate averaging time and form
of the index. With regard to the
averaging time and form, the
Administrator notes that in previous
reviews, a 24-hour averaging time was
selected and the form was defined as the
3-year average of annual 90th percentile
values. As at the time of proposal, the
Administrator recognizes that the
available information continues to
provide support for the short-term
nature of visibility effects. He further
recognizes that no new information is
available in this reconsideration to
inform his conclusions regarding
averaging time, and therefore, he
considers past analyses of 24-hour and
subdaily PM2.5 light extinction to inform
his conclusions on averaging time. As
described in the proposal (88 FR 5659,
January 27, 2023) and in responding to
comments in section V.B.3 above, prior
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
analyses demonstrated that there are
strong correlations between 24-hour and
subdaily (i.e., 4-hour average) PM2.5
light extinction, indicating that a 24hour averaging time is an appropriate
surrogate for the subdaily time periods
associated with when individuals
experience visibility impairment and
that a longer averaging time may also be
less influenced by atypical conditions
and/or atypical instrument performance.
The Administrator also notes that the
CASAC did not provide advice or
recommendations with regard to the
averaging time of the visibility index,
although some public commenters
referenced CASAC advice in past
reviews that a subdaily standard based
on daylight hours would better reflect
the public welfare effects of public
perceptions of visibility impairment
than a 24-hour standard. However, in
considering the available scientific and
quantitative information, as well as the
CASAC’s advice in their reviews of the
2019 draft PA and 2021 draft PA, the
Administrator concludes that the 24hour averaging time continues to be
appropriate for the visibility index
because it is an appropriate surrogate for
subdaily time periods and results in a
more stable target.
With regard to the form of the
visibility index, the Administrator notes
the approach in other NAAQS that a
multi-year percentile form offers greater
stability to the air quality management
process by reducing the possibility that
statistically unusual indicator values
will lead to transient violations of the
standard. He recognizes that using a 3year average provides stability from the
occasional effects of inter-annual
meteorological variability (including
relative humidity) that can result in
unusually high pollution levels for a
particular year (88 FR 5659, January 27,
2023) and recognizes that a stable
standard contributes to the benefits of
the NAAQS by ensuring that attainment
strategies are designed to address nontransient problems and achieve durable
air quality improvements. For these
reasons, he concludes that a 3-year
average continues to be appropriate.
In considering the percentile that
would be appropriate with the 3-year
average, the Administrator recognizes
that there is very little new information
available in this reconsideration to
inform selection of an alternative form
of the visibility index and that the
appropriate form requires the exercise of
public welfare policy judgment. In
selecting the appropriate target level of
protection for the visibility index, the
Administrator is required to assess
when visibility impairment becomes
adverse to public welfare, weighing both
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
the degree of visibility impairment (in
dv) and the frequency of such
impairment (through the form). As with
the mass-based PM air quality standard,
the target level of protection for the
visibility index must be selected in
conjunction with the form to determine
the appropriate stringency. In so doing,
consistent with approaches in past
reviews, the Administrator first notes
that the Regional Haze Program targets
the 20% most impaired days for
improvements in visual air quality in
Class I areas, which are the days above
the 80th percentile form of the visibility
index. The Administrator concludes
that a percentile form set at the 80th
percentile would not be likely to
sufficiently improve visual air quality
on the worst days based on the visibility
index. In considering the information
available in past reviews regarding the
form of the visibility index, as well as
the analysis of alternative forms based
on recent air quality discussed above,
the Administrator notes that a 90th
percentile form would represent the
median of the distribution of the 20%
most impaired days, and meeting a
visibility index with a 90th percentile
form would reasonably be expected to
lead to improvements in visual air
quality for days both above and below
the 90th percentile (88 FR 5660, January
27, 2023). In reaching his conclusion
that a 90th percentile would
appropriately achieve improved air
quality both above and below that
percentile, the Administrator took into
consideration assessments of air quality
data and potential alternative
percentiles for the form. The
Administrator further notes that,
consistent with the conclusions in the
2011 PA and 2020 PA, the 2022 PA
concluded that there is no new
information from public preference
studies that would suggest that a 90th
percentile form is not appropriate. The
Administrator also considers air quality
analyses described above in responding
to public comments regarding the
percentile form of the visibility index.
In particular, the Administrator notes
that while a higher percentile form (i.e.,
95th or 98th) would somewhat further
limit the number of days with peak PMrelated light extinction, the differences
in the 3-year averages of estimated light
extinction for the 90th, 95th, and 98th
percentile forms are small. For example,
he notes that for the original IMPROVE
equation, in areas that meet the current
24-hour PM2.5 standard, all sites have
light extinction estimates for a 90th
percentile form at or below 26 dv, and
for a 95th or 98th percentile form light
extinction estimates are at or below 29
PO 00000
Frm 00139
Fmt 4701
Sfmt 4700
16339
dv.176 He further notes that, in most
locations when estimating light
extinction based on the original
IMPROVE equation, the difference
between a 95th or 98th percentile form
and a 90th percentile form is generally
less than 3 dv.177 Moreover, the
Administrator concludes that a 90th
percentile form achieves a very high
degree of control but appropriately
targets the group of worst days, rather
than the few very worst days. Based on
the available information and these
analyses, the Administrator concludes
that the information does not indicate
that it would be appropriate to consider
limiting the occurrence of days with
peak PM-related light extinction to a
greater degree, nor did the CASAC
provide advice or recommendations
related to the form of the visibility
index. Therefore, the Administrator
judges that it remains appropriate to
define a visibility index in terms of a 24hour averaging time and form based on
the 3-year average of annual 90th
percentile values.
With regard to the level of the
visibility index, as at the time of
proposal, the Administrator continues
to recognize that there is very little new
information available to inform his
judgment regarding the range of levels
of visibility impairment judged to be
acceptable by at least 50% of study
participants in the visibility preference
studies,178 and therefore, the range of 20
to 30 dv identified in the 2022 PA
remains appropriate for considering the
level of the visibility index. The
Administrator also recognizes that the
uncertainties and limitations associated
with the public preferences identified in
the 2012 and 2020 reviews continue to
persist, and that these limitations and
uncertainties contributed to the
decisions in 2012 and 2020 that a level
at the upper end of the range (i.e., 30 dv)
was selected. The Administrator
specifically notes that, while the studies
176 Gantt, B., and Hagan, N. (2023). Analysis of
Percentile Forms of the Visibility Index.
Memorandum to the Rulemaking Docket for the
Review of the National Ambient Air Quality
Standards for Particulate Matter (EPA–HQ–OAR–
2015–0072). Available at: https://
www.regulations.gov/docket/EPA-HQ-OAR-20150072.
177 Gantt, B., and Hagan, N. (2023). Analysis of
Percentile Forms of the Visibility Index.
Memorandum to the Rulemaking Docket for the
Review of the National Ambient Air Quality
Standards for Particulate Matter (EPA–HQ–OAR–
2015–0072). Available at: https://
www.regulations.gov/docket/EPA-HQ-OAR-20150072.
178 For reasons stated above and described in the
2022 PA and proposal, the Administrator does not
find it appropriate to use the most recent preference
study based on the Grand Canyon study area (Malm
et al., 2019) for purposes of identifying a target level
of protection for the visibility index.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16340
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
are methodologically similar, there are a
number of factors that can influence
comparability across the studies and
that the available studies may not
capture the full range of visibility
preferences in the U.S. population, as
described in more detail in section
V.D.3 of the 2022 proposal (88 FR 5659–
5660, January 27, 2023). The
Administrator also notes the CASAC’s
advice in their review of the 2021 draft
PA that there are a limited number of
visibility preference studies available to
inform the Administrator’s judgment
regarding the appropriate target level of
protection for the visibility index
(Sheppard, 2022a, p. 21 of consensus
responses). In considering the available
information, including uncertainties
and limitation, and the CASAC’s advice,
the Administrator proposed to conclude
that it is appropriate to consider a target
level of protection for the visibility
index within the range of 20 to 30 dv,
and that establishing a target level of
protection at the upper end of the range
was appropriate. In so doing, the
Administrator proposed to conclude
that the protection provided by a
visibility index based on estimated light
extinction, a 24-hour averaging time,
and a 90th percentile form, averaged
over 3 years, set to a level of 30 dv
would be requisite to protect public
welfare with regard to visibility
impairment.
However, at the time of proposal, the
Administrator recognized that the
available evidence on visibility
impairment generally reflects a
continuum and that the public
preference studies do not provide
information about the specific level for
which visibility impairment would be
‘‘acceptable’’ or ‘‘unacceptable’’ across
the country, and that alternative target
levels of protection could be supported.
At that time, in soliciting public
comments, the Administrator
recognized that other interpretations,
assessments, and judgments based on
the available welfare effects evidence for
this reconsideration could be possible
(88 FR 5662, January 27, 2023).
With regard to the appropriate target
level of protection for the visibility
index, the Administrator first notes that
while the public preference studies
were conducted in several geographical
areas across the U.S., and they provide
insight into regional preferences for
visibility impairment, none of the
studies identify a specific level of
visibility impairment that would be
perceived as ‘‘acceptable’’ or
‘‘unacceptable’’ across the whole U.S.
population. He also noted that there
have been significant questions about
how to set a standard for visibility that
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
is neither overprotective nor
underprotective for some areas of the
U.S. As described in the proposal (88 FR
5660, January 27, 2023), in establishing
the Regional Haze Program to improve
visibility in Class I areas, Congress
noted that ‘‘as a matter of equity, the
national ambient air quality standards
cannot be revised to adequately protect
visibility in all areas of the country.’’
H.R. Rep. 95–294 at 205. For the reasons
noted above, in reaching his proposed
decision regarding visibility
impairment, the Administrator
recognized that he is not seeking to set
a standard that would eliminate
visibility impairment in Class I areas,
but significant uncertainties remain
regarding how to judge when visibility
impairment becomes adverse to public
welfare across the range of daily outdoor
activities for Americans across the
country.
In reaching final conclusions
regarding the available information,
along with the CASAC’s advice and
public comments, the Administrator
again considers what constitutes an
appropriate target level of protection,
and in particular considers whether a
target level of protection below 30 dv is
warranted. In so doing, he first notes the
variability in public preferences of
visibility impairment as demonstrated
by the available public preferences,
which support a range of potential target
levels of protection for the visibility
index from 20 to 30 dv. He also notes
that this range informed the 2012 and
2020 then-Administrators final
decisions that a target level of protection
at the upper end of the range (i.e., 30 dv)
would be most appropriate, given the
uncertainties and limitations associated
with the public preference studies. As
described in in section V.B.3 above in
responding to public comments, the
Administrator recognizes that a number
of factors can influence public
preferences across studies, in particular
due to the types of scenes depicted in
the images as well as the distances at
which the objects of interest are located
from the camera. Furthermore, the
Administrator recognizes the small
number of public preference studies
currently available makes precise
interpretations of their results
challenging for determining a nationally
appropriate target level of visibility
protection. The Administrator also
recognizes that the CASAC, in their
review of 2021 draft PA, reiterated that
PM-visibility relationships are regionspecific based on aerosol composition,
and that several public commenters
emphasized the importance of the sight
path distance in the images when
PO 00000
Frm 00140
Fmt 4701
Sfmt 4700
considering how to interpret the public
preference studies.
In this reconsideration, the
Administrator judges that in
determining when visibility impairment
becomes adverse to public welfare for
purposes of the secondary NAAQS,
while continuing to recognize that
substantial uncertainties remain and
that there is relatively limited new
information regarding public
preferences of visibility impairment, it
is important to balance the weight
placed on uncertainties with the
strength of the scientific evidence. In so
doing, the Administrator first concludes
that, consistent with previous reviews
and his proposed decision, it remains
appropriate to consider a target level of
protection within the range of 20 to 30
dv. However, in further considering the
available scientific and quantitative
information, CASAC advice, and public
comments, he further concludes that in
selecting a target level within that range
it is appropriate to place weight on both
the middle of the range, as supported by
the study in Phoenix, AZ, as well as the
upper end, as supported by the
Washington, DC, study. In so doing, he
notes that the Washington, DC, and
Phoenix, AZ, studies employ similar
methodologies that are subject to fewer
uncertainties than older public
preference studies (including their use
of WinHaze to reduce uncertainties in
the preference solicitations) although he
does note that the Phoenix, AZ, study
yielded the best results of the four
public preference studies in terms of the
least noisy preference results and the
most representative selection of
participants. Further, the Administrator
judges that this approach would take
into account scenes that are similar to
both the Washington, DC, study and
Phoenix, AZ, study, which would be
more representative of the ‘‘typical’’
scenes encountered across more areas of
the U.S. than an approach that places
weight on just one study or on studies
conducted in certain geographical areas
of the country. In considering this
information, along with the
uncertainties and limitations of the
public preference studies, the
Administrator judges that it would be
appropriate to select a target level of
protection based on placing equal
weight on the upper end of the range
(i.e., 30 dv) and the middle of the range
(i.e., 24 dv based on the Phoenix, AZ,
study) in order to provide protection
against visibility impairment in
different geographical areas of the U.S.
For these reasons, the Administrator
concludes that a visibility index with a
target level of protection of 27 dv,
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
defined in terms of estimated light
extinction, with a 24-hour averaging
time and a 3-year, 90th percentile form,
would provide adequate protection
against PM-related visibility effects. In
reaching this conclusion, the
Administrator judges that such a target
level of protection balances the
information from these two key public
preference studies in such a way
appropriately weighs both near-field
and more distant landscape features that
may be of importance to public
perceptions of visibility.
In further considering the appropriate
target level of protection for the
visibility index, the Administrator again
recognizes the complexity of the
relationship between PM and light
extinction which is dependent on a
number of factors, including PM
composition, size fraction, and age of
the particles in ambient air, as well as
relative humidity. As noted in
responding to comments above, these
factors can vary geographically across
the U.S. and local or regional
meteorological conditions can also vary
spatially and temporally. These factors
are critical inputs to the IMPROVE
equation and can influence the resulting
estimated light extinction such that it is
not a straightforward comparison
between estimated light extinction in
one area of the country versus another.
Moreover, the Administrator recognizes
that there is variability in estimated
light extinction depending on the
version of the IMPROVE equation that is
used. As described in more detail in the
2022 PA and the proposal, and in
reaching his decisions on the indicator
of the visibility index above, the
Administrator notes that the 2022 PA
concluded that one version of the
IMPROVE equation is not more accurate
or precise in estimating light extinction,
and that difference in locations may
support the selection of inputs into the
IMPROVE equation or of the appropriate
IMPROVE equation to estimate light
extinction on a regional basis rather
than on a national basis.
In considering the available
information, including variations in
both public preferences of visibility
impairment and estimates of light
extinction using one or more IMPROVE
equation, as well as the CASAC’s advice
in their review of the 2019 draft PA and
2021 draft PA and public comments, the
Administrator judges that a target level
of protection of 27 dv would be
appropriate. In so doing, he concludes
that a target level of protection above 27
dv would not provide adequate
protection against PM-related visibility
impairment based on the 50%
acceptability values when both the
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
Washington, DC, and Phoenix, AZ,
studies are considered. However, he
also notes that when considering the
50% acceptability values from studies
conducted in different areas of the U.S.
and with different scenes and images
depicted, the available public
preference studies do not provide a
‘‘bright line’’ at and above which
visibility impairment is considered
adverse to public welfare. He further
recognizes that, as discussed just above,
there are a number of region-specific
factors that can influence light
extinction, and thereby influence
visibility impairment, as well as
variations in public preferences of
visibility impairment based on the
available studies, that complicate
selection of a single target level of
protection that would be appropriate for
a national visibility index. While the
Administrator recognizes that the
uncertainties and limitations associated
with public preferences of visibility and
estimating light extinction have
persisted over the last several PM
NAAQS reviews, he also recognizes that
in reaching conclusions regarding the
appropriate target level of protection for
the visibility index also involves public
welfare policy judgments regarding how
to appropriately consider the particular
uncertainties around identifying when
visibility impairment becomes adverse
to public welfare, and the limitations on
relying on the public preference studies.
The Administrator also places weight
on the high degree of spatial and
temporal variability in PM composition
and relative humidity across the U.S. in
considering a target level of protection.
This approach of establishing a target
level of protection that takes into
account 50% acceptability values from
both eastern and western sites is a more
appropriate basis for determining the
requisite level of protection against
known or anticipated adverse effects on
public welfare across diverse locations,
i.e., a standard that is neither more nor
less stringent than necessary
nationwide. Specifically, the
Administrator judges that a target level
of protection for the visibility index
focused on maintaining estimated light
extinction between the upper end of the
range of the target levels of protection
(i.e., 30 dv based on the Washington,
DC, study) and the middle of the range
(i.e., 24 dv based on the Phoenix, AZ,
study) to be more appropriate for a
nationwide standard to protect against
visibility impairment compared to a
value derived from one location or one
type of scene alone. For these reasons,
in selecting a target level of protection,
the Administrator concludes that a
PO 00000
Frm 00141
Fmt 4701
Sfmt 4700
16341
target level of protection somewhere
between the upper end and middle of
the range is appropriate because he
judges that this approach, in
conjunction with the Regional Haze
program, is sufficient, but not more
stringent than necessary, to protect
against adverse effects on public
welfare. Thus, he concludes a secondary
24-hour PM2.5 NAAQS should be
evaluated based on its ability to provide
protection against visibility impairment
associated with estimated light
extinction of 27 dv based on estimated
light extinction, a 24-hour averaging
time, and a 90th percentile form,
averaged over 3 years.
Having concluded that it is
appropriate to identify a target level of
protection in terms of a visibility index
based on estimated light extinction as
described above, the Administrator next
considers the degree of protection from
visibility impairment afforded by the
current secondary PM standards. He
considers the updated analyses of PMrelated visibility impairment presented
in the 2022 PA (U.S. EPA, 2022b,
section 5.3.1.2) and described in section
V.B.1.a of the proposal, and notes that
the results of the analyses are consistent
with the results from the 2012 and 2020
reviews.
Taking into consideration the full
body of scientific evidence and
technical information concerning the
known and anticipated effects of PM on
visibility impairment, the Administrator
concludes that the current secondary
PM2.5 and PM10 standards are requisite
to protect against PM-related visibility
impairment. While the inclusion of the
coarse fraction had a relatively modest
impact on calculated light extinction in
the analyses presented in the 2022 PA,
he recognizes the continued importance
of the PM10 standard given the potential
for larger impacts in locations with
higher coarse particle concentrations,
such as in the southwestern U.S., for
which only a few sites met the criteria
for inclusion in the analyses in the 2022
PA (U.S. EPA, 2019a, section 13.2.4.1;
U.S. EPA, 2022b, section 5.3.1.2).
With regard to the adequacy of the
secondary 24-hour PM2.5 standard, the
Administrator notes that, in their review
of the 2021 draft PA, the CASAC stated
that ‘‘[i]f a value of 20–25 deciviews is
deemed to be an appropriate visibility
target level of protection, then a
secondary 24-hour PM2.5 standard in the
range of 25–35 mg/m3 should be
considered’’ (Sheppard, 2022a, p. 21 of
consensus responses). The
Administrator recognizes that the
CASAC recommended that the
Administrator provide additional
justification for a visibility index target
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16342
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
of 30 dv but did not specifically
recommend that he choose an
alternative level for the visibility index.
The Administrator carefully considered
the advice of CASAC and the public
comments and concluded that a lower
target level of visibility was appropriate
in order to properly reflect both a
broader set of studies and a broader
range of vistas that were the subject of
those studies. However, in their review
of the 2021 draft PA, the CASAC
recognized that even a visibility index
target in the range of 20–25 dv could
still warrant retention of the current
secondary 24-hour PM2.5 standard. The
Administrator also considers the advice
from the CASAC in their review of the
2019 draft PA, who ‘‘recogniz[ed] that
uncertainties. . .remain about the best
way to evaluate’’ PM-related visibility
effects (Cox, 2019b, p. 13 consensus
responses). The Administrator
considered the CASAC’s advice,
together with the available scientific
evidence and quantitative information,
in reaching his conclusions.
The Administrator recognizes that
conclusions regarding the appropriate
weight to place on the scientific and
technical information examining PMrelated visibility impairment, including
how to consider the range and
magnitude of uncertainties inherent in
that information, is a public welfare
policy judgment left to the
Administrator. In reaching his final
decision in 2020, the thenAdministrator noted that the available
evidence regarding visibility effects had
changed very little since the 2012
review, specifically recognizing that, as
evaluated in the 2019 ISA, there were
no new visibility studies that were
conducted in the U.S. and there was
little new information available with
regard to acceptable levels of visibility
impairment in the U.S. (85 FR 82742,
December 18, 2020). As such, the thenAdministrator concluded that the
protection provided by a standard
defined in terms of a PM2.5 visibility
index, with a 24-hour averaging time, a
90th percentiles form averaged over
three years, set at a level of 30 dv, was
requisite to protect public welfare
against visibility impairment (85 FR
82743, December 18, 2020). He also
recognized that there was some new
information to inform quantitative
analyses of light extinction, but that the
results of the analyses conducted in the
2020 PA were consistent with those
from the 2012 review. The thenAdministrator recognized that the
analyses demonstrated that the 3-year
visibility metric was at or below about
30 dv in all areas that met the current
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
secondary 24-hour PM2.5 standard, and
was below 25 dv in most of those areas
(85 FR 82743, December 18, 2020).
Therefore, the Administrator judged that
the secondary 24-hour PM2.5 standard
provided sufficient protection for visual
air quality of 30 dv, which he judged
appropriate (88 FR 82744, December 18,
2020). In this reconsideration, the ISA
Supplement evaluated newly available
studies on public preferences for
visibility impairment and/or
development methodologies or
conducted quantitative analyses of light
extinction. In considering the available
scientific and quantitative information,
including that newly available in this
reconsideration, the current
Administrator reached the same
preliminary conclusions in the notice of
proposed rulemaking regarding the 3year visibility index and the current
secondary PM standards as the thenAdministrator in the 2020 final
decision. However, in light of public
comments on the proposal, the
Administrator has further considered
the available scientific evidence and
information, as well as the CASAC’s
advice regarding visibility effects in
their review of the 2021 draft PA. In so
doing, the Administrator judges that it
is appropriate to place more weight on
certain aspects of the evidence that he
had placed less weight on in reaching
his proposed conclusions (i.e., he
focused on the both the middle and the
upper end of the range of the 50%
acceptability values from the available
public preference studies). As such, the
Administrator notes his conclusion on
the appropriate visibility index (i.e.,
with a 24-hour averaging time; a 3-year,
90th percentile form; and a level of 27
dv), which takes into account the
regional variations in public preferences
and equations for estimating light
extinction, and his conclusions
regarding the quantitative analyses of
the relationship between the visibility
index and the current secondary 24hour PM2.5 standard. In so doing, the
Administrator concludes that the
current secondary standards provide
requisite protection against PM-related
visibility effects.
With respect to climate effects, as at
the time of proposal, the Administrator
recognizes that a number of
improvements and refinements have
been made to climate models since the
time of the 2012 review. However,
despite continuing research and the
strong evidence supporting a causal
relationship with climate effects (U.S.
EPA, 2019a, section 13.3.9), the
Administrator notes that there are still
significant limitations in quantifying the
PO 00000
Frm 00142
Fmt 4701
Sfmt 4700
contributions of the direct and indirect
effects of PM and PM components on
climate forcing (U.S. EPA, 2022b,
sections 5.3.2.1.1 and 5.5). He also
recognizes that models continue to
exhibit considerable variability in
estimates of PM-related climate impacts
at regional scales (e.g., ∼100 km),
compared to simulations at the global
scale (U.S. EPA, 2022b, sections
5.3.2.1.1 and 5.5). Moreover, the effects
of PM on climate are diverse as well as
uncertain. Depending on the
circumstances, the radiative forcing
effects of PM in the atmosphere can
vary, such that positive forcing could
result in warming of the Earth’s surface,
whereas a negative forcing could result
in cooling (U.S. EPA, 2019a, section
13.3.2.2). The resulting uncertainty
leads the Administrator to conclude that
the scientific information available in
this reconsideration remains insufficient
to quantify, with confidence, the
impacts of ambient PM on climate in the
U.S. (U.S. EPA, 2022b, section 5.3.2.2.1)
and that there is not an adequate
scientific basis to link attainment of any
particular PM concentration in ambient
air in the U.S. to specific climate effects.
Consequently, the Administrator judges
that there is insufficient information at
this time to revise the current secondary
PM standards or to promulgate a
distinct secondary standard to address
PM-related climate effects.
With respect to materials effects, the
Administrator notes that the available
evidence continues to support the
conclusion that there is a causal
relationship with PM deposition (U.S.
EPA, 2019a, section 13.4). He recognizes
that deposition of particles in the fine or
coarse fractions can result in physical
damage and/or impaired aesthetic
qualities. Particles can contribute to
materials damage by adding to the
effects of natural weathering processes
and by promoting the corrosion of
metals, the degradation of painted
surfaces, the deterioration of building
materials, and the weakening of material
components. While some recent
evidence on materials effects of PM is
available in the 2019 ISA, the
Administrator notes that this evidence
is primarily from studies conducted
outside of the U.S. in areas where PM
concentrations in ambient air are higher
than those observed in the U.S. (U.S.
EPA, 2019a, section 13.4). Given the
limited amount of information on the
quantitative relationships between PM
and materials effects in the U.S., and
uncertainties in the degree to which
those effects could be adverse to the
public welfare, the Administrator judges
that the available scientific information
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
remains insufficient to quantify, with
confidence, the public welfare impacts
of ambient PM on materials and that
there is insufficient information at this
time to revise the current secondary PM
standards or to promulgate a distinct
secondary standard to address PMrelated materials effects.
Taken together, the Administrator
concludes that the scientific and
quantitative information for PM-related
non-ecological welfare effects (i.e.,
visibility, climate, and materials),179
along with the uncertainties and
limitations, supports the current level of
protection provided by the secondary
PM standards as being requisite to
protect against known and anticipated
adverse effects on public welfare. For
visibility impairment, this conclusion
reflects his consideration of the
evidence for PM-related light extinction,
together with his consideration of
updated air quality analyses of the
relationship between the visibility index
and the current secondary 24-hour PM2.5
standard and the protection provided by
the current secondary PM2.5 and PM10
standards. For climate and materials
effects, this conclusion reflects his
judgment that, although it remains
important to maintain secondary PM2.5
and PM10 standards to provide some
degree of control over long- and shortterm concentrations of both fine and
coarse particles, it is appropriate not to
change the existing secondary standards
at this time and that it is not appropriate
to establish any distinct secondary PM
standards to address PM-related climate
and materials effects at this time. As
such, the Administrator recognizes that
current suite of secondary standards
(i.e., the 24-hour PM2.5, 24-hour PM10,
and annual PM2.5 standards) together
provide such control for both fine and
coarse particles and long- and shortterm visibility and non-visibility (e.g.,
climate and materials) effects related to
PM in ambient air. His conclusions on
the secondary standards are consistent
with advice from the CASAC, which
noted substantial uncertainties remain
in the scientific evidence for climate
and materials effects, as well as the
majority of public comments on the
secondary PM standards. Thus, based
on his consideration of the evidence and
analyses for PM-related welfare effects,
as described above, and his
consideration of CASAC advice and
public comments on the secondary
standards, the Administrator concludes
179 As noted earlier, other welfare effects of PM,
such as ecological effects, are being considered in
the separate, on-going review of the secondary
NAAQS for oxides of nitrogen, oxides of sulfur and
PM.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
that it is appropriate not to change those
standards (i.e., the current 24-hour and
annual PM2.5 standards, 24-hour PM10
standard) at this time.
C. Decision on the Secondary PM
Standards
For the reasons discussed above and
taking into account information and
assessments presented in the 2019 ISA,
ISA Supplement, and 2022 PA, advice
from the CASAC, and consideration of
public comments, the Administrator
concludes that the current secondary
PM standards are requisite to protect
public welfare from known or
anticipated adverse effects and is not
changing the standards at this time.
VI. Interpretation of the NAAQS for PM
The EPA is finalizing revisions on
data calculations in appendix K for
PM10 and appendix N for PM2.5.
Revisions to appendix K make the PM10
data handling procedures for the 24hour PM10 standards more consistent
with those of other NAAQS pollutants
and codify existing practices. Revisions
to appendix N update references to the
revision(s) of the standards and change
data handling provisions related to
combining data from nearby monitoring
sites to codify existing practices that are
currently being implemented as the EPA
standard operating procedures.
A. Amendments to Appendix K:
Interpretation of the NAAQS for
Particulate Matter
The EPA proposed to modify its data
handling procedures for the 24-hour
PM10 standard in appendix K to part 50
(88 FR 5662, January 27, 2023). The
proposed modifications include: (1)
Revising design value calculations to be
on a site-level basis, (2) codifying site
combinations to maintain a continuous
data record, and (3) clarifying daily
validity requirements for continuous
monitors. The purpose of these
modifications is to make the data
handling procedures for the 24-hour
PM10 standard more consistent with
those of other NAAQS pollutants and
codify existing practices that are
currently being implemented as EPA
standard operating procedures.
The EPA received few comments on
these proposed appendix K revisions,
the majority of which were supportive.
One commenter was not supportive of
the proposed appendix K revision to
site-level PM10 design values, asserting
that it would amount to an imposition
of a more stringent PM10 standard due
to the potential high bias of FEMs. The
EPA disagrees with this assertion
because site-level design values would
combine data from any high biased FEM
PO 00000
Frm 00143
Fmt 4701
Sfmt 4700
16343
with other monitors at the site rather
than calculate a monitor-level design
value with data solely from that highbiased FEM. The EPA tested the impact
of calculating site-level PM10 design
values for the 2019–2021 period by
assigning the lowest parameter
occurrence code as the primary monitor
and calculating site-level design values.
Most resulting site-level design values
were either identical to or in-between
the multiple monitor-level design values
at the site. Combining data from two or
more monitors also has the benefit of
increasing the number of valid sample
days at many sites. For the 2019–2021
test period, approximately 10% of the
sites with more than one monitor went
from having multiple invalid design
values to a single valid design value.
One commenter was not supportive of
a footnote in the preamble of the NPRM
stating that in the absence of a
designated primary monitor at a given
site, the default primary monitor would
be one with the most complete data
record (88 FR 5662, January 27, 2023).
Because the procedure for calculating
PM10 design values on a site-level basis
being finalized here will require
monitoring agencies to designate a
primary monitor for each site in their
annual network plans (88 FR 5694,
January 27, 2023; App. K, 1.0(b)), the
EPA agrees with the commenter that
this footnote was unnecessary.
Therefore, the EPA is finalizing these
appendix K revisions as proposed.
B. Amendments to Appendix N:
Interpretation of the NAAQS for PM2.5
The EPA proposed to modify its data
handling procedures for the annual and
24-hour PM2.5 standards in appendix N
to part 50 (88 FR 5663, January 27,
2023). These proposed revisions
include: (1) Updating references to the
revisions of the standards rather than
stating the specific level, and (2)
codifying site combinations to maintain
a continuous data record. The purpose
of both modifications is to codify
existing practices that are currently
being implemented as the EPA standard
operating procedures.
The EPA received few comments on
these revisions in the proposed rule,
with most supportive of the appendix N
revisions.
Although the EPA did not propose or
request comment on this issue, one
commenter suggested that appendix N
be revised to only allow data from the
primary monitor to be used in PM2.5
NAAQS designations asserting that it
would add flexibility. The EPA
disagrees with the commenter’s
assertion that this would add flexibility
because it could force agencies to run
E:\FR\FM\06MRR3.SGM
06MRR3
16344
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
their FRMs on a daily schedule or
potentially lead to invalid design values
if manual sampling interruptions or
laboratory issues impact FRM data
completeness. This change would also
be undesirable because it could reduce
by two-thirds the number of days used
in calculations for the annual and 24hour PM2.5 design values at many sites.
Therefore, the EPA is finalizing these
appendix N revisions as proposed.
ddrumheller on DSK120RN23PROD with RULES3
VII. Amendments to Ambient
Monitoring and Quality Assurance
Requirements
The EPA is finalizing revisions to
ambient air monitoring requirements for
PM to improve the usefulness of and
appropriateness of data used in
regulatory decision making. These
changes focus on ambient monitoring
requirements found in 40 CFR parts 50
(appendix L), 53, and 58 with associated
appendices (A, B, C, D, and E). These
changes include addressing updates in
the approval of reference and equivalent
methods, updates in quality assurance
statistical calculations to account for
lower concentration measurements,
updates to support improvements in PM
methods, a revision to the PM2.5
network design to account for at-risk
populations, and updates to the Probe
and Monitoring Path Siting Criteria for
NAAQS pollutants. The EPA also took
comment on how to incorporate data
from next generation technologies into
Agency efforts. A summary of the
comments received is included in this
section.
A. Amendment to 40 CFR Part 50
(Appendix L): Reference Method for the
Determination of Fine Particulate Matter
as PM2.5 in the Atmosphere—Addition
of the Tisch Cyclone as an Approved
Second Stage Separator
The EPA proposed a change to the
FRM for PM2.5 (40 CFR part 50,
appendix L), the addition of an
alternative PM2.5 particle size separator
to that of the Well Impactor Ninety-Six
(WINS) and the Very Shape Cut Cyclone
(VSCC) size separators (88 FR 5663,
January 27, 2023). The new separator is
the TE–PM2.5C cyclone manufactured by
Tisch Environmental Inc.,180 Cleves
Ohio, which has been shown to have
performance equivalent to that of the
originally specified WINS impactor with
regards to aerodynamic cutpoint and
PM2.5 concentration measurement. In
addition, the new TE–PM2.5C has a
significantly longer service interval than
the WINS and is comparable to that of
the VSCC separator. Generally, the TE–
180 Mention of commercial names does not
constitute EPA endorsement.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
PM2.5C is also physically
interchangeable with the WINS and
VSCC where both are manufactured for
the same sampler. The proposed change
would allow either the WINS, VSCC, or
TE–PM2.5C to be used in a PM2.5 FRM
sampler. As is the case for the WINS
and VSCC, the TE–2.5C is now also an
approved size separator for candidate
PM2.5 FEMs. Currently, the EPA has
designated one PM2.5 sampler
configured with TE–PM2.5C separator as
a Class II PM2.5 equivalent method and
one as a PM10-2.5 equivalent method.
Upon promulgation of this change to
appendix L, these instruments would be
redesignated as PM2.5 and PM10-2.5
FRMs, respectively. Owners of such
samplers should contact the sampler
manufacturer to receive a new reference
method label for the samplers.
The EPA received only one comment
regarding this proposed change, which
was supportive. Therefore, the EPA is
finalizing this change to Appendix L as
proposed.
B. Issues Related to 40 CFR Part 53
(Reference and Equivalent Methods)
The EPA proposed to clarify the
regulations associated with FRM and
FEM applications for review by the EPA
(88 FR 5664, January 27, 2023).
Revisions were also proposed in
instances where current regulatory
specifications are no longer pertinent
and require updating. In addition, the
EPA proposed to correct a compiled a
list of noted minor errors in the
regulations associated with the testing
requirements and acceptance criteria for
FRMs and FEMs in part 53. These errors
are typically not associated with the
content of Federal Register documents
but often relate to transcription errors
and typographical errors in the
electronic CFR (eCFR) and printed
versions of the CFR.
1. Update to Program Title and Delivery
Address for FRM and FEM Applications
The EPA proposed a change to 40 CFR
53.4(a) to update the delivery address
for FRM and FEM Applications and
Modification Requests, as well as
update the name of the program
responsible for their review (88 FR
5664, January 27, 2023). These revisions
are due solely to organizational changes
and do not affect the structure or role of
the Reference and Equivalent Methods
Designation Program in reviewing new
FRM and FEM application requests and
requests to modify existing designated
instruments. The EPA received no
comments on this revision and,
therefore, the EPA is finalizing this
revision as proposed.
PO 00000
Frm 00144
Fmt 4701
Sfmt 4700
2. Requests for Delivery of a Candidate
FRM or FEM Instrument
The EPA proposed a change to 40 CFR
53.4(d), which currently allows the EPA
to request only candidate PM2.5 FRMs
and Class II or Class III equivalent
methods for testing purposes as part of
the applicant review process (88 FR
5664, January 27, 2023). The EPA
proposed to revise this section to enable
requesting any candidate FRM, FEM, or
a designated FRM or FEM associated
with a Modification Request, regardless
of NAAQS pollutant type or metric. The
EPA received no comments on these
revisions; therefore, the EPA is
finalizing this revision as proposed.
3. Amendments to Requirements for
Submission of Materials in 40 CFR
53.4(b)(7) for Language and Format
The EPA proposed a change to 40 CFR
53.4(b)(7) to specify that all written
FRM and FEM application materials
must be submitted to the EPA in English
in MS Word format and that submitted
data must be submitted in MS Excel
format (88 FR 5664, January 27, 2023).
The EPA received no comments on
these revisions; therefore, the EPA is
finalizing this section as proposed.
4. Amendment to Designation of
Reference and Equivalent Methods
The EPA proposed a change to 40 CFR
53.8(a) to clarify the terms of new FRM
and FEM methods to ensure that
candidate samplers and analyzers are
not publicly announced, marketed, or
sold until the EPA’s approval has been
formally announced in the Federal
Register (88 FR 5664, January 27, 2023).
The EPA received no comments on
these revisions; therefore, the EPA is
finalizing this section as proposed.
5. Amendment to One Test Field
Campaign Requirement for Class III
PM2.5 FEMs
The EPA proposed a change to 40 CFR
53.35(b)(1)(ii)(D) that involves field
comparability tests for candidate Class
III PM2.5 FEMs, including the
requirement that a total of five field
campaigns must be conducted at four
separate sites, A, B, C, and D (88 FR
5664, January 27, 2023). The existing
Site D specifications require that the site
‘‘shall be in a large city east of the
Mississippi River, having
characteristically high sulfate
concentrations and high humidity
levels.’’ However, dramatic decreases in
ambient sulfate concentration make it
difficult for applicants to routinely meet
the high sulfate concentration
requirement. Therefore, the EPA
proposed to revise the Site D
specifications to read ‘‘shall be in a large
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
city east of the Mississippi River, having
characteristically high humidity levels.’’
Only one comment was received on this
proposed revision, which was
supportive. Therefore, the EPA is
finalizing the revision to 40 CFR
53.35(b)(1)(ii)(D), as proposed.
ddrumheller on DSK120RN23PROD with RULES3
6. Amendment to Use of Monodisperse
Aerosol Generator
The EPA proposed a change to 40 CFR
53.61(g), 53.62(e), and Table F–1 that
involves the wind tunnel evaluation of
candidate PM10 inlets and candidate
PM2.5 fractionators under static
conditions, which requires the
generation and use of monodisperse
calibration aerosols of specified
aerodynamic sizes (88 FR 5664, January
27, 2023). In the current regulations, the
TSI Incorporated Vibrating Orifice
Aerosol Generator (VOAG) is the only
monodisperse generator that is
approved for this purpose. However,
TSI Incorporated no longer
manufacturers nor supports the VOAG.
Therefore, a commercially available
monodisperse aerosol generator (Model
1520 Fluidized Monodisperse Aerosol
Generator, MSP Corporation,
Shoreview, MN) has been added to list
of approved generators for this purpose.
No comments were received on this
revision; therefore, the EPA is finalizing
this revision as proposed.
7. Corrections to 40 CFR Part 53
(Reference and Equivalent Methods)
Certain provisions of 40 CFR 53.14,
Modification of a reference or
equivalent method, incorrectly state an
EPA response deadline of 30 days for
receipt of modification materials in
response to an EPA notice. Per a 2015
amendment (80 FR 65460, 65416, Oct.
26, 2015), all EPA response deadlines
for modifications of reference or
equivalent methods are 90 days from
day of receipt. Thus, the EPA proposed
a correction to specify the correct 90day deadline (88 FR 5664, January 27,
2023).
Requirements for Reference and
Equivalent Methods for Air Monitoring
of Criteria Pollutants identifies the
applicable 40 CFR part 50 appendices
and 40 CFR part 53 subparts for each
criteria pollutant. The four rows in the
section for PM10–2.5 erroneously do not
include the footnote instruction that the
aforementioned pollutant alternative
Class III requirements may be
substituted in regard to Appendix O to
Part 50—Reference Method for the
Determination of Coarse Particulate
Matter as PM10–2.5 in the Atmosphere.
Table B–1 specifies that the
interference equivalent for each
interferent is ±0.005 ppm for both the
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
standard-range and lower-range limits,
with the exception of nitric oxide (NO)
for the lower-range limit per note 4.
When testing the lower range of SO2, the
limit for NO is ±0.003 ppm, therefore,
an incorrect lower limit (±0.0003) is
currently stated in note 4 for this
exception to the SO2 lower range limit.
Thus, the EPA proposed a correction to
Table B–1 to specify the correct limit in
note 4 (88 FR 5664, January 27, 2023).
After the EPA received an inquiry
regarding the interaction of NO and O3,
the EPA investigated the interferent
testing requirements stated by 40 CFR
part 53, subpart B. The EPA has
determined that during the 2011 SO2
amendment and subsequent 2015 O3
amendment, several typographical
errors were introduced into Table B–3,
the most significant of which is the
omission of note 3, which instructs the
applicant to not mix the pollutant with
the interferent. Thus, the EPA proposed
revisions to Table B–3 to correct these
errors (88 FR 5664, January 27, 2023).
Additionally, appendix A to subpart B
of part 53 provides figures depicting
optional forms for reporting test results.
Figure B–3 lists an incorrect formula:
the lower detectible limit section is
missing the proper operator in the LDL
calculation formula and Figure B–5 lists
an incorrect calculation metric, and
there is a typesetting error in the
calculation of the standard deviation.
The EPA proposed to correct the
typesetting errors and noted other errors
to be corrected in several formulas
provided throughout § 53.43 (88 FR
5664, January 27, 2023).
The EPA proposed a revision to 40
CFR 53.43(a)(2)(xvi), 53.43(b)(2)(iv), and
53.43(b)(2)(iv) to correct typographical
errors in equations.
The EPA proposed a revision to Table
C–4 of part 53 Subpart C (88 FR 5700).
This change is related to field
comparability tests of candidate PM2.5,
PM10–2.5, and PM10 FEMs, which
requires testing at wide range of ambient
concentrations. For this reason, Table
C–4 specifies a minimum number of
valid sample sets to be conducted at
specified high concentrations. However,
due to the dramatic decrease in ambient
PM concentrations in the past two
decades, these number of valid test days
at high concentrations has been difficult
to achieve. Accordingly, the EPA
proposed to revise the testing
specifications for high concentration
events in Table C–4 to reflect current
levels of ambient PM for all three PM
metrics. In addition to the revision of
the ambient PM concentration
specifications to Table C–4, there are
also several entry errors that required
correction.
PO 00000
Frm 00145
Fmt 4701
Sfmt 4700
16345
The EPA received no comments on
these proposed revisions; therefore, the
EPA is finalizing the changes as
proposed.
C. Changes to 40 CFR Part 58 (Ambient
Air Quality Surveillance)
1. Quality Assurance Requirements for
Monitors Used in Evaluations for
National Ambient Air Quality Standards
In the proposal, the EPA described
how we evaluated the quality system as
part of the PM NAAQS reconsideration
(88 FR 5665, January 27, 2023). In this
section, the EPA identified several areas
for improvement in steadily declining
average ambient PM2.5 concentrations
across the country and the final decision
to revise primary annual PM2.5 NAAQS
described in section II above. We
assessed PM2.5 concentration data across
a range of values to determine if any
changes to the statistical calculations
used to evaluate the data quality in the
PM2.5 network were warranted. This
section describes the EPA’s assessment,
comments received, and the EPA’s final
decisions on the proposed changes.
Other changes in this section include
clarifications and other improvements
that will facilitate consistency and the
operation of quality assurance programs
by State, local, and Tribal (SLT)
agencies nationwide.
a. Quality System Requirements
The EPA reconsidered the appendix
A, section 2.3.1.1 goal for acceptable
measurement uncertainty (88 FR 5665,
January 27, 2023) for automated and
manual PM2.5 methods for total bias.
The existing total bias goal is an upper
90 percent confidence limit for the
coefficient of variation (CV) of 10
percent and ±10 percent for total bias.
The intent of the proposal was to
investigate if this bias goal is still
realistic given updated precision and
bias statistic. The EPA received one
comment that bias reevaluation may be
premature, since the final NAAQS
standard had not yet been determined at
the time of the proposal. The EPA
acknowledges this comment but
clarifies that the proposed new bias
statistic was evaluated at a range of
levels including the range of proposed
PM2.5 standards in the technical
memorandum, ‘‘Task 16 on PEP/NPAP
Task Order: Bias and Precision DQOs
for the PM2.5 Ambient Air Monitoring
Network.’’ 181 Considering the
181 Noah, G. (2023). Task 16 on PEP/NPAP Task
Order: Bias and Precision DQOs for the PM2.5
Ambient Air Monitoring Network. Memorandum to
the Rulemaking Docket for the Review of the
National Ambient Air Quality Standards for
E:\FR\FM\06MRR3.SGM
Continued
06MRR3
16346
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
justification in the technical
memorandum and the lack of adverse
comments regarding this part of the
proposal, the EPA is retaining the
appendix A, section 2.3.1.1, goal for
acceptable measurement uncertainty for
automated and manual PM2.5 methods
for total bias.
The EPA also proposed to update and
clarify ambient air monitoring
requirements found in 40 CFR part 58,
appendix A, section 2.6.1 pertaining to
EPA Protocol Gas standards used for
ambient air monitoring and the Ambient
Air Protocol Gas Verification Program
(PGVP) (88 FR 5665, January 27, 2023).
The EPA proposed to revise appendix A
to clarify that in order to participate in
the Ambient Air PGVP, producers of
Protocol Gases must adhere to the
requirements of 40 CFR 75.21(g), and
only regulatory ambient air monitoring
programs may submit cylinders for
assay verification to the EPA Ambient
Air PGVP. The EPA received mixed
comments in support of and in
opposition to this proposed revision.
The sole commenter opposing the
proposed revision indicated that the
proposed PGVP requirements would be
additional and is concerned with an
increased resource burden. But the EPA
responds that the PGVP requirements
that were proposed to be added are
consistent with the existing PGVP
requirements in 40 CFR 75.21(g), and
PGVP has been defined as a regulatory
requirement since 2016 (81 FR 17263,
March 28, 2016), so the proposed part
58 changes are not ‘‘additional’’ to
existing regulations. After consideration
of the comments, the EPA is finalizing
the update and clarification of ambient
air monitoring requirements found in
appendix A, section 2.6.1 pertaining to
EPA Protocol Gas standards used for
ambient air monitoring and the Ambient
Air PGVP as proposed.
b. Measurement Quality Check
Requirements
The EPA proposed to remove section
3.1.2.2 from appendix A, which allows
NO2 compressed gas standards to be
used to generate audit standards (88 FR
5665, January 27, 2023). The EPA
ddrumheller on DSK120RN23PROD with RULES3
Particulate Matter (EPA–HQ–OAR–2015–0072).
Available at: https://www.regulations.gov/docket/
EPA-HQ-OAR-2015-0072.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
received one comment supporting this
change. As a result of the comment
received and other general supportive
comments regarding quality assurance,
the EPA is finalizing the removal of
section 3.1.2.2 from appendix A as
proposed.
The EPA proposed to revise the
requirement in Appendix A, section
3.1.3.3 changing the National
Performance Audit Program (NPAP)
requirement for annual verification of
gaseous standards to the ORDrecommended certification periods
identified in Table 2–3 of the EPA
Traceability Protocol for Assay and
Certification of Gaseous Calibration
Standards (appendix A, section 6.0(4))
(88 FR 5665). The EPA received one
comment supporting this change. As a
result of the comment received and
other general supportive comments
regarding quality assurance, the EPA is
finalizing the updated NPAP gaseous
certification requirement in section
3.1.3.3 as proposed.
The EPA proposed to adjust the
minimum value required by appendix
A, section 3.2.4, to be considered valid
sample pairs for the PM2.5 Performance
Evaluation Program (PEP) from 3 mg/m3
to 2 mg/m3 (88 FR 5665, January 27,
2023). The EPA received comments in
support and against the change. In the
only opposing comment, the commenter
expressed concern that the method
detection limit (MDL) for PM2.5 is 2 mg/
m3. The commenter also indicated that
the MDL ‘‘typically has minimal value
per the definition of the MDL.’’ 40 CFR
part 50, appendix L states, ‘‘The lower
detection limit of the mass
concentration measurement range is
estimated to be approximately 2 mg/m3,
based on noted mass changes in field
blanks in conjunction with the 24 m3
nominal total air sample volume
specified for the 24-hour sample.’’ The
EPA notes that field blanks currently
average less than 10 mg nationally, and
when divided by the 24 m3 nominal
total air sample volume specified for a
24-hour sample, the result is 0.4 mg/m3.
The appendix L MDL referenced by the
commenter was part of the 1997 PM
NAAQS rulemaking (62 FR 38652, July
18, 1997); current data shows that the
MDL is substantially lower than the
EPA’s original estimate. After review of
PO 00000
Frm 00146
Fmt 4701
Sfmt 4700
the comments, and in consideration of
the recently calculated detection limit
for the PM2.5 FRM that is substantially
lower than our original estimate,182 the
EPA is finalizing the revised minimum
value for valid sample pairs for the
PM2.5 Performance Evaluation Program
(PEP) from 3 mg/m3 to 2 mg/m3 in
appendix A, section 3.2.4 as proposed.
c. Calculations for Data Quality
Assessments
The EPA proposed to change
Equations 6 and 7 of appendix A,
section 4.2.1 that are used to calculate
the Collocated Quality Control Sampler
Precision Estimate for PM10, PM2.5 and
Pb (88 FR 5666, January 27, 2023). The
proposed new statistics are designed to
address the high imprecision values that
result from using these calculations to
compare low concentrations that are
now more routinely observed in the
networks. The EPA received several
comments in support of this change in
general, but some commenters indicated
that they believed there was an error in
the new calculation that may result in
high imprecision from the calculation of
the equation. The EPA reviewed the
technical memorandum and confirmed
that a multiplier of 100 was
unintentionally left in the proposed
relative difference equation, Equation 6.
Also, equation 6 was corrected from a
normalized percent difference to a
normalized relative percent difference
that is appropriate for comparing
collocated pairs at low concentrations.
The technical memorandum titled
‘‘Task 16 on PEP/NPAP Task Order:
Bias and Precision DQOs for the PM2.5
Ambient Air Monitoring Network’’ has
been amended to correct the error and
is included in the docket for this
action.183
Equation 6 as proposed at 88 FR 5666
(January 27, 2023) was:
182 See the EPA’s PM
2.5 Data Quality Dashboard
available at https://sti-r-shiny.shinyapps.io/QVA_
Dashboard/.
183 Noah, G. (2023). Task 16 on PEP/NPAP Task
Order: Bias and Precision DQOs for the PM2.5
Ambient Air Monitoring Network. Memorandum to
the Rulemaking Docket for the Review of the
National Ambient Air Quality Standards for
Particulate Matter (EPA–HQ–OAR–2015–0072).
Available at: https://www.regulations.gov/docket/
EPA-HQ-OAR-2015-0072.
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
16347
meas- audit
Si
= - - - - - ; : = = - X 100
.Jaudit
And the corrected Equation 6 is:
X--Y.
=
t-
l
l
J(Xi - Yi)/2
'
Equation 7 is below and is unchanged.
= 100 *
CV90NAAQS
As a result of the positive comments
received and the correction to the
equation made in response to some
comments, the EPA is finalizing the
updated Equation 6 as described and is
finalizing Equation 7 as proposed for the
calculation of the Collocated Quality
Control Sampler Precision Estimate for
PM10, PM2.5, and Pb in section 4.2.1.
The EPA proposed to update the
appendix A, section 4.2.5, Equation 8,
calculation for the Performance
Evaluation Program Bias Estimate for
k-1
kx_Ef=1 tf-O:f:.1 tJ2 X
2k(k-1)
NAAQS Concentration*x:.1 ,k_ 1
PM2.5 (88 FR 5666–67, January 27,
2023). Because average ambient PM
concentrations across the nation have
steadily declined since the
promulgation of the PM2.5 standard, the
EPA proposed to replace the current
percent difference equation with a
relative difference equation. The EPA
received several comments in support of
this change in general, but some
commenters identified a potential error
in the new calculation that resulted in
an artificially high estimate, which they
100 *
~!1
s-
"'i=l
I
n.JNAAQS concentration
h
were
do not support. The EPA reviewed the
technical memorandum and discovered
that a multiplier of 100 was left in the
new relative difference equation used in
the bias equation. The technical
memorandum, ‘‘Task 16 on PEP/NPAP
Task Order: Bias and Precision DQOs
for the PM2.5 Ambient Air Monitoring
Network’’ has been amended to correct
the error and is included in the
docket.184 The proposed Equation 8
proposed at 88 FR 5667 (January 27,
2023) was:
meas- audit
St=
-- - x 100
-~
vaudit
and the corrected Equation 8 is:
t
n.JNAAQS concentration
where
St
meas - audit
=----'1audit
The EPA proposed to update the
references and hyperlinks in appendix
A, section 6 (88 FR 5667, January 27,
2023) to provide accuracy in identifying
and locating essential supporting
once a month, once a quarter, once
every six months, or distributed over all
four quarters depending on the check)
between checks and encourages
monitoring organizations to perform
data assessments at regular intervals.
The EPA received two comments
regarding this proposed footnote. One
commenter indicated that this change is
inconsistent with the QA Handbook for
Air Pollution Measurement Systems:
‘‘Volume II: Ambient Air Quality
Monitoring Program QA Handbook.’’
The EPA agrees with the commenter;
because the QA Handbook is guidance,
184 Noah, G. (2023). Task 16 on PEP/NPAP Task
Order: Bias and Precision DQOs for the PM2.5
Ambient Air Monitoring Network. Memorandum to
the Rulemaking Docket for the Review of the
National Ambient Air Quality Standards for
Particulate Matter (EPA–HQ–OAR–2015–0072).
Available at: https://www.regulations.gov/docket/
EPA-HQ-OAR-2015-0072.
d. References
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
PO 00000
Frm 00147
Fmt 4701
Sfmt 4700
E:\FR\FM\06MRR3.SGM
06MRR3
ER06MR24.032
i=
documentation and delete references to
historical documents that do not
represent current practices. The EPA
received only favorable comments, and
as a result, the EPA is finalizing the
updated the references and hyperlinks
in appendix A, section 6, as proposed.
The EPA also proposed to add a
footnote to Table A–1 of part 58,
appendix A—Minimum Data
Assessment Requirements for NAAQS
Related Criteria Pollutant Monitors (88
FR 5669, January 27, 2023). The
proposed footnote clarifies the
allowable time (i.e., every two weeks,
As a result of the supportive
comments received and the correction
to the equation in response to some
comments, the EPA is updating and
finalizing Equation 8 as described for
the calculation for the Performance
Evaluation Program Bias Estimate for
PM2.5, in section 4.2.5.
ddrumheller on DSK120RN23PROD with RULES3
Ln 1 s
ER06MR24.031
100 x
16348
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
the EPA will revise it after this action
is finalized to be consistent with the
updated CFR provision. Another
commenter does not support the
addition of the footnote due to concerns
about limiting flexibility. In response,
the EPA reiterates that the proposed
revision is intended to clarify intent and
does not make any changes to the
required frequencies or acceptance
criteria for data assessment. A ‘‘weight
of evidence’’ narrative is still found in
40 CFR part 58, appendix A, section
1.2.3. As a result of the comments
received and the rationale discussed
above, the EPA is finalizing the addition
of the new footnote to Table A–1 of part
58, appendix A—Minimum Data
Assessment Requirements for NAAQS
Related Criteria Pollutant Monitors as
proposed.
2. Quality Assurance Requirements for
Prevention of Significant Deterioration
(PSD) Air Monitoring
The EPA proposed to revise appendix
B, Quality Assurance Requirements for
Prevention of Significant Deterioration
(PSD) Air Monitoring (88 FR 5667,
January 27, 2023), in parallel to the
proposal to revise appendix A. Thus,
this section of the proposal included
similar detail and proposed revisions
related to evaluating quality system
statistical calculations for PM2.5,
clarifications and other improvements
that would facilitate consistency and the
operation of quality assurance programs
for PSD by SLT agencies nationwide.
ddrumheller on DSK120RN23PROD with RULES3
a. Quality System Requirements
The EPA reconsidered the goal in
appendix B, section 2.3.1.1 for
acceptable measurement uncertainty for
automated and manual PM2.5 methods
for total bias (88 FR 5668, January 27,
2023).185 The current total bias goal is
an upper 90 percent confidence limit for
the coefficient of variation (CV) of 10
percent and ±10 percent for total bias.
The EPA’s intent was to investigate if
this goal is still realistic given updated
precision and bias statistics. The EPA
received one comment that bias
reevaluation may be premature, since
the final NAAQS standard had not yet
185 In the proposal, in section VII.C.2 Quality
Assurance Requirements for Prevention of
Significant Deterioration (PSD) Air Monitoring (88
FR 5667–69), the EPA inadvertently referred to
‘‘appendix A’’ in the section rather than the correct
‘‘appendix B.’’ The EPA’s intent to have proposed
changes to appendix B on these pages is made clear
by the section header, the Table of Contents on page
5559, and the proposed regulatory text for appendix
B on pages 5707–08. See, e.g., id. at p.5668
(preamble erroneously states that the EPA proposed
to change appendix A, section 2.6.1); id. at p.5668
(preamble erroneously states that the EPA proposed
to adjust the minimum value required by appendix
A, section 3.2.4).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
been determined at the time of the
proposal. The EPA acknowledges this
comment but clarifies that the proposed
new bias statistic was evaluated at a
range of levels including the proposed
range of PM2.5 standards in the technical
memorandum, ‘‘Task 16 on PEP/NPAP
Task Order: Bias and Precision DQOs
for the PM2.5 Ambient Air Monitoring
Network.’’ 186 Considering the
justification in the technical
memorandum and the lack of adverse
comments regarding the substantive
proposal, the EPA is retaining the
appendix B, section 2.3.1.1, goal for
acceptable measurement uncertainty for
automated and manual PM2.5 methods
for total bias.
The EPA also proposed to update and
clarify ambient air monitoring
requirements found in 40 CFR part 58,
appendix B, section 2.6.1 pertaining to
EPA Protocol Gas standards used for
ambient air monitoring and the Ambient
Air PGVP (88 FR 5668, January 27,
2023). The EPA proposed to revise
appendix B to clarify that in order to
participate in the Ambient Air PGVP,
producers of Protocol Gases must
adhere to the requirements of 40 CFR
75.21(g), and only regulatory ambient
air monitoring programs may submit
cylinders for assay verification to the
EPA Ambient Air PGVP. The EPA
received comments in support of and in
opposition to this proposed revision.
The commenter opposing the revision
indicated that the proposed PGVP
requirements would be additional and is
concerned with an increased resource
burden. However, the EPA disagrees
with the commenter because that the
proposed PGVP requirements are
consistent with the existing PGVP
requirements in 40 CFR 75.21(g). PGVP
has been defined as a regulatory
requirement since 2016 (81 FR 17263,
March 28, 2016), so the proposed part
58 changes are not ‘‘additional’’ to
existing regulations. After consideration
of the comments, the EPA is finalizing
the update and clarification of ambient
air monitoring requirements found in
appendix B, section 2.6.1 pertaining to
EPA Protocol Gas standards used for
ambient air monitoring and the Ambient
Air PGVP as proposed.
186 Noah, G. (2023). Task 16 on PEP/NPAP Task
Order: Bias and Precision DQOs for the PM2.5
Ambient Air Monitoring Network. Memorandum to
the Rulemaking Docket for the Review of the
National Ambient Air Quality Standards for
Particulate Matter (EPA–HQ–OAR–2015–0072).
Available at: https://www.regulations.gov/docket/
EPA-HQ-OAR-2015-0072.
PO 00000
Frm 00148
Fmt 4701
Sfmt 4700
b. Measurement Quality Check
Requirements
The EPA proposed to remove section
3.1.2.2 from appendix B, which allows
NO2 compressed gas standards to be
used to generate audit standards (88 FR
5668, January 27, 2023). The EPA
received one comment supporting this
change. As a result of the comment
received and other general supportive
comments regarding quality assurance,
the EPA is finalizing the removal of
section 3.1.2.2 from appendix B as
proposed.
The EPA proposed to revise the
requirement in Appendix B, section
3.1.3.3 changing the National
Performance Audit Program (NPAP)
requirement for annual verification of
gaseous standards to the ORDrecommended certification periods
identified in Table 2–3 of the EPA
Traceability Protocol for Assay and
Certification of Gaseous Calibration
Standards (appendix B, section 6.0(4))
(88 FR 5668, January 27, 2023). The EPA
received one comment supporting this
change. As a result of the comment
received and other general supportive
comments regarding quality assurance,
the EPA is finalizing the updated NPAP
gaseous certification requirement in
section 3.1.3.3 as proposed.
The EPA proposed to adjust the
minimum value required by appendix
B, section 3.2.4, to be considered valid
sample pairs for the PM2.5 Performance
Evaluation Program (PEP) from 3 mg/m3
to 2 mg/m3 (88 FR 5668, January 27,
2023). The EPA received comments in
support and against the change. In the
only opposing comment, the commenter
expressed concern that the method
detection limit (MDL) for PM2.5 is 2 mg/
m3. The commenter also indicated that
the MDL ‘‘typically has minimal value
per the definition of the MDL.’’ 40 CFR
part 50, appendix L states, ‘‘The lower
detection limit of the mass
concentration measurement range is
estimated to be approximately 2 mg/m3,
based on noted mass changes in field
blanks in conjunction with the 24 m3
nominal total air sample volume
specified for the 24-hour sample’’. The
EPA notes that field blanks currently
average less than 10 mg nationally, and
when divided by the 24 m3 nominal
total air sample volume specified for a
24-hour sample, the result is 0.4 mg/m3.
The appendix L MDL referenced by the
commenter was part of the 1997 PM
NAAQS rulemaking more than 20 years
ago (62 FR 38652, July 18, 1997); current
data shows that the MDL is substantially
lower than EPA’s original estimate.
After review of the comments, and in
consideration of the recently calculated
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
detection limit for the PM2.5 FRM that
is substantially lower than our original
estimate, the EPA is revising the
minimum value for valid sample pairs
for the PM2.5 Performance Evaluation
Program (PEP) from 3 mg/m3 to 2 mg/m3
in appendix B, section 3.2.4 as
proposed.
c. Calculations for Data Quality
Assessments
The EPA proposed to change
Equations 6 and 7 of appendix B,
section 4.2.1 used for calculating the
Collocated Quality Control Sampler
Precision Estimate for PM10, PM2.5 and
Pb (88 FR 5707, January 27, 2023).
These new statistics are designed to
address the high imprecision values that
result from using these calculations to
compare low concentrations that are
now more routinely observed in the
networks. The EPA received several
comments in support of this change in
general, but a couple commenters
indicated that there could be an error in
the new calculation that resulted in high
imprecision from the calculation of the
equation. The EPA reviewed the
technical memorandum and discovered
16349
that a multiplier of 100 was
unintentionally left in the proposed
relative difference equation, Equation 6.
Also, equation 6 was corrected from a
normalized percent difference to a
normalized relative percent difference
that is appropriate for comparing
collocated pairs at low concentrations.
The technical memorandum titled
‘‘Task 16 on PEP/NPAP Task Order:
Bias and Precision DQOs for the PM2.5
Ambient Air Monitoring Network’’ was
amended to correct the error and is
included in the docket.187
Equation 6 in the proposal (88 FR 5668, January 27, 2023) was:
Si
=
meas-audit
-----;::::==-- X 100
✓audit
And the corrected Equation 6 is:
Equation 7 is below and is unchanged.
= 100 *
k-1
kxl;f=1 tr-O:f=1 ti)2 X
2k(k-1)
NAAQS Concentration•xi.1 ,1c_ 1
As a result of the positive comments
received and the correction to the
equation made in response to those
comments, the EPA is finalizing the
update to Equation 6 and retaining
Equation 7 as proposed for the
calculation of the Collocated Quality
Control Sampler Precision Estimate for
PM10, PM2.5 and Pb in section 4.2.1.
The EPA proposed to update the
appendix B, section 4.2.5, Equation 8,
calculation for the Performance
Evaluation Program Bias Estimate for
PM2.5 (88 FR 5668–59, January 27,
2023). Because average ambient PM
concentrations across the nation have
steadily declined since the
promulgation of the PM2.5 standard, the
EPA proposed to replace the current
percent difference equation with a
relative difference equation. The EPA
received several comments in support of
this change in general, but some
commenters identified a potential error
in the new calculation that resulted in
an artificially high estimate, which they
do not support. The EPA reviewed the
technical memorandum and discovered
that a multiplier of 100 was left in the
new relative difference equation used in
the bias equation. The technical
memorandum, ‘‘Task 16 on PEP/NPAP
Task Order: Bias and Precision DQOs
for the PM2.5 Ambient Air Monitoring
Network’’ has been amended to correct
the error and is included in the docket.
The proposed Equation 8 (88 FR 5669,
January 27, 2023) was:
187 Noah, G. (2023). Task 16 on PEP/NPAP Task
Order: Bias and Precision DQOs for the PM2.5
Ambient Air Monitoring Network. Memorandum to
the Rulemaking Docket for the Review of the
National Ambient Air Quality Standards for
Particulate Matter (EPA–HQ–OAR–2015–0072).
Available at: https://www.regulations.gov/docket/
EPA-HQ-OAR-2015-0072.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
PO 00000
Frm 00149
Fmt 4701
Sfmt 4700
E:\FR\FM\06MRR3.SGM
06MRR3
ER06MR24.033
ddrumheller on DSK120RN23PROD with RULES3
CV90NAAQS
16350
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
100 *
:I;!t 1 S·
,= '
where
n.jNAAQS concentration
St
- audit
= meas
..Jaiiiilf x 100
audit
and the corrected Equation 8 is:
ddrumheller on DSK120RN23PROD with RULES3
As a result of the supportive
comments received and the correction
to the equation in response to some
comments, the EPA is updating and
finalizing Equation 8 as described for
the calculation for the Performance
Evaluation Program Bias Estimate for
PM2.5, in section 4.2.5.
Lni=l s t
n,JNAAQS concentration
where s,
received and the rationale discussed
above, the EPA is adding the new
footnote to Table B–1 of part 58,
appendix B—Minimum Data
Assessment Requirements for NAAQS
Related Criteria Pollutant PSD Monitors
as proposed.
d. References
3. Amendments to PM Ambient Air
Quality Methodology
The EPA proposed to update the
references and hyperlinks in appendix
B, section 6 (88 FR 5669, January 27,
2023) to provide accuracy in identifying
and locating essential supporting
documentation and delete references to
historical documents that do not
represent current practices. The EPA
received only favorable comments, and
as a result, the EPA is finalizing the
updated the references and hyperlinks
in appendix B, section 6, as proposed.
The EPA also proposed to add a
footnote to Table B–1 of part 58,
appendix B—Minimum Data
Assessment Requirements for NAAQS
Related Criteria Pollutant PSD Monitors
(88 FR 5669, January 27, 2023). The
proposed footnote clarifies the
allowable time (i.e., every two weeks,
once a month, once a quarter, once
every six months, or distributed over all
four quarters depending on the check)
between checks and encourages
monitoring organizations to perform
data assessments at regular intervals.
The EPA received two comments
regarding this proposal. One commenter
indicated that this change is
inconsistent with the QA Handbook.
The EPA agrees with the commenter;
because the QA Handbook is guidance,
the EPA will revise it after this action
is finalized to be consistent with the
updated CFR provision. Another
commenter does not support the
addition of the footnote due to concerns
about limiting flexibility. In response,
the EPA reiterates that the proposed
revision is intended to clarify intent and
does not make any changes to the
required frequencies or acceptance
criteria for data assessment. A ‘‘weight
of evidence’’ narrative is still found in
40 CFR part 58, appendix B, section
1.2.3. As a result of the comments
a. Revoking Approved Regional
Methods (ARMs)
The EPA proposed to remove
provisions for approval and use of
Approved Regional Methods (ARMs)
throughout parts 50 and 58 of the CFR
(88 FR 5669, January 27, 2023). ARMs
are continuous PM2.5 methods that have
been approved specifically within a
State or local air agency monitoring
network for purposes of comparison to
the NAAQS and to meet other
monitoring objectives. Currently, there
are no approved ARMs. There are,
however, more than a dozen approved
Federal Equivalent Methods (FEMs) for
PM2.5. These approved FEMs are eligible
for comparison to the NAAQS and to
meet other monitoring objectives.
The EPA received comments from
multiple State air programs in support
of the proposal to remove provisions for
approval and use of ARMs. One
commenter cites that there are multiple
FEMs available for monitoring agencies
to work with and that the agency was
never able to get a candidate ARM to
meet the requirements for approval.
With the availability of multiple FEMs
that now work in the monitoring
agency’s network, the commenting
agency does not anticipate the need to
ever pursue an ARM in the future and,
therefore, suggests that the ARM
provision is no longer needed. Another
commenter strongly supported the
proposed changes to remove the ARM
provisions. The EPA also received
comments from a few agencies that
supported retaining the ARM provisions
instead. One commenter cited the need
to consider the rapid advancement of
various new technologies and that, in
some cases, approved continuous FEMs
may have shortcomings, meaning that
losing the ability to propose an ARM in
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
PO 00000
Frm 00150
Fmt 4701
Sfmt 4700
meas - audit
=----,/audit
the future may limit useful alternative
options to monitoring agencies. Another
commenter suggested that the removal
of the ARM would take away the ability
and right to use locally derived
correction factors.
After considering the comments for
and against removing the provisions for
ARMs, the EPA believes it is most
appropriate to remove the ARM
provisions. As described in the
proposal, when the EPA first proposed
the process for approving and using
ARMs, there were no continuous FEMs
approved. There are now over a dozen
approved PM2.5 continuous FEMs and
no approved ARMs. Therefore, the EPA
is finalizing the removal of ARMs
throughout 40 CFR parts 50 and 58 as
proposed.
b. Calibration of PM Federal Equivalent
Methods (FEMs)
The EPA proposed to modify its
specifications for PM FEMs in appendix
C to Part 58 (88 FR 5670–73, January 27,
2023). Specifically, the EPA proposed
that valid State, local, and Tribal (SLT)
air monitoring data from Federal
Reference Methods (FRMs) generated in
routine networks and submitted to the
EPA may be used to improve the PM
concentration measurement
performance of approved FEMs. This
approach, initiated by instrument
manufacturers, would be implemented
as a national solution in factory
calibrations of approved FEMs through
a firmware update. This could apply to
any PM FEM methods (i.e., PM10, PM2.5,
and PM10–2.5).
The EPA proposed this modification
because there are some approved PM
FEMs that are not currently meeting bias
measurement quality objectives (MQOs)
when evaluating data nationally as
described in the 2022 PA (U.S. EPA,
2022b, section 2.2.3.1), meaning that an
update to factory calibrations may be
appropriate; however, there is no clearly
defined process to update the
calibration of FEMs. While there are
several types of data available to use as
the reference for such updates (e.g.,
routinely operated FRMs, audit program
FRMs, and chemical speciation sampler
E:\FR\FM\06MRR3.SGM
06MRR3
ER06MR24.034
100 x
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
data), we proposed to use routinely
operated SLT FRMs as the basis of
comparison upon which to calibrate
FEMs. The goal of updating factory
calibrations would be to increase the
number of routinely operating FEMs
meeting bias MQOs across the networks
in which they are operated. While there
are other approaches that could improve
data comparability between PM FEMs
and collocated FRMs, the EPA believes
that the proposed modification to
calibrate PM FEMs represents the most
reliable approach to update FEM factory
calibrations, since the existing FRM
network data that meet MQOs would be
used to set updated factory calibrations.
While the Agency proposed to add
this language to more expressly define
a process to update factory calibrations
of approved PM FEMs, the EPA believes
that the existing rules for updating
approved FRMs and FEMs found at 40
CFR 53.14 may also continue to be
utilized for this purpose, as appropriate.
40 CFR 53.14 allows instrument
manufactures to submit to the EPA a
‘‘Modification of a reference or
equivalent method.’’ Submitting a
modification request may be appropriate
to ensure an approved FEM continues to
meet 40 CFR 53.9, ‘‘Conditions of
designation.’’ Specifically, 40 CFR
53.9(c) requires that, ‘‘Any analyzer,
PM10 sampler, PM2.5 sampler, or
PM10–2.5 sampler offered for sale as part
of an FRM or FEM shall function within
the limits of the performance
specifications referred to in § 53.20(a),
§ 53.30(a), § 53.35, § 53.50, or § 53.60, as
applicable, for at least 1 year after
delivery and acceptance when
maintained and operated in accordance
with the manual referred to in
§ 53.4(b)(3).’’ Thus, instrument
manufacturers are encouraged to seek
improvements to their approved FEM
methods as needed to continue to meet
data quality needs as operated across
the network.
There are several technical
components to EPA’s proposed
modification, including: the reference
data to be used in the calibrations;
implementing as a national solution in
factory calibrations of approved FEMs
through firmware updates; application
to any PM FEM methods (i.e., PM10,
PM2.5, and PM10–2.5); the appropriate
range of data to be used to develop and
test new factory calibrations, from just
the most representative concentrations
up to all available concentrations; the
representative set of geographic
locations that can be used; whether
outliers may be included or not
included; that new factory calibrations
should be developed using data from at
least 2 years and tested on data from a
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
separate year or years; that updates to
factory calibrations can occur as often as
needed; that calibrations should be
evaluated by monitoring agencies as
part of routine data assessments, e.g.,
during certification of data and 5-year
assessments; the EPA’s recognition that
only data from existing operating sites is
available; and finally, that an updated
factory calibration does not have to
work with the original field study data
submitted that led to the original FEM
designation.
With the proposed modification, the
EPA solicited input on these technical
issues as well as the overall approach
and any alternatives that could lead to
more sites meeting the bias MQO with
automated FEMs, especially for those
sites that are near the level of the
primary annual PM2.5 NAAQS, as
proposed to be revised in section II
above. In response, the EPA received
comments from about two dozen
entities, most of which were SLT air
programs or Multi-Jurisdictional
Organizations (MJOs) comprised of
these entities.
Overall, there was broad and strong
support from a majority of commenters
for the proposed requirement to use
FRM data generated in routine networks
and submitted to the EPA to update
factory calibrations included as part of
approved FEMs. There were a smaller
number of critical comments on the
proposed process as well as some
commentors that supported the
proposed requirement but also provided
additional suggestions for the EPA’s
consideration. Below, we address each
of the areas on which the EPA requested
comment regarding the calibration of
PM FEMs, as well as a few additional
areas where multiple commenters
offered input on other areas related to
our proposal.
A majority of the commenters on the
proposed PM FEM calibration process
support the process to use valid State,
local, and Tribal FRM data generated in
routine networks and submitted to the
EPA to improve the PM concentration
measurement performance of approved
FEMs. Some commenters suggested that
this action is needed to ensure that data
reported from FRMs and FEMs are
comparable and correction methods
applied to data from FEM monitors are
defensible across the national PM
monitoring network. Others stated that
they agree with the EPA that this is a
critical step in the right direction to
account for the discrepancies between
PM2.5 FRM data and PM2.5 FEM data.
Some commented that applying
corrections includes a recognition that,
while different measurement principles
may produce differences in the resulting
PO 00000
Frm 00151
Fmt 4701
Sfmt 4700
16351
data, having an approach that
minimizes bias is extremely important.
Finally, some stated their belief that a
correction factor is necessary to preserve
data integrity with the FRM.
The EPA also received comments
suggesting ways that the PM FEM
correction could be performed,
including through detailed analysis of
data; by having PM FEM instrument
manufacturers evaluate nationally
available valid FRM data to update
factory calibrations; and, by having the
instrument manufacturers implement
calibration adjustments at the factory.
The EPA also received supportive
comments on the PM FEMs calibration
relating to comparability to the NAAQS.
For example, a commenter stated that it
is important to ensure bias MQOs are
met for FEMs run at sites potentially
affected by revised standards as well as
the need to accurately designate areas as
attaining or not attaining the NAAQS.
There were comments supporting the
correction of PM FEM data as helping
the EPA and SLT monitoring programs
continue to evolve toward more
automated methods. For example, one
commenter appreciates the EPA’s
support for the ongoing move from
filter-based PM2.5 FRMs to use of
continuous FEMs, stating that they
concur with the EPA’s assessment that
there is monitoring bias between FRMs
and FEMs, and commending EPA for
recognizing ongoing data quality issues
for FEMs and for taking action to
improve these issues in collaboration
with instrument manufacturers and SLT
agencies.
A small number of commenters were
critical of the proposed FEM calibration
approach. One commenter noted that
EPA should further examine the
handling of FEM PM2.5 data when used
for comparison to the NAAQS. In
response, we note that monitoring
agencies and the EPA will continue to
examine the comparability and use of
FEM data used in comparison to the
NAAQS. Another commenter suggested
that the calibration process for a
designated PM monitor should not be
altered following Class III designation
approval. The EPA disagrees as we
believe it is appropriate for FEMs to be
calibrated with routinely operated
FRMs, because doing so is an efficient
way to work towards FEM data meeting
the bias MQO across the networks in
which the FEMs are currently being
operated. Also, having continuous PM
FEMs meeting bias MQOs allows the
use of the data in a variety of other ways
that manually operated FRMs samplers
cannot support. Another commenter
stated that, if a particular FEM
designated make or model of
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16352
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
instruments fails to meet MQOs, then
that make or model should be removed
from the designations altogether. The
EPA agrees and clarifies that the
modification would not prevent removal
of FEM designation from a make or
model of instrument under the existing
40 CFR 53.11—Cancellation of reference
or equivalent method designation. This
may be appropriate if there are no other
solutions to improve the method such
that it achieves bias MQOs.
A few commenters provided specific
recommendations for how the
regulatory language could be improved.
These included comments that the new
regulatory language proposed for 40
CFR part 58, appendix C, section 2.2
must ensure consistency and
transparency when requesting changes
to the factory calibration; that the EPA
should incorporate binding regulatory
language in 40 CFR part 58, appendix C,
section 2.2 (i.e., it currently lacks
‘‘shall’’ or ‘‘must’’) to ensure the
language is not open to inconsistency
and does not provide unique deference
to instrument manufacturers without a
mechanism for transparent
communication of the changes being
made and the supporting technical
analysis. A commenter also requested
that the EPA define the core
requirements needed to ensure all
requests for updating factory
calibrations are required to follow the
same process, using data of the same
known quality, and evaluating the
effectiveness of the resulting correction
factors consistently.
In response to these comments, while
the EPA agrees that the proposed
regulatory language for 40 CFR part 58,
appendix C, section 2.2 must ensure
consistency and transparency when
entities request changes to factory
calibrations, the EPA disagrees that the
regulations cannot also provide some
flexibility. For example, we believe that
a degree of flexibility is appropriate
regarding whether outliers in the data to
be used for factory calibration should or
should not be included, the range of
data to be included, and in utilizing
collocated FRM and FEM data for
updated calibrations from a
representative set of geographic areas in
which it is produced. The EPA believes
that the proposal defined the core
requirements needed to ensure all
requests for updating FEM factory
calibrations will follow the same
process, using data of the same known
quality and evaluating the effectiveness
of the resulting correction factors
consistently.
In its proposal, the EPA identified
that while there are several types of data
available to use as the reference for FEM
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
calibration updates, including data from
routinely operated FRMs, audit program
FRMs, and PM2.5 chemical speciation
samplers, the EPA proposed to use
routinely operated State, local, and
Tribal FRMs as the basis of comparison
upon which to calibrate FEMs (88 FR
5670–71, January 27, 2023).
Importantly, routine SLT agency FRM
data form the largest portion of the
monitored air quality data used in
epidemiologic studies that are being
used to inform proposed decisions
regarding the adequacy of the public
health protection afforded by the
primary PM2.5 NAAQS, as discussed in
section II above.
Overall, there was broad and strong
support for utilizing collocated FRM
data from routine SLT networks to
provide calibrations of the continuous
FEMs. For example, several commenters
agree that valid SLT air monitoring data
generated in routine networks and
submitted to the EPA will improve the
PM concentration measurement
performance of approved FEMs.
Another commenter provided support
for PM FEM instrument manufacturers
to evaluate nationally available valid
FRM data as well as other data sets such
as the performance evaluation audit
program to update factory calibrations.
The EPA believes that the routinely
operated PM FRMs represent the best
and largest source of data to calibrate
continuous PM FEMs, and that
performance evaluation audit program
data should be kept independent of the
calibration process. This will mean that
assessments of the routine monitoring
operations, including both the FRM and
any future updated PM FEMs, will
appropriately remain independent in
evaluating whether updated methods
are meeting bias MQOs. The EPA is,
therefore, finalizing its approach to use
routinely operated SLT FRMs as the
basis of comparison upon which to
calibrate continuous PM FEMs as
proposed.
Regarding the EPA’s proposed
requirement to utilize factory
calibrations (88 FR 5670–71, January 27,
2023), several commenters agreed that
factory calibrations provide the best
option to improve PM FEMs. For
example, one commenter stated that the
correction factors are necessary to
preserve data integrity with the FRM,
and they support the proposal that the
approach be initiated by instrument
manufacturers and implemented as a
national solution through firmware
updates.
Regarding the proposed requirement
that calibrations be initiated by
instrument manufacturers (88 FR 5671,
January 27, 2023), most commenters
PO 00000
Frm 00152
Fmt 4701
Sfmt 4700
were supportive of the proposed
approach that recalibration of FEM PM
instruments be initiated by instrument
manufacturers. For example, one
commenter stated they support allowing
instrument companies submit
improvements to their existing FEMs, as
vendors should be encouraged to
improve their methods. Another
commenter noted that having a
methodology initiated by the
manufacturer will have nationwide
consistency. A few of commenters
recommended that SLT air agencies
should have the additional ability to
petition the EPA Administrator to
initiate factory calibrations of FEMs to
better meet MQOs when data collected
by their agencies indicate disparities,
because the monitoring agencies are
responsible for the quality of the data
from the specific makes and models of
instrumentation used in their networks.
While the EPA believes that, in most
cases, the instrument companies should
be the ones to initiate the process for
calibration of FEMs to routinely
operated FRMs, we agree with the
commenters who suggested that other
options should be available, including
allowing monitoring agencies or MJOs
to work independently or together to
pursue improvements to designated
FEMs. However, the EPA believes that
any such improvements initiated by
monitoring agencies or MJOs should
still be facilitated through the
responsible instrument company. Also,
any such effort to improve data quality
should be employed across all the
networks in which the methods are
operated and not limited to the
networks operated by the agency(s)
pursuing such improvements.
Regarding how frequently factory
calibrations should be updated, our
proposal identified that it would be
most appropriate to not define a specific
time period for updates; rather, updates
should be based on whether or not
quality data is being produced across a
given network (88 FR 5672, January 27,
2023). Regarding this issue, one
commenter recommended that
instrument manufacturers be required to
evaluate and, if necessary, adjust PM
FEMs factory calibrations on an ongoing
basis at regular intervals. The EPA notes
that while it does not have the authority
to require instrument companies to
evaluate the quality of data from
operating FEMs under 40 CFR part 58,
the EPA does routinely participate in
conferences and workshops and makes
assessments of data quality specific to
instrument makes and models publicly
available. The EPA also regularly
summarizes relevant FRM and FEM data
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
quality in documents such as the 2022
PA (U.S. EPA, 2022b). Therefore,
consistent with the proposal, we are not
finalizing any specifics regarding how
frequently factory calibrations should be
updated but commit to continue to
routinely provide information to SLT
agencies regarding FEM data quality.
The EPA proposed that the calibration
of FEMs could apply to any of the PM
FEM method indicators (i.e., PM10,
PM2.5, and PM10–2.5) (88 FR 5670,
January 27, 2023). The EPA received
only supportive comments. All
comments that included a discussion of
three PM metrics support their
inclusion for calibration of PM FEMs.
Therefore, the EPA is finalizing the
inclusion of all three PM indicators (i.e.,
PM10, PM2.5, and PM10–2.5) as proposed.
The EPA proposed that either all data
available or a range of data up to 125%
of the 24-hour NAAQS for the PM
indicator of interest may be used to
establish new factory calibrations, (88
FR 5671–73, January 27, 2023). The EPA
received many comments supportive of
the proposal and one comment offering
a different approach on the range of data
to use. One commenter recommends
that the EPA should consider using all
‘‘validated’’ data because how these
instruments behave under normal
operating ranges may be just as
important as how they behave when
monitoring conditions are low or
elevated, and that the full range of data
should be used when determining the
appropriate level of the standard, just as
the full range of data is used in
determining if an area is attaining the
standard. In response to this comment,
the EPA believes that making
allowances for some flexibilities will
increase the likelihood of instrument
companies pursuing such
improvements. Also, even though there
is flexibility, the EPA will still be able
to evaluate the appropriateness of a
range of concentration data included as
part of each application submitted.
Also, the EPA notes that in certain
circumstances, States do petition the
EPA to set aside data under the
Exceptional Events Rule (§ 50.14,
‘‘Treatment of air quality monitoring
data influenced by exceptional events’’).
Where approved, exceptional event data
are set aside from use in regulatory
decisions. Thus, there is a process to set
aside certain high concentration data for
certain purposes. Therefore, the EPA is
finalizing the provision that factory
calibrations may be based on a range of
valid data as proposed.
The EPA solicited comment on the
representative set of geographic
locations to use in the calibration of
FEMs compared to collocated FEMs (88
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
FR 5671, January 27, 2023). Most
commenters were supportive of the
approach of using representative sites in
SLT networks from across the country.
For example, several commenters
provided their support for PM FEM
instrument manufacturers to evaluate
nationally available valid FRM data to
update factory calibrations. Commenters
disagreeing with a national geographic
approach preferred to allow local
solutions to correct data. For example,
one commenter suggested having a local
or regional option because PM
instruments are impacted by, and
respond differently to, a variety of local
factors, including relative humidity,
temperature, concentration levels, and
particle composition. The EPA agrees
that there are challenges in the response
of PM FEMs to a variety of local factors;
however, this can be true of many
methods and are not specific to PM
FEMs and, therefore, does not provide a
reason to reject this approach in this
instance. Another commenter stated that
the proposed national correction factor
is a ‘‘flawed concept,’’ suggesting that it
is ‘‘widely understood throughout the
monitoring community that monitors
perform best with a local correction
factor.’’ This commentor offered no
record or citation supporting this point.
The EPA counters that while monitoring
agencies may statistically correct data
from a PM continuous monitor for AQI
purposes (40 CFR part 58, appendix G),
there are both examples of well
performing statistically corrected PM
continuous monitors being used for AQI
purposes; however, without proper
attention and updates, there are also
examples of poorly performing ones.
Finally, another commenter believes
that a national correction factor cannot
possibly incorporate data to represent
all the scenarios across the nation that
have an impact on monitor performance
and data quality. Although the EPA
agrees that there are a variety of local
scenarios that could affect monitor
performance, the overall benefits of
having nationally consistent
measurement of PM concentrations and
national calibration of data outweigh the
potential advantages of locally specific
calibrations.
Several commenters also disagreed
with using local and regional
calibrations of data, including some
monitoring agencies that asserted being
unable to reinvest in the operation of
FRMs that would be required to locally
calibrate their own PM FEMs. Further,
every approved PM FEM method
designated today is effectively
calibrated through demonstration of
field testing in the areas in which it was
PO 00000
Frm 00153
Fmt 4701
Sfmt 4700
16353
required to be tested (40 CFR
53.35(b)(1)). Moreover, the EPA
proposed to require instrument
manufacturers to demonstrate that they
can improve the number of sites
meeting bias MQOs by initiating a
recalibration of an FEM. Thus, the use
of a national set of sites where the
methods are operated is essentially a
fine-tuning of the PM FEMs
performance across all sites where it is
used.
After considering all the comments
received, the EPA believes it is
appropriate to finalize as proposed with
a representative set geographic locations
at SLT sites to calibrate PM FEMs.
Identification of such sites would be
made by the applicant of the planned
updated calibration, subject to EPA
approval, and submitted to the EPA in
accordance with the requirements and
application instructions in 40 CFR part
58, appendix C, sections 2.2 and 2.7.
The EPA encourages early
communication between an applicant
seeking a method update and the EPA
to facilitate the most appropriate sites
are included in any updated application
of the methods calibration.
The EPA proposed that instrument
companies may, but are not required to,
check for and exclude any potential
outliers that may exist in the validated
State, local, and Tribal agency network
data available from AQS that would be
used to establish new factory
calibrations. The EPA received two
comments regarding potential outlier
approaches. One commenter disagreed
with the proposed approach and instead
recommended the use of all ‘‘validated’’
data, because how these instruments
behave under normal operating ranges
may be just as important as how they
behave when monitoring conditions are
low or elevated. The EPA acknowledges
this point; however, the proposal on
outliers allows flexibility in using
standard outlier tests if needed to
include or exclude such data as part of
the calibration process. Ultimately, the
true test of success for an updated
method calibration will be that a higher
number of sites are meeting bias MQOs
in the areas in which the method is
used, which will include all routine
valid data including any potential
outliers. Another commenter asserted
concerns with the ability of instrument
manufacturers to analyze data within
individual monitoring agencies. The
EPA disagrees with the commenter
because decisions whether to include or
exclude outliers should be flexible and
made on a case-by-case basis. Moreover,
the expected substantially larger dataset
from routinely operated collocated
FRMs and FEMs compared to what was
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16354
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
used in the original FEM designation
testing (§ 53.35 Test procedure for Class
II and Class III methods for PM2.5 and
PM10–2.5) will minimize the effect of any
potential outliers.
In contrast to these two comments,
the EPA received many comments
supportive of the proposed outlier
approach overall. Therefore, the EPA is
finalizing this part of the proposal that
instrument companies may, but are not
required to, check for and exclude any
potential outliers that may exist in the
validated State, local, and Tribal agency
network data available from AQS that
would be used to establish new factory
calibrations.
Several commenters offered input on
statistical criteria and initial testing
requirements for approval of candidate
PM FEMs and the role of instrument
manufacturers in this process. The EPA
did not propose any changes related to
these issues; however, these comments
have been considered below.
One commenter suggested that data
quality objectives, bias, and precision
estimators for different monitoring
methods should be based on averages at
both national and regional levels for
purposes of comparison. Another
commenter asked to strengthen the
criteria for Class 3 Equivalency
standards for candidate PM
instrumentation. On testing
requirements, one commenter
recommended that the EPA consider
updating the 40 CFR part 53 process for
approving FEMs so that the testing
process more closely reflects the
regulatory deployment and data
handling that generates NAAQScomparable data. Another commenter
asked that the results from ‘‘summer’’
and ‘‘winter’’ field evaluations not be
averaged together because it allows
agencies to minimize the error of biased
instruments by averaging poor results
with data often biased in the other
direction. The same commenter also
recommended that candidate
instruments data sets should not be
averaged together as is done currently
where data from triplicate instruments
are averaged for each day. Another
commenter asked that the EPA require
FEM field comparability tests in the
northwest (e.g., in EPA Region 10) in
areas where particulate derived from
biomass predominates to ensure that
certified instruments will perform
reliably in regions influenced by these
sources. Related to the different
measurement principles and the
instrument companies’ role in PM
FEMs, one commenter noted that FEMs
may never align perfectly with the
FRMs due to the use of different
measurement principles. Another
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
commenter asked that manufacturers of
FEM instruments be held accountable
for ensuring that they continue to meet
FEM criteria, whether through
calibration updates and/or follow-up
evaluations. Another commenter
suggested that instrument
manufacturers should be required to
further evaluate the FEM monitoring
data at defined intervals including, but
not limited to, the 2-year and 5-year
approval anniversaries.
The EPA did not propose to make
modifications to the statistical criteria or
testing requirements; however, we did
solicit comment on any alternatives that
would lead to more sites meeting the
bias MQO with automated FEMs,
especially for those sites that are near
the level of the primary annual PM2.5
NAAQS as proposed (88 FR 5672–73,
January 27, 2023). While the comments
requesting that the statistical criteria be
strengthened may have merit, doing so
would not address the large inventory of
already deployed PM FEMs used
throughout the country. Also, without
performing a detailed Data Quality
Objective (DQO) design process, it is
unclear how changing one or more
statistical criteria would help improve
the number of sites meeting the bias
MQO now or in the future. Similarly,
while the comments asking for changes
to the locations of testing may also have
merit, the EPA believes this could be a
deterrent for instrument manufactures
to seek additional improvements since
more testing would be required, at least
for candidate methods. Regarding the
comment on the different measurement
principles, the EPA concurs that
different measurement principles may
never align perfectly. Also, the EPA
notes that the Agency has longstanding
goals for acceptable measurement
uncertainty of automated and manual
PM2.5 methods in 40 CFR part 58,
appendix A, section 2.3.1.1. Therefore,
while having different measurement
principles align is useful, meeting the
goal for acceptable measurement
uncertainty is the objective.
Regarding the comments related to the
instrument companies’ role in PM
FEMs, the EPA notes that FEMs are
already required to meet 40 CFR 53.9,
‘‘Conditions of designation.’’
Specifically, 40 CFR 53.9(c) requires
that, ‘‘Any analyzer, PM10 sampler,
PM2.5 sampler, or PM10–2.5 sampler
offered for sale as part of an FRM or
FEM shall function within the limits of
the performance specifications referred
to in § 53.20(a), § 53.30(a), § 53.35,
§ 53.50, or § 53.60, as applicable, for at
least 1 year after delivery and
acceptance when maintained and
operated in accordance with the manual
PO 00000
Frm 00154
Fmt 4701
Sfmt 4700
referred to in § 53.4(b)(3).’’ The EPA
does not have the authority to require
instrument manufacturers to further
evaluate the FEM monitoring data at
defined intervals, including but not
limited to the 2-year and 5-year
approval anniversaries, as one
commenter suggested.
In addition to these few
recommendations, the EPA received
many comments supportive of the
proposal that valid State, local, and
Tribal air monitoring data from FRMs
generated in routine networks and
submitted to the EPA may be used to
improve the PM concentration
measurement performance of approved
FEMs; therefore, consistent with the
proposal we are not finalizing any
updates to the statistical criteria, testing
requirements, or requirements on
instrument manufactures as proposed.
The EPA proposed that any new
factory calibration should be developed
using data from at least 2 years and
tested on a separate year(s) of data (88
FR 5672, January 27, 2023). Comments
on this part of the proposal were
generally supportive. One commenter
requested that at least a 3-year dataset,
rather than the proposed 2 years, be
used for a representative design value
comparison of the FEM and FRM
datasets to be evaluated. Another
commenter pointed out that as large a
data set as possible should be used, but
EPA should not limit it to only data
collected by instruments that have
operated for more than 2 years.
In response to these comments, the
EPA notes the broad support for the
proposal as written. Also, the EPA notes
that the 2-year period for using data to
develop a factory calibration is a
minimum, and that more years may be
used as appropriate. Therefore, the EPA
is finalizing its approach that any new
factory calibration should be developed
using data from at least 2 years and
tested on a separate year(s) of data as
proposed.
The EPA proposed several aspects of
the FEM calibration on which we did
not receive specific comments,
including a provision that FEM methods
should be evaluated by monitoring
agencies as part of routine data
assessments, such as during certification
of data and 5-year assessments; the fact
that the EPA recognizes only data from
existing operating sites are available for
use in factory calibrations; and
recognition that an updated factory
calibration does not have to work with
the original field study data submitted
that led to the designation as an FEM.
With the broad general support from
commenters summarized above, the
EPA is finalizing each of these
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
individual aspects of the FEM
calibration as proposed.
In the proposal, the EPA identified
that we should expect a lag between the
date when an already designated
method is approved with a new factory
calibration as an updated method by the
EPA and when it can be implemented
in the field. The EPA solicited comment
on how to approach the data produced
during this lag. Commenters provided
input not only on how to address data
during the lag, but also regarding how
to address data already collected prior
to a method update that has the
potential to be used in regulatory
decision making, particularly where
such collected data do not meet the bias
MQO. In response to this solicitation of
comment, there was a consistent
recommendation that calibrations of
data associated with method updates
should be applied to all relevant PM
data prior to the EPA using it for
designations under a final NAAQS.
While the EPA appreciates these
comments and recognizes their support
for retroactive data correction, at this
time and following this final rule,
monitoring agencies should continue to
report PM FEM data as measured. This
component of this final rule is focused
only on revising 40 CFR part 53,
appendix C to implement an updated
calibration for approved PM FEMs. The
issue of how prior and future
monitoring data will be used in the
implementation of this NAAQS, such as
for designations, and for air quality
regulatory programs is outside the scope
of this rulemaking and will therefore be
addressed by the EPA in a subsequent
relevant action or actions.
The EPA received comments on
whether updates to PM FEM methods
should be required to be implemented
or there would flexibility in when and
if a monitoring agency implemented
them. The commenters asked that EPA
be flexible in allowing the use of
updated method correction factors
intended to improve the data
comparability between the FRMs and
FEMs.
In most cases, the EPA expects that
updating the FEMs will result in
improved data quality and more sites
meeting bias MQOs; however, the EPA
is not finalizing an update requirement
in this action. Monitoring agencies can
assess their data and make decisions on
an update based on whether they are
meeting the bias MQOs. Such decisions
on whether or not to update a method
may efficiently be included in those
agencies’ annual monitoring network
plans under 40 CFR 58.10, ‘‘Annual
monitoring network plan and periodic
assessment,’’ which are already subject
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
to EPA Regional office approval. In
some circumstances, it is possible the
original PM FEM may be revised in a
manner where only the updated method
has an active approved designation. In
these cases, monitoring agencies would
need to address updating their PM FEM
in a timely manner.
The EPA solicited input on any
alternative approaches that could lead
to more sites meeting the bias MQO
with automated PM FEMs, especially for
those sites that are near the level of the
primary annual PM2.5 NAAQS as
proposed to be revised in section II
above. A few commentors provided
input on potential options for
alternative approaches and several
others offered input on how a local or
regional calibration of an FEM could
work. Among alternative approaches,
one commenter suggested that
manufacturers of FEMs could provide
settings that would allow for
adjustments to make FEM data more
‘‘FRM-like.’’ Another commenter
suggested working with the
manufacturers of FEM equipment to
diagnose the cause of the bias and then
to address it appropriately.
The EPA received several comments
on how to implement a local or regional
calibration of FEMs. One commenter
suggested that EPA could allow for SLT
agencies to adjust FEM data to be more
‘‘FRM-like’’ prior to submitting data to
AQS. Another commenter suggested
using a rolling 3-month linear regression
based on a comparison of FEM data to
PM2.5 levels measured by a 1-in-6-day
FRM. Another commenter
recommended that the EPA allow the
application of a correction factor that is
from an area with a similar climate and
other conditions. Another commenter
suggested that, for metropolitan
statistical areas (MSAs) where the recalibrated FEMs still do not meet
equivalency criteria, monitoring
agencies should be able to use the
rolling linear regression technique to
further calibrate the FEMs within an
MSA. Another commenter suggested
that developing a simple linear
regression could establish the
relationship between FEM data and
FRM data and be used to adjust the FEM
data at each site where they are
collocated. Another commenter
suggested that averaging the results
within a MSA and applying it on an
MSA basis with the previous 2 years of
data could provide an adjustment
method for sites without a collocated
FRM. Another commenter identified
that a regional correction factor
potentially could improve instrument
accuracy to biomass sources, which are
PO 00000
Frm 00155
Fmt 4701
Sfmt 4700
16355
a large component of PM in many
communities.
Among the alternative approaches
suggested, having settings that would
allow for adjustments to make FEM data
more ‘‘FRM-like’’ has merit, but
assuming this was within a PM FEM
itself, it would need to be separately
incorporated into each make and model
of FEM. If EPA were to pursue this
alternative approach, the suggestion
could be incorporated into a future
regulatory action as a potential
condition of designation because,
without having the opportunity to
thoughtfully consider how every step of
such an approach would need to work,
including what such requirements
would look like and how potential
settings adjustments would be made, it
is not appropriate for the EPA to require
the availability of such settings now, nor
would it address the inventory of
currently available PM FEMs already
operating.
Regarding the suggestion that the EPA
and SLTs should work with the
manufacturers of FEM equipment to
diagnose the cause of any biases and
then to address them appropriately, the
EPA supports this recommendation, but
does not believe a regulatory change is
required to allow the monitoring
community (EPA and SLTs) to work
with instrument manufacturers in this
way.
Regarding the several comments on
how to implement a local or regional
calibration of FEMs, the EPA
acknowledges the desire for this
flexibility but believes that any such
provisions for local or regional
calibration of FEMs would need to be
thoroughly thought out and proposed
for consideration across the monitoring
community. While several commenters
support such an approach, the EPA also
received adverse comments on the
potential for local and regional
calibration of PM FEMs instead of
national. Most of the criticism of local
and regional calibration of PM FEMs
centered on both the lack of existing
operating PM FRMs in commenters’
networks and monitoring agencies’
inability to staff the higher number of
operating FRMs that would have to be
collocated with PM FEMs to calibrate.
Thus, the commenters that oppose local
and regional calibrations of data prefer
to utilize the national calibration of
FEM data as proposed. Acknowledging
all of these viewpoints, the EPA believes
that it would not be appropriate to
institute such an approach at this time.
As discussed throughout this section,
this final rule, the EPA is embarking on
a new national approach to calibration
of FEMs where valid State, local, and
E:\FR\FM\06MRR3.SGM
06MRR3
16356
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
Tribal air monitoring data from FRMs
generated in routine networks and
submitted to the EPA may be used to
improve the PM concentration
measurement performance of approved
FEMs. The EPA and the community of
SLT monitoring agencies can further
consider other solutions to improving
PM FEM methods, including local and
regional scale calibration of FEMs, in a
future review of the PM NAAQS.
In summary, the EPA is finalizing its
proposal to allow valid State, local, and
Tribal air monitoring data from PM
FRMs and FEMs generated in routine
networks and submitted to the EPA to
update factory calibrations included as
part of approved FEMs (40 CFR part 58,
appendix C, sections 2.2 and 2.7). This
approach, which will typically be
initiated by instrument manufacturers
but can also be spurred by monitoring
agencies, MJOs of monitoring agencies,
and the EPA itself, is to be implemented
as a national solution in factory
calibrations of approved FEMs through
a firmware update, subject to EPA
approval. FEM calibrations can apply to
any PM FEM methods (i.e., PM10, PM2.5,
and PM10–2.5). As part of this process,
the EPA is finalizing that a range of data
based on the most representative
concentrations up to all available
concentrations may be used in
developing and testing a new factory
calibration; that a representative set of
geographic locations can be used; that
outliers may be included or not
included; that a new factory calibration
should be developed using data from at
least 2 years and tested on a separate
year(s) of data; that updates to factory
calibrations can occur as often as
needed and should be evaluated by
monitoring agencies as part of routine
data assessments such as during
certification of data and 5-year
assessments; that the EPA recognizes
only data from existing operating sites is
available; and that an updated factory
calibration does not have to work with
the original field study data submitted
that led to the designation as an FEM.
The EPA is finalizing this approach as
proposed with the intention of having
more sites meet the bias MQOs with
automated PM FEMs.
4. Revisions to the PM2.5 Monitoring
Network Design Criteria To Address AtRisk Communities
To enhance protection of air quality
in communities subject to
disproportionate air pollution risk,
particularly in light of the proposed
range for a revised primary annual PM2.5
standard, the EPA proposed to modify
the PM2.5 monitoring network design
criteria to include an environmental
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
justice (EJ) factor that accounts for
proximity of at-risk populations (i.e.,
those identified in the 2019 ISA and ISA
Supplement as being at increased risk of
adverse health effects from PM2.5
exposures to sources of concern),
consistent with the statutory
requirement that the NAAQS protect the
health of at-risk populations (88 FR
5673, January 27, 2023). Specifically,
the EPA proposed to modify the existing
requirement at 40 CFR part 58,
appendix D, section 4.7.1(b)(3)): ‘‘For
areas with additional required SLAMS,
a monitoring station is to be sited in an
area of poor air quality,’’ to additionally
address at-risk communities with a
focus on anticipated exposures from
local sources of emissions. The
scientific evidence evaluated in the
2019 ISA and ISA Supplement indicates
that sub-populations at potentially
greater risk from PM2.5 exposures
include children, lower socioeconomic
status (SES) 188 populations, minority
populations (particularly Black
populations), and people with certain
preexisting diseases (particularly
cardiovascular disease and asthma). The
EPA proposed that communities with
relatively higher proportions of subpopulations at greater risk from PM2.5
exposure within the jurisdiction of a
State or local monitoring agency should
be considered ‘‘at-risk communities’’ for
these purposes.
The PM2.5 network design criteria
have led to a robust national network of
PM2.5 monitoring stations. These
monitoring stations are largely in CoreBased Statistical Areas (CBSAs) 189
across the country that include many
PM2.5 monitoring sites in at-risk
communities. Many of the
epidemiologic studies evaluated in the
2019 ISA and ISA Supplement,
including those that provide evidence of
disparities in PM2.5 exposure and health
risk in minority populations and lowSES populations, often use data from
these existing PM2.5 monitoring sites.
However, we anticipate that with the
more protective annual NAAQS
finalized in section II above,
188 SES is a composite measure that includes
metrics such as income, occupation, and education,
and can play a role in populations’ access to
healthy environments and healthcare.
189 Metropolitan and Micropolitan Statistical
Areas are collectively referred to as ‘‘Core-Based
Statistical Areas.’’ Metropolitan statistical areas
have at least one urbanized area of 50,000 or more
population, plus adjacent territory that has a high
degree of social and economic integration with the
core as measured by commuting ties. Micropolitan
statistical areas are a set of statistical areas that have
at least one urban cluster of at least 10,000 but less
than 50,000 population, plus adjacent territory that
has a high degree of social and economic
integration with the core as measured by
commuting ties.
PO 00000
Frm 00156
Fmt 4701
Sfmt 4700
characterizing localized air quality
issues around local emission sources
may become even more important. The
EPA believes that adding a network
design requirement to locate monitors in
at-risk communities will improve our
characterization of exposures for at-risk
communities where localized air quality
issues may contribute to air pollution
exposures. Requiring that PM2.5
monitoring stations be sited in at-risk
communities will allow other methods
to be operated alongside PM2.5
measurements to support multiple
monitoring objectives per 40 CFR part
58, appendix D, section 1.1. The EPA
believes that it is appropriate to
formalize the monitoring network’s
characterization of PM2.5 concentrations
in communities at increased risk to
provide such areas with the level of
protection intended with the PM2.5
NAAQS. The addition of this
requirement will also lead to enhanced
local data that will allow air quality
regulators help communities reduce
exposures and inform future
implementation and reviews of the
NAAQS.
The EPA received comments
concerning the proposed requirement to
modify the PM2.5 monitoring network
design criteria to include an EJ factor
that accounts for the proximity of
populations at increased risk of adverse
health effects from PM2.5 exposures to
sources of concern. Commenters
included State, local, and Tribal air
agencies and multijurisdictional
organizations (MJOs) comprised of those
agencies; industry and industry groups;
other Federal, State, and local
government entities; public health,
medical, and environmental
nongovernmental organizations (NGOs);
and private citizens. The EPA proposed
to require that sites located in at-risk
communities (particularly those whose
air quality is potentially affected by
local sources of concern) should
nonetheless meet the requirements to be
considered representative of ‘‘areawide’’
air quality as this is consistent with all
other minimally required sites. There
were several other technical
components of the proposed
requirement for which we asked for
comment, including: how to identify atrisk communities; the PM sources of
concern important to consider; the
datasets that can be used to identify
communities with high exposures; the
most useful measurement methods to
collocate with PM2.5 in at-risk
communities; and the timeline to
implement any new or moved sites.
Overall, most commenters were very
supportive of the EPA’s proposed
modification to the PM2.5 monitoring
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
network design criteria to include an EJ
factor that accounts for proximity of
populations at increased risk of adverse
health effects from PM2.5 exposures to
sources of concern. A few commenters
offered detailed supporting comments.
For example, one commenter
recommended targeting investment in
regulatory monitors in EJ communities,
opining that there is presently a lack of
equitable distribution of these monitors
in low-income and minority
communities. Another commenter
supports the inclusion of an EJ factor in
PM2.5 monitoring network design
criteria as a means to assess whether
disparities in exposure are reduced in
the future. The EPA appreciates the
support for the proposed requirement
and acknowledges the desirability of a
goal to assess if disparities in exposure
are reduced in the future as a result of
these monitoring efforts.
Some commenters were generally
supportive of the proposed requirement
but suggested that the EPA should recast
the approach in a more specific way or
offered additional examples of sources
of concern. For example, one
commenter stated that PM2.5 emissions
from residential and commercial wood
burning result in localized hotspots that
are often not revealed by community air
monitoring. Another commenter asked
that the EPA adopt a strategy to monitor
EJ communities near both larger wellknown point sources of PM2.5 and along
traffic corridors as well as smaller
sources that, when taken together, may
create a large amount of emissions and
health harms in the area. Another
commenter stated that the national
network of monitors operated by the
EPA captures data used for generalized
modeling, but overall monitoring is not
as granular as one would expect,
especially in urban areas. For instance,
the commenter suggested that EPA
could monitor suspected ‘‘hot spots’’
(e.g., residential development adjacent
to highways and active construction
sites) to better manage and mitigate
PM2.5 pollution at their sites of origin,
and that more extensive and granular
monitoring data would also facilitate
essential research and inform future
evaluations and adjustments of the
NAAQS. The EPA acknowledges these
comments identifying other sources of
concern, and we address these and other
potential sources of concern below.
Among adverse comments, a few
commenters stated that ‘‘at-risk
communities’’ is not well defined. The
EPA disagrees and directs those
commenters to the numerous places
where this definition is covered,
including in Section II.B.2 of the
proposal where we explained the term
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
related to a variety of at-risk populations
(88 FR 5591–92, January 27, 2023) as
well as section 12.5 of the 2019 ISA
(U.S. EPA, 2019a) and section 3.3.3 of
the ISA Supplement (U.S. EPA, 2022a).
Other commenters oppose the addition
of the proposed monitoring because
they feel it would reduce flexibility for
agencies in deciding where they should
site monitors, advocating that
monitoring agencies should be afforded
maximum flexibility to identify where
to site monitors for at-risk areas.
Because the EPA recognizes the
challenges cited by these commenters
related to establishing new ambient air
monitoring stations, the EPA is
finalizing the modified requirement on
PM2.5 monitoring network design
criteria intended to address at-risk
communities that allows flexibility
regarding which EJ communities should
be monitored. Finally, one commenter
asked that the EPA clarify a specific
metric to judge how to site monitors in
at-risk communities. Instead, the EPA
believes it is appropriate for agencies to
recommend what they believe to be the
most important things to consider for
their sites to meet the PM2.5 network
design requirements and, thus, applying
a new metric could take away from local
priorities for at-risk communities.
A few commenters asked that the EPA
require more monitoring than proposed.
One commenter stated that it would be
more beneficial to overburdened
communities if air monitoring were
required in all at-risk communities. A
few commenters asked that EPA require
additional monitoring for attainment of
PM2.5 NAAQS in EJ communities. In
response to these comments, the EPA
supports the SLT agencies’ initiatives to
conduct additional monitoring beyond
the minimum monitoring requirements
and network design criteria. In addition,
the EPA supports agencies’ use of
alternative datasets such as sensors and
sensors networks, satellites, and other
non-regulatory monitoring where
appropriate for non-regulatory data
uses. The EPA notes that many
monitoring agencies already operate
more monitoring sites than are
minimally required, and we expect this
to continue as agencies consider siting
monitors in at-risk communities.
However, the EPA also received
substantial concerns from monitoring
agencies about their resource
constraints, including staffing to
support any potential new monitoring.
The EPA also notes that the existing and
robust network of almost 1,000 PM2.5
sites nationally is designed to continue
to protect all populations at the level of
the NAAQS discussed in section II of
this final action by always having at
PO 00000
Frm 00157
Fmt 4701
Sfmt 4700
16357
least one site in the area of expected
maximum concentration for each CBSA
where monitoring is required. As a
result of the revisions to the annual
PM2.5 NAAQS being finalized in this
action, a small number of new
monitoring sites will also be required
under EPA’s current minimum
monitoring requirements. With the
monitoring network design changes
finalized in this rule, many of these
existing and new sites will form an
important sub-component of the PM2.5
network by better characterizing air
quality in at-risk communities,
particularly with respect to sources of
concern.
The EPA concludes that the
requirements in this final rule for siting
of monitoring in at-risk communities
will meaningfully improve the PM2.5
monitoring network and its
characterization of air quality in at-risk
communities, without placing
substantial new resource burdens on
States and their monitoring agencies
that would be associated with
requirements for additional monitoring
sites. Therefore, the EPA is finalizing
this part of the proposed action without
requiring additional monitoring sites
beyond what would be associated with
the revised annual PM2.5 NAAQS
described in section II as they pertain to
the minimum requirements associated
with Table D–5 of Appendix D to Part
58—PM2.5 Minimum Monitoring
Requirements.
A few commenters asked that the EPA
enhance monitoring in smaller cities
and rural areas. One commenter asked
for the EPA to extend the proposed
monitoring network to Micropolitan
Statistical Areas with populations of
10,000–50,000 and to rural areas.
Another commenter pointed out that
current air quality monitoring networks
focus on urban and densely populated
areas; therefore, rural areas are often not
captured in this existing monitoring
infrastructure, despite well-documented
examples of high PM concentration in
rural communities. The commenter
believes this results in inadequate
assessment of air pollution exposures
for a substantial segment of the U.S.
population. The EPA disagrees that
there needs to be additional
requirements for small CBSA’s and rural
areas. Regarding these comments, the
EPA points out that we have a longstanding requirement for each State to
monitor at background and transport
sites (40 CFR part 58, appendix D,
section 4.7.3—Requirement for PM2.5
Background and Transport Sites). Also,
if an agency deems it appropriate to do
so, monitoring coverage of rural areas
can be accomplished with other tools
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16358
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
such as sensors and sensors networks,
satellites, and other non-regulatory
monitoring. Although there may be
short-term high exposures in rural areas,
there is no evidence that long-term
averages are higher in rural areas
compared to urban areas with
significantly higher density of
populations and emissions. For smaller
cities or rural areas that may have
concentrations near the level of the
PM2.5 NAAQS finalized in section II
above, monitoring agencies are
encouraged to monitor and address
emissions as appropriate.
Some commenters disagree that the
proposed revision to the PM2.5
monitoring network design criteria to
address at-risk communities is needed.
One commenter stated that including an
EJ factor is not necessary because the
current network is designed to protect
all citizens. Another commenter stated
that EJ factors could be cumbersome to
implement. Another commenter
asserted the proposal to add SLAMS in
at-risk communities with higher PM2.5
concentrations might create more
granular data and provide for a greater
margin of safety for those communities
and monitors in such a way that data
from those areas could misrepresent the
larger area represented by the network.
In response to the comment on the
current network protecting all citizens,
the EPA agrees that by measuring in the
community with the highest
concentration of PM2.5 we protect other
citizens; however, as stated in the
proposal, the EPA believes that adding
a requirement for sites with an EJ factor
near sources of concern will enhance
the overall network to the benefit of all
citizens. Also, we anticipate that with
the more protective annual NAAQS
finalized in section II above,
characterizing localized air quality
issues will become even more important
around local emission sources. As for EJ
factors being cumbersome to implement,
the EPA disagrees because there are
many such locations already operating
successfully in the current network.
Regarding the comment that sites in atrisk communities may misrepresent the
larger area represented by a particular
network, the EPA notes that pursuant to
40 CFR part 58, minimally required sites
in a given network are to represent areawide air quality; therefore, sites in atrisk communities, by definition, would
be representative of the communities
within the network in which they are
sited for the level of protection intended
under the annual PM2.5 NAAQS.
In the proposal, the EPA identified
that, in light of the evidence of
increased risk to at-risk communities, it
would be appropriate to better
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
characterize exposures for communities
in proximity to local sources of concern
(88 FR 5673–76, January 27, 2023).
Thus, the EPA proposed that enhanced
networks should include representation
of at-risk communities living near
emission sources of concern (e.g., major
ports, rail yards, airports, industrial
areas, or major transportation corridors).
The EPA requested comment on the
types of sources of concern most
important to consider. In addition to
supporting the types of sources the EPA
identified in the proposal, commenters
also identified several additional
localized sources such as railroads,
stationary sources, transportation
facilities, and communities with high
numbers of wood stoves.
A few commenters suggested the
inclusion of sources that are often
considered line and/or area sources,
e.g., traffic corridors and emissions from
federally regulated facilities, military
installations, and national forests.
Commenters also identified other
sources usually associated with longrange transport such as smoke from
wildfire and prescribed fires and longdistance transport of PM, for example
from Saharan dust and other
international transport. As explained in
the proposal, the site with the highest
expected PM2.5 is already required to
have a monitor by our long-standing
requirement that monitors be placed
‘‘. . . in the area of expected maximum
concentration’’ (§ 58.1 and appendix D,
section 4.7.1(b)(1)). The EPA expects
that both sites with the expected
maximum concentration and sites
specifically placed in at-risk
communities would be impacted by any
long-range transport in the area.
Therefore, the EPA believes any
emphasis on the sources of concern
should prioritize localized sources,
including point, area, and line sources
of concern impacting the at-risk
community of interest. Therefore, based
upon the comments, the EPA is
finalizing a broader example list of
sources of concern to include localized
sources such as point sources and
transportation facilities, since these are
the most commonly expected additional
sources of concern. In response to the
other sources of concern suggested by
commenters, the EPA notes that while it
has provided examples, the siting of
monitors in EJ communities would not
be limited to these examples. Thus, the
revised set of examples would include
‘‘a major industrial area, point source(s),
port, rail yard, airport, or other
transportation facility or corridor.’’ In
finalizing this modified list of examples,
the EPA is not looking to prioritize one
PO 00000
Frm 00158
Fmt 4701
Sfmt 4700
type of source category over another;
rather, we intend to further illustrate the
types of localized sources of pollution
that might impact at-risk communities
such that the siting of monitors nearby
may be appropriate.
One commenter noted that the
proposal may have unintentionally
taken out the requirement related to
specific design criteria for PM2.5 in 40
CFR part 58, appendix D, 4.7.1(b)(3)
that, for an area with a requirement for
an additional SLAMS monitor, it should
‘‘be sited in an area of poor air quality.’’
Thus, the language as proposed neither
requires that such monitors be sited in
areas of poor air quality, nor does it
require that the monitor be sited in an
area that is anticipated to experience
poor air quality from unspecified (and
thus potentially relatively insignificant)
sources in the area. The EPA agrees that
this was not our intention; the EPA
wants to protect populations in at-risk
communities by ensuring they are
protected by the NAAQS when there are
sources of concern that may be
impacting them (i.e., not insignificant
sources). Thus, the EPA is reinstating
this requirement in the network design
language and combining it with the
examples of the types of localized
sources of concern: ‘‘For areas with
additional required SLAMS, a
monitoring station is to be sited in an atrisk community with poor air quality,
particularly where there are anticipated
effects from sources in the area (e.g., a
major industrial area, point source(s),
port, rail yard, airport, or other
transportation facility or corridor).’’
To ensure minimally required
monitoring sites appropriately represent
exposures in at-risk communities, the
EPA proposed that sites represent ‘‘areawide’’ air quality near local sources of
concern (88 FR 5674, January 27, 2023).
Sites representing ‘‘area-wide’’ air
quality are those monitors sited at
neighborhood, urban, and regional
scales, as well as those monitors sited at
either micro- or middle-scale that are
identified as being representative of
many such locations in the same
Metropolitan Statistical Area (MSA).190
Most existing—as well as new or moved
sites—are expected to be neighborhoodscale, which means that the monitoring
stations would typically represent
conditions throughout some reasonably
homogeneous urban sub-region with
dimensions of a few kilometers per part
58, appendix D, section 4.7.1(c)(3).
Additionally, as described in § 58.30,
190 MSA means a CBSA associated with at least
one urbanized area of 50,000 population or greater.
The central-county, plus adjacent counties with a
high degree of integration, comprise the area.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
sites representing ‘‘area-wide’’ air
quality have a long-standing
applicability to both the annual and 24hour PM2.5 NAAQS. Our proposed
requirement for siting monitors in
communities representing ‘‘area-wide’’
air quality is consistent with other
network design objectives pursuant to
which we seek to have monitors located
where people live, work, and play.
The EPA received a few comments on
its proposed requirement that minimally
required sites represent ‘‘area-wide’’ air
quality. One commenter stated that the
inclusion of a provision for EJ would
narrow the location of monitors to
certain communities that may not best
represent ‘‘areawide’’ air quality.
Another commenter asked the EPA to
consider removing requirements that
sites be area-wide, since 24-hour and
annual averaging times would miss
short, elevated pollution events. A
couple commenters had concerns with
the difference in the scale of
representation between EJ monitors
using small scale and other NAAQS
monitors using area-wide scale, in that
area-wide scale would not protect those
most at risk. However, another
commenter agreed with the EPA that
sites representing at-risk communities
should represent area-wide air quality.
In addition to these comments, the EPA
received many comments with support
for its proposed modifications to the
network design criteria as whole.
Regarding whether narrowing the
location to certain communities may not
best represent ‘‘area-wide’’ air quality,
the EPA notes that sites are either
identified as being area-wide or not; the
EPA did not suggest it was seeking a
best ‘‘area-wide’’ location. In response
to the comment that area-wide site may
miss short, elevated pollution events,
the EPA is aware that there can be local,
short-term spikes in PM2.5
concentrations. However, the network
design criteria associated with
minimally required sites is applicable to
both the annual and 24-hour PM2.5
NAAQS, and the EPA believes it is
appropriate to continue to ensure all
minimally required sites have the most
utility and remain applicable to both
forms of the PM2.5 NAAQS. The
identification of unique micro- and
middle-scale sites was directed at
discretionary efforts of any monitoring
agency, with the recognition that such
sites, (i.e., relatively unique micro-scale,
or localized hot spot, or unique middlescale impact sites), are not applicable to
the annual NAAQS as described in
§ 58.30—Special consideration for data
comparison to the NAAQS.
After considering all the comments on
this topic, the EPA is finalizing this part
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
of the modification to the network
design criteria to maintain, consistent
with our long-standing network design
criteria, that all minimally required sites
are to represent area-wide air quality.
In addition to using data from the
robust network of almost 1,000 PM2.5
sites for NAAQS and AQI purposes,
having a stable network of long-term
sites is especially valuable to examine
trends and to inform long-term health
and epidemiology studies that support
reviews of the PM NAAQS. Therefore,
while we proposed to add a PM2.5
network design criterion to address atrisk communities, many sites are likely
already in valuable locations meeting
one of the existing network design
criteria (i.e., being in an area-wide area
of expected maximum concentration or
collocated with near-road sites) and
supporting multiple monitoring
objectives. Also, in many communities,
there may already be sites meeting the
network design criterion we proposed
for at-risk communities. Thus,
acknowledging the value of having longterm data from a consistent set of
network sites, the EPA believes that
moving sites should be minimized,
especially in MSAs with a small number
of sites. However, because a small
number of new sites are expected to be
required due to the existing minimum
monitoring requirements (40 CFR part
58, appendix D, Table D–5) 191 and the
revised primary annual PM2.5 NAAQS
detailed in section II, and because sites
occasionally have to be moved—due to,
for example, loss of access to a site or
a site no longer meeting siting criteria—
the EPA believes it is appropriate to
prioritize establishing sites in at-risk
communities near sources of concern,
whenever new sites are established,
whether because it is a new site or a
replacement for a prior site that must be
moved. The EPA accordingly proposed
that annual monitoring network plans
(40 CFR 58.10(a)(1)) and 5-year
assessments (40 CFR 58.10(d)) that
include any of the few new sites that
will be required include a commitment
to examine the ability of existing and
proposed sites to support air quality
characterization for areas with at-risk
populations in the community and the
objective discussed herein.
In the proposal, the EPA identified
that assessing and prioritizing at-risk
communities for monitoring can be
191 Gantt, B. (2022). Analyses of Minimally
Required PM2.5 Sites Under Alternative NAAQS.
Memorandum to the Rulemaking Docket for the
Review of the National Ambient Air Quality
Standards for Particulate Matter (EPA–HQ–OAR–
2015–0072). Available at: https://
www.regulations.gov/docket/EPA-HQ-OAR-20150072.
PO 00000
Frm 00159
Fmt 4701
Sfmt 4700
16359
accomplished through several
approaches (88 FR 5675). The most
critical aspect of prioritizing which
communities to monitor is their
representation of the at-risk populations
described earlier in this section. The
other major consideration is whether the
community is near a source or sources
of concern. While many CBSAs have
one or more sources of concern
described above, some CBSAs will not
have a quantity of emissions from
sources of concern that result in an
elevated level of measured PM2.5
concentrations in surrounding
communities. The siting criteria to be
‘‘in the area of expected maximum
concentration,’’ § 58.1 & appendix D,
section 4.7.1(b)(1) ensures there is a
monitoring site in the community with
the highest exposure in each CBSA with
a monitoring requirement. Some CBSAs
may also have a requirement to
collocate a PM2.5 monitor at a near-road
NO2 station. Therefore, the EPA believes
that for cases where an additional PM2.5
site is required, we should include a
criterion that the site be in an at-risk
community when there are no sources
of concern identified in that CBSA, or
such sources do exist but are not
expected to lead to elevated levels of
measured PM2.5 concentrations.
In its proposal, the EPA highlighted
that tools such as the EPA’s
EJSCREEN 192 are available to identify
the at-risk communities intended for
monitoring as part of the proposed
revision to the PM2.5 network design
criteria (88 FR 5675–76, January 27,
2023). The EPA solicited comment on
other tools and/or datasets that can be
utilized to identify at-risk communities.
In addition to support for using
EJSCREEN, commenters identified
several other options to identify at-risk
communities intended for monitoring as
part of the proposed revision to the
PM2.5 network design criteria. Among
similar tools, one commentor suggesting
using CalEnviroScreen.193 Commenters
also identified different options for
models including InMAP,194 satellitederived models that can be employed to
help identify EJ communities, and
hybrid models. A few commenters also
suggested using sensors and sensor
networks such as the BlueSky 195 and
PurpleAir 196 sensors.
The EPA supports the use of other
State and local tools designed to help
identify the at-risk communities that
192 See:
https://www.epa.gov/ejscreen.
https://oehha.ca.gov/calenviroscreen.
https://inmap.run/#home.
195 Mention of commercial names does not
constitute EPA endorsement.
196 Mention of commercial names does not
constitute EPA endorsement.
193 See:
194 See:
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16360
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
should be monitored to meet the revised
network design criteria. The EPA
additionally agrees with commenters
that the use of models as well as sensors
and sensor networks may be appropriate
and helpful in identifying the most
appropriate at-risk communities in
which to locate monitors.
For at-risk communities, monitoring
agencies need data that can best inform
where there may be elevated levels of
exposures from sources of concern.
While we use FRMs and FEMs to
determine compliance with the NAAQS,
data from these methods will only be
available at existing sites. However,
there are several additional datasets
available that may be useful in
evaluating the potential for elevated
levels of exposure to communities near
sources of concern. In the proposal, EPA
identified potential non-regulatory
monitoring datasets such as CSN,
IMPROVE, and AQI non-regulatory
PM2.5 continuous monitors; modeling
data that utilizes emission inventory
and meteorological data; emerging
sensor networks such as those that
comprise EPA and the USFS’s Fire and
Smoke Map; 197 and satellites that
measure radiance and, with
computational algorithms, can be used
to estimate PM2.5 from aerosol optical
depth (AOD) (88 FR 5675–76, January
27, 2023). The EPA solicited comment
on datasets most useful to identify
communities with high exposures for
PM2.5 NAAQS (i.e., annual or 24-hour).
In addition to providing information
about datasets that can inform the
NAAQS comparison, commenters
additionally identified several types of
datasets that may be useful to identify
where there may be elevated levels of
exposures from sources of concern.
These datasets include satellite
measurements, sensors, and sensor
network data, which may all be useful
to find hot spots in communities.
Commenters also identified EJScreen
and CalEnviroScreen, which are
screening and mapping tools that utilize
several datasets. Another commenter
stated that to better understand
exposure differences in disadvantaged
communities, shorter measurement
intervals should be measured and
reported.
In considering the datasets identified
in the proposal as well as the ones
commenters provided, the EPA believes
all the datasets have value to help
inform where there may be elevated
levels of exposures from sources of
concern. However, each of them may
also have limitations and, therefore,
users should be careful not to rely solely
197 See:
https://fire.airnow.gov/.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
on one dataset versus another for all
purposes. Fortunately, many of the
available datasets are becoming easier to
work with and more accessible, which
will allow interested parties and
monitoring agencies the opportunity to
efficiently review the datasets and
determine best applicability. For all of
these reasons, the EPA is not finalizing
a requirement to use a specific dataset
or tool to identify at risk communities;
however, whatever datasets a
monitoring agency elects to use, its plan
to use such data for purposes of meeting
the network design requirements will be
subject to EPA approval as part of the
40 CFR 58.10 annual monitoring
network plan. Regarding the comment
recommending shorter measurement
intervals in measuring and reporting
data to better understand exposure
differences in disadvantaged
communities, the EPA agrees and
generally supports use of continuous
methods. While we generally support
use of continuous methods, approved
filter-based technologies and methods
also provide valuable air quality
information. Therefore, the EPA is not
requiring the use of automated
continuous methods beyond what is
already required in 40 CFR part 58,
appendix D, section 4.7.2—Requirement
for Continuous PM2.5 Monitoring.
The monitoring methods appropriate
for use at required PM2.5 sites in at-risk
communities are FRMs and automated
continuous FEMs (88 FR 5675–76,
January 27, 2023). These are the
methods eligible to compare to the PM2.5
NAAQS, which is the primary objective
for collecting this data. There are several
other monitoring objectives that would
benefit from the use of automated
continuous FEMs. For example, having
hourly data available from automated
continuous FEMs would allow sites to
provide data in near-real time to support
forecasting and near real-time reporting
of the AQI. Automated continuous
methods are also useful to support
evaluation of other methods such as
low-cost sensors. When used in
combination with on-site wind speed
and wind direction measurements,
automated FEMs can provide useful
pollution roses, which help in
identifying the origin of emissions that
affect a community. Additionally, when
collocated with continuous carbon
methods such as an aethalometer,
automated FEMs can help identify
potential local carbon sources
contributing to increased exposure in
the community. While either FRMs or
automated FEMs may be used at a site
for comparison to the PM2.5 NAAQS, the
EPA supports use of automated
PO 00000
Frm 00160
Fmt 4701
Sfmt 4700
continuous FEMs at sites in at-risk
communities.
The EPA requested comment on the
measurement methods most useful to
collocate with PM2.5 in at-risk
communities (88 FR 5675–76, January
27, 2023), and a few commenters
provided input. One commenter
recommended that the EPA should
employ supplemental technologies and
systems to increase coverage of the
regulatory monitoring network and
obtain more complete data to further
protect public health and address
environmental injustice in air pollution
exposure. Another commenter
recommended that the EPA invest in
community-led monitoring and mobile
air quality monitoring with a goal of
recording block-level variabilities in
data. And another commenter cited the
value of community-deployed PM2.5
monitoring.
The EPA appreciates the comments
provided on the measurement methods
most useful to collocate with PM2.5
monitoring sites in at-risk communities.
Because the use of methods beyond the
required PM2.5 FRMs or FEMs or other
criteria pollutant measurements meeting
a NAAQS monitoring requirement is
voluntary, the establishment of PM2.5
NAAQS comparable sites in at-risk
communities will allow for
collaboration at multiple levels. The
EPA strongly encourages such
collaboration with impacted
communities, and the measurement
methods discussed here should be
considered for use as appropriate.
In the proposal, the EPA identified
that, to meet the revised network design
criteria, there will be only a few new
sites required,198 plus any potentially
moved sites in cases where an existing
site lease is lost or otherwise requires
relocation (88 FR 5675–76, January 27,
2023). To handle these new or relocated
sites, the EPA proposed to build upon
our existing regulatory process for
selecting and approving these sites
under 40 CFR 58.10 (88 FR 5676,
January 27, 2023). In the proposal, we
stated it would be appropriate to
provide at least 12 months from the
effective date of the final rule to allow
monitoring agencies to initiate planning
to implement these measures by seeking
input from communities and other
interested parties and considering
whether to revise their PM2.5 networks
198 Gantt, B. (2022). Analyses of Minimally
Required PM2.5 Sites Under Alternative NAAQS.
Memorandum to the Rulemaking Docket for the
Review of the National Ambient Air Quality
Standards for Particulate Matter (EPA–HQ–OAR–
2015–0072). Available at: https://
www.regulations.gov/docket/EPA-HQ-OAR-20150072.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
or explain how their existing networks
meet the objectives of the proposed
modification to the network design
criteria. Thus, the EPA proposed that
monitoring agencies should address
their approach to the question of
whether any new or moved sites are
needed and identify the potential
communities in which the agencies are
considering adding monitoring, if
applicable, as well as identifying how
they intend to meet the revised criteria
for PM2.5 network design to address atrisk communities in the agencies’
annual monitoring network plans due to
each applicable EPA Regional office no
later than July 1, 2024 (see 40 CFR
58.10). Specifics on the resulting new or
moved sites for PM2.5 network design to
address at-risk communities were
proposed to be detailed in the annual
monitoring network plans due to each
applicable EPA Regional office no later
than July 1, 2025 (40 CFR 58.10). The
EPA proposed that any new or moved
sites would be required to be
implemented and fully operational no
later than 24 months from the date of
approval of a plan or January 1, 2027,
whichever comes first, but the EPA
solicited comment on whether less time
is needed (e.g., 12 months from plan
approval and/or January 1, 2026).
The EPA received a few comments on
its proposed timeline for monitoring
agencies to identify, propose, and
ultimately bring any new or moved sites
online. One commenter asked that the
timeline give states more time to start or
move sites. A few commenters asked
that the EPA only require meeting a
timeline for identifying whether any
new or moved sites are needed after the
EPA has provided the monitoring
agencies with guidance on the priority
of the potential at-risk communities.
One of those commenters further
requests that the EPA allow at least 24
months from the date of approval of a
§ 58.10 monitoring plan identifying any
relocation of monitoring sites or
establishment of new monitoring sites to
implement any changes to the network,
citing the need for more time to work
with local officials, procure monitoring
equipment, and contract for services, all
of which can cause significant delays in
establishing a monitoring site. Another
commenter asked that the EPA remain
attentive to the challenges that States,
and air agencies face regarding
recruiting and retaining the specialized
staff needed to support their existing
regulatory monitoring networks and the
capital resources needed to implement
and sustain new monitoring stations in
areas that are clearly meeting the
existing PM NAAQS or any revised PM
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
NAAQS. Another commenter stated that
the July 1, 2024, timeline for a network
evaluation this complex is insufficient,
noting that they submit their draft
annual monitoring network plan for
public review and comment in midApril for 30 days. Because the final plan
is due July 1 and must include all
comments and responses and describe
any changes based on those comments,
the timeline does not take these
requirements into consideration by
allowing for the more extensive
assessment of changes that may be
needed to meet the proposed new
monitoring requirements. The
commenter stated that it would be
appropriate to provide at least 12
months from the effective date of this
final rule for monitoring agencies to
initiate planning to implement these
measures, seek input, consider revisions
to their PM2.5 networks, and explain
how their existing networks meets the
objectives of the final rule. The
commenter notes that that SLT agencies
should be provided a minimum of 18
months after the final recommendation
is published to add this information to
their § 58.10 annual monitoring network
plans. Another commenter encourages
the EPA to retain the proposed deadline
for any newly required monitoring
stations in at-risk communities to be
operational (i.e., 24 months after the
July 2025 network plan approval or
January 1, 2027, whichever is earlier).
While the need for this data is urgent,
the commenter stated that the process
for procuring instrumentation, securing
leases, and building permits, and other
logistics in constructing new monitoring
sites can take a significant amount of
time, some of which are outside of
agencies’ control.
As stated earlier, the EPA received
strong support for our proposal to
modify the PM2.5 monitoring network
design criteria to include an EJ factor
that accounts for proximity of
populations at increased risk of adverse
health effects from PM2.5 exposures to
sources of concern from a wide range of
commenters. A few commenters support
the timeline proposed, a few others
support starting any new or moved sites
sooner than proposed, while other
commenters asked for more time or
offered conditions regarding how to
establish an appropriate timeline.
The EPA disagrees with the
commenter that suggested the EPA
should only require agencies to meet a
timeline to identify whether any new or
moved sites are needed after the EPA
has provided the monitoring agencies
with guidance on the priority of the
potential at-risk communities, because
the regulatory text provides all the
PO 00000
Frm 00161
Fmt 4701
Sfmt 4700
16361
guidance required for agencies to begin
this process. As we explained above, the
EPA does not anticipate that many new
or moved sites will be required based on
the final rule because we think most
sites are already in suitable locations
and long-term sites are highly valued.
Also, monitoring agencies have
discretion to provide to the EPA their
recommendations regarding how they
intend to meet the modifications to the
PM2.5 monitoring network design
criteria to include an EJ factor that
accounts for proximity of populations at
increased risk of adverse health effects
from PM2.5 exposures to sources of
concern. Overall, the EPA believes that
having sites in the areas of expected
maximum concentrations will best
ensure that all communities are
protected. Since there may be multiple
choices for sites in EJ areas near sources
of concern, the EPA acknowledges that
there may be many locations that can
meet the revised PM2.5 network design
criteria. While, as we explained earlier,
we want such sites to also be in areas
of poor air quality, the sites in the area
of maximum concentration will ensure
that all communities are protected, there
can be more flexibility afforded in the
selection amongst at-risk communities
to meet the revised requirements, since
any alternative at-risk communities
would already be protected.
The EPA considered both the
concerns and support for the timeline
proposed and clarifies that the
component of the proposed requirement
regarding the need to identify potential
new sites or an intention to move sites
to be included in the annual monitoring
network plan due to EPA on July 1,
2024, would be satisfied with a
statement of intent to pursue a new site
per the revised network design criteria
and in consideration of the minimum
monitoring requirements. While
monitoring agencies may provide as
much detail as they deem appropriate
regarding the revised PM2.5 network
design criteria in their annual
monitoring network plans due on July 1,
2024, there is no expectation that any
details on site-specific information
would be included at that stage. We
encourage agencies to provide their
initial thinking on the communities they
are most interested in monitoring
pursuant to the revised network design
criteria. Therefore, the EPA is finalizing
the timeline as proposed, including the
provision that monitoring agencies
report their intention to add or move
sites, where required, in their annual
monitoring network plans due to each
applicable EPA Regional office no later
than July 1, 2024 (40 CFR 58.10). The
E:\FR\FM\06MRR3.SGM
06MRR3
16362
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
monitoring agencies will then provide
specifics on any new or moved sites for
PM2.5 network design to address at-risk
communities in the annual monitoring
network plans due to each applicable
EPA Regional office no later than July 1,
2025 (40 CFR 58.10). And any new or
moved sites shall be implemented and
fully operational no later than 24
months from the date of approval of a
§ 58.10 plan, or January 1, 2027,
whichever comes first.
In summary, the EPA is finalizing
modifications to the PM2.5 network
design criteria to include an EJ factor to
address at-risk communities with a
focus on exposures from sources of
concern in areas of poor air quality.
While this modification to the PM2.5
network design requires sites to be
located in at-risk communities,
particularly those whose air quality is
potentially affected by local sources of
concern, such sites must still meet the
requirement for being considered ‘‘areawide’’ air quality. In finalizing this
modification to the PM2.5 network
design requirement, the EPA is making
two changes in the final rule response
to the comments received. First, the
EPA is broadening our examples of
‘‘sources of concern’’ to include
localized sources such as point sources
and major transportation facilities or
corridors. Second, the EPA is reinstating
‘‘poor air quality’’ in our requirement
for the modified network design criteria,
meaning the revised PM2.5 network
design requirement now states: ‘‘For
areas with additional required SLAMS,
a monitoring station is to be sited in an
at-risk community with poor air quality,
particularly where there are anticipated
effects from sources in the area (e.g., a
major industrial area, point source(s),
port, rail yard, airport, or other
transportation facility or corridor).’’ All
other aspects of the PM2.5 network
design requirements are being finalized
as proposed.
5. Revisions to Probe and Monitoring
Path Siting Criteria
The EPA proposed changes to
monitoring requirements in the
Appendix E—Probe and Monitoring
Path Siting Criteria for Ambient Air
Quality Monitoring (88 FR 5676–78,
January 27, 2023). Since 2006, the EPA
finalized multiple rule revisions to
establish siting requirements for
PM10–2.5 and O3 monitoring sites (71 FR
2748, January 17, 2006), Near-Road NO2
monitoring sites (75 FR 6535, February
9, 2010), Near-Road CO monitoring sites
(76 FR 54342, August 31, 2011), and
Near-Road PM2.5 monitoring sites (78 FR
3285, January 15, 2013). Through these
previous revisions to the regulatory text,
VerDate Sep<11>2014
20:35 Mar 05, 2024
Jkt 262001
some requirements were inadvertently
omitted, and, over time, the clarity of
this appendix was reduced through
those omissions that, in a few instances,
led to unintended and conflicting
regulatory requirements. The EPA
proposed to reinstate portions of
previous Probe and Monitoring Path
Siting Criteria Requirements from
previous rulemakings, where
appropriate, to restore the original
intent.
The EPA only received a few
comments on the proposed rulemaking
pertaining to the proposed changes
regarding probe and monitoring path
siting criteria for ambient air quality
monitoring, most of which were
supportive of the proposed revisions.
One commenter noted that the image for
Figure E–1 in Appendix E to part 58 was
distorted and of extremely poor quality,
rendering the text in places almost
unreadable (88 FR 5712, January 27,
2023). The EPA makes several
references to Figure E–1, which
provides detailed information needed
for assessing a range of acceptable probe
distances from roadways based on a
monitor’s spatial scale. The commenter
also stated that a higher quality image
is needed for the figure so that agencies
can fully interpret the figure to the
extent that EPA requires. The EPA
agrees with the commenter that a higher
quality image for Figure E–1 is
important and needed. Based on this
comment, the EPA is finalizing the
revision to Figure E–1 to clearly
communicate the requirements of
appendix E.
The EPA is revising appendix E in its
entirety as proposed (88 FR 5709–5717,
January 27, 2023) for clarity and as
described in detail below.
a. Separate Section for Open Path
Monitoring Requirements
The EPA proposed to relocate all open
path monitor siting criteria
requirements to a separate section in
appendix E from those requirements for
siting samplers and monitors that utilize
probe inlets (88 FR 5676, January 27,
2023). Separate sections for these
distinct monitoring method types allows
the EPA to more clearly articulate
minimum technical siting requirements
for each.
The EPA received one supportive
comment to adopt this change and
received no adverse comments. Another
commenter stated the regulatory text of
the proposal improves the clarity of the
appendix but encouraged the EPA to
break the summary tables down further
into more manageable components
(perhaps by pollutant). The commenter
stated that summary tables for the
PO 00000
Frm 00162
Fmt 4701
Sfmt 4700
proposed appendix continue to be a
‘‘jumbled mess of regulatory
requirements.’’ The EPA agrees that the
summary tables E–3 and E–6 in the
proposal could be improved further.
Also, the EPA found that footnote 3 of
Table E–6 in the proposed rule was
incomplete and corrected this editorial
error.
Therefore, the EPA is making editorial
changes to both summary tables E–3 and
E–6 and finalizing the remainder of the
language as proposed with the open
path monitor siting criteria
requirements placed into a separate
section of the appendix.
b. Distance Precision for Spacing Offsets
The EPA proposed to require that
when rounding is performed to assess
compliance with these siting
requirements, the distance
measurements will be rounded such as
to retain at least two significant figures
(88 FR 5676, January 27, 2023). The EPA
proposed to communicate this rounding
requirement in the regulatory text using
footnotes in the tables of this appendix.
The EPA received two supportive
comments and no adverse comments
regarding this proposed change. While
supportive of the proposal, one of the
two supporting comments suggested it
would be clearer if EPA explicitly
defined a decimal in the distance values
and round to the nearest tenths place for
these assessments. The EPA disagrees
with this recommendation because in
some cases it would be more restrictive
and burdensome than the proposed
requirement that was intended to
provide both clarity and flexibility.
Therefore, the EPA is finalizing the
language as proposed.
c. Summary Table of Probe Siting
Criteria
The EPA proposed to provide
additional specificity and flexibility to
the summary table for probe siting
criteria by changing the ‘‘>’’ (greater
than) symbols to ‘‘≥’’ (greater than or
equal to) symbols in the summary table
E–4 (88 FR 5676, January 27, 2023).
Because one commenter pointed out to
the EPA that in the prior version of the
rule there was no table E–4, as a clerical
matter, we have renumbered this
summary table to table E–3 in the final
rule. This proposed minor revision to
the summary table more clearly
expresses the EPA’s intent that the
distance offsets provided in the
summary tables in appendix E are
acceptable for NAAQS compliance
monitoring.
The EPA received one comment
supporting the proposal. The EPA
received no adverse comments. Because
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
one commenter pointed out to the EPA
that in the prior version of the rule there
was no table E–4, as a clerical matter,
we have renumbered this summary table
to table E–3 in the final rule. Therefore,
the EPA is updating the table numbering
and otherwise finalizing the tables as
proposed.
ddrumheller on DSK120RN23PROD with RULES3
d. Spacing From Minor Sources
The EPA proposed to clarify and
provide flexibility regarding siting
monitors near minor sources by
changing a requirement to a goal (88 FR
5676–77, January 27, 2023). To
accomplish this, the EPA proposed to
replace the ‘‘must’’ in the regulation
with a ‘‘should.’’ While the EPA
proposed to change this requirement to
a goal, the EPA reiterated in the
proposal that it recommends that sites
with minor sources be avoided
whenever practicable and probe inlets
should be spaced as far from minor
sources as possible when alternative
monitoring stations are not suitable.
The EPA received one comment
supporting the proposed revision and
received no adverse comments.
Therefore, the EPA is finalizing the
language as proposed.
e. Spacing From Obstructions and Trees
The EPA proposed to clarify and
redefine that the minimum arc required
to be free of obstructions for a probe
inlet or monitoring path is 270-degrees
and that probe inlets must be no closer
than 10-meters to the driplines of any
trees (88 FR 5677, January 27, 2023).
These changes were proposed because
of inconsistencies introduced into the
rule with the 2006 rulemaking. Both are
discussed in more detail in the
following sections.
The majority of comments received
were supportive of these proposed siting
amendments and clarifications. Two
commenters were not supportive of this
proposal. One adverse comment focused
on the potential that site modifications
would be required if the minimum arc
required to be free of obstructions for a
probe inlet is 270-degrees. The second
adverse comment pertained to the
proposal to clarify distance
requirements from tree driplines. The
commenter stated they would expect
significant challenges in meeting the
proposed 20-meter tree dripline
distance. This comment is not a
substantive negative comment because
the 20-meter distance provided in the
proposal is a goal and not a
requirement. As such, monitoring
organizations should not expect
additional challenges in meeting the
probe siting requirements. One
supportive commenter on the 270-
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
degree minimum arc proposal also
requested that the EPA acknowledge
that some cases exist where monitoring
is desired or necessary to protect the
public health, but siting criteria cannot
be met.
Based on the only two negative
comments received from monitoring
agencies or organizations, one of which
was not substantive, the EPA believes
most sites already meet these proposed
requirements related to the arc and
distance from dripline. However, the
EPA also acknowledges that there may
be limited cases where this proposed
revision may require site modifications,
and some sites may not be able to be
achieve the proposed siting
requirements, even with modifications
to the site. For cases where long-term
trend sites or monitors that determine
the design value for their area cannot
reasonably meet these regulatory siting
requirements, the EPA encourages
monitoring organizations to work with
their respective EPA Regional offices to
determine if a waiver from this siting
criteria would be appropriate under
appendix E, section 10.
These siting requirements are
discussed in more detail below in
sections VII.B.5.f and VII.B.5.h.
f. Reinstating Minimum 270-Degree Arc
and Clarified 180-Degree Arc
The EPA proposed to correct
identified inconsistencies in the 270degree requirement for unrestricted
airflow to the probe inlet by reinstating
the requirement stated in appendix E,
paragraph 4(b), and to clarify that the
continuous 180-degree minimum arc of
unrestricted airflow provision is
reserved for monitors sited on the side
of a building or wall to comply with
network design criteria requirements
specified in appendix D of part 58 (88
FR 5677, January 27, 2023).
The EPA received two comments
regarding this proposal, with one being
supportive and one being negative. The
adverse comment focused on the
potential that site modifications would
be required if this revision was made.
The commenter supporting the proposal
also requested that the EPA
acknowledge that some cases exist
where monitoring is desired or
necessary to protect the public health,
but siting criteria cannot be met. The
EPA agrees with both commenters and
acknowledges that there does exist
limited cases where this proposal would
require site modifications and some
sites may not be able to be achieve the
proposed siting requirement even with
modifications to the site. For these
cases, and especially when long-term
trend sites or monitors that determine
PO 00000
Frm 00163
Fmt 4701
Sfmt 4700
16363
the design value for their area cannot
reasonably meet these regulatory siting
requirements, the EPA encourages
monitoring organizations to work with
their respective EPA Regional Offices to
determine if a waiver from this siting
criteria is appropriate through the
provisions found in Section 10 of this
appendix.
Based on the EPA only receiving a
single negative comment regard the 270degree and 180-degree provisions the
EPA thinks most sites already meet
these proposed requirements.
Additionally, as stated above, the EPA
is also retaining waiver provisions from
these siting requirements for the
remaining cases that can be exercised
when appropriate. Therefore, the EPA is
finalizing the language as proposed.
g. Obstacles That Act as Obstructions
The EPA proposed to clarify the
definitions of ‘‘obstructions’’ and
‘‘obstacles’’ in the regulatory text (88 FR
5677, January 27, 2023). Stating that,
‘‘[o]bstructions to the air flow of the
probe inlet are those obstacles that are
horizontally closer than twice the
vertical distance the obstacle protrudes
above the probe inlet and can be
reasonably thought to scavenge reactive
gases or to restrict the airflow for any
pollutant,’’ the EPA proposed to
reiterate that the EPA does not generally
consider objects or obstacles such as flag
poles or site towers used for NOy
convertors and meteorological sensors,
etc., to be deemed obstructions.
The EPA received one comment
supporting the proposal and received no
adverse comments. Therefore, the EPA
is finalizing the definitions as proposed.
h. 10-Meter Tree Dripline Requirement
The EPA proposed to reconcile the
conflicting requirements in 5(a) and the
prior table E–4 footnote 3 by clarifying
that the probe inlet must always be no
closer than 10 meters to the tree dripline
(88 FR 5677, January 27, 2023). The EPA
also proposed to reinstate the goal ‘‘that
monitor probe inlets should be at least
20-meters from the driplines of trees,’’ a
goal that was inadvertently omitted
during previous rule revisions. In
addition, the EPA proposed to clarify
that if a tree or group of trees is
considered an ‘‘obstruction,’’ section
4(a) will apply.
As described above, the majority of
comments received were supportive of
the EPA proposed amendments and
clarification, with two commenters
focused on the possibility that
monitoring agencies may not be able to
meet the revised siting requirements.
Specific to the proposed dripline
requirement, the EPA reiterates that the
E:\FR\FM\06MRR3.SGM
06MRR3
16364
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
20-meter tree dripline offset is not a
requirement, but rather a goal.
Monitoring programs should as much as
practicable attempt to meet this 20meter tree dripline offset goal but are
only required to be at least 10 meters
removed from tree driplines. If these
requirements cannot be met, the EPA
encourages monitoring organizations to
contact their respective EPA Regional
offices to determine if a waiver from this
siting criteria would be appropriate
under appendix E, section 10.
Another commenter recommended
that the proposal should also include an
elevation specification. For instance, if
a monitor is on the roof of a shelter, a
tree below that roof should not be
considered an obstruction no matter the
distance to the dripline. The EPA
considers this scenario to occur in
practice only rarely. The EPA agrees
that when the overall tree height is less
than the height of the probe inlet, the
tree is not obstructing the airflow to the
probe inlet. However, a tree in such
proximity to the probe inlet in many
cases is not likely to remain at a height
lower than the probe inlet. The EPA
considers a scenario such as this to be
best addressed in the waiver provisions
of this appendix due both to the rarity
of this occurring as well as the need for
the EPA to periodically reassess
whether tree growth has adversely
impacted the site conditions.
For these reasons, the EPA is
finalizing the language as proposed.
ddrumheller on DSK120RN23PROD with RULES3
i. Spacing Requirement for Microscale
Monitoring
The EPA proposed to require that
microscale sites for any pollutant shall
have no trees or shrubs blocking the
line-of-sight fetch between the monitor’s
probe inlet and the source under
investigation (88 FR 5677, January 27,
2023). This proposed revision would
bring consistency between near-road
monitoring stations and other
microscale monitoring.
The EPA received one comment on
this proposed requirement expressing
concerns regarding its practicality and
legality. The commenter stated agencies
may at times want to site a monitor
close to a source, but the closest
location will have trees in the line of
sight on private property. Additionally,
in some cases, the trees may have been
planted for the purpose of reducing offproperty emissions from a source such
as a Concentrated Animal Feeding
Operation (CAFO). The commenter
further stated that the proposal
mandates that State agencies order the
removal of trees from private property to
collect valid data.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
The EPA disagrees that the proposed
requirement is impractical or unlawful.
The proposed requirement would not
require, mandate, or otherwise empower
monitoring agencies to force the
removal of trees on private property.
The EPA agrees with the commenter
that trees may at times be planted as
part of control strategies to reduce
offsite emissions and thus protect the
public, but the EPA disagrees with the
commenter that the trees must be
removed to perform ambient air
monitoring in these locations. Rather, if
trees or shrubs block the line-of-sight
fetch between the monitor’s probe inlet
and the source under investigation, it is
the EPA’s position that, for most cases,
a microscale designation does not
accurately reflect the monitoring scale
for this location, and instead the EPA
would recommend that the monitoring
scale be designated to a more
representative monitoring scale such as
middle scale or neighborhood scale.
Moreover, for cases where long-term
trend sites or monitors that determine
the design value for an area cannot
reasonably meet this regulatory siting
requirement, the EPA encourages
monitoring organizations to work with
their respective EPA Regional offices to
determine if a waiver from this siting
criteria may be appropriate under
appendix E, section 10.
For these reasons, the EPA is
finalizing the language as proposed.
j. Waiver Provisions
The EPA proposed to maintain the
appendix E, section 10 waiver
provisions in the current regulation for
siting criteria, but to modify section 10.3
to require that waivers from the probesiting criteria must be reevaluated and
renewed minimally every 5 years (88 FR
5677–78, January 27, 2023).
The EPA received one comment
supporting the proposal and no adverse
comments. Therefore, the EPA is
finalizing the language as proposed.
k. Acceptable Probe Materials
The EPA proposed to expand the list
of acceptable probe materials for
sampling reactive gases in appendix E,
section 9, from just borosilicate glass
and fluorinated ethylene propylene
(FEP) Teflon®, or their equivalents. The
EPA proposed to add polyvinylidene
fluoride (PVDF), also known as Kynar®,
polytetrafluoroethylene (PTFE), and
perfluoroalkoxy (PFA) to the list of
approved materials for efficiently
transporting gaseous criteria pollutants,
and the use of NafionTM upstream of
ozone analyzers (88 FR 5678, January
27, 2023). Mention of trade names or
PO 00000
Frm 00164
Fmt 4701
Sfmt 4700
commercial products does not
constitute endorsement.
The EPA received two comments
supporting the proposal and received no
adverse comments. Therefore, the EPA
is finalizing the language as proposed.
D. Incorporating Data From Next
Generation Technologies
In the proposal, the EPA requested
comment on how to incorporate data
from next generation technologies into
Agency efforts (88 FR 5678–80, January
27, 2023). The near real-time integration
of data from PM2.5 continuous monitors,
sensors, and satellites has allowed the
EPA to use data in certain informational
applications such as EPA and USFS’s
Fire and Smoke Map.199 This mapping
product uses Application Program
Interfaces (APIs) where data sets are
automatically shared on prespecified
computer servers. Given the success of
the Fire and Smoke Map, the EPA
indicated interest in exploring the use of
next-generation technologies to develop
additional approaches, products, and
applications to help address important
non-regulatory air quality data needs.
Therefore, the EPA solicited comment
on the most important data uses and
data sets to consider in such future
initiatives. Such approaches and/or
products could utilize historical or near
real-time data. The EPA sought this
input and prioritization on use of next
generation technologies to help improve
the utility of data to better support air
quality management to improve public
health and the environment.
The EPA received comments from
about two dozen entities on its request
for comments on how to incorporate
data from next generation technologies.
The entities that provided comment
included federal agencies;
representatives of industry and industry
groups; public health, medical, and
environmental organizations; State,
local and related multi-state
organizations involved in air program
management; Tribes and Tribal
organizations involved in air program
management; and other State and local
governments.
While there were some differences
across commenters, a majority of the
commenters support use of next
generation data for non-regulatory
purposes, but not for regulatory decision
making due to their inherent
uncertainties and limitations. The EPA
also received comments from some
environmental organizations support
using alternative data for regulatory
decision making.
199 Available
E:\FR\FM\06MRR3.SGM
at https://fire.airnow.gov/.
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
Many commenters pointed out that
they are already successfully using
sensor data and networks in
supplemental and informational
applications and support further
expansion of these capabilities. Across
many commenters, there was support
for using next generation data as ‘‘fit for
purpose,’’ filling in gaps, finding hot
spots, identifying and addressing EJ
concerns, and evaluating and informing
network siting. The EPA acknowledges
the successful examples of sensor data
and networks for non-regulatory
purposes. A few commenters support
expanding the use of sensor data to
provide real-time AQI; the EPA is
interested in this use of next generation
data as well. A few commenters pointed
the need for the EPA to work closely
with them and their communities to
understand and use next generation
data, while others expressed a desire for
help developing best practices around
collecting and using next generation
data, developing products with data
analysis/visualization, and developing
appropriate QA/QC for sensor data. The
EPA acknowledges each of these
requests and expects to continue to
work closely with SLTs and other
stakeholders to understand and develop
information on the collection and use of
next generation data.
A few commenters offered more
detailed comments. Some recommended
that the EPA repropose implementation
provisions related to next generation
technologies with greater clarity to
provide for meaningful comment. For
example, the use of low-cost sensor and
satellite data could be used in drawing
nonattainment area boundaries or
identifying sources for emissions
control, but doing so would be such a
significant change from prior EPA
policy that it warrants a more specific
proposal, beyond the scope of this
request for comment. In response to this
comment, the EPA notes it did not
propose or change the use of nonregulatory measurement data as part of
this proposal, but instead opened an
opportunity to comment about the use
of next generation technologies.
Another commenter stated that while
low-cost sensor data can be invaluable
for some purposes, the potentially
overwhelming amount of data produced
by sensors may present additional
challenges to communities without the
resources or expertise to analyze it. Cost
is another concern associated with some
next generation technologies of which
some communities may not be aware, as
the initial cost of the sensor alone is not
indicative of the total cost of operation,
which can include costs of internet
access and servers. The EPA appreciates
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
the need to consider all the costs of
implementing and maintaining sensor
data.
Another commenter stated that having
a dense sensor network collocated with
FRMs and FEMs could help ensure
timely maintenance of the regulatory
measurements in the event there
appears to be a divergence of data. The
EPA appreciates the comment that
emphasizes how sensors could be used
to complement the FRM and FEM data
with regard to ensuring timely
maintenance.
Another commentor strongly opposes
incorporating sensor data into any EPA
systems unless robust quality assurance
(QA) practices are widely established
and managed by qualified personnel.
The EPA agrees that QA is necessary,
and notes that the ‘‘fit for purpose’’
aspect of using sensor data will inform
the appropriate QA associated with the
intended use of such data.
In summary, the EPA invited
comment on how we should consider
incorporating data from next generation
technologies into our air monitoring
efforts. In seeking comment on this
topic, the EPA did not propose to add,
edit, or delete any regulatory language
associated with the PM NAAQS. The
EPA received comments from a variety
of entities that largely support using
next generation data for a variety of
purposes that supplement, but cannot
replace, the measurement data from
monitoring methods required (i.e.,
FRMs and FEMs) for regulatory decision
making. Across many commenters, there
was support for using next generation
technologies and data as ‘‘fit for
purpose,’’ filling in gaps, finding hot
spots, identifying, and addressing EJ
concerns, and evaluating and informing
network siting. Quality assurance of the
data will be an important component in
the use of next generation technology
data. The EPA will consider these
comments as it continues its work with
the co-regulated community comprised
of SLT agencies and other stakeholders
to understand and use next generation
data and joint efforts to manage the
nation’s ambient air.
VIII. Clean Air Act Implementation
Requirements for the Revised Primary
Annual PM2.5 NAAQS
The EPA’s revision to the primary
annual PM2.5 NAAQS discussed in
section II above triggers a number of
implementation related activities that
were described in the NPRM. The two
most immediate implementation
impacts following a final new or revised
NAAQS are related to stationary source
permitting and the initial area
designations process. Permitting
PO 00000
Frm 00165
Fmt 4701
Sfmt 4700
16365
implications are discussed below in
section VIII.E. With regard to initial area
designations, the EPA is separately
issuing a memorandum regarding the
Initial Area Designations for the Revised
Primary Annual Fine Particle National
Ambient Air Quality Standard
Memorandum (the ‘‘Annual PM2.5
NAAQS Designations Memorandum’’)
that will provide information about the
statutory schedule for the designations
process. For other implementation
related implications, please refer back to
the NPRM section VIII.
The NPRM also referred to the PM2.5
State Implementation Plan (SIP)
Requirements Rule (81 FR 58010,
August 24, 2016), which specifies
planning requirements for areas
designated as nonattainment for
purposes of the PM2.5 NAAQS and
includes a number of key
recommendations for areas to consider
implications of environmental justice
through the attainment planning
process, consistent with the
identification of at-risk groups in the
2019 ISA and ISA Supplement and the
statutory requirement to protect the
health of at-risk groups. As stated in the
NPRM, State and local air agencies are
encouraged to consider how they might
develop implementation plans that
encourage early emission reductions.
A. Designation of Areas
As discussed in section II, with
respect to the PM2.5 NAAQS, the EPA is
finalizing: (1) Revisions to the level of
the primary annual PM2.5 NAAQS and
retaining the current primary 24-hour
PM2.5 NAAQS (section II.B.4); and (2) no
change to the current secondary annual
and 24-hour PM2.5 NAAQS at this time
(section V.B.4). Upon promulgation of a
new or revised NAAQS, States and the
EPA must initiate the process for initial
designations.
The timeline for initial area
designations begins with promulgation
of the revised primary annual PM2.5
NAAQS, as stated in the CAA section
107(d)(1)(B)(i). Through this process,
which provides for input from States
and others at various stages, the EPA
identifies areas of the country that either
meet or do not meet the revised primary
annual PM2.5 NAAQS, along with the
nearby areas contributing to NAAQS
violations. The following includes
additional information regarding the
designations process described in the
CAA.
Section 107(d)(1) of the CAA states
that, ‘‘By such date as the Administrator
may reasonably require, but not later
than 1 year after promulgation of a new
or revised national ambient air quality
standard for any pollutant under section
E:\FR\FM\06MRR3.SGM
06MRR3
16366
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
109, the Governor of each State shall
. . . submit to the Administrator a list
of all areas (or portions thereof) in the
State’’ and make recommendations for
whether the EPA should designate those
areas as nonattainment, attainment, or
unclassifiable.200 The CAA provides the
EPA with discretion to require States to
submit their designations
recommendations within a reasonable
amount of time not exceeding one
additional year.201 Section 107(d)(1)(A)
of the CAA also states that ‘‘the
Administrator may not require the
Governor to submit the required list
sooner than 120 days after promulgating
a new or revised national ambient air
quality standard.’’ Section
107(d)(1)(B)(i) further provides, ‘‘Upon
promulgation or revision of a NAAQS,
the Administrator shall promulgate the
designations of all areas (or portions
thereof) . . . as expeditiously as
practicable, but in no case later than 2
years from the date of promulgation.
Such period may be extended for up to
one year in the event the Administrator
has insufficient information to
promulgate the designations.’’ With
respect to the NAAQS setting process,
courts have interpreted the term
‘‘promulgation’’ to be signature and
widespread dissemination of a final
rule.202
If the EPA agrees with the
designations recommendation of the
State, then it may proceed to promulgate
the designations for such areas. If,
however, the EPA disagrees with the
State’s recommendation, then the EPA
may elect to make modifications to the
recommended designations. By no later
than 120 days prior to promulgating the
final designations, the EPA is required
to notify States of any intended
modifications to the State designation
recommendations for any areas or
portions thereof, including the
boundaries of areas, as the EPA may
deem necessary. States then have an
opportunity to comment on the EPA’s
intended modification and tentative
designation decision. If a State elects
not to provide designation
recommendations for any area, then the
EPA must itself promulgate the
designation that it deems appropriate.
200 While the CAA says ‘‘designating’’ with
respect to the Governor’s letter, in the full context
of the CAA section it is clear that the Governor
actually makes a recommendation to which the EPA
must respond via a specified process if the EPA
does not accept it.
201 In certain circumstances in which the
Administrator has insufficient information to
promulgate area designations within two years from
the promulgation of the NAAQS, CAA section
107(d)(1)(B)(i) provides that the EPA may extend
the designations schedule by up to one year.
202 API v. Costle, 609 F.2d 20 (D.C. Cir. 1979).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
While section 107(d) of the CAA
specifically addresses the designations
process for States, the EPA intends to
follow the same process for Tribes to the
extent practicable, pursuant to section
301(d) of the CAA regarding Tribal
authority, and the Tribal Authority Rule
(63 FR 7254, February 12, 1998). To
provide clarity and consistency in doing
so, the EPA issued a guidance
memorandum to our Regional Offices on
working with Tribes during the
designations process.203
Consistent with the process used in
previous area designations efforts, the
EPA will evaluate each area on a caseby-case basis considering the specific
facts and circumstances unique to the
area to support area boundary decisions
for the revised standard. The EPA
intends to issue a designations
memorandum which will provide
information regarding the designations
process. In broad overview, the EPA has
historically used area-specific analyses
to support nonattainment area boundary
recommendations and final boundary
determinations by evaluating factors
such as air quality data, emissions and
emissions-related data (e.g., population
density and degree of urbanization,
traffic and commuting patterns),
meteorology, geography/topography,
and jurisdictional boundaries. We
expect to follow a similar process when
establishing area designations for this
revised PM2.5 NAAQS. CAA section
107(d) explicitly requires that the EPA
designate as nonattainment not only the
area that is violating the pertinent
standard, but also those nearby areas
that contribute to the violation in the
violating area. In the PM2.5 NAAQS
Designations Memorandum, the EPA
intends to include information
regarding consideration of federal land
boundaries that may be fully or partially
included within the bounds of a county
otherwise identified as nonattainment.
As with past revisions of the PM2.5
NAAQS, the EPA intends to make the
designations decisions for the revised
primary annual PM2.5 NAAQS based on
the most recent three years of qualityassured, certified air quality data in the
EPA’s Air Quality System (AQS).
Accordingly, the EPA recommends that
States base their initial area designation
recommendations on the most current
available three years of complete and
certified air quality data at the time of
the recommendations. The EPA will
then base the final designations on the
most recent three consecutive years of
complete, certified air quality
monitoring data available at the time of
final designations.204
Monitoring data are currently
available from numerous existing PM2.5
Federal Equivalent Methods (FEM) and
Federal Reference Methods (FRM) sites
to determine violations of the revised
primary annual PM2.5 NAAQS. As
described in section VII.C.3.b, the EPA
took comment on how to deal with
cases where an FEM is approved by the
EPA with an update and when it can be
implemented in the field. The EPA took
comment on how to approach the data
produced during this lag and received
input from over a dozen commenters.
The commenters asked that the EPA be
flexible in allowing the use of updated
method correction factors intended to
improve the data comparability between
the FRMs and FEMs. The EPA will
address any data correction issues
between the FRMs and FEMs through a
future Notice of Data Availability
(NOA).
Consistent with past practice and as
noted in the NPRM, the EPA intends to
provide additional information
concerning the designations process,
including information about the
schedule and recommendations for
determining area boundaries in the
forthcoming Annual PM2.5 NAAQS
Designations Memorandum. Other
topics addressed in this memorandum
include the schedule for preparing and
submitting exceptional events initial
notification and exceptional events
demonstrations relevant to the
designations process, and information
related to wildfire and prescribed fire on
wildlands as it pertains to initial area
designations, as well as addressing
back-correction of PM FEM data when
a method has an approved factory
calibration as part of a method update.
The Annual PM2.5 NAAQS Designations
Memorandum is intended to assist
States and Tribes in formulating their
area recommendations.205
As discussed in the proposal, the
‘‘Treatment of Data Influenced by
Exceptional Events; Final Rule,’’ (81 FR
68216, October 3, 2016) and codified at
40 CFR 50.1, 40 CFR 50.14, and 40 CFR
51.930, contains instructions and
requirements for air agencies that may
203 ‘‘Guidance to Regions for Working with Tribes
during the National Ambient Air Quality Standards
(NAAQS) Designations Process,’’ December 20,
2011, Memorandum from Stephen D. Page to
Regional Air Directors, Regions 1–X available at
https://www.epa.gov/sites/default/files/2017-02/
documents/12-20-11_guidance_to_regions_for_
working_with_tribes_naaqs_designations.pdf.
204 In certain circumstances in which the
Administrator has insufficient information to
promulgate area designations within two years from
the promulgation of a new or revised NAAQS, CAA
section 107(d)(1)(B)(i) provides the EPA may extend
the designations schedule by up to one year.
205 See: https://www.epa.gov/particle-pollutiondesignations.
PO 00000
Frm 00166
Fmt 4701
Sfmt 4700
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
flag air quality data for certain days in
the Air Quality System due to potential
impacts from exceptional events (i.e.,
such as prescribed fires on wildland,
wildfires, or high wind dust storms).
Accordingly, for purposes of initial area
designations for a new or revised
NAAQS, an air agency may submit to
the EPA an exceptional events
demonstration with supporting
information and analyses for each
monitoring site and day the air agency
claims the EPA should exclude from
design value calculations for
designations purposes.
The EPA has provided tools to assist
air agencies in preparing adequate
exceptional events demonstrations.206
Further, the EPA will continue to work
with air agencies as they identify
exceptional events that may influence
decisions related to the initial area
designations process, and to prepare
and submit exceptional events
demonstrations if appropriate.
Importantly, air quality monitoring data
may be influenced by emissions from
prescribed fires on wildland and
wildfires. The EPA’s Exceptional Events
Rule provides for both of these types of
events to be considered as exceptional
events, provided the affected air
agencies submit exceptional events
demonstrations that meet the procedural
and technical requirements of the EPA’s
Exceptional Events Rule. To that end,
the EPA has issued guidance addressing
development of exceptional events
demonstrations for both wildfire and
prescribed fires on wildland.207 In light
of the growing frequency and severity of
wildfire events, and expected increases
in the application of prescribed fire as
a means to achieve long-term reductions
in high severity wildfire risk and
associated smoke impacts, the EPA
seeks to ensure that the Agency’s
exceptional events process provides an
efficient and clear pathway for
excluding data that may be affected by
such events in a manner that is
consistent with the Clean Air Act and
the public health objectives of the
NAAQS. Accordingly, the EPA is
continuing to explore opportunities to
develop additional tools that could
206 See the EPA’s Exceptional Events homepage at
https://www.epa.gov/air-quality-analysis/treatmentair-quality-data-influenced-exceptional-eventshomepage-exceptional.
207 See EPA’s ‘‘Final Guidance on the Preparation
of Exceptional Events Demonstrations for Wildfire
Events that May Influence Ozone Concentrations
and EPA’s Exceptional Events Guidance: Prescribed
Fire on Wildland that May Influence Ozone and
Particulate Matter Concentrations,’’ found on EPA’s
Exceptional Events homepage at https://
www.epa.gov/air-quality-analysis/treatment-airquality-data-influenced-exceptional-eventshomepage-exceptional.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
assist air agencies in preparing
exceptional events demonstrations for
wildfires and prescribed fires on
wildland. In addition, EPA intends to
continue engaging with the U.S.
Department of Agriculture, U.S.
Department of the Interior, air agencies,
and other stakeholders on these issues.
For more information regarding the
exceptional events demonstration
submission deadlines for the area
designations process, please see Table 2
to 40 CFR 50.14(c)(2)(vi)—‘‘Schedule for
Initial Notification and Demonstration
Submission for Data Influenced by
Exceptional Events for Use in Initial
Area Designations.’’
B. Section 110(a)(1) and (2)
Infrastructure SIP Requirements
As discussed in the NPRM, the CAA
directs States to address basic SIP
requirements to implement, maintain,
and enforce the NAAQS. Under CAA
sections 110(a)(1) and (2), states are
required to have State implementation
plans that provide the necessary air
quality management infrastructure that
provides for the implementation,
maintenance, and enforcement of the
NAAQS. After the EPA promulgates a
new or revised NAAQS, States are
required to make a new SIP submission
to establish that they meet the necessary
structural requirements for such new or
revised NAAQS or make changes to do
so. The EPA refers to this type of SIP
submission as an ‘‘infrastructure SIP
submission.’’ Under CAA section
110(a)(1), all States are required to make
these infrastructure SIP submissions
within three years after the effective
date of a new or revised primary
standard. While the CAA authorizes the
EPA to set a shorter time for States to
make these SIP submissions, the EPA is
requiring submission of infrastructure
SIPs within three years of the effective
date of this revised primary annual
PM2.5 NAAQS.
The EPA has provided general
guidance to States concerning its
interpretation of these requirements of
CAA section 110(a)(1) and (2) in the
context of infrastructure SIP
submissions for a new or revised
NAAQS.208 The EPA encourages States
to use this guidance when developing
their infrastructure SIPs for this revised
primary annual PM2.5 NAAQS.
As a reminder, the EPA notes that
States are not required to address
nonattainment plan requirements for
purposes of the revised primary annual
208 See ‘‘Guidance on Infrastructure State
Implementation Plan (SIP) Elements under Clean
Air Act Sections 110(a)(1) and 110(a)(2)’’ September
2013, Memorandum from Stephen D. Page to
Regional Air Directors, Regions 1–10.
PO 00000
Frm 00167
Fmt 4701
Sfmt 4700
16367
PM2.5 NAAQS on the same schedule as
infrastructure SIP requirements. The
EPA interprets the CAA such that two
elements identified in section 110(a)(2)
are not subject to the 3-year submission
deadline of section 110(a)(1) and thus
States are not required to address them
in the context of an infrastructure SIP
submission. The elements pertain to
part D, in title I of the CAA, which
addresses additional SIP requirements
for nonattainment areas. Therefore, for
the reasons explained below, the
following section 110(a)(2) elements are
considered by the EPA to be outside the
scope of infrastructure SIP actions: (1)
The portion of section 110(a)(2)(C),
programs for enforcement of control
measures and for construction or
modification of stationary sources that
applies to permit programs applicable in
designated nonattainment areas (known
as ‘‘nonattainment new source review’’)
under part D; and (2) section
110(a)(2)(I), which requires a SIP
submission pursuant to part D, in its
entirety.
Accordingly, the EPA does not expect
States to address the requirement for a
new or revised NAAQS in the
infrastructure SIP submissions to
include regulations or emissions limits
developed specifically for attaining the
relevant standard in areas designated
nonattainment for the revised primary
annual PM2.5 NAAQS. States are
required to submit infrastructure SIP
submissions for the revised primary
annual PM2.5 NAAQS before they will
be required to submit nonattainment
plan SIP submissions to demonstrate
attainment with the same NAAQS.
States are required to submit
nonattainment plan SIP submissions to
provide for attainment and maintenance
of a revised primary annual PM2.5
NAAQS within 18 months from the
effective date of nonattainment area
designations as required under CAA
section 189(a)(2)(B). The EPA reviews
and acts upon these later SIP
submissions through a separate process.
For this reason, the EPA does not expect
States to address new nonattainment
area emissions controls per section
110(a)(2)(I) in their infrastructure SIP
submissions.
One of the required infrastructure SIP
elements is that each State SIP must
contain adequate provisions to prohibit,
consistent with the provisions of title I
of the CAA, emissions from within the
State that will significantly contribute to
nonattainment in, or interfere with
maintenance by, any other State of the
primary or secondary NAAQS.209 This
209 CAA
E:\FR\FM\06MRR3.SGM
section 110(a)(2)(D)(i)(I).
06MRR3
16368
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
element is often referred to as the ‘‘good
neighbor’’ or ‘‘interstate transport’’
provision.210 The provision has two
prongs: significant contribution to
nonattainment (prong 1), and
interference with maintenance (prong
2). The EPA and States must give
independent significance to prong 1 and
prong 2 when evaluating downwind air
quality problems under CAA section
110(a)(2)(D)(i)(I).211 Further, case law
has established that the EPA and States
must implement requirements to meet
interstate transport obligations in
alignment with the applicable statutory
attainment schedule of the downwind
areas impacted by upwind-state
emissions.212 Thus, the EPA anticipates
that States will need to address
interstate transport obligations
associated with this revised PM
NAAQS, in alignment with the
provisions of subpart 4 of part D of the
CAA, as discussed in more detail in
section VIII.C below. Specifically, States
must implement any measures required
to address interstate transport
obligations as expeditiously as
practicable and no later than the next
statutory attainment date, i.e., for this
NAAQS revision as expeditiously as
practicable, but no later than the end of
the sixth calendar year following
nonattainment area designations. See
CAA section 188(c). States may find it
efficient to make SIP submissions to
address the interstate transport
provisions separately from other
infrastructure SIP elements.
Each State has the authority and
responsibility to review its air quality
management program’s existing SIP
provisions in light of a new or revised
NAAQS to determine if any revisions
are necessary to implement the new or
revised NAAQS. Most States have
revised and updated their SIPs in recent
years to address requirements associated
with other revised NAAQS. For certain
infrastructure elements, some States
may believe they already have adequate
State regulations adopted and approved
into the SIP to address a particular
requirement with respect to the revised
primary annual PM2.5 NAAQS.
If a State determines that existing SIPapproved provisions are adequate in
light of this revised primary annual
PM2.5 NAAQS with respect to a given
infrastructure SIP element (or sub210 CAA section 110(a)(2)(D)(i)(II) also addresses
certain interstate effects that states must address
and thus is also sometimes referred to as relating
to ‘‘interstate transport.’’
211 See North Carolina v. EPA, 531 F.3d 896, 909–
11 (D.C. Cir. 2008).
212 See id. 911–13. See also Wisconsin v. EPA,
938 F.3d 303, 313–20 (D.C. Cir. 2019); Maryland v.
EPA, 958 F.3d 1185, 1203–04 (D.C. Cir. 2020).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
element), then the State may make an
infrastructure SIP submission
‘‘certifying’’ that the existing State’s
existing EPA approved SIP already
contains provisions that address one or
more specific section 110(a)(2)
infrastructure elements.213 In the case of
such a submission, the State does not
have to include a copy of the relevant
provision (e.g., rule or statute) itself.
Rather, this certification submission
should provide citations to the SIPapproved State statutes, regulations, or
non-regulatory measures, as
appropriate, in or referenced by the
already EPA-approved SIP that meet
particular infrastructure SIP element
requirements. The State’s infrastructure
SIP submission should also include an
explanation as to how the State has
determined that those existing
provisions meet the relevant
requirements.
Like any other SIP submission, that
State can make such an infrastructure
SIP submission certifying that it has
already met some or all of the applicable
requirements only after it has provided
reasonable notice and opportunity for
public hearing. This ‘‘reasonable notice
and opportunity for public hearing’’
requirement for infrastructure SIP
submissions is to meet the requirements
of CAA sections 110(a) and 110(l).
Under the EPA’s regulations at 40 CFR
part 51, if a public hearing is held, an
infrastructure SIP submission must
include a certification by the State that
the public hearing was held in
accordance with the EPA’s procedural
requirements for public hearings. See 40
CFR part 51, appendix V, section 2.1(g),
and see 40 CFR 51.102.
In consultation with the EPA’s
Regional office, a State should follow all
applicable EPA regulations governing
infrastructure SIP submissions in 40
CFR part 51—e.g., subpart I (Review of
New Sources and Modifications),
subpart J (Ambient Air Quality
Surveillance), subpart K (Source
Surveillance), subpart L (Legal
Authority), subpart M
(Intergovernmental Consultation),
subpart O (Miscellaneous Plan Content
Requirements), subpart P (Protection of
Visibility), and subpart Q (Reports). For
the EPA’s general criteria for
infrastructure SIP submissions, refer to
40 CFR part 51, appendix V, Criteria for
Determining the Completeness of Plan
Submissions. For additional information
on infrastructure SIP submission
requirements, refer to the EPA’s 2013
guidance entitled ‘‘Guidance on
213 A ‘‘certification’’ approach would not be
appropriate for the interstate pollution control
requirements of CAA section 110(a)(2)(D)(i).
PO 00000
Frm 00168
Fmt 4701
Sfmt 4700
Infrastructure State Implementation
Plan (SIP) Elements under Clean Air Act
Sections 110(a)(1) and 110(a)(2).’’ The
EPA recommends that States
electronically submit their
infrastructure SIPs to the EPA through
the State Plan Electronic Collaboration
System (SPeCS),214 an online system
available through the EPA’s Central Data
Exchange.
C. Implementing Revised Primary
Annual PM2.5 NAAQS in Nonattainment
Areas
As discussed in the NPRM, the EPA
issued a SIP Requirements Rule for
implementing the PM2.5 NAAQS (81 FR
58010, August 24, 2016) (PM2.5 SIP
Requirements Rule). It provides
guidance and establishes additional
regulatory requirements for States
regarding development of attainment
plans for nonattainment areas for the
1997, 2006, and 2012 revisions of the
PM2.5 NAAQS. The guidance and
regulations in the SIP Requirements
Rule also apply to any States for which
the EPA promulgates nonattainment
area designations for the new revised
primary annual PM2.5 NAAQS.
The PM2.5 SIP Requirements Rule
provides comprehensive information
regarding nonattainment plan
requirements including, among other
things: nonattainment area emissions
inventories; policies regarding PM2.5
precursor pollutants (i.e., SO2, NOX,
VOC, and ammonia); control strategies
(such as reasonably available control
measures and reasonably available
control technology for direct PM2.5 and
relevant precursors); air quality
modeling; attainment demonstrations;
reasonable further progress
requirements; quantitative milestones;
and contingency measures. Information
provided in the PM2.5 SIP Requirements
Rule is supplemented by other EPA
documents, including guidance on
emissions inventory development (80
FR 8787, February 19, 2015; U.S. EPA,
2017), optional PM2.5 precursor
demonstrations (U.S. EPA, 2019b),215
and guidance on air quality modeling
for meeting air quality goals for the
ozone and PM2.5 NAAQS and regional
haze program (U.S. EPA, 2018b).
As stated in the NPRM, the PM2.5 SIP
Requirements Rule provides
recommendations to States regarding
consideration of environmental justice
in the context of PM2.5 attainment
214 https://cdx.epa.gov/.
215 Provides guidance on developing
demonstrations under section 189(e) intended to
show that a certain PM2.5 precursor in a particular
nonattainment area does not significantly
contribute to PM2.5 concentrations that exceed the
standard.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
planning. Some of the considerations for
States include: (1) Identifying areas with
overburdened communities where more
ambient monitoring may be warranted;
(2) targeting emissions reductions that
may be needed to attain the PM2.5
NAAQS; and (3) increasing
opportunities for meaningful
involvement for overburdened
populations (see 88 FR 5558, 5684,
January 27, 2023; 80 FR 58010, 58136,
August 25, 2016). In light of the
identification of at-risk populations for
this reconsideration, the EPA
encourages States to consider these and
other factors as part of their attainment
plan SIP development process.
The PM2.5 SIP Requirements Rule
outlines some examples of how States
can elect to implement these
recommendations.216 For instance,
States can use modeling and screening
tools to better understand where sources
of PM2.5 or PM2.5 precursor emissions
are located and identify areas that may
be candidates for additional ambient
monitoring. Furthermore, once these
target areas are identified, States can
prioritize direct PM2.5 or PM2.5
precursor control measures and
enforcement strategies in these areas to
reduce ambient PM2.5 and achieve the
NAAQS. As articulated in the NPRM
and the PM2.5 SIP Requirements Rule,
the EPA recognizes that States have
flexibility under the CAA to concentrate
State resources on controlling sources of
PM2.5 emissions in light of
environmental justice considerations
(see 88 FR 5558, 5684, January 27, 2023;
81 FR 58010, 58137, August 24, 2016).
Moreover, States can establish
opportunities to bolster meaningful
involvement in a number of ways, such
as communicating in appropriate
languages, ensuring access to draft SIPs
and other information, and developing
enhanced notice-and-comment
opportunities, as appropriate (see 88 FR
5558, 5684, January 27, 2023; 80 FR
58010, 58136, August 25, 2016).
As previously mentioned, the PM2.5
SIP Requirements Rule provides
guidance and regulatory requirements
for remaining nonattainment areas for
the 1997, 2006, and 2012 revisions of
the PM2.5 NAAQS, as well as for
nonattainment areas designated
pursuant to any future revisions of the
PM2.5 NAAQS, including the revised
annual PM2.5 NAAQS being finalized in
this action. The EPA is not making any
changes to the current PM2.5 SIP
Requirements Rule.
216 For more information on the EPA’s
recommendations and examples, see 81 FR 58010,
58137, August 24, 2016.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
D. Implementing the Primary and
Secondary PM10 NAAQS
As summarized in sections III.B.4 and
V.B.4 above, the EPA is retaining the
current primary and secondary 24-hour
PM10 NAAQS to protect against the
health effects associated with short-term
exposures to thoracic coarse particles
and against the welfare effects
considered in this reconsideration (i.e.,
visibility, climate, and materials effects).
The EPA is retaining the existing
implementation strategy for meeting the
CAA requirements for the PM10
NAAQS. States and emissions sources
should continue to follow the existing
regulations and guidance for
implementing the current standards.217
E. Prevention of Significant
Deterioration and Nonattainment New
Source Review Programs for the Revised
Primary Annual PM2.5 NAAQS
The CAA, at parts C and D of title I,
contains preconstruction review and
permitting programs applicable to new
major stationary sources and major
modifications of existing major sources.
The preconstruction review of each new
major stationary source and major
modification applies on a pollutantspecific basis, and the requirements that
apply for each pollutant depend on
whether the area in which the source is
situated is designated as attainment (or
unclassifiable) or nonattainment for that
pollutant. In areas designated
attainment or unclassifiable for a
pollutant, the Prevention of Significant
Deterioration (PSD) requirements under
part C apply to construction at major
sources. In areas designated
nonattainment for a pollutant, the
Nonattainment New Source Review
(NNSR) requirements under part D
apply to construction at major sources.
Collectively, those two sets of permit
requirements are commonly referred to
as the ‘‘major New Source Review’’ or
‘‘major NSR’’ programs.
Until the EPA designates an area with
respect to the revised primary annual
PM2.5 NAAQS, the NSR provisions
applicable under an area’s current
designation for the 1997, 2006, and 2012
PM2.5 NAAQS would continue to apply.
See 40 CFR 51.166(i)(2) and 52.21(i)(2).
That is, for areas designated as
217 CAA Sections 110(a) and 172 contain general
nonattainment planning provisions, regarding the
public review, adoption, submittal, and content of
implementation plans. CAA Section 189 specifies
additional plan provisions for particulate matter
nonattainment areas. General Preamble for the
Implementation of Title I of the Clean Air Act
Amendments of 1990 provides a detailed
discussion of the EPA’s interpretation of the Title
I requirements (57 FR 13498, April 16, 1992; 59 FR
41998, August 16, 1994).
PO 00000
Frm 00169
Fmt 4701
Sfmt 4700
16369
attainment/unclassifiable for the 1997,
2006, and 2012 PM2.5 NAAQS, PSD will
apply to new major stationary sources
and major modifications that trigger
major source permitting requirements
for PM2.5. For areas designated
nonattainment for the 1997, 2006, or
2012 PM2.5 NAAQS, NNSR
requirements will apply for new major
stationary sources and major
modifications that trigger major source
permitting requirements for PM2.5.
When the initial area designations for
this revised primary annual PM2.5
NAAQS become effective, those
designations will further determine
whether PSD or NNSR applies to PM2.5
in a particular area, depending on the
designation status. New major sources
and major modifications will be subject
to the PSD program requirements for
PM2.5 if they are located in an area that
does not have a current nonattainment
designation under CAA section 107 for
PM2.5.218
Under the PSD program, the permit
applicant must demonstrate that the
new or modified source emissions
increase does not cause or contribute to
a NAAQS violation. In 2017, the EPA
revised the Guideline on Air Quality
Models (published as appendix W to 40
CFR part 51) to address primary and
secondary PM2.5 impacts in making this
demonstration. The EPA has since
provided associated technical guidance,
models and tools, such as the recent
‘‘Final Guidance for Ozone and Fine
Particulate Matter Permit Modeling’’
(July 29, 2022).219 Additionally, in light
of this NAAQS revision, the EPA is
updating its guidance that provides
recommended significant impact levels
(SILs) for PM2.5 and expects that an
updated SIL for the revised primary
annual PM2.5 NAAQS will be available
218 40
CFR 51.166(i)(2) and 52.21(i)(2).
July 29, 2022, the EPA issued ‘‘Final
Guidance for Ozone and Fine Particulate Matter
Permit Modeling,’’ available at https://
www.epa.gov/system/files/documents/2022-07/
Guidance_for_O3_PM25_Permit_Modeling.pdf. This
guidance provides the EPA’s recommendations for
how a stationary source seeking a PSD permit may
demonstrate that it will not cause or contribute to
a violation of the National Ambient Air Quality
Standards for Ozone and PM2.5 and PSD increments
for PM2.5, as required under section 165(a)(3) of the
Clean Air Act and 40 CFR 51.166(k) and 52.21(k).
The EPA has also previously issued two technical
guidance documents for use in conducting these
demonstrations: ‘‘Guidance on the Development of
Modeled Emission Rates for Precursors (MERPs) as
a Tier 1 Demonstration Tool for Ozone and PM2.5
under the PSD Permitting Program,’’ available at
https://www.epa.gov/sites/default/files/2020-09/
documents/epa-454_r-19-003.pdf, and ‘‘Guidance
on the Use of Models for Assessing the Impacts of
Emissions from Single Sources on the Secondarily
Formed Pollutants: Ozone and PM2.5,’’ available at
https://www.epa.gov/sites/default/files/2020-09/
documents/epa-454_r-16-005.pdf.
219 On
E:\FR\FM\06MRR3.SGM
06MRR3
16370
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
on or before the effective date of the
final NAAQS.
The statutory requirements for a PSD
permit program set forth under part C of
title I of the CAA (sections 160 through
169) are addressed by the EPA’s PSD
regulations found at 40 CFR 51.166
(minimum requirements for an
approvable PSD SIP) and 40 CFR 52.21
(PSD permitting program for permits
issued under the EPA’s Federal
permitting authority). These regulations
already apply to PM2.5 in areas that are
designated attainment or unclassifiable
for PM2.5 whenever a proposed new
major source or major modification
triggers PSD requirements for PM2.5.
For PSD, a ‘‘major stationary source’’
is one with the potential to emit 250
tons per year (tpy) or more of any
regulated NSR pollutant, unless the new
or modified source is classified under a
list of 28 source categories contained in
the statutory definition of ‘‘major
emitting facility’’ in section 169(1) of
the CAA. For those 28 source categories,
a ‘‘major stationary source’’ is one with
the potential to emit 100 tpy or more of
any regulated NSR pollutant. A ‘‘major
modification’’ is a physical change or a
change in the method of operation of an
existing major stationary source that
results, first, in a significant emissions
increase of a regulated NSR pollutant
and, second, in a significant net
emissions increase of that pollutant. See
40 CFR 51.166(b)(2)(i), 40 CFR
52.21(b)(2)(i). The EPA PSD regulations
define the term ‘‘regulated NSR
pollutant’’ to include any pollutant for
which a NAAQS has been promulgated
and any pollutant identified by the EPA
as a constituent or precursor to such
pollutant. See 40 CFR 51.166(b)(49), 40
CFR 52.21(b)(50). These regulations
identify SO2 and NOX as precursors to
PM2.5 in attainment and unclassifiable
areas. See 40 CFR 51.166(b)(49)(i)(b), 40
CFR 52.21(b)(50)(i)(b).220 Thus, for
PM2.5, the PSD program currently
requires the review and control of
emissions of direct PM2.5 emissions and
SO2 and NOX (as precursors to PM2.5),
absent a demonstration otherwise for
NOX. Among other things, for each
regulated NSR pollutant emitted or
increased in a significant amount, the
220 Sulfur dioxide is a precursor to PM
2.5 in all
attainment and unclassifiable areas. NOX is
presumed to be a precursor to PM2.5 in all
attainment and unclassifiable areas, unless a state
or the EPA demonstrates that emissions of NOX
from sources in a specific area are not a significant
contributor to that area’s ambient PM2.5
concentrations. VOC is presumed not to be a
precursor to PM2.5 in any attainment or
unclassifiable area, unless a state or the EPA
demonstrates that emissions of VOC from sources
in a specific area are a significant contributor to that
area’s ambient PM2.5 concentrations.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
PSD program requires a new major
stationary source or a major
modification to apply the ‘‘best
available control technology’’ (BACT) to
limit emissions and to conduct an air
quality impact analysis to demonstrate
that the proposed major stationary
source or major modification will not
cause or contribute to a violation of any
NAAQS or PSD increment.221 See CAA
section 165(a)(3) and (4), 40 CFR
51.166(j) and (k), 40 CFR 52.21(j) and
(k). The PSD requirements may also
include, in appropriate cases, an
analysis of potential adverse impacts on
Class I areas. See CAA sections 162(a)
and 165(d), 40 CFR 51.166(p); 40 CFR
52.21(p)).222 The EPA developed the
Guideline on Air Quality Models and
other documents to, among other things,
provide methods and guidance for
demonstrating that increased emissions
from construction will not cause or
contribute to exceedances of the PM2.5
NAAQS and PSD increments for
PM2.5.223
Upon the effective date of the revised
primary annual PM2.5 NAAQS, the
demonstration required under CAA
Section 165(a)(3), and the associated
regulations, must include the revised
primary annual PM2.5 NAAQS. In past
NAAQS revision rules, including the
2012 PM2.5 NAAQS (78 FR 3086,
January 15, 2013) and 2015 Ozone
NAAQS (80 FR 65292, October 26,
2015), the EPA included limited
provision that exempted certain sources
with pending PSD permit applications
(those that had reached a particular
stage in the permitting process at the
time the revised NAAQS was
promulgated or became effective) from
the requirement to demonstrate that the
proposed emissions increases would not
cause or contribute to a violation of the
221 By establishing the maximum allowable level
of ambient pollutant concentration increase in a
particular area, an increment defines ‘‘significant
deterioration’’ of air quality in that area. Increments
are defined by the CAA as maximum allowable
increases in ambient air concentrations above a
baseline concentration and are specified in the PSD
regulations by pollutant and area classification
(Class I, II and III). 40 CFR 51.166(c), 40 CFR
52.21(c); 75 FR 64864 (October 20, 2010).
222 Congress established certain Class I areas in
section 162(a) of the CAA, including international
parks, national wilderness areas, and national parks
that meet certain criteria. Such Class I areas, known
as mandatory Federal Class I areas, are afforded
special protection under the CAA. In addition,
States and Tribal governments may establish Class
I areas within their own political jurisdictions to
provide similar special air quality protection.
223 See 40 CFR part 51, appendix W; 82 FR 5182
(January 17, 2017); See also U.S. EPA, 2021d. The
EPA provided an initial version of the 2021
guidance for public comment on February 10, 2020.
Upon consideration of the comments received, and
consistent with Executive Order 13990, the EPA
revised the initial draft guidance and posted the
revised version for additional public comment.
PO 00000
Frm 00170
Fmt 4701
Sfmt 4700
revised NAAQS.224 In August 2019, the
U.S. Court of Appeals for the D.C.
Circuit vacated the exemption provision
in the PSD rules for the 2015 Ozone
NAAQS, finding that the provision
contradicted ‘‘Congress’s ‘express policy
choice’ not to allow construction which
will ‘cause or contribute to’
nonattainment of ‘any’ effective
NAAQS, regardless of when they are
adopted or when a permit was
completed.’’ Murray Energy Corp. v.
EPA, 936 F.3d 597, 627 (D.C. Cir.
2019).225 Based on that court decision,
the EPA is not establishing any PSD
permitting exemption provision in this
action. Some commenters requested that
the EPA provide the same kind of relief
for pending PSD permit applications by
extending the effective date of this new
revised NAAQS beyond the 60 days that
the EPA has traditionally used for such
rules. Such comments are addressed in
the Response to Comments portion of
this action. The EPA is making this
revised primary annual PM2.5 NAAQS
effective in 60 days.
The EPA anticipates that the existing
PM2.5 air quality in some areas will not
be in attainment with the revised
primary annual PM2.5 NAAQS, and the
EPA will designate these areas as
nonattainment at a later date, consistent
with the designation process described
in the preceding sections. However,
until such nonattainment designation
occurs, proposed new major sources and
major modifications located in any area
currently designated attainment or
unclassifiable for all preexisting PM2.5
NAAQS will continue to be subject to
the PSD program requirements for
PM2.5. Any proposed major stationary
source or major modification triggering
PSD requirements for PM2.5 that does
not receive its PSD permit by the
effective date of a new nonattainment
designation for the area where the
source would locate would then be
required to satisfy applicable NNSR
preconstruction permit requirements for
PM2.5.
In areas where air pollution exceeds
the level of the revised primary annual
PM2.5 NAAQS, a PSD permit applicant
must demonstrate that the source or
modification will not cause or
224 This exemption was referred to as
‘‘grandfathering’’ in the 2015 Ozone NAAQS and
the D.C. Circuit’s Murray Energy Corp. decision on
that exemption. See 80 FR 65292, 65431 (October
26, 2015); Murray Energy Corp. v. EPA, 936 F.3d
597, 627 (D.C. Cir. 2019). The EPA refers to this
‘‘grandfathering’’ provision in this action as an
exemption provision.
225 While the specifics of this case involved the
2015 ozone NAAQS, the case was based upon an
interpretation of CAA section 165(a) and therefore
applies equally to any PSD permitting exemption
provision for a new or revised NAAQS.
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
contribute to a violation of the NAAQS.
Section 165(a)(3)(B) of the CAA states
that a proposed source may not
construct unless it demonstrates that it
will not cause or contribute to a
violation of any NAAQS. This statutory
requirement is implemented through a
provision contained in the PSD
regulations at 40 CFR 51.166(k) and
52.21(k).226 If a source cannot make this
demonstration, or if its initial air quality
impact analysis shows that the source’s
impact would cause or contribute to a
violation, the reviewing authority may
not issue a PSD permit to that source.
However, a PSD permit applicant may
be able to make this demonstration if it
compensates for the adverse impact that
would otherwise cause or contribute to
a violation of the NAAQS. In contrast to
the NSR requirements for nonattainment
areas, the PSD regulations do not
explicitly specify remedial actions that
a prospective source must take to
address such a situation, but the EPA
has historically recognized that sources
applying for PSD permits may utilize
offsetting reductions in emissions as
part of the required PSD demonstration
under CAA section 165(a)(3)(B).227
Part D of title I of the CAA includes
preconstruction review and permitting
requirements applicable to new major
stationary sources and major
modifications located in areas
designated nonattainment for a
pollutant for which the EPA has
established a NAAQS (i.e., a criteria
pollutant). The relevant part D
requirements are typically referred to as
226 40 CFR 51.166(k) states that SIPs must require
that the owner or operator of the proposed source
or modification demonstrate that allowable
emission increases from the proposed source or
modification, in conjunction with all other
applicable emissions increases or reductions
(including secondary emissions), would not cause
or contribute to air pollution in violation of: (i) Any
national ambient air quality standard in any air
quality control region; or (ii) any applicable
maximum allowable increase over the baseline
concentration in any area.
227 See, e.g., Memorandum from Stephen D. Page,
Director, Office of Air Quality Planning and
Standards to Regional Air Division Directors,
Guidance Concerning Implementation of the 1-hour
SO2 NAAQS for the Prevention of Significant
Deterioration Program. August 23, 2010. Office of
Air Quality Planning and Standards U.S. EPA,
Research Triangle Park. Available at: https://
www.epa.gov/sites/default/files/2015-07/
documents/appwso2.pdf; 44 FR 3274, 3278, January
16, 1979; See also In re Interpower of New York,
Inc., 5 E.A.D. 130, 141 (EAB 1994) (describing an
EPA Region 2 PSD permit that relied in part on
offsets to demonstrate the source would not cause
or contribute to a violation of the NAAQS). 52 FR
24634, 24684, July 1, 1987; 78 FR 3085, 3261–62,
January 15, 2013. The EPA has recognized the
ability of sources to obtain offsets in the context of
PSD though the PSD provisions of the Act do not
expressly reference offsets as the NNSR provisions
of the Act do. See 80 FR 65292, 65441, October 26,
2015.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
the nonattainment NSR (NNSR)
program. The EPA’s regulations for the
NNSR program are contained in 40 CFR
51.165 and 52.24 and part 51, appendix
S. Specifically, the EPA has developed
minimum program requirements for a
NNSR program that is approvable in a
SIP, and those requirements, which
include requirements for PM2.5, are
contained in 40 CFR 51.165. In addition,
40 CFR part 51, appendix S, contains
requirements constituting an interim
NNSR program. This interim program
enables NNSR permitting in
nonattainment areas by States that lack
a SIP-approved NNSR permitting
program during the time between the
date of the relevant designation and the
date that the EPA approves into the SIP
a NNSR program. See 40 CFR part 51,
appendix S, section I; 40 CFR 52.24(k).
For NNSR, ‘‘major stationary source’’
is generally defined as a source with the
potential to emit at least 100 tpy of the
regulated NSR pollutant for which the
area is designated nonattainment. In
some cases, however, the CAA and the
NNSR regulations define ‘‘major
stationary source’’ for NNSR in terms of
a lower rate dependent on the pollutant
and degree of nonattainment in the area.
For purposes of the PM2.5NAAQS, in
addition to the general threshold level
of 100 tpy in Moderate PM2.5
nonattainment areas, a lower major
source threshold of 70 tpy applies in
Serious PM2.5 nonattainment areas
pursuant to subpart 4 of part D, title I
of the CAA. See 40 CFR
51.165(a)(1)(iv)(A)(1)(vii) and (viii); 40
CFR part 51, appendix S, II.A.4(i)(a)(7)
and (8).
Under the NNSR program, direct
PM2.5 emissions and emissions of each
PM2.5 precursor are considered
separately in accordance with the
applicable major source threshold. For
example, the threshold for Serious PM2.5
nonattainment areas is 70 tpy of direct
PM2.5, as well as for the PM2.5
precursors SO2, NOX, VOC, and
ammonia.228 See 40 CFR
51.165(a)(1)(iv)(A)(1)(vii) and (viii); 40
CFR part 51, appendix S, II.A.4.(i)(a)(7)
and (8). A source qualifies as major for
nonattainment NSR in a PM2.5
nonattainment area if it emits or has the
potential to emit direct PM2.5 or any
228 All of these pollutants are identified as
precursors to PM2.5 in NNSR regulations. See 40
CFR 51.165(a)(1)(xxxvii)(C)(2). No significant
emission rate is established by the EPA for
ammonia, and states are required to define
‘‘significant’’ for ammonia for their respective areas
unless the state pursues the optional precursor
demonstration to exclude ammonia from planning
requirements. See 40 CFR 51.165(a)(1)(x)(F); 40 CFR
51.165(a)(13).
PO 00000
Frm 00171
Fmt 4701
Sfmt 4700
16371
PM2.5 precursor in an amount equal to
or greater than the applicable threshold.
For modifications, NNSR applies to
proposed physical changes or changes
in the method of operation of an
existing stationary source where (1) the
source is major for the nonattainment
pollutant (or a precursor for that
pollutant) and (2) the physical change or
change in the method of operation of a
major stationary source results, first, in
a significant emissions increase of a
regulated NSR pollutant and, second, in
a significant net emissions increase of
that same nonattainment pollutant (or
same precursor for that pollutant). See
40 CFR 51.165(a)(1)(v)(A); 40 CFR part
51, appendix S, II.A.5.(i). For example,
to qualify as a major modification for
SO2 (as a PM2.5 precursor) in a Moderate
PM2.5 nonattainment area, the existing
source would have to have the potential
to emit 100 tpy or more of SO2, and the
project would have to result in an
increase in SO2 emissions of 40 tpy or
more. See 40 CFR 51.165(a)(1)(x)(A).
New major stationary sources and
major modifications for PM2.5 subject to
NNSR must comply with the ‘‘lowest
achievable emission rate’’ (LAER), as
defined in the CAA and NNSR rules.
Such sources must also perform other
analyses and obtain emission offsets, as
required under section 173 of the CAA
and applicable regulations.
Following the promulgation of this
revised primary annual PM2.5 NAAQS,
some new areas may be designated
nonattainment for PM2.5. Where a State
does not have an existing NNSR
program or where the current NNSR
program does not apply to PM2.5, that
State will be required to submit the
necessary SIP revisions to ensure that
new major stationary sources and major
modifications for PM2.5 or a PM2.5
precursor undergo preconstruction
review pursuant to the NNSR program.
States with designated nonattainment
areas for the revised primary annual
PM2.5 NAAQS are required to make SIP
submissions to meet nonattainment plan
requirements within 18 months from the
effective date of designations, as
required under CAA section
189(a)(2)(B). States that have existing
NNSR program requirements that
cannot be interpreted to apply at the
time of designation to the revised
primary annual PM2.5 NAAQS may, in
the interim, issue permits in accordance
with the applicable nonattainment
permitting requirements contained in 40
CFR part 51, appendix S, which would
apply to the revised primary annual
PM2.5 NAAQS upon its effective date.
See 73 FR 28321, 28340, May 16, 2008.
Finally, the EPA has released several
documents that discuss air permitting
E:\FR\FM\06MRR3.SGM
06MRR3
16372
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
and environmental justice, including,
for example, a memorandum 229 and
attached permitting principles.230 The
EPA recommends that PSD and NNSR
permitting authorities review this
memorandum and the principles and
consider applying them in their air
permitting actions as appropriate to
help identify, analyze, and address
environmental justice concerns in those
air permitting actions to help ensure
that the NAAQS achieve their intended
health benefits for at-risk populations.
F. Transportation Conformity Program
Transportation conformity is required
under CAA section 176(c) to ensure that
transportation plans, transportation
improvement programs (TIPs) and
federally supported highway and transit
projects will not cause or contribute to
any new air quality violation, increase
the frequency or severity of any existing
violation, or delay timely attainment or
any required interim emissions
reductions or other milestones.
Transportation conformity applies to
areas that are designated as
nonattainment or nonattainment areas
that have been redesignated to
attainment with an approved CAA
section 175A maintenance plan (i.e.,
maintenance areas) for transportationrelated criteria pollutants: carbon
monoxide, ozone, NO2, PM2.5, and PM10.
Transportation conformity for the
revised primary annual PM2.5 NAAQS
does not apply until one year after the
effective date of nonattainment
designations for that NAAQS. See CAA
section 176(c)(6) and 40 CFR 93.102(d)).
The EPA’s Transportation Conformity
Rule 231 establishes the criteria and
procedures for determining whether
transportation activities conform to the
SIP. No changes are being made to the
transportation conformity rule in this
final rulemaking. The EPA notes that
the transportation conformity rule
already addresses the PM2.5 and PM10
NAAQS. However, in the future, the
EPA intends to review the need to issue
or revise guidance describing how the
current conformity rule applies in
nonattainment and maintenance areas
ddrumheller on DSK120RN23PROD with RULES3
229 Memorandum
from Joseph Goffman, Principal
Deputy Assistant Administrator, Office of Air and
Radiation, to Air and Radiation Division Directors,
‘‘Principles for Addressing Environmental Justice in
Air Permitting’’ (December 22, 2022), available at
https://www.epa.gov/caa-permitting/ej-airpermitting-principles-addressing-environmentaljustice-concerns-air.
230 Id., Attachment, ‘‘EJ in Air Permitting:
Principles for Addressing Environmental Justice
Concerns in Air Permitting’’ (December 2022),
available at https://www.epa.gov/caa-permitting/ejair-permitting-principles-addressingenvironmental-justice-concerns-air.
231 40 CFR part 93, subpart A.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
for the revised primary annual PM2.5
NAAQS, as needed.
G. General Conformity Program
The conformity requirement under
CAA section 176(c) ensures that federal
activities implemented by federal
agencies will not interfere with a State’s
ability to attain and maintain the
NAAQS. Under CAA 176(c)(1), the
requirement prohibits Federal agencies
from approving, permitting, licensing,
or funding activities that do not conform
to the purpose of the applicable SIP for
the control and prevention of air
pollution. See CAA 176(c)(1)(A). Under
CAA 176(c)(1)(B), conformity to an
implementation plan means that federal
activities will not cause or contribute to
any new violations of the NAAQS,
increase the frequency or severity of any
existing NAAQS violation, or delay
timely attainment or any required
interim emissions reductions or other
milestones contained in the applicable
SIP.
The general conformity program 232
implements CAA section 176(c)(4)(A),
and the criteria and procedures for
determining conformity of federal
activities to the applicable SIP are
established under 40 CFR part 93
subpart B, sections 93.150 through
93.165. General Conformity applies to
federal activities that (1) would cause
emissions of relevant criteria or
precursor pollutants to originate within
nonattainment areas or areas that have
been redesignated to attainment with an
approved CAA section 175A
maintenance plan (i.e., maintenance
areas), as set forth under 40 CFR 93.153,
and (2) are not Federal Highway
Administration (FHWA) or Federal
Transit Administration (FTA)
transportation projects as defined in 40
CFR 93.101 under the transportation
conformity requirements. See 40 CFR
93.153. General conformity for the
revised primary annual PM2.5 NAAQS
does not apply until one year after the
effective date of the nonattainment
designation for that NAAQS. See 40
CFR 93.153(k).
With regard to issues regarding
prescribed fires, which were addressed
earlier in this action, here is some
additional information regarding
prescribed fires and General Conformity
regulations. Under the General
Conformity regulations at 40 CFR
93.153(c)(4), a conformity evaluation is
not required to support a decision by a
federal agency to conduct or carry out
prescribed burning when the burn is
consistent with the terms of a land
management plan or other plan that
232 40
PO 00000
CFR part 93 subpart B.
Frm 00172
Fmt 4701
Sfmt 4700
includes the prescribed burn at issue,
where the overall plan that includes the
burn was previously evaluated under 40
CFR part 93 subpart B by the
responsible federal agency, and the
agency found the plan conforms under
CAA paragraphs 176(c)(1)(A) and (1)(B).
This assumes the burn at issue will be
conducted by meeting any conditions
specified as necessary for meeting
conformity in the agency’s decision to
approve the plan. Alternatively, a
presumption of conformity applies also
under 40 CFR 93.153(i)(2) for prescribed
fires conducted in accordance with a
Smoke Management Program that meets
the requirements of the EPA’s 1998
Interim Air Quality Policy on Wildland
and Prescribed Fires or an equivalent
replacement EPA policy. The preamble
to the Exceptional Events Rule explains
that the EPA adapted language
associated with the six basic
components of a certifiable Smoke
Management Program for exceptional
events purposes from the 1998 Interim
Air Quality Policy on Wildland and
Prescribed Fires (see, e.g., 81 FR 68216,
68252 (including footnote 75), 68256,
October 2, 2016). The Exceptional
Events Rule at 40 CFR 50.14(a)(3)(ii)(A)
also indicates that certain requirements
within the Exceptional Events Rule can
be satisfied if a prescribed fire is
conducted under a certified Smoke
Management Program or using
appropriate basic smoke management
practices such as those identified in
Table 1 to 40 CFR 50.14 (see e.g., 81 FR
68216, 68250–68257, 68277–68278,
October 3, 2016).
No changes are being made to the
general conformity regulations in this
final rulemaking and the EPA notes that
the courts recognize the regulations
constitute control for the established
PM2.5 and PM10 NAAQS. However, in
the future, the EPA intends to review
the need to issue or revise guidance
describing how the current General
Conformity regulations apply within
nonattainment and maintenance areas
for the revised primary annual PM2.5
NAAQS, as needed.233
IX. Statutory and Executive Order
Reviews
Additional information about these
statutes and Executive orders can be
233 Further, the EPA’s current Unified Agenda
and Regulatory Plan includes its intention to issue
a proposed rule to amend the General Conformity
Regulations. The EPA intends to address in that
regulatory action topics regarding prescribed fire,
including consideration of smoke management
approaches such as those discussed in the
Exceptional Events Rule, among other topics. See,
e.g., https://www.reginfo.gov/public/do/eAgenda
ViewRule?pubId=202310&RIN=2060-AV28.
E:\FR\FM\06MRR3.SGM
06MRR3
16373
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
found at https://www.epa.gov/lawsregulations/laws-and-executive-orders.
A. Executive Order 12866: Regulatory
Planning and Review and Executive
Order 14094: Modernizing Regulatory
Review
This action is ‘‘significant regulatory
action’’ as defined under section 3(f)(1)
of Executive Order 12866, as amended
by Executive Order 14094. Accordingly,
the EPA submitted this action to the
Office of Management and Budget
(OMB) for review. Documentation of
any changes made in response to the
Executive Order 12866 review is
available in the docket. The EPA
prepared an illustrative analysis of the
potential costs and benefits associated
with this action. This analysis,
‘‘Regulatory Impact Analysis for the
Reconsideration of the National
Ambient Air Quality Standards for
Particulate Matter,’’ is available in the
Regulatory Impact Analysis (RIA)
docket (EPA–HQ–OAR–2019–0587) and
briefly summarized below. However, the
CAA and judicial decisions make clear
that the economic and technical
feasibility of attaining ambient
standards are not to be considered in
setting or revising NAAQS, although
such factors may be considered in the
development of State plans to
implement the standards. Accordingly,
although an RIA has been prepared, the
results of the RIA have not been
considered in issuing this final rule.
The RIA estimates the costs and
monetized human health benefits in
2032, after implementing existing and
expected regulations and assessing
emissions reductions to meet the
current primary annual and 24-hour
particulate matter NAAQS (12/35 mg/
m3), associated with applying national
control strategies for the revised annual
and 24-hour standard levels of 9/35 mg/
m3, as well as the following less and
more stringent alternative standard
levels: (1) A less stringent alternative
annual standard level of 10 mg/m3 in
combination with the current 24-hour
standard (i.e., 10/35 mg/m3), (2) a more
stringent alternative annual standard
level of 8 mg/m3 in combination with the
current 24-hour standard (i.e., 8/35 mg/
m3), and (3) a more stringent alternative
24-hour standard level of 30 mg/m3 in
combination with an annual standard
level of 10 mg/m3 (i.e., 10/30 mg/m3).
Table 3 provides a summary of the
estimated monetized benefits, costs, and
net benefits associated with applying
national control strategies toward
reaching the revised and alternative
standard levels.
TABLE 3—ESTIMATED MONETIZED BENEFITS, COSTS, AND NET BENEFITS OF THE ILLUSTRATIVE CONTROL STRATEGIES
APPLIED TOWARD THE PRIMARY REVISED AND ALTERNATIVE ANNUAL AND DAILY STANDARD LEVELS OF 10/35 μg/m3,
10/30 μg/m3, 9/35 μg/m3, AND 8/35 μg/m3 IN 2032 FOR THE U.S.
[Millions of 2017$]
10/35
10/30
9/35
8/35
Benefits a ...........................
Costs b ...............................
$8,500 and $17,000 ..........
$200 ..................................
$10,000 and $21,000 ........
$340 ..................................
$22,000 and $46,000 ........
$590 ..................................
$48,000 and $99,000.
$1,500.
Net Benefits ................
$8,300 and $17,000 ..........
$9,900 and $21,000 ..........
$22,000 and $46,000 ........
$46,000 and $97,000.
Notes: Rows may not appear to add correctly due to rounding. We provide a snapshot of costs and benefits in 2032, using the best available
information to approximate social costs and social benefits recognizing uncertainties and limitations in those estimates. The estimated costs and
monetized human health benefits associated with applying national control strategies do not fully account for all the emissions reductions needed
to reach the final and more stringent alternative standard levels for some standard levels analyzed.
a We assume that there is a cessation lag between the change in PM exposures and the total realization of changes in mortality effects. Specifically, we assume that some of the incidences of premature mortality related to PM2.5 exposures occur in a distributed fashion over the 20
years following exposure, which affects the valuation of mortality benefits at different discount rates. Similarly, we assume there is a cessation
lag between the change in PM exposures and both the development and diagnosis of lung cancer. The benefits are associated with two point
estimates from two different epidemiologic studies, and we present the benefits calculated at a real discount rate of 3 percent. The monetized
benefits exclude additional health and welfare benefits that could not be quantified.
b The costs are annualized using a 7 percent interest rate.
ddrumheller on DSK120RN23PROD with RULES3
B. Paperwork Reduction Act (PRA)
This action does not impose any new
information collection burden under the
PRA. OMB has previously approved the
information collection activities
contained in the existing regulations
and has assigned OMB control number
2060–0084. The data collected through
this information collection consist of
ambient air concentration
measurements for the seven air
pollutants with national ambient air
quality standards (i.e., ozone, sulfur
dioxide, nitrogen dioxide, lead, carbon
monoxide, PM2.5 and PM10), ozone
precursors, air toxics, meteorological
variables at a select number of sites, and
other supporting measurements.
Accompanying the pollutant
concentration data are quality
assurance/quality control data and air
monitoring network design information.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
The EPA and others (e.g., State and local
air quality management agencies, tribal
entities, environmental organizations,
academic institutions, industrial groups)
use the ambient air quality data for
many purposes including informing the
public and other interested parties of an
area’s air quality, judging an area’s air
quality in comparison with the
established health or welfare standards,
evaluating an air quality management
agency’s progress in achieving or
maintaining air pollutant levels below
the national and local standards,
developing and revising State
Implementation Plans (SIPs), evaluating
air pollutant control strategies,
developing or revising national control
policies, providing data for air quality
model development and validation,
supporting enforcement actions,
documenting episodes and initiating
episode controls, assessing air quality
PO 00000
Frm 00173
Fmt 4701
Sfmt 4700
trends, and conducting air pollution
research.
C. Regulatory Flexibility Act (RFA)
I certify that this action will not have
a significant economic impact on a
substantial number of small entities
under the RFA. This action will not
impose any requirements on small
entities. Rather, this final rule
establishes national standards for
allowable concentrations of PM in
ambient air as required by section 109
of the CAA. See also American Trucking
Associations v. EPA, 175 F.3d 1027,
1044–45 (D.C. Cir. 1999) (NAAQS do
not have significant impacts upon small
entities because NAAQS themselves
impose no regulations upon small
entities), rev’d in part on other grounds,
Whitman v. American Trucking
Associations, 531 U.S. 457 (2001).
E:\FR\FM\06MRR3.SGM
06MRR3
16374
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
D. Unfunded Mandates Reform Act
(UMRA)
This action does not contain an
unfunded mandate of $100 million or
more as described in the Unfunded
Mandates Reform Act (UMRA), 2 U.S.C.
1531–1538, and does not significantly or
uniquely affect small governments.
Furthermore, as indicated previously, in
setting a NAAQS the EPA cannot
consider the economic or technological
feasibility of attaining ambient air
quality standards, although such factors
may be considered to a degree in the
development of State plans to
implement the standards. See also
American Trucking Associations v.
EPA, 175 F. 3d at 1043 (noting that
because the EPA is precluded from
considering costs of implementation in
establishing NAAQS, preparation of the
RIA pursuant to the Unfunded
Mandates Reform Act would not furnish
any information that the court could
consider in reviewing the NAAQS).
E. Executive Order 13132: Federalism
This action will not have substantial
direct effects on the States, on the
relationship between the National
Government and the States, or on the
distribution of power and
responsibilities among the various
levels of government. However, the EPA
recognizes that States will have a
substantial interest in this action and
any future revisions to associated
requirements.
ddrumheller on DSK120RN23PROD with RULES3
F. Executive Order 13175: Consultation
and Coordination With Indian Tribal
Governments
This action does not have Tribal
implications, as specified in Executive
Order 13175. It does not have a
substantial direct effect on one or more
Indian Tribes as Tribes are not obligated
to adopt or implement any NAAQS. In
addition, Tribes are not obligated to
conduct ambient monitoring for PM or
to adopt the ambient monitoring
requirements of 40 CFR part 58. Thus,
Executive Order 13175 does not apply
to this action. However, consistent with
the EPA Policy on Consultation and
Coordination with Indian Tribes, the
EPA offered consultation to all 574
Federally Recognized Tribes during the
development of this action. Although no
Tribes requested consultation, the EPA
provided informational meetings
including an informational meeting
with the Pueblo de San Ildefonso and
provided information on the monthly
National Tribal Air Association calls.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
G. Executive Order 13045: Protection of
Children From Environmental Health
Risks and Safety Risks
Executive Order 13045 directs federal
agencies to include an evaluation of the
health and safety effects of the planned
regulation on children in federal health
and safety standards and explain why
the regulation is preferable to
potentially effective and reasonably
feasible alternatives. This action is
subject to Executive Order 13045
because it is a significant regulatory
action under section 3(f)(1) of Executive
Order 12866, and the EPA believes that
the environmental health or safety risk
addressed by this action may have a
disproportionate effect on children.
Accordingly, we have evaluated the
environmental health or safety effects of
PM exposures on children. The
protection offered by these standards
may be especially important for
children because childhood represents a
lifestage associated with increased
susceptibility to PM-related health
effects. Because children have been
identified as a susceptible population,
we have carefully evaluated the
environmental health effects of
exposure to PM pollution among
children. Children make up a
substantial fraction of the U.S.
population, and often have unique
factors that contribute to their increased
risk of experiencing a health effect due
to exposures to ambient air pollutants
because of their continuous growth and
development. As described in the 2019
Integrated Science Assessment, children
may be particularly at risk for health
effects related to ambient air PM2.5
exposures compared with adults
because they have (1) a developing
respiratory system, (2) increased
ventilation rates relative to body mass
compared with adults, and (3) an
increased proportion of oral breathing,
particularly in boys, relative to adults.
More detailed information on the
evaluation of the scientific evidence and
policy considerations pertaining to
children, including an explanation for
why the Administrator judges the
revised standards to be requisite to
protect public health, including the
health of children, with an adequate
margin of safety, are contained in
section II.A.2. ‘‘Overview of the Health
Effects Evidence’’, section II.A.2.b
‘‘Public Health Implications and At-Risk
Populations’’ and II.B ‘‘Conclusions on
the Primary PM2.5 Standards’’ of this
preamble. Copies of all documents have
been placed in the public docket for this
action. The Administrator judges that
revising the primary annual PM2.5
standard to a level of 9.0 mg/m3 and
PO 00000
Frm 00174
Fmt 4701
Sfmt 4700
retaining the primary 24-hour PM2.5
standard provides requisite public
health protection with an adequate
margin of safety, including for children.
Furthermore, the Policy on Children’s
Health also applies to this action.
Information on how the Policy was
applied is described in section II.A.2
‘‘Overview of the Health Effects
Evidence’’, section II.A.2.b ‘‘Public
Health Implications and At-Risk
Populations’’ and II.B ‘‘Conclusions on
the Primary PM2.5 Standards’’ of this
preamble.
H. Executive Order 13211: Actions
Concerning Regulations That
Significantly Affect Energy Supply,
Distribution or Use
This action is not a ‘‘significant
energy action’’ because it is not likely to
have a significant adverse effect on the
supply, distribution, or use of energy.
The purpose of this action is to revise
level of the primary annual PM2.5
NAAQS. The action does not prescribe
specific pollution control strategies by
which these ambient standards and
monitoring revisions will be met. Such
strategies will be developed by States on
a case-by-case basis, and the EPA cannot
predict whether the control options
selected by States will include
regulations on energy suppliers,
distributors, or users. Thus, the EPA
concludes that this action does not
constitute a significant energy action as
defined in Executive Order 13211.
I. National Technology Transfer and
Advancement Act (NTTAA)
This rulemaking involved
environmental monitoring or
measurement. The EPA has decided it
will continue to use the existing
indicators for fine (PM2.5) and coarse
(PM10) particles. The indicator for fine
particles is measured using the
Reference Method for the Determination
of Fine Particulate Matter as PM2.5 in the
Atmosphere (appendix L to 40 CFR part
50), which is known as the PM2.5 FRM,
and the indicator for coarse particles is
measured using the Reference Method
for the Determination of Particulate
Matter as PM10 in the Atmosphere
(appendix J to 40 CFR part 50), which
is known as the PM10 FRM.
To the extent feasible, the EPA
employs a Performance-Based
Measurement System (PBMS), which
does not require the use of specific,
prescribed analytic methods. The PBMS
is defined as a set of processes wherein
the data quality needs, mandates or
limitations of a program or project are
specified and serve as criteria for
selecting appropriate methods to meet
those needs in a cost-effective manner.
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
ddrumheller on DSK120RN23PROD with RULES3
It is intended to be more flexible and
cost effective for the regulated
community; it is also intended to
encourage innovation in analytical
technology and improved data quality.
Though the FRM defines the particular
specifications for ambient monitors,
there is some variability with regard to
how monitors measure PM, depending
on the type and size of PM and
environmental conditions. Therefore, it
is not practically possible to fully define
the FRM in performance terms to
account for this variability.
Nevertheless, our approach in the past
has resulted in multiple brands of
monitors being approved as FRM for
PM, and we expect this to continue.
Also, the FRMs described in 40 CFR
part 50 and the equivalency criteria
described in 40 CFR part 53, constitute
a performance-based measurement
system for PM, since methods that meet
the field testing and performance
criteria can be approved as FEMs. Since
finalized in 2006 (71 FR 61236, October
17, 2006) the new field and performance
criteria for approval of PM2.5 continuous
FEMs has resulted in the approval of 13
approved FEMs. In summary, for
measurement of PM2.5 and PM10, the
EPA relies on both FRMs and FEMs,
with FEMs relying on a PBMS approach
for their approval. The EPA is not
precluding the use of any other method,
whether it constitutes a voluntary
consensus standard or not, as long as it
meets the specified performance
criteria.
J. Executive Order 12898: Federal
Actions To Address Environmental
Justice in Minority Populations and
Low-Income Populations and Executive
Order 14096: Revitalizing Our Nation’s
Commitment to Environmental Justice
for All
The EPA believes that the human
health or environmental conditions
associated with the primary PM2.5
NAAQS that exist prior to this action
result in or have the potential to result
in disproportionate and adverse human
health or environmental effects on
communities with environmental justice
concerns. There is strong evidence for
racial and ethnic disparities in PM2.5
exposures and PM2.5-related health risk,
as assessed in the 2019 Integrated
Science Assessment and with even more
evidence available since the literature
cutoff date for the 2019 Integrated
Science Assessment and evaluated in
the Supplement to the 2019 Integrated
Science Assessment. There is strong
evidence demonstrating that Black and
Hispanic populations, in particular,
have higher PM2.5 exposures than nonHispanic White populations. Black
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
populations or individuals that live in
predominantly Black neighborhoods
experience higher PM2.5 exposures, in
comparison to non-Hispanic White
populations. There is also consistent
evidence across multiple studies that
demonstrate increased risk of PM2.5related health effects, with the strongest
evidence for health risk disparities for
mortality. There is also evidence of
health risk disparities for both Hispanic
and non-Hispanic Black populations
compared to non-Hispanic White
populations for cause-specific mortality
and incident hypertension.
Socioeconomic status (SES) is a
composite measure that includes
metrics such as income, occupation, or
education, and can play a role in access
to healthy environments as well as
access to healthcare. SES may be a
factor that contributes to differential risk
from PM2.5-related health effects.
Studies assessed in the 2019 Integrated
Science Assessment and Supplement to
the 2019 Integrated Science Assessment
provide evidence that lower SES
communities are exposed to higher
concentrations of PM2.5 compared to
higher SES communities. Studies using
composite measures of neighborhood
SES consistently demonstrated a
disparity in both PM2.5 exposure and the
risk of PM2.5-related health outcomes.
There is some evidence that supports
associations larger in magnitude
between mortality and long-term PM2.5
exposures for those with low income or
living in lower income areas compared
to those with higher income or living in
higher income neighborhoods.
Additionally, evidence supports
conclusions that lower SES is associated
with cause-specific mortality and
certain health endpoints (i.e., HI and
CHF), but less so for all-cause or total
(non-accidental) mortality.
The EPA believes that this action is
likely to reduce existing
disproportionate and adverse effects on
communities with environmental justice
concerns.
The EPA additionally identified and
addressed environmental justice
concerns by providing opportunities for
public input on the proposed decisions.
The EPA held a multi-day virtual public
hearing for the public to provide oral
testimony and there was a 60-day public
comment period for the proposed
action. As described in section II.A.3
above, the EPA conducted a risk
assessment to support this action that
included an at-risk analysis that
evaluates exposure and PM2.5 mortality
risk for older adults (e.g., 65 years and
older), stratified for White, Black, Asian,
Native American, Non-Hispanic, and
Hispanic individuals. This at-risk
PO 00000
Frm 00175
Fmt 4701
Sfmt 4700
16375
analysis found that compared to a
primary annual PM2.5 standard with a
level of 12.0 mg/m3, meeting a revised
annual standard with a level of 9.0 mg/
m3 is estimated to reduce PM2.5associated health risks in the 30 study
areas controlled by the annual standard
by about 22–28% and is expected to
reduce disparities in exposure and risk
among these populations.
The information supporting this
Executive Order review is is contained
in sections II.A.2, II.B.3.a, II.B.3.c, II.B.2,
and II.B.4. of this preamble and also in
the 2019 Integrated Science Assessment,
Supplement to the 2019 Integrated
Science Assessment, and 2022 Policy
Assessment. The EPA has carefully
evaluated the potential impacts on
minority populations and low SES
populations as discussed in sections
II.A.2, II.A.3, II.B.2, and II.B.4 of this
preamble. The 2019 Integrated Science
Assessment, Supplement to the
Integrated Science Assessment, and
2022 Policy Assessment contain the
evaluation of the scientific evidence,
quantitative risk analyses and policy
considerations that pertain to these
populations. These documents are
available in the public docket for this
action.
K. Congressional Review Act (CRA)
This action is subject to the CRA, and
the EPA will submit a rule report to
each House of the Congress and to the
Comptroller General of the United
States. This action meets the criteria set
forth in 5 U.S.C. 804(2).
References
Abt Associates, Inc. (2001). Assessing public
opinions on visibility impairment due to
air pollution: Summary report. U.S.
Environmental Protection Agency.
Research Triangle Park, NC. Available at:
https://www3.epa.gov/ttn/naaqs/
standards/pm/data/vis_rpt_final.pdf.
Abt Associates, Inc. (2005). Particulate matter
health risk assessment for selected urban
areas: Draft report. EPA Contract No. 68–
D–03–002. U.S. Environmental
Protection Agency. Research Triangle
Park, NC. Available at: https://
www3.epa.gov/ttn/naaqs/standards/pm/
data/PMrisk20051220.pdf.
ATS (2000). What Constitutes an Adverse
Health Effect of Air Pollution? American
Journal of Respiratory and Critical Care
Medicine 161(2): 665–673.
BBC Research & Consulting (2003). Phoenix
area visibility survey. Denver, CO.
Available at: https://
www.regulations.gov/document/EPAHQ-OAR-2015-0072-0089.
Behbod, B, Urch, B, Speck, M, Scott, JA, Liu,
L, Poon, R, Coull, B, Schwartz, J,
Koutrakis, P, Silverman, F and Gold, DR
(2013). Endotoxin in concentrated coarse
and fine ambient particles induces acute
systemic inflammation in controlled
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16376
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
human exposures. Occupational and
Environmental Medicine 70(11): 761–
767.
Bell, ML, Ebisu, K, Peng, RD, Walker, J,
Samet, JM, Zeger, SL and Dominic, F
(2008). Seasonal and regional short-term
effects of fine particles on hospital
admissions in 202 U.S. counties, 1999–
2005. American Journal of Epidemiology
168(11): 1301–1310.
Bellavia, A, Urch, B, Speck, M, Brook, RD,
Scott, JA, Albetti, B, Behbod, B, North,
M, Valeri, L, Bertazzi, PA, Silverman, F,
Gold, D and Baccarelli, AA (2013). DNA
hypomethylation, ambient particulate
matter, and increased blood pressure:
Findings from controlled human
exposure experiments. Journal of the
American Heart Association 2(3):
e000212.
Bennett, JE, Tamura-Wicks, H, Parks, RM,
Burnett, RT, Pope, CA, Bechle, MJ,
Marshall, JD, Danaei, G and Ezzati, M
(2019). Particulate matter air pollution
and national and county life expectancy
loss in the USA: A spatiotemporal
analysis. PLoS Medicine 16(7):
e1002856.
Besson, P, Mun˜oz, C, Ramı´rez-Sagner, G,
Salgado, M, Escobar, R and Platzer, W
(2017). Long-Term Soiling Analysis for
Three Photovoltaic Technologies in
Santiago Region. IEEE Journal of
Photovoltaics 7(6): 1755–1760.
Bourdrel, T, Annesi-Maesano, I, Alahmad, B,
Maesano, CN and Bind, MA (2021). The
impact of outdoor air pollution on
COVID–19: a review of evidence from in
vitro, animal, and human studies.
30(159).
Bra¨uner, EV, M2014
19:03 Mar 05, 2024
Jkt 262001
United States. Atmospheric Environment
175: 25–32.
Chen, H, Burnett, RT, Copes, R, Kwong, JC,
Villeneuve, PJ, Goldberg, MS, Brook, RD,
van Donkelaar, A, Jerrett, M, Martin, RV,
Brook, JR, Kopp, A and Tu, JV (2016).
Ambient fine particulate matter and
mortality among survivors of myocardial
infarction: population-based cohort
study. Environmental Health
Perspectives 124(9): 1421–1428.
Correia, AW, Pope, CA, III, Dockery, DW,
Wang, Y, un, Ezzati, M and Dominici, F
(2013). Effect of air pollution control on
life expectancy in the United States: an
analysis of 545 U.S. counties for the
period from 2000 to 2007. Epidemiology
24(1): 23–31.
Corrigan, AE, Becker, MM, Neas, LM, Cascio,
WE and Rappold, AG (2018). Fine
particulate matters: The impact of air
quality standards on cardiovascular
mortality. Environmental research 161:
364–369.
Cox, LA. (2019a). Letter from Louis Anthony
Cox, Jr., Chair, Clean Air Scientific
Advisory Committee, to Administrator
Andrew R. Wheeler. Re: CASAC Review
of the EPA’s Integrated Science
Assessment for Particulate Matter
(External Review Draft—October 2018).
April 11, 2019. EPA–CASAC–19–002.
Office of the Administrator, Science
Advisory Board U.S. EPA HQ,
Washington DC. Available at: https://
casac.epa.gov/ords/sab/r/sab_apex/
casac/0?report_
id=1069&request=APPLICATION_
PROCESS%3DREPORT_
DOC&session=7184955370570.
Cox, LA. (2019b). Letter from Louis Anthony
Cox, Jr., Chair, Clean Air Scientific
Advisory Committee, to Administrator
Andrew R. Wheeler. Re: CASAC Review
of the EPA’s Policy Assessment for the
Review of the National Ambient Air
Quality Standards for Particulate Matter
(External Review Draft—September
2019). December 16, 2019. EPA–CASAC–
20–001. Office of the Administrator,
Science Advisory Board U.S. EPA HQ,
Washington DC. Available at: https://
casac.epa.gov/ords/sab/r/sab_apex/
casac/0?report_
id=1073&request=APPLICATION_
PROCESS%3DREPORT_
DOC&session=6224161457429.
DeFlorio-Barker, S, Crooks, J, Reyes, J and
Rappold, AG (2019). Cardiopulmonary
effects of fine particulate matter
exposure among older adults, during
wildfire and non-wildfire periods, in the
United States 2008–2010. Environmental
health perspectives 127(3): 037006.
DHEW (1969). Air Quality Criteria for
Particulate Matter. National Air
Pollution Control Administration.
Washington, DC U.S. Department of
Health. January 1969.
Di, Q, Amini, H, Shi, L, Kloog, I, Silvern, R,
Kelly, J, Sabath, MB, Choirat, C,
Koutrakis, P and Lyapustin, A (2019). An
ensemble-based model of PM2.5
concentration across the contiguous
United States with high spatiotemporal
resolution. Environment International
130: 104909.
PO 00000
Frm 00176
Fmt 4701
Sfmt 4700
Di, Q, Dai, L, Wang, Y, Zanobetti, A, Choirat,
C, Schwartz, JD and Dominici, F (2017a).
Association of short-term exposure to air
pollution with mortality in older adults.
JAMA: Journal of the American Medical
Association 318(24): 2446–2456.
Di, Q, Kloog, I, Koutrakis, P, Lyapustin, A,
Wang, Y and Schwartz, J (2016).
Assessing PM2.5 exposures with high
spatiotemporal resolution across the
Continental United States.
Environmental Science and Technology
50(9): 4712–4721.
Di, Q, Wang, Y, Zanobetti, A, Wang, Y,
Koutrakis, P, Choirat, C, Dominici, F and
Schwartz, JD (2017b). Air pollution and
mortality in the Medicare population.
New England Journal of Medicine
376(26): 2513–2522.
Dominici, F, Schwartz, J, Di, Q, Braun, D,
Choirat, C and Zanobetti, A (2019).
Assessing adverse health effects of longterm exposure to low levels of ambient
air pollution: Phase 1. Health Effects
Institute. Boston, MA. Available at:
https://www.healtheffects.org/system/
files/dominici-rr-200-report.pdf.
Ely, DW, Leary, JT, Stewart, TR and Ross, DM
(1991). The establishment of the Denver
Visibility Standard. Colorado
Department of Health. Denver, Colorado.
Erickson, AC, Christidis, T, Pappin, A, Brook,
JR, Crouse, DL, Hystad, P, Li, C, Martin,
RV, Meng, J, Pinault, L, von Donkelaar,
A, Weichenthal, S, Tjepkema, M,
Burnett, RT and Brauer, M (2020).
Disease assimilation: The mortality
impacts of fine particulate matter on
immigrants to Canada. Health Reports
31(3): 14–26.
Eum, K, Suh, HH, Pun, V and Manjourides,
J (2018). Impact of long-term temporal
trends in fine particulate matter (PM2.5)
on association of annual PM2.5 exposure
and mortality: an analysis of over 20
million Medicare beneficiaries.
Environmental Epidemiology 2(2): e009.
Fiore, AM, Naik, V and Leibensperger, EM
(2015). Air quality and climate
connections. Journal of the Air and
Waste Management Association 65(6):
645–685.
Frank, N. (2012). Memorandum to PM
NAAQS Review Docket (EPA–HQ–OAR–
2007–0492) regarding the Differences
between maximum and composite
monitor annual PM2.5 design values by
CBSA. Dec 14, 2012. Docket ID No. EPA–
HQ–OAR–2007–0492. Office of Air
Quality Planning and Standards
Research Triangle Park, NC. Available at:
https://www.regulations.gov/document/
EPA-HQ-OAR-2007-0492-10099.
Franklin, M, Zeka, A and Schwartz, J (2007).
Association between PM2.5 and all-cause
and specific-cause mortality in 27 US
communities. Journal of Exposure
Science and Environmental
Epidemiology 17(3): 279–287.
Gantt, B, Owen, RC and Watkins, N (2021).
Characterizing Nitrogen Oxides and Fine
Particulate Matter near Major Highways
in the United States Using the National
Near-Road Monitoring Network.
Environmental science & technology
55(5): 2831–2838.
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
Ghio, AJ, Hall, A, Bassett, MA, Cascio, WE
and Devlin, RB (2003). Exposure to
concentrated ambient air particles alters
hematologic indices in humans.
Inhalation Toxicology 15(14): 1465–
1478.
Ghio, AJ, Kim, C and Devlin, RB (2000).
Concentrated ambient air particles
induce mild pulmonary inflammation in
healthy human volunteers. American
Journal of Respiratory and Critical Care
Medicine 162(3): 981–988.
Greven, S, Dominici, F and Zeger, S (2011).
An Approach to the Estimation of
Chronic Air Pollution Effects Using
Spatio-Temporal Information. Journal of
the American Statistical Association
106(494): 396–406.
Gr2014
19:03 Mar 05, 2024
Jkt 262001
States associated with reduced coal
emissions between 2005 and 2012.
Epidemiology 30(4): 477–485.
Hutchinson, JA, Vargo, J, Milet, M, French,
NH, Billmire, M, Johnson, J and Hoshiko,
S (2018). The San Diego 2007 wildfires
and Medi-Cal emergency department
presentations, inpatient hospitalizations,
and outpatient visits: An observational
study of smoke exposure periods and a
bidirectional case-crossover analysis.
PLoS medicine 15(7): e1002601.
IPCC (2013). Climate change 2013: The
physical science basis. Contribution of
working group I to the fifth assessment
report of the Intergovernmental Panel on
Climate Change. T. F. Stocker, D. Qin, G.
K. Plattner et al. Cambridge University
Press. Cambridge, UK.
Kloog, I, Ridgway, B, Koutrakis, P, Coull, BA
and Schwartz, JD (2013). Long- and
short-term exposure to PM2.5 and
mortality: Using novel exposure models.
Epidemiology 24(4): 555–561.
Krewski, D, Jerrett, M, Burnett, RT, Ma, R,
Hughes, E, Shi, Y, Turner, MC, Pope, CA,
III, Thurston, G, Calle, EE, Thun, MJ,
Beckerman, B, Deluca, P, Finkelstein, N,
Ito, K, Moore, DK, Newbold, KB,
Ramsay, T, Ross, Z, Shin, H and
Tempalski, B (2009). Extended follow-up
and spatial analysis of the American
Cancer Society study linking particulate
air pollution and mortality. ISSN 1041–
5505, HEI Research Report 140. Health
Effects Institute. Boston, MA. Available
at: https://www.healtheffects.org/system/
files/Krewski140Statement.pdf.
Laden, F, Schwartz, J, Speizer, FE and
Dockery, DW (2006). Reduction in fine
particulate air pollution and mortality:
extended follow-up of the Harvard Six
Cities study. American Journal of
Respiratory and Critical Care Medicine
173(6): 667–672.
Lavigne, E, Burnett, RT and Weichenthal, S
(2018). Association of short-term
exposure to fine particulate air pollution
and mortality: effect modification by
oxidant gases. Scientific Reports 8(1):
16097.
Lee, M, Koutrakis, P, Coull, B, Kloog, I and
Schwartz, J (2015). Acute effect of fine
particulate matter on mortality in three
Southeastern states from 2007–2011.
Journal of Exposure Science and
Environmental Epidemiology 26(2): 173–
179.
Lepeule, J, Laden, F, Dockery, D and
Schwartz, J (2012). Chronic exposure to
fine particles and mortality: an extended
follow-up of the Harvard Six Cities study
from 1974 to 2009. Environmental
Health Perspectives 120(7): 965–970.
Lippmann, M, Chen, LC, Gordon, T, Ito, K
and Thurston, GD (2013). National
Particle Component Toxicity (NPACT)
Initiative: Integrated epidemiologic and
toxicologic studies of the health effects
of particulate matter components:
Investigators’ Report. 177. Health Effects
Institute. Boston, MA.
Liu, C, Chen, R, Sera, F, Vicedo-Cabrera, AM,
Guo, Y, Tong, S, Coelho, M, Saldiva,
PHN, Lavigne, E, Matus, P, Valdes
Ortega, N, Osorio Garcia, S, Pascal, M,
PO 00000
Frm 00177
Fmt 4701
Sfmt 4700
16377
Stafoggia, M, Scortichini, M, Hashizume,
M, Honda, Y, Hurtado-Dı´az, M, Cruz, J,
Nunes, B, Teixeira, JP, Kim, H, Tobias,
˚ stro¨m, C,
A, ´In˜iguez, C, Forsberg, B, A
Ragettli, MS, Guo, YL, Chen, BY, Bell,
ML, Wright, CY, Scovronick, N, Garland,
RM, Milojevic, A, Kysely´, J, Urban, A,
Orru, H, Indermitte, E, Jaakkola, JJK,
Ryti, NRI, Katsouyanni, K, Analitis, A,
Zanobetti, A, Schwartz, J, Chen, J, Wu,
T, Cohen, A, Gasparrini, A and Kan, H
(2019). Ambient Particulate Air Pollution
and Daily Mortality in 652 Cities. New
England Journal of Medicine 381(8):
705–715.
Lowenthal, DH and Kumar, N (2004).
Variation of mass scattering efficiencies
in IMPROVE. Journal of the Air and
Waste Management Association (1990–
1992) 54(8): 926–934.
Lowenthal, DH and Kumar, N (2016).
Evaluation of the IMPROVE Equation for
estimating aerosol light extinction.
Journal of the Air and Waste
Management Association 66(7): 726–737.
Lucking, AJ, Lundba¨ck, M, Barath, SL, Mills,
NL, Sidhu, MK, Langrish, JP, Boon, NA,
Pourazar, J, Badimon, JJ, Gerlofs-Nijland,
ME, Cassee, FR, Boman, C, Donaldson,
K, Sandstrom, T, Newby, DE and
Blomberg, A (2011). Particle traps
prevent adverse vascular and
prothrombotic effects of diesel engine
exhaust inhalation in men. Circulation
123(16): 1721–1728.
Malm, WC and Hand, JL (2007). An
examination of the physical and optical
properties of aerosols collected in the
IMPROVE program. Atmospheric
Environment 41(16): 3407–3427.
Malm, WC, Schichtel, B, Molenar, J, Prenni,
A and Peters, M (2019). Which visibility
indicators best represent a population’s
preference for a level of visual air
quality? Journal of the Air & Waste
Management Association 69(2): 145–161.
Malm, WC, Sisler, JF, Huffman, D, Eldred,
RA and Cahill, TA (1994). Spatial and
seasonal trends in particle concentration
and optical extinction in the United
States. Journal of Geophysical Research
99(D1): 1347–1370.
Mauad, T, Rivero, DH, de Oliveira, RC,
Lichtenfels, AJ, Guimaraes, ET, de
Andre, PA, Kasahara, DI, Bueno, HM and
Saldiva, PH (2008). Chronic exposure to
ambient levels of urban particles affects
mouse lung development. American
Journal of Respiratory and Critical Care
Medicine 178(7): 721–728.
Mie, G (1908). Beitrage zur Optik truber
Medien, speziell kolloidaler
Metallosungen [Optics of cloudy media,
especially colloidal metal solutions].
Annalen der Physik 25(3): 377–445.
Miller, KA, Siscovick, DS, Sheppard, L,
Shepherd, K, Sullivan, JH, Anderson, GL
and Kaufman, JD (2007). Long-term
exposure to air pollution and incidence
of cardiovascular events in women. New
England Journal of Medicine 356(5):
447–458.
Myhre, G, Shindell, D, Bre´on, FM, Collins,
W, Fuglestvedt, J, Huang, J, Koch, D,
Lamarque, JF, Lee, D, Mendoza, B,
Nakajima, T, Robock, A, Stephens, G,
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16378
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
Takemura, T and Zhang, H, Eds. (2013).
Anthropogenic and natural radiative
forcing. Cambridge University Press
Cambridge, UK.
NHLBI (2017). ‘‘NHLBI fact book, fiscal year
2012: Disease statistics.’’ Retrieved
August 23, 2017, from https://
web.archive.org/web/20170711012213/
https://www.nhlbi.nih.gov/about/
documents/factbook/2012/chapter4.
Orr, A, AL Migliaccio, C, Buford, M, Ballou,
S and Migliaccio, CT (2020). Sustained
effects on lung function in community
members following exposure to
hazardous pm2. 5 levels from wildfire
smoke. Toxics 8(3): 53.
Pitchford, M, Maim, W, Schichtel, B, Kumar,
N, Lowenthal, D and Hand, J (2007).
Revised algorithm for estimating light
extinction from IMPROVE particle
speciation data. Journal of the Air and
Waste Management Association 57(11):
1326–1336.
Pope, CA, III, Burnett, RT, Thurston, GD,
Thun, MJ, Calle, EE, Krewski, D and
Godleski, JJ (2004). Cardiovascular
mortality and long-term exposure to
particulate air pollution: epidemiological
evidence of general pathophysiological
pathways of disease. Circulation 109(1):
71–77.
Pope, CA, III, Ezzati, M and Dockery, DW
(2009). Fine-particulate air pollution and
life expectancy in the United States. New
England Journal of Medicine 360(4):
376–386.
Pruitt, E. (2018). Memorandum from E. Scott
Pruitt, Administrator, U.S. EPA to
Assistant Administrators. Back-to-Basics
Process for Reviewing National Ambient
Air Quality Standards. May 9, 2018.
Office of the Administrator U.S. EPA
HQ, Washington DC. Available at:
https://www.epa.gov/criteria-airpollutants/back-basics-processreviewing-national-ambient-air-qualitystandards.
Pryor, SC (1996). Assessing public perception
of visibility for standard setting
exercises. Atmospheric Environment
30(15): 2705–2716.
Puett, RC, Hart, JE, Yanosky, JD, Spiegelman,
D, Wang, M, Fisher, JA, Hong, B and
Laden, F (2014). Particulate matter air
pollution exposure, distance to road, and
incident lung cancer in the Nurses’
Health Study cohort. Environmental
Health Perspectives 122(9): 926–932.
Pun, VC, Kazemiparkouhi, F, Manjourides, J
and Suh, HH (2017). Long-term PM2.5
exposures and respiratory, cancer and
cardiovascular mortality in American
older adults. American Journal of
Epidemiology 186(8): 961–969.
Raaschou-Nielsen, O, Andersen, ZJ, Beelen,
R, Samoli, E, Stafoggia, M, Weinmayr, G,
Hoffmann, B, Fischer, P,
Nieuwenhuijsen, MJ, Brunekreef, B, Xun,
WW, Katsouyanni, K, Dimakopoulou, K,
Sommar, J, Forsberg, B, Modig, L, Oudin,
A, Oftedal, B, Schwarze, PE, Nafstad, P,
¨ stenson, CG,
De Faire, U, Pedersen, NL, O
Fratiglioni, L, Penell, J, Korek, M,
Pershagen, G, Eriksen, KT, S2014
19:03 Mar 05, 2024
Jkt 262001
Bueno-De-mesquita, B, Key, TJ, De
Hoogh, K, Concin, H, Nagel, G, Vilier, A,
Grioni, S, Krogh, V, Tsai, MY, Ricceri, F,
Sacerdote, C, Galassi, C, Migliore, E,
Ranzi, A, Cesaroni, G, Badaloni, C,
Forastiere, F, Tamayo, I, Amiano, P,
Dorronsoro, M, Trichopoulou, A, Bamia,
C, Vineis, P and Hoek, G (2013). Air
pollution and lung cancer incidence in
17 European cohorts: Prospective
analyses from the European Study of
Cohorts for Air Pollution Effects
(ESCAPE). The Lancet Oncology 14(9):
813–822.
Ramanathan, G, Yin, F, Speck, M, Tseng, CH,
Brook, JR, Silverman, F, Urch, B, Brook,
RD and Araujo, JA (2016). Effects of
urban fine particulate matter and ozone
on HDL functionality. Particle and Fibre
Toxicology 13(1): 26.
Ryan, PA, Lowenthal, D and Kumar, N
(2005). Improved light extinction
reconstruction in interagency monitoring
of protected visual environments.
Journal of the Air and Waste
Management Association 55(11): 1751–
1759.
Sacks, JD, Ito, K, Wilson, WE and Neas, LM
(2012). Impact of covariate models on the
assessment of the air pollution-mortality
association in a single- and
multipollutant context. American
Journal of Epidemiology 176(7): 622–
634.
Sanders, NJ, Barreca, AI and Neidell, MJ
(2020a). Estimating causal effects of
particulate matter regulation on
mortality. Epidemiology (Cambridge,
Mass.) 31(2): 160.
Sanders, NJ, Barreca, AI and Neidell, MJ
(2020b). Estimating Causal Effects of
Particulate Matter Regulation on
Mortality. Epidemiology 31(2): 160–167.
Schwartz, J, Austin, E, Bind, MA, Zanobetti,
A and Koutrakis, P (2015). Estimating
causal associations of fine particles with
daily deaths in Boston. American Journal
of Epidemiology 182(7): 644–650.
Schwartz, J, Bind, MA and Koutrakis, P
(2017). Estimating causal effects of local
air pollution on daily deaths: Effect of
low levels. Environmental Health
Perspectives 125(1): 23–29.
Schwartz, J, Wei, Y, Di, Q, Dominici, F and
Zanobetti, A (2021). A national
difference in differences analysis of the
effect of PM2. 5 on annual death rates.
Environmental Research 194: 110649.
Schwartz, JD, Wang, Y, Kloog, I, YitshakSade, M, Dominici, F and Zanobetti, A
(2018). Estimating the Effects of PM2.5 on
Life Expectancy Using Causal Modeling
Methods. Environmental Health
Perspectives 126(12): 127002.
Sheppard, EA. (2022a). Letter from Elizabeth
A. (Lianne) Sheppard, Chair, Clean Air
Scientific Advisory Committee, to
Administrator Michael S. Regan. Re:
CASAC Review of the EPA’s Policy
Assessment for the Review of the
National Ambient Air Quality Standards
for Particulate Matter (External Review
Draft—October 2021). March 18, 2022.
EPA–CASAC–22–002. Office of the
Administrator, Science Advisory Board
U.S. EPA HQ, Washington DC. Available
PO 00000
Frm 00178
Fmt 4701
Sfmt 4700
at: https://casac.epa.gov/ords/sab/r/sab_
apex/casac/0?report_
id=1094&request=APPLICATION_
PROCESS%3DREPORT_
DOC&session=7184955370570.
Sheppard, EA. (2022b). Letter from Elizabeth
A. (Lianne) Sheppard, Chair, Clean Air
Scientific Advisory Committee, to
Administrator Michael S. Regan. Re:
CASAC Review of the EPA’s Supplement
to the 2019 Integrated Science
Assessment for Particulate Matter
(External Review Draft—October 2021).
EPA–CASAC–22–001. Office of the
Administrator, Science Advisory Board
U.S. EPA HQ, Washington DC. Available
at: https://casac.epa.gov/ords/sab/r/sab_
apex/casac/0?report_
id=1093&request=APPLICATION_
PROCESS%3DREPORT_
DOC&session=10813926997922.
Shi, L, Zanobetti, A, Kloog, I, Coull, BA,
Koutrakis, P, Melly, SJ and Schwartz, JD
(2016). Low-concentration PM2.5 and
mortality: estimating acute and chronic
effects in a population-based study.
Environmental Health Perspectives
124(1): 46–52.
Shin, HH, Gogna, P, Maquiling, A, Parajuli,
RP, Haque, L and Burr, B (2021).
Comparison of hospitalization and
mortality associated with short-term
exposure to ambient ozone and PM2.5 in
Canada. Chemosphere 265: 128683.
Sivagangabalan, G, Spears, D, Masse, S, Urch,
B, Brook, RD, Silverman, F, Gold, DR,
Lukic, KZ, Speck, M, Kusha, M, Farid, T,
Poku, K, Shi, E, Floras, J and
Nanthakumar, K (2011). The effect of air
pollution on spatial dispersion of
myocardial repolarization in healthy
human volunteers. Journal of the
American College of Cardiology 57(2):
198–206.
Smith, AE and Howell, S (2009). An
assessment of the robustness of visual air
quality preference study results. CRA
International. Washington, DC. Available
at: https://www.regulations.gov/
document/EPA-HQ-ORD-2007-05170085.
Statistics Canada (2023). Census Profile. 2021
Census of Population. Statistics Canada
Catalogue no. 98–316–X2021001.
Ottawa. Released March 29, 2023.
Available at: https://www12.statcan.
gc.ca/census-recensement/2021/dp-pd/
prof/index.cfm?Lang=E.
Thurston, GD, Kipen, H, Annesi-Maesano, I,
Balmes, J, Brook, RD, Cromar, K, De
Matteis, S, Forastiere, F, Forsberg, B,
Frampton, MW, Grigg, J, Heederik, D,
Kelly, FJ, Kuenzli, N, Laumbach, R,
Peters, A, Rajagopalan, ST, Rich, D, Ritz,
B, Samet, JM, Sandstrom, T, Sigsgaard,
T, Sunyer, J and Brunekreef, B (2017). A
joint ERS/ATS policy statement: what
constitutes an adverse health effect of air
pollution? An analytical framework.
European Respiratory Journal 49(1):
1600419.
Tong, H, Rappold, AG, Caughey, M,
Hinderliter, AL, Bassett, M, Montilla, T,
Case, MW, Berntsen, J, Bromberg, PA,
Cascio, WE, Diaz-Sanchez, D, Devlin, RB
and Samet, JM (2015). Dietary
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
supplementation with olive oil or fish oil
and vascular effects of concentrated
ambient particulate matter exposure in
human volunteers. Environmental
Health Perspectives 123(11): 1173–1179.
U.S. EPA (2004a). Air Quality Criteria for
Particulate Matter. (Vol I of II). Office of
Research and Development. Research
Triangle Park, NC. U.S. EPA. EPA–600/
P–99–002aF. October 2004. Available at:
https://nepis.epa.gov/Exe/ZyPURL.
cgi?Dockey=P100LFIQ.txt.
U.S. EPA (2004b). Air Quality Criteria for
Particulate Matter. (Vol II of II). Office of
Research and Development. Research
Triangle Park, NC. U.S. EPA. EPA–600/
P–99–002bF. October 2004. Available at:
https://nepis.epa.gov/Exe/ZyPURL.
cgi?Dockey=P100LG7Q.txt.
U.S. EPA (2005). Review of the National
Ambient Air Quality Standards for
Particulate Matter: Policy Assessment of
Scientific and Technical Information,
OAQPS Staff Paper. Office of Air Quality
Planning and Standards. Research
Triangle Park, NC. U.S. EPA. EPA–452/
R–05–005a. December 2005. Available at:
https://nepis.epa.gov/Exe/ZyPURL.
cgi?Dockey=P1009MZM.txt.
U.S. EPA (2008). Integrated Review Plan for
the National Ambient Air Quality
Standards for Particulate Matter. Office
of Research and Development, National
Center for Environmental Assessment;
Office of Air Quality Planning and
Standards, Health and Environmental
Impacts Division. Research Triangle
Park, NC. U.S. EPA. EPA 452/R–08–004.
March 2008. Available at: https://nepis.
epa.gov/Exe/ZyPURL.cgi?Dockey
=P1001FB9.txt.
U.S. EPA (2009a). Particulate Matter National
Ambient Air Quality Standards: Scope
and Methods Plan for Health Risk and
Exposure Assessment. Office of Air
Quality Planning and Standards, Health
and Environmental Impacts Division.
Research Triangle Park, NC. U.S. EPA.
EPA–452/P–09–002. February 2009.
Available at: https://nepis.epa.gov/Exe/
ZyPURL.cgi?Dockey=P100FLWP.txt.
U.S. EPA (2009b). Particulate Matter National
Ambient Air Quality Standards: Scope
and Methods Plan for Urban Visibility
Impact Assessment. Office of Air Quality
Planning and Standards, Health and
Environmental Impacts Division.
Research Triangle Park, NC. U.S. EPA.
EPA–452/P–09–001. February 2009.
Available at: https://nepis.epa.gov/Exe/
ZyPURL.cgi?Dockey=P100FLUX.txt.
U.S. EPA (2009c). Integrated Science
Assessment for Particulate Matter (Final
Report). Office of Research and
Development, National Center for
Environmental Assessment. Research
Triangle Park, NC. U.S. EPA. EPA–600/
R–08–139F. December 2009. Available
at: https://cfpub.epa.gov/ncea/risk/
recordisplay.cfm?deid=216546.
U.S. EPA (2010a). Quantitative Health Risk
Assessment for Particulate Matter (Final
Report). Office of Air Quality Planning
and Standards, Health and
Environmental Impacts Division.
Research Triangle Park, NC. U.S. EPA.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
EPA–452/R–10–005. June 2010.
Available at: https://nepis.epa.gov/Exe/
ZyPURL.cgi?Dockey=P1007RFC.txt.
U.S. EPA (2010b). Particulate Matter UrbanFocused Visibility Assessment (Final
Document). Office of Air Quality
Planning and Standards, Health and
Environmental Impacts Division.
Research Triangle Park, NC. U.S. EPA.
EPA–452/R–10–004. July 2010. Available
at: https://nepis.epa.gov/Exe/
ZyPURL.cgi?Dockey=P100FO5D.txt.
U.S. EPA (2011). Policy Assessment for the
Review of the Particulate Matter National
Ambient Air Quality Standards. Office of
Air Quality Planning and Standards,
Health and Environmental Impacts
Division. Research Triangle Park, NC.
U.S. EPA. EPA–452/R–11–003. April
2011. Available at: https://nepis.epa.gov/
Exe/ZyPURL.cgi?
Dockey=P100AUMY.txt.
U.S. EPA (2012). Responses to Significant
Comments on the 2012 Proposed Rule on
the National Ambient Air Quality
Standards for Particulate Matter (June 29,
2012; 77 FR 38890). Research Triangle
Park, NC. U.S. EPA. Docket ID No. EPA–
HQ–OAR–2007–0492. Available at:
https://www3.epa.gov/ttn/naaqs/
standards/pm/data/20121214rtc.pdf.
U.S. EPA (2015). Preamble to the integrated
science assessments. U.S. Environmental
Protection Agency, Office of Research
and Development, National Center for
Environmental Assessment, RTP
Division. Research Triangle Park, NC.
U.S. EPA. EPA/600/R–15/067. November
2015. Available at: https://cfpub.epa.
gov/ncea/isa/recordisplay.cfm
?deid=310244.
U.S. EPA (2016). Integrated review plan for
the national ambient air quality
standards for particulate matter. Office of
Air Quality Planning and Standards.
Research Triangle Park, NC. U.S. EPA.
EPA–452/R–16–005. December 2016.
Available at: https://www3.epa.gov/ttn/
naaqs/standards/pm/data/201612-finalintegrated-review-plan.pdf.
U.S. EPA (2017). Emissions Inventory
Guidance for Implementation of Ozone
and Particulate Matter National Ambient
Air Quality Standards (NAAQS) and
Regional Haze Regulations. Office of Air
Quality Planning and Standards, Office
of Air and Radiation. Research Triangle
Park, NC. U.S. EPA. U.S. EPA–454/B–
17–002. Available at: https://
www.epa.gov/sites/default/files/2017-07/
documents/ei_guidance_may_2017_
final_rev.pdf.
U.S. EPA (2018a). Technical Assistance
Document (TAD) for the Reporting of
Daily Air Quality—the Air Quality Index
(AQI). U.S. Environmental Protection
Agency, Office of Air Quality Planning
and Standards. Research Triangle Park,
NC. U.S. EPA. EPA 454/B–18–007.
September 2018. Available at: https://
www.airnow.gov/sites/default/files/202005/aqi-technical-assistance-documentsept2018.pdf.
U.S. EPA (2018b). Modeling Guidance for
Demonstrating Air Quality Goals for
Ozone, PM2.5, and Regional Haze. Office
PO 00000
Frm 00179
Fmt 4701
Sfmt 4700
16379
of Air Quality Planning and Standards,
Air Quality Policy Division. Research
Triangle Park, NC. U.S. EPA. EPA 454/
R–18–009. November 2018. Available at:
https://www.epa.gov/sites/default/files/
2020-10/documents/o3-pm-rh-modeling_
guidance-2018.pdf.
U.S. EPA (2019a). Integrated Science
Assessment (ISA) for Particulate Matter
(Final Report). U.S. Environmental
Protection Agency, Office of Research
and Development, National Center for
Environmental Assessment. Washington,
DC. U.S. EPA. EPA/600/R–19/188.
December 2019. Available at: https://
www.epa.gov/naaqs/particulate-matterpm-standards-integrated-scienceassessments-current-review.
U.S. EPA (2019b). PM2.5 Precursor
Demonstration Guidance. Office of Air
Quality Planning and Standards, Air
Quality Policy Division. Research
Triangle Park, NC. U.S. EPA. EPA–454/
R–19–004. May 2019. Available at:
https://nepis.epa.gov/Exe/ZyPDF.cgi/
P100YD1Q.PDF?
Dockey=P100YD1Q.PDF.
U.S. EPA (2020a). Responses to significant
comments on the 2020 proposed rule on
the National Ambient Air Quality
Standards for particulate matter (April
30, 2020; 85 FR 24094). EPA–HQ–OAR–
2015–0072. Available at: https://
www.epa.gov/sites/production/files/
2020-12/documents/pm_naaqs_
response_to_comments_final.pdf.
U.S. EPA (2020b). Policy Assessment for the
Review of the National Ambient Air
Quality Standards for Particulate Matter.
Office of Air Quality Planning and
Standards, Health and Environmental
Impacts Division. Research Triangle
Park, NC. U.S. EPA. EPA–452/R–20–002.
January 2020. Available at: https://
www.epa.gov/system/files/documents/
2021-10/final-policy-assessment-for-thereview-of-the-pm-naaqs-01-2020.pdf.
U.S. EPA (2021a). Supplement to the 2019
Integrated Science Assessment for
Particulate Matter (External Review
Draft). U.S. Environmental Protection
Agency, Office of Research and
Development, Center for Public Health
and Environmental Assessment.
Research Triangle Park, NC. U.S. EPA.
EPA/600/R–21/198. December 2019.
Available at: https://www.epa.gov/
naaqs/particulate-matter-pm-standardsintegrated-science-assessments-currentreview.
U.S. EPA (2021b). Comparative Assessment
of the Impacts of Prescribed Fire Versus
Wildfire (CAIF): A Case Study in the
Western U.S. U.S. Environmental
Protection Agency. Washington, DC. U.S.
EPA. EPA/600/R–21/197.
U.S. EPA (2021c). Policy Assessment for the
Review of the National Ambient Air
Quality Standards for Particulate Matter
(External Review Draft). Office of Air
Quality Planning and Standards, Health
and Environmental Impacts Division.
Research Triangle Park, NC. U.S. EPA.
EPA–452/P–21–001. October 2021.
Available at: https://www.epa.gov/
system/files/documents/2022-05/
E:\FR\FM\06MRR3.SGM
06MRR3
ddrumheller on DSK120RN23PROD with RULES3
16380
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
Final%20Policy%20Assessment
%20for%20the%20Reconsideration
%20of%20the%20PM%20NAAQS_
May2022_0.pdf.
U.S. EPA (2021d). Guidance for Ozone and
Fine Particulate Matter Permit Modeling.
U.S. Environmental Protection Agency,
Office of Air Quality Planning and
Standards, Air Quality Assessment
Division. Research Triangle Park, NC.
U.S. EPA. EPA–454/P–21–001.
September 2021.
U.S. EPA (2022a). Policy Assessment for the
Reconsideration of the National Ambient
Air Quality Standards for Particulate
Matter. Office of Air Quality Planning
and Standards, Health and
Environmental Impacts Division.
Research Triangle Park, NC. U.S. EPA.
EPA–452/R–22–004. May 2022.
Available at: https://www.epa.gov/
system/files/documents/2022-05/
Final%20Policy%20Assessment%20for
%20the%20Reconsideration
%20of%20the%20PM%20NAAQS_
May2022_0.pdf.
U.S. EPA (2022b). Supplement to the 2019
Integrated Science Assessment for
Particulate Matter (Final Report). U.S.
Environmental Protection Agency, Office
of Research and Development, Center for
Public Health and Environmental
Assessment. Research Triangle Park, NC.
U.S. EPA. EPA/600/R 22/028. May 2022.
Available at: https://www.epa.gov/
naaqs/particulate-matter-pm-standardsintegrated-science-assessments-currentreview.
Urch, B, Speck, M, Corey, P, Wasserstein, D,
Manno, M, Lukic, KZ, Brook, JR, Liu, L,
Coull, B, Schwartz, J, Gold, DR and
Silverman, F (2010). Concentrated
ambient fine particles and not ozone
induce a systemic interleukin-6 response
in humans. Inhalation Toxicology 22(3):
210–218.
Van de Hulst, H (1981). Light scattering by
small particles. Dover Publications, Inc.
New York.
van Donkelaar, A, Martin, RV, Li, C and
Burnett, RT (2019). Regional estimates of
chemical composition of fine particulate
matter using a combined geosciencestatistical method with information from
satellites, models, and monitors.
Environmental Science & Technology
53(5).
Vieira, JL, Guimaraes, GV, de Andre, PA,
Cruz, FD, Nascimento Saldiva, PH and
Bocchi, EA (2016a). Respiratory filter
reduces the cardiovascular effects
associated with diesel exhaust exposure
a randomized, prospective, double-blind,
controlled study of heart failure: the
FILTER–HF trial. JACC: Heart Failure
4(1): 55–64.
Vieira, JL, Guimaraes, GV, de Andre, PA,
Nascimento Saldiva, PH and Bocchi, EA
(2016b). Effects of reducing exposure to
air pollution on submaximal
cardiopulmonary test in patients with
heart failure: Analysis of the
randomized, double-blind and controlled
FILTER–HF trial. International Journal of
Cardiology 215: 92–97.
Wang, Y, Shi, L, Lee, M, Liu, P, Di, Q,
Zanobetti, A and Schwartz, JD (2017).
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
Long-term exposure to PM2.5 and
mortality among older adults in the
Southeastern US. Epidemiology 28(2):
207–214.
Ward-Caviness, CK, Weaver, AM, Buranosky,
M, Pfaff, ER, Neas, LM, Devlin, RB,
Schwartz, J, Di, Q, Cascio, WE and DiazSanchez, D (2020). Associations between
long-term fine particulate matter
exposure and mortality in heart failure
patients. Journal of the American Heart
Association 9(6): e012517.
Wei, Y, Yazdi, MD, Di, Q, Requia, WJ,
Dominici, F, Zanobetti, A and Schwartz,
J (2021). Emulating causal dose-response
relations between air pollutants and
mortality in the Medicare population.
Environmental Health: A Global Access
Science Source 20(1): 53.
Wu, X, Braun, D, Schwartz, J,
Kioumourtzoglou, MA and Dominici, F
(2020). Evaluating the impact of longterm exposure to fine particulate matter
on mortality among the elderly. Science
Advances 6(29): eaba5692.
Wyatt, LH, Devlin, RB, Rappold, AG, Case,
MW and Diaz-Sanchez, D (2020). Low
levels of fine particulate matter increase
vascular damage and reduce pulmonary
function in young healthy adults.
Particle and fibre toxicology 17(1): 1–12.
Yorifuji, T, Kashima, S and Doi, H (2016).
Fine-particulate air pollution from diesel
emission control and mortality rates in
Tokyo: a quasi-experimental study.
Epidemiology 27(6): 769–778.
Zanobetti, A and Schwartz, J (2009). The
effect of fine and coarse particulate air
pollution on mortality: A national
analysis. Environmental Health
Perspectives 117(6): 1–40.
Zhang, Z, Wang, J, Kwong, JC, Burnett, RT,
van Donkelaar, A, Hystad, P, Martin, RV,
Bai, L, McLaughlin, J and Chen, H
(2021). Long-term exposure to air
pollution and mortality in a prospective
cohort: The Ontario Health Study.
Environment International 154: 106570.
List of Subjects
40 CFR Part 50
Environmental protection, Air
pollution control, Carbon monoxide,
Lead, Nitrogen dioxide, Ozone,
Particulate matter, Sulfur oxides.
40 CFR Part 53
Environmental protection,
Administrative practice and procedure,
Air pollution control, Reporting and
recordkeeping requirements.
40 CFR Part 58
Environmental protection,
Administrative practice and procedure,
Air pollution control, Intergovernmental
relations, Reporting and recordkeeping
requirements.
Michael S. Regan,
Administrator.
For the reasons set forth in the
preamble, chapter I of title 40 of the
PO 00000
Frm 00180
Fmt 4701
Sfmt 4700
Code of Federal Regulations is amended
as follows:
PART 50—NATIONAL PRIMARY AND
SECONDARY AMBIENT AIR QUALITY
STANDARDS
1. The authority citation for part 50
continues to read as follows:
■
Authority: 42 U.S.C. 7401, et seq.
■
2. Add § 50.20 to read as follows:
§ 50.20 National primary ambient air
quality standards for PM2.5.
(a) The national primary ambient air
quality standards for PM2.5 are 9.0
micrograms per cubic meter (mg/m3)
annual arithmetic mean concentration
and 35 mg/m3 24-hour average
concentration measured in the ambient
air as PM2.5 (particles with an
aerodynamic diameter less than or equal
to a nominal 2.5 micrometers) by either:
(1) A reference method based on
appendix L to this part and designated
in accordance with part 53 of this
chapter; or
(2) An equivalent method designated
in accordance with part 53 of this
chapter.
(b) The primary annual PM2.5
standard is met when the annual
arithmetic mean concentration, as
determined in accordance with
appendix N to this part, is less than or
equal to 9.0 mg/m3.
(c) The primary 24-hour PM2.5
standard is met when the 98th
percentile 24-hour concentration, as
determined in accordance with
appendix N to this part, is less than or
equal to 35 mg/m3.
■ 3. Amend appendix K to part 50 by:
■ a. In section 1.0 revising paragraph
(b);
■ b. In section 2.3 adding paragraph (d);
and
■ c. In section 3.0 adding paragraphs (a)
and (b).
The revision and additions read as
follows:
Appendix K to Part 50—Interpretation
of the National Ambient Air Quality
Standards for Particulate Matter
1.0
General
*
*
*
*
*
(b) The terms used in this appendix are
defined as follows:
Average refers to the arithmetic mean of
the estimated number of exceedances per
year, as per section 3.1 of this appendix.
Collocated monitors refer to two or more
air measurement instruments for the same
parameter (e.g., PM10 mass) operated at the
same site location, and whose placement is
consistent with part 53 of this chapter. For
purposes of considering a combined site
record in this appendix, when two or more
monitors are operated at the same site, one
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
monitor is designated as the ‘‘primary’’
monitor with any additional monitors
designated as ‘‘collocated.’’ It is implicit in
these appendix procedures that the primary
monitor and collocated monitor(s) are all
reference or equivalent methods; however, it
is not a requirement that the primary and
collocated monitors utilize the same specific
sampling and analysis method.
Combined site data record is the data set
used for performing computations in this
appendix and represents data for the primary
monitors augmented with data from
collocated monitors according to the
procedure specified in section 3.0(a) of this
appendix.
Daily value for PM10 refers to the 24-hour
average concentration of PM10 calculated or
measured from midnight to midnight (local
time).
Exceedance means a daily value that is
above the level of the 24-hour standard after
rounding to the nearest 10 mg/m3 (i.e., values
ending in 5 or greater are to be rounded up).
Expected annual value is the number
approached when the annual values from an
increasing number of years are averaged, in
the absence of long-term trends in emissions
or meteorological conditions.
Primary monitors are suitable monitors
designated by a State or local agency in their
annual network plan as the default data
source for creating a combined site data
record. If there is only one suitable monitor
at a particular site location, then it is
presumed to be a primary monitor.
Year refers to a calendar year.
*
2.3
ddrumheller on DSK120RN23PROD with RULES3
*
*
*
*
*
Data Requirements
*
*
*
*
(d) 24-hour average concentrations will be
computed from submitted hourly PM10
concentration data for each corresponding
day of the year and the result will be stored
in the first, or start, hour (i.e., midnight, hour
‘0’) of the 24-hour period. A 24-hour average
concentration shall be considered valid if at
least 75 percent of the hourly averages (i.e.,
18 hourly values) for the 24-hour period are
available. In the event that fewer than all 24
hourly average concentrations are available
(i.e., fewer than 24 but at least 18), the 24hour average concentration shall be
computed on the basis of the hours available
using the number of available hours within
the 24-hour period as the divisor (e.g., the
divisor is 19 if 19 hourly values are
available). 24-hour periods with 7 or more
missing hours shall also be considered for
computations in this appendix if, after
substituting zero for all missing hourly
concentrations, the resulting 24-hour average
daily value exceeds the level of the 24-hour
standard specified in § 50.6 after rounding to
the nearest 10 mg/m3.
*
*
*
*
*
3.0 Computational Equations for the 24Hour Standards
(a) All computations shown in this
appendix shall be implemented on a sitelevel basis. Site level concentration data shall
be processed as follows:
VerDate Sep<11>2014
20:48 Mar 05, 2024
Jkt 262001
(1) The default dataset for PM10 mass
concentrations for a site shall consist of the
measured concentrations recorded from the
designated primary monitor(s). All daily
values produced by the primary monitor are
considered part of the site record.
(2) If a daily value is not produced by the
primary monitor for a particular day, but a
value is available from a single collocated
monitor, then that collocated monitor value
shall be considered part of the combined site
data record. If daily value data is available
from two or more collocated monitors, the
average of those collocated values shall be
used as the daily value. The data record
resulting from this procedure is referred to as
the ‘‘combined site data record.’’
(b) In certain circumstances, including but
not limited to site closures or relocations,
data from two nearby sites may be combined
into a single site data record for the purpose
of calculating a valid design value. The
appropriate Regional Administrator may
approve such combinations if the Regional
Administrator determines that the measured
concentrations do not differ substantially
between the two sites, taking into
consideration factors such as distance
between sites, spatial and temporal patterns
in air quality, local emissions and
meteorology, jurisdictional boundaries, and
terrain features.
*
*
*
*
*
4. Amend appendix L to part 50 by
revising section 7.3.4 and adding
section 7.3.4.5 to read as follows:
■
Appendix L to Part 50—Reference
Method for the Determination of Fine
Particulate Matter as PM2.5 in the
Atmosphere
*
*
*
*
*
7.3.4 Particle size separator. The sampler
shall be configured with one of the three
alternative particle size separators described
in this section. One separator is an impactortype separator (WINS impactor) described in
sections 7.3.4.1, 7.3.4.2, and 7.3.4.3 of this
appendix. One alternative separator is a
cyclone-type separator (VSCCTM) described
in section 7.3.4.4 of this appendix. The other
alternative separator is also a cyclone-type
separator (TE–PM2.5C) described in section
7.3.4.5 of this appendix.
*
*
*
*
*
7.3.4.5 A second cyclone-type separator is
identified as a Tisch TE–PM2.5C Cyclone
particle size separator specified as part of
EPA-designated reference method RFPS–
1014–219 and as manufactured by Tisch
Environmental Incorporated, 145 S. Miami
Avenue, Village of Cleves, Ohio 45002.
*
*
*
*
*
5. Amend appendix N to part 50 by:
a. In section 1.0 revising paragraph (a);
b. In section 3.0 adding paragraph
(d)(3);
■ c. In section 4.1 revising paragraph (a);
and
■ d. In section 4.2 revising paragraph
(a).
■
■
■
PO 00000
Frm 00181
Fmt 4701
Sfmt 4700
16381
The addition and revisions read as
follows.
Appendix N to Part 50—Interpretation
of the National Ambient Air Quality
Standards for PM2.5
1.0
General
(a) This appendix explains the data
handling conventions and computations
necessary for determining when the national
ambient air quality standards (NAAQS) for
PM2.5 are met, specifically the primary and
secondary annual and 24-hour PM2.5 NAAQS
specified in §§ 50.7, 50.13, 50.18, and 50.20.
PM2.5 is defined, in general terms, as particles
with an aerodynamic diameter less than or
equal to a nominal 2.5 micrometers. PM2.5
mass concentrations are measured in the
ambient air by a Federal Reference Method
(FRM) based on appendix L to this part, as
applicable, and designated in accordance
with part 53 of this chapter or by a Federal
Equivalent Method (FEM) designated in
accordance with part 53 of this chapter. Only
those FRM and FEM measurements that are
derived in accordance with part 58 of this
chapter (i.e., that are deemed ‘‘suitable’’)
shall be used in comparisons with the PM2.5
NAAQS. The data handling and computation
procedures to be used to construct annual
and 24-hour NAAQS metrics from reported
PM2.5 mass concentrations, and the
associated instructions for comparing these
calculated metrics to the levels of the PM2.5
NAAQS, are specified in sections 2.0, 3.0,
and 4.0 of this appendix.
*
*
*
*
*
3.0 Requirements for Data Use and Data
Reporting for Comparisons With the NAAQS
for PM2.5
*
*
*
*
*
(d) * * *
(3) In certain circumstances, including but
not limited to site closures or relocations,
data from two nearby sites may be combined
into a single site data record for the purpose
of calculating a valid design value. The
appropriate Regional Administrator may
approve such site combinations if the
Regional Administrator determines that the
measured concentrations do not differ
substantially between the two sites, taking
into consideration factors such as distance
between sites, spatial and temporal patterns
in air quality, local emissions and
meteorology, jurisdictional boundaries, and
terrain features.
*
4.1
*
*
*
*
Annual PM2.5 NAAQS
(a) Levels of the primary and secondary
annual PM2.5 NAAQS are specified in
§§ 50.7, 50.13, 50.18, and 50.20 as applicable.
*
4.2
*
*
*
*
Twenty-Four-Hour PM2.5 NAAQS
(a) Levels of the primary and secondary 24hour PM2.5 NAAQS are specified in §§ 50.7,
50.13, 50.18, and 50.20 as applicable.
*
E:\FR\FM\06MRR3.SGM
*
*
06MRR3
*
*
16382
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
PART 53—AMBIENT AIR MONITORING
REFERENCE AND EQUIVALENT
METHODS
6. The authority citation for part 53
continues to read as follows:
■
Authority: Sec. 301(a) of the Clean Air Act
(42 U.S.C. 1857g(a)), as amended by sec.
15(c)(2) of Pub. L. 91–604, 84 Stat. 1713,
unless otherwise noted.
Subpart A—General Provisions
7. Amend § 53.4 by:
a. Revising paragraph (a);
b. Adding paragraph (b)(7); and
c. Revising paragraph (d).
The revisions and addition read as
follows:
■
■
■
■
§ 53.4 Applications for reference or
equivalent method determinations.
(a) Applications for FRM or FEM
determinations and modification
requests of existing designated
instruments shall be submitted to: U.S.
Environmental Protection Agency,
Director, Center for Environmental
Measurement and Modeling, Reference
and Equivalent Methods Designation
Program (MD–D205–03), 109 T.W.
Alexander Drive, P.O. Box 12055,
Research Triangle Park, North Carolina
27711 (commercial delivery address:
4930 Old Page Road, Durham, North
Carolina 27703).
*
*
*
*
*
(b) * * *
(7) All written materials for new FRM
and FEM applications and modification
requests must be submitted in English
in MS Word format. For any calibration
certificates originally written in a nonEnglish language, the original nonEnglish version of the certificate must
be submitted to EPA along with a
version of the certificate translated to
English. All laboratory and field data
associated with new FRM and FEM
applications and modification requests
must be submitted in MS Excel format.
§ 53.8 Designation of reference and
equivalent methods.
All worksheets in MS Excel must be
unprotected to enable full inspection as
part of the application review process.
*
*
*
*
*
(d) For candidate reference or
equivalent methods or for designated
instruments that are the subject of a
modification request, the applicant, if
requested by EPA, shall provide to EPA
a representative sampler or analyzer for
test purposes. The sampler or analyzer
shall be shipped free on board (FOB)
destination to Director, Center for
Environmental Measurements and
Modeling, Reference and Equivalent
Methods Designation Program (MD
D205–03), U.S. Environmental
Protection Agency, 4930 Old Page Road,
Durham, North Carolina 27703,
scheduled to arrive concurrently with or
within 30 days of the arrival of the other
application materials. This sampler or
analyzer may be subjected to various
tests that EPA determines to be
necessary or appropriate under § 53.5(f),
and such tests may include special tests
not described in this part. If the
instrument submitted under this
paragraph (d) malfunctions, becomes
inoperative, or fails to perform as
represented in the application before the
necessary EPA testing is completed, the
applicant shall be afforded the
opportunity to repair or replace the
device at no cost to the EPA. Upon
completion of EPA testing, the sampler
or analyzer submitted under this
paragraph (d) shall be repacked by EPA
for return shipment to the applicant,
using the same packing materials used
for shipping the instrument to EPA
unless alternative packing is provided
by the applicant. Arrangements for, and
the cost of, return shipment shall be the
responsibility of the applicant. The EPA
does not warrant or assume any liability
for the condition of the sampler or
analyzer upon return to the applicant.
■ 8. Amend § 53.8 by revising paragraph
(a) to read as follows:
(a) A candidate method determined
by the Administrator to satisfy the
applicable requirements of this part
shall be designated as an FRM or FEM
(as applicable) by and upon publication
of the designation in the Federal
Register. Applicants shall not publicly
announce, market, or sell the candidate
sampler and analyzer as an approved
FRM or FEM (as applicable) until the
designation is published in the Federal
Register.
*
*
*
*
*
9. Amend § 53.14 by revising
paragraphs (c)(4), (5), and (6) to read as
follows:
■
§ 53.14 Modification of a reference or
equivalent method.
*
*
*
*
*
(c) * * *
(4) Send notice to the applicant that
additional information must be
submitted before a determination can be
made and specify the additional
information that is needed (in such
cases, the 90-day period shall
commence upon receipt of the
additional information).
(5) Send notice to the applicant that
additional tests are necessary and
specify which tests are necessary and
how they shall be interpreted (in such
cases, the 90-day period shall
commence upon receipt of the
additional test data).
(6) Send notice to the applicant that
additional tests will be conducted by
the Administrator and specify the
reasons for and the nature of the
additional tests (in such cases, the 90day period shall commence 1 calendar
day after the additional tests are
completed).
*
*
*
*
*
■ 10. Revise table A–1 to subpart A of
part 53 to read as follows:
TABLE A–1 TO SUBPART A OF PART 53—SUMMARY OF APPLICABLE REQUIREMENTS FOR REFERENCE AND EQUIVALENT
METHODS FOR AIR MONITORING OF CRITERIA POLLUTANTS
Pollutant
SO2 ............
Reference or equivalent
Reference ...........................
ddrumheller on DSK120RN23PROD with RULES3
Equivalent ..........................
CO ..............
O3 ...............
NO2 ............
VerDate Sep<11>2014
Reference ...........................
Equivalent ..........................
Reference ...........................
Equivalent ..........................
Reference ...........................
Equivalent ..........................
19:03 Mar 05, 2024
Manual or automated
Manual ...............................
Automated ..........................
Manual ...............................
Automated ..........................
Automated ..........................
Manual ...............................
Automated ..........................
Automated ..........................
Manual ...............................
Automated ..........................
Automated ..........................
Manual ...............................
Automated ..........................
Jkt 262001
PO 00000
Frm 00182
Applicable
appendix of
part 50 of this
chapter
A–2
A–1
A–1
A–1
C
C
C
D
D
D
F
F
F
Fmt 4701
Sfmt 4700
Applicable subparts of this part
A
B
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
................
✓
✓
................
✓
✓
................
✓
✓
................
✓
E:\FR\FM\06MRR3.SGM
C
✓
✓
✓
✓
✓
✓
✓
✓
06MRR3
D
E
F
16383
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
TABLE A–1 TO SUBPART A OF PART 53—SUMMARY OF APPLICABLE REQUIREMENTS FOR REFERENCE AND EQUIVALENT
METHODS FOR AIR MONITORING OF CRITERIA POLLUTANTS—Continued
Pollutant
Pb ...............
PM10-Pb .....
Reference or equivalent
Reference ...........................
Equivalent ..........................
Reference ...........................
Equivalent ..........................
PM10 ...........
Reference ...........................
Equivalent ..........................
PM2.5 ..........
Reference
Equivalent
Equivalent
Equivalent
Reference
Equivalent
Equivalent
Equivalent
PM10–2.5 .....
1 Some
...........................
Class I ..............
Class II .............
Class III ............
...........................
Class I ..............
Class II .............
Class III ............
Manual or automated
Manual ...............................
Manual ...............................
Automated ..........................
Manual ...............................
Manual ...............................
Automated ..........................
Manual ...............................
Manual ...............................
Automated ..........................
Manual ...............................
Manual ...............................
Manual ...............................
Automated ..........................
Manual ...............................
Manual ...............................
Manual ...............................
Automated ..........................
Applicable
appendix of
part 50 of this
chapter
G
G
G
Q
Q
Q
J
J
J
L
L
L1
L1
L,2 O
L,2 O
L,2 O
1 L, 1 2 O
Applicable subparts of this part
A
B
C
✓
✓
................
................
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
................
................
................
................
................
................
................
................
................
................
................
................
................
✓
✓
................
✓
✓
................
✓
2✓
✓
................
✓
2✓
✓
D
E
✓
✓
✓
................
................
................
................
................
................
................
................
✓
✓
✓
✓
✓
✓
✓
✓
F
12✓
1✓
1,2 ✓
1✓
requirements may apply, based on the nature of each particular candidate method, as determined by the Administrator.
Class III requirements may be substituted.
2 Alternative
Table B–1 to Subpart B of Part 53—
Performance Limit Specifications for
Automated Methods
*
*
*
*
*
Subpart B—Procedures for Testing
Performance Characteristics of
Automated Methods for SO2, CO, O3,
and NO2
11. Amend table B–1 to subpart B of
part 53 by revising footnote 4 to read as
follows:
ddrumheller on DSK120RN23PROD with RULES3
■
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
4 For
nitric oxide interference for the SO2
ultraviolet fluorescence (UVF) method,
PO 00000
Frm 00183
Fmt 4701
Sfmt 4700
interference equivalent is ±0.003 ppm for the
lower range.
*
*
*
*
*
12. Revise table B–3 to subpart B of
part 53 to read as follows:
■
E:\FR\FM\06MRR3.SGM
06MRR3
19:03 Mar 05, 2024
PO 00000
Jkt 262001
Frm 00184
Ultraviolet fluorescence .....................................................................
Flame photometric .............................................................................
Gas chromatography .........................................................................
Spectrophotometric-wet chemical (pararosanaline) ..........................
Electrochemical .................................................................................
Conductivity .......................................................................................
Spectrophotometric-gas phase, including DOAS ..............................
Ethylene Chemiluminescence ...........................................................
NO-chemiluminescence ....................................................................
Electrochemical .................................................................................
Spectrophotometric-wet chemical (potassium iodide) ......................
Spectrophotometric-gas phase, including ultraviolet absorption and
DOAS.
Non-dispersive Infrared .....................................................................
Gas chromatography with flame ionization detector .........................
Electrochemical .................................................................................
Catalytic combustion-thermal detection ............................................
IR fluorescence .................................................................................
Mercury replacement-UV photometric ..............................................
Chemiluminescent .............................................................................
Spectrophotometric-wet chemical (azo-dye reaction) .......................
Electrochemical .................................................................................
Spectrophotometric-gas phase .........................................................
Analyzer type 2
..............
..............
..............
..............
..............
..............
..............
..............
0.2
..............
..............
..............
..............
0.2
0.2
0.2
..............
..............
..............
..............
..............
..............
Hydrochloric
acid
..............
..............
..............
0.1
..............
..............
3 0.1
..............
3 0.1
3 0.1
..............
..............
..............
0.1
0.1
0.1
..............
..............
..............
3 0.1
3 0.1
..............
Ammonia
Sulfur
dioxide
Nitrogen
dioxide
Nitric
oxide
Carbon
dioxide
Ethylene
Ozone
M-xylene
Water
vapor
Carbon
monoxide Methane
3 Do
2 Analyzer
1 Concentrations
Ethane
Naphthalene
................
................
................
................
................
................
................
................
................
................
..............
..............
..............
..............
..............
..............
0.5
0.5
0.5
0.5
................
................
................
................
................
................
4 0.1
4 0.1
4 0.1
4 0.1
..............
..............
0.5
..............
..............
..............
0.5
0.5
0.5
0.5
750
................
................
750
750
................
................
750
750
................
................
................
0.2
0.2
................
0.2
................
................
................
................
..............
..............
..............
..............
..............
..............
..............
0.5
0.5
0.5
4 10 .............. .............. ......................
................
20,000
4 10 ..............
................
20,000
0.5 ......................
4 10 .............. .............. ......................
................
20,000
4
................
20,000
10
5.0
0.5 ......................
4 10 ..............
................
20,000
0.5 ......................
4
................ ..............
10 ..............
0.5 ......................
................
20,000 ................ .............. .............. ......................
................ .............. ................ .............. .............. ......................
................
20,000
50 .............. .............. ......................
................
20,000
50 .............. .............. ......................
4 0.14
6 0.05
0.5
0.5 ................ ................
0.5
0.2
20,000 ................ .............. ..............
4 0.14 ................ ..............
0.01
750 ................ .............. ................ 3 20,000
50 .............. .............. ......................
4 0.14 ................ ..............
0.1
750 ................ .............. ................ 3 20,000
50 .............. .............. ......................
4 0.14
0.1
0.5 ..............
750 ................
0.5 ................ .............. ................ .............. .............. ......................
4
3
0.1
0.14
0.5
0.5 ................
0.2
0.5 ................
20,000 ................ .............. .............. ......................
4 0.14
................
0.5 ..............
750 ................ .............. ................ .............. ................ .............. .............. ......................
4
................
0.14
0.5
0.5 ................ ................
0.5
0.2 .............. ................ .............. .............. ......................
3 0.1 .............. ................ ..............
4 0.08 ................
3 20,000 ................ .............. .............. ......................
750 ................
3 0.1 ..............
4 0.08 ................
3 20,000 ................ .............. .............. ......................
0.5 ..............
750 ................
4 0.08 ................
3 20,000 ................ .............. .............. ......................
................
0.5
0.5 .............. ................ ................
3 0.5 ................ ................
4 0.08 ................ .............. ................ .............. .............. ......................
................
0.5
0.5
3 0.5 ................ ................
4 0.08
................
0.5
0.5
0.02
20,000 ................ .............. .............. ......................
5 0.1
Hydrogen
sulfide
[Parts per million]
TABLE B–3 TO SUBPART B OF PART 53—INTERFERENT TEST CONCENTRATION 1
of interferent listed must be prepared and controlled to ±10 percent of the stated value.
types not listed will be considered by the Administrator as special cases.
not mix interferent with the pollutant.
4 Concentration of pollutant used for test. These pollutant concentrations must be prepared to ±10 percent of the stated value.
5 If candidate method utilizes an elevated-temperature scrubber for removal of aromatic hydrocarbons, perform this interference test.
6 If naphthalene test concentration cannot be accurately quantified, remove the scrubber, use a test concentration that causes a full-scale response, reattach the scrubber, and evaluate response for interference.
CO ...............
CO ...............
CO ...............
CO ...............
CO ...............
CO ...............
NO2 ..............
NO2 ..............
NO2 ..............
NO2 ..............
VerDate Sep<11>2014
SO2 ..............
SO2 ..............
SO2 ..............
SO2 ..............
SO2 ..............
SO2 ..............
SO2 ..............
O3 ................
O3 ................
O3 ................
O3 ................
O3 ................
Pollutant
ddrumheller on DSK120RN23PROD with RULES3
16384
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
Fmt 4701
Sfmt 4700
E:\FR\FM\06MRR3.SGM
06MRR3
16385
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
Appendix A to Subpart B of Part 53—
Optional Forms for Reporting Test
Results
*
*
*
*
LDL Interference Test Data
Applicant llllllllllllllll
*
TEST
READING or
CALCUIATION
PARAMETER
LOWER
DmCTABLE UMIT
1
2
4
3
5
TEST NUMBER
7
9
8
6
10
12
11
14
13
15
Bz
Bi.
LDL= Bi.-Bz
Rs
Ra.
IE=Ra.-R,
Rz
Ra
IE=Ra-R,
R.
Rs
IE=Re-R.
R.
R,.
IE=Ra-R.
1
2
3*
INTERFERENCE
EQUIVALENT
Analyzer llllllllllllllll
Date llllllllllllllllll
Pollutant llllllllllllllll
Figure B–3 to Appendix A to Subpart B of
Part 53—Form for Test Data and
Calculations for Lower Detectable Limit
(LDL) and Interference Equivalent (IE) (see
§ 53.23(c) and (d))
13. Amend appendix A to subpart B
of part 53 by revising figures B–3 and
B–5 to read as follows:
■
4*
Rs
5*
Re
IE=R.,-R.
TOTAL*
z'.11e,1
..
1=1
*If required.
*
*
*
*
Analyzer llllllllllllllll
Date llllllllllllllllll
Pollutant llllllllllllllll
Figure B–5 to Appendix A to Subpart B of
Part 53—Form for Calculating Zero Drift,
Span Drift and Precision (see § 53.23(e))
*
Calculation of Zero Drift, Span Drift, and
Precision
Applicant llllllllllllllll
TEST
12
HOUR
ZERO
DRIFT
24
HOUR
TESTDAY(n)
CALCUIATION
PARAMmR
12ZD
1
2
3
4
5
6
7
8
9
10
11
U
13
14
15
=c....-c....
Z =(Lt+ LJ/2
24ZD•Z.-Z...,
24ZD=Z'.-Z',..,
SPAN
DRIFT
24
SD,. = S,. -S..-i
S..-1
HOUR
X100%
SD,.= S,.-S',._i
P20 =%STANDARD
DEVIATION of (P,...P,)
URL
PREC- (P20)
1------------+---+----+---t---+---+---t---,t---+--+----+---t---+---t,---+---t
ISION 80%
P20 =%STANDARD
URL
DEVIATION of (P.,."P12)
(Pao)
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
PO 00000
Frm 00185
Fmt 4701
Sfmt 4725
E:\FR\FM\06MRR3.SGM
06MRR3
ER06MR24.036
X100%
20%
ER06MR24.035
ddrumheller on DSK120RN23PROD with RULES3
S',._1
16386
*
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
*
*
*
§ 53.35 Test procedure for Class II and
Class III methods for PM2.5 and PM10¥2.5.
*
Subpart C—Procedures for
Determining Comparability Between
Candidate Methods and Reference
Methods
*
14. Amend § 53.35 by revising
paragraph (b)(1)(ii)(D) to read as follows:
■
*
*
(b) * * *
(1) * * *
(ii) * * *
*
*
(D) Site D shall be in a large city east
of the Mississippi River, having
characteristically high humidity levels.
*
*
*
*
*
15. Revise table C–4 to subpart C of
part 53 to read as follows:
■
TABLE C–4 TO SUBPART C OF PART 53—TEST SPECIFICATIONS FOR PM10, PM2.5, AND PM10–2.5 CANDIDATE
EQUIVALENT METHODS
ddrumheller on DSK120RN23PROD with RULES3
PM2.5
PM10–2.5
Specification
PM10
Class I
Class II
Class III
Class II
Acceptable concentration range (Rj), μg/
m 3.
Minimum number of
test sites.
Minimum number of
candidate method
samplers or analyzers per site.
Number of reference
method samplers
per site.
Minimum number of
acceptable sample
sets per site for
PM10 methods:
Rj < 20 μg/m3 ......
Rj > 20 μg/m3 ......
Total .............
Minimum number of
acceptable sample
sets per site for
PM2.5 and PM10–2.5
candidate equivalent
methods:
Rj < 15 μg/m3 for
24-hr or Rj < 8
μg/m3 for 48-hr
samples..
Rj > 15 μg/m3 for
24-hr or Rj > 8
μg/m3 for 48-hr
samples.
Each season ........
Total, each
site.
Precision of replicate
reference method
measurements, PRj
or RPRj, respectively; RP for Class
II or III PM2.5 or
PM10–2.5, maximum.
Precision of PM2.5 or
PM10–2.5 candidate
method, CP, each
site.
Slope of regression relationship.
Intercept of regression
relationship, μg/m3.
5–300 .......................
3–200 .......................
3–200 .......................
3–200 .......................
3–200 .......................
3–200.
2 ...............................
1 ...............................
2 ...............................
4 ...............................
2 ...............................
4.
3 ...............................
3 ...............................
31 .............................
31 .............................
31 .............................
3.1
3 ...............................
3 ...............................
31 .............................
31 .............................
31 .............................
3.1
3 ...............................
3 ...............................
10 .............................
..................................
..................................
..................................
..................................
..................................
..................................
..................................
..................................
..................................
..................................
..................................
..................................
..................................
3 ...............................
3 ...............................
3 ...............................
3 ...............................
3.
..................................
3 ...............................
3 ...............................
3 ...............................
3 ...............................
3.
..................................
..................................
10 .............................
10 .............................
23 .............................
23 .............................
23 .............................
23 .............................
5 μg/m3 or 7%. ........
2 μg/m3 or 5%. ........
10%2 ........................
23 .............................
23 (46 for two-season sites).
10%2 ........................
10%2 ........................
23.
23 (46 for two-season sites).
10%.2
..................................
..................................
10%2 ........................
15%2 ........................
15%2 ........................
15%.2
1 ±0.10 .....................
1 ±0.05 .....................
1 ±0.10 .....................
1 ±0.10 .....................
1 ±0.10 .....................
1 ±0.12.
0 ±5 ..........................
0 ±1 ..........................
Between: 13.55—
(15.05 × slope),
but not less than—
1.5; and 16.56—
(15.05 × slope),
but not more than
+1.5.
Between: 15.05—
(17.32 × slope),
but not less than—
2.0; and 15.05—
(13.20 × slope),
but not more than
+2.0.
Between: 62.05—
(70.5 × slope), but
not less than—3.5;
and 78.95—(70.5
× slope), but not
more than +3.5.
Between: 70.50—
(82.93 × slope),
but not less than—
7.0; and 70.50—
(61.16 × slope),
but not more than
+7.0.
≥ 0.97 .......................
≥ 0.97 .......................
≥ 0.93—for CCV ≤ 0.4;
≥ 0.85 + 0.2 × CCV—for 0.4 ≤ CCV ≤ 0.5;
≥ 0.95—for CCV ≥ 0.5
Correlation of reference method and
candidate method
measurements.
1 Some
missing daily measurement values may be permitted; see test procedure.
as the root mean square over all measurement sets.
2 Calculated
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
PO 00000
Frm 00186
Fmt 4701
Sfmt 4700
E:\FR\FM\06MRR3.SGM
06MRR3
Class III
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
formula in paragraph (c)(2)(iv) to read as
follows:
Subpart D—Procedures for Testing
Performance Characteristics of
Methods for PM10
§ 53.43
(a) * * *
(2) * * *
*
(iv) * * *
*
*
(c) * * *
(2) * * *
*
*
(xvi) * * *
Test procedures.
16. Amend § 53.43 by revising the
formula in paragraph (a)(2)(xvi) and the
■
16387
17. Amend § 53.51 by revising
paragraph (d)(2) to read as follows:
■
§ 53.51 Demonstration of compliance with
design specifications and manufacturing
and test requirements.
ddrumheller on DSK120RN23PROD with RULES3
*
*
*
*
*
(d) * * *
(2) VSCC and TE–PM2.5C separators.
For samplers and monitors utilizing the
BGI VSCC or Tisch TE–PM2.5C particle
size separators specified in sections
7.3.4.4 and 7.3.4.5 of appendix L to part
50 of this chapter, respectively, the
respective manufacturers shall identify
the critical dimensions and
manufacturing tolerances for the
separator, devise appropriate test
procedures to verify that the critical
dimensions and tolerances are
maintained during the manufacturing
process, and carry out those procedures
on each separator manufactured to
verify conformance of the manufactured
products. The manufacturer shall also
maintain records of these tests and their
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
Subpart F—Procedures for Testing
Performance Characteristics of Class II
Equivalent Methods for PM2.5
18. Amend § 53.61 by revising
paragraph (g) introductory text, the first
sentence of paragraph (g)(1), the first
sentence of (g)(1)(i), (g)(2)(i) and adding
paragraph (g)(2)(iii) to read as follows:
■
§ 53.61
Test conditions.
*
*
*
*
*
(g) Vibrating Orifice Aerosol
Generator (VOAG) and Flow-Focusing
Monodisperse Aerosol Generator
(FMAG) conventions. This section
prescribes conventions regarding the
use of the vibrating orifice aerosol
generator (VOAG) and the flow-focusing
monodisperse aerosol generator (FMAG)
for the size-selective performance tests
outlined in §§ 53.62, 53.63, 53.64, and
53.65.
PO 00000
Frm 00187
Fmt 4701
Sfmt 4700
(1) Particle aerodynamic diameter.
The VOAG and FMAG produce nearmonodisperse droplets through the
controlled breakup of a liquid jet. * * *
(i) The physical diameter of a
generated spherical particle can be
calculated from the operational
parameters of the VOAG and FMAG as:
*
*
*
*
*
(2) * * *
(i) Solid particle tests performed in
this subpart shall be conducted using
particles composed of ammonium
fluorescein. For use in the VOAG or
FMAG, liquid solutions of known
volumetric concentration can be
prepared by diluting fluorescein powder
(C2OH12O5, FW = 332.31, CAS 2321–07–
5) with aqueous ammonia. Guidelines
for preparation of fluorescein solutions
of the desired volume concentration
(Cvol) are presented in Vanderpool and
Rubow (1988) (Reference 2 in appendix
A to this subpart). For purposes of
converting particle physical diameter to
aerodynamic diameter, an ammonium
fluorescein particle density of 1.35 g/
cm3 shall be used.
*
*
*
*
*
(iii) Calculation of the physical
diameter of the particles produced by
the VOAG and FMAG requires
E:\FR\FM\06MRR3.SGM
06MRR3
ER06MR24.038
Subpart E—Procedures for Testing
Physical (Design) and Performance
Characteristics of Reference Methods
and Class I and Class II Equivalent
Methods for PM2.5 or PM10–2.5
test results and submit evidence that
this procedure is incorporated into the
manufacturing procedure, that the test is
or will be routinely implemented, and
that an appropriate procedure is in
place for the disposition of units that
fail this tolerance tests.
*
*
*
*
*
ER06MR24.037
if C¯j is above 80 mg/m3.
ER06MR24.039
if C¯j is below 80 mg/m3, or
16388
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
knowledge of the liquid solution’s
volume concentration (Cvol). Because
uranine is essentially insoluble in oleic
acid, the total particle volume is the
sum of the oleic acid volume and the
uranine volume. The volume
concentration of the liquid solution
shall be calculated as:
Equation 5 to Paragraph (g)(2)(iii)
Where:
Vu = uranine volume, ml;
Voleic = oleic acid volume, ml;
Vsol = total solution volume, ml;
Mu = uranine mass, g;
Pu = uranine density, g/cm3;
Moleic = oleic acid mass, g; and
Poleic = oleic acid density, g/cm3.
*
*
*
*
19. The authority citation for part 58
continues to read as follows:
Authority: 42 U.S.C. 7403, 7405, 7410,
7414, 7601, 7611, 7614, and 7619.
Subpart A—General Provisions
20. Amend § 58.1 by:
a. Removing the definition for
‘‘Approved regional method (ARM)’’;
and
■ b. Revising the definition for
‘‘Traceable.’’
The revision reads as follows:
■
■
Definitions.
ddrumheller on DSK120RN23PROD with RULES3
*
*
*
*
Traceable means a measurement
result from a local standard whereby the
result can be related to the International
System of Units (SI) through a
documented unbroken chain of
calibrations, each contributing to the
measurement uncertainty. Traceable
measurement results must be compared
and certified, either directly or via not
more than one intermediate standard, to
a National Institute of Standards and
Technology (NIST)-certified reference
standard. Examples include but are not
limited to NIST Standard Reference
Material (SRM), NIST-traceable
Reference Material (NTRM), or a NISTcertified Research Gas Mixture (RGM).
Traceability to the SI through other
National Metrology Institutes (NMIs) in
addition to NIST is allowed if a
Declaration of Equivalence (DoE) exists
between NIST and that NMI.
*
*
*
*
*
■
21. Amend § 58.10 by:
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
Vsol
§ 58.10 Annual monitoring network plan
and periodic network assessment.
■
Subpart B—Monitoring Network
(Mu/ Pu)+ (Moleic/ PoleiJ
a. Revising paragraphs (a)(1) and
(b)(10) and (13);
■ b. Adding paragraph (b)(14); and
■ c. Revising paragraph (d).
The revisions and addition read as
follows:
*
*
Vu+ Voleic
V.
sol
■
PART 58—AMBIENT AIR QUALITY
SURVEILLANCE
§ 58.1
=
(a)(1) Beginning July 1, 2007, the
State, or where applicable local, agency
shall submit to the Regional
Administrator an annual monitoring
network plan which shall provide for
the documentation of the establishment
and maintenance of an air quality
surveillance system that consists of a
network of SLAMS monitoring stations
that can include FRM and FEM
monitors that are part of SLAMS, NCore,
CSN, PAMS, and SPM stations. The
plan shall include a statement of
whether the operation of each monitor
meets the requirements of appendices
A, B, C, D, and E to this part, where
applicable. The Regional Administrator
may require additional information in
support of this statement. The annual
monitoring network plan must be made
available for public inspection and
comment for at least 30 days prior to
submission to the EPA and the
submitted plan shall include and
address, as appropriate, any received
comments.
*
*
*
*
*
(b) * * *
(10) Any monitors for which a waiver
has been requested or granted by the
EPA Regional Administrator as allowed
for under appendix D or appendix E to
this part. For those monitors where a
waiver has been approved, the annual
monitoring network plan shall include
the date the waiver was approved.
*
*
*
*
*
(13) The identification of any PM2.5
FEMs used in the monitoring agency’s
network where the data are not of
sufficient quality such that data are not
to be compared to the national ambient
air quality standards (NAAQS). For
required SLAMS where the agency
identifies that the PM2.5 Class III FEM
does not produce data of sufficient
quality for comparison to the NAAQS,
PO 00000
Frm 00188
Fmt 4701
Sfmt 4700
the monitoring agency must ensure that
an operating FRM or filter-based FEM
meeting the sample frequency
requirements described in § 58.12 or
other Class III PM2.5 FEM with data of
sufficient quality is operating and
reporting data to meet the network
design criteria described in appendix D
to this part.
(14) The identification of any site(s)
intended to address being sited in an atrisk community where there are
anticipated effects from sources in the
area as required in section 4.7.1(b)(3) of
appendix D to this part. An initial
approach to the question of whether any
new or moved sites are needed and to
identify the communities in which they
intend to add monitoring for meeting
the requirement in this paragraph
(b)(14), if applicable, shall be submitted
in accordance with the requirements of
section 4.7.1(b)(3) of appendix D to this
part, which includes submission to the
EPA Regional Administrator no later
than July 1, 2024. Specifics on the
resulting proposed new or moved sites
for PM2.5 network design to address atrisk communities, if applicable, would
need to be detailed in annual
monitoring network plans due to each
applicable EPA Regional office no later
than July 1, 2025. The plan shall
provide for any required sites to be
operational no later than 24 months
from date of approval of a plan or
January 1, 2027, whichever comes first.
*
*
*
*
*
(d) The State, or where applicable
local, agency shall perform and submit
to the EPA Regional Administrator an
assessment of the air quality
surveillance system every 5 years to
determine, at a minimum, if the network
meets the monitoring objectives defined
in appendix D to this part, whether new
sites are needed, whether existing sites
are no longer needed and can be
terminated, and whether new
technologies are appropriate for
incorporation into the ambient air
monitoring network. The network
assessment must consider the ability of
existing and proposed sites to support
air quality characterization for areas
with relatively high populations of
E:\FR\FM\06MRR3.SGM
06MRR3
ER06MR24.040
Cvol
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
susceptible individuals (e.g., children
with asthma) and other at-risk
populations, and, for any sites that are
being proposed for discontinuance, the
effect on data users other than the
agency itself, such as nearby States and
Tribes or health effects studies. The
State, or where applicable local, agency
must submit a copy of this 5-year
assessment, along with a revised annual
network plan, to the Regional
Administrator. The assessments are due
every 5 years beginning July 1, 2010.
*
*
*
*
*
■ 22. Amend § 58.11 by revising
paragraphs (a)(2) and (e) to read as
follows:
ddrumheller on DSK120RN23PROD with RULES3
§ 58.11
Network technical requirements.
(a) * * *
(2) Beginning January 1, 2009, State
and local governments shall follow the
quality assurance criteria contained in
appendix A to this part that apply to
SPM sites when operating any SPM site
which uses an FRM or an FEM and
meets the requirements of appendix E to
this part, unless the Regional
Administrator approves an alternative to
the requirements of appendix A with
respect to such SPM sites because
meeting those requirements would be
physically and/or financially
impractical due to physical conditions
at the monitoring site and the
requirements are not essential to
achieving the intended data objectives
of the SPM site. Alternatives to the
requirements of appendix A may be
approved for an SPM site as part of the
approval of the annual monitoring plan,
or separately.
*
*
*
*
*
(e) State and local governments must
assess data from Class III PM2.5 FEM
monitors operated within their network
using the performance criteria described
in table C–4 to subpart C of part 53 of
this chapter, for cases where the data are
identified as not of sufficient
comparability to a collocated FRM, and
the monitoring agency requests that the
FEM data should not be used in
comparison to the NAAQS. These
assessments are required in the
monitoring agency’s annual monitoring
network plan described in § 58.10(b) for
cases where the FEM is identified as not
of sufficient comparability to a
collocated FRM. For these collocated
PM2.5 monitors, the performance criteria
apply with the following additional
provisions:
(1) The acceptable concentration
range (Rj), mg/m3 may include values
down to 0 mg/m3.
(2) The minimum number of test sites
shall be at least one; however, the
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
number of test sites will generally
include all locations within an agency’s
network with collocated FRMs and
FEMs.
(3) The minimum number of methods
shall include at least one FRM and at
least one FEM.
(4) Since multiple FRMs and FEMs
may not be present at each site, the
precision statistic requirement does not
apply, even if precision data are
available.
(5) All seasons must be covered with
no more than 36 consecutive months of
data in total aggregated together.
(6) The key statistical metric to
include in an assessment is the bias
(both additive and multiplicative) of the
PM2.5 continuous FEM(s) compared to a
collocated FRM(s). Correlation is
required to be reported in the
assessment, but failure to meet the
correlation criteria, by itself, is not
cause to exclude data from a continuous
FEM monitor.
■ 23. Amend § 58.12 by revising
paragraph (d)(1):
§ 58.12
Operating schedules.
*
*
*
*
*
(d) * * *
(1)(i) Manual PM2.5 samplers at
required SLAMS stations without a
collocated continuously operating PM2.5
monitor must operate on at least a 1-in3 day schedule unless a waiver for an
alternative schedule has been approved
per paragraph (d)(1)(ii) of this section.
(ii) For SLAMS PM2.5 sites with both
manual and continuous PM2.5 monitors
operating, the monitoring agency may
request approval for a reduction to 1-in6 day PM2.5 sampling or for seasonal
sampling from the EPA Regional
Administrator. Other requests for a
reduction to 1-in-6 day PM2.5 sampling
or for seasonal sampling may be
approved on a case-by-case basis. The
EPA Regional Administrator may grant
sampling frequency reductions after
consideration of factors (including but
not limited to the historical PM2.5 data
quality assessments, the location of
current PM2.5 design value sites, and
their regulatory data needs) if the
Regional Administrator determines that
the reduction in sampling frequency
will not compromise data needed for
implementation of the NAAQS.
Required SLAMS stations whose
measurements determine the design
value for their area and that are within
plus or minus 10 percent of the annual
NAAQS, and all required sites where
one or more 24-hour values have
exceeded the 24-hour NAAQS each year
for a consecutive period of at least 3
years are required to maintain at least a
1-in-3 day sampling frequency until the
PO 00000
Frm 00189
Fmt 4701
Sfmt 4700
16389
design value no longer meets the criteria
in this paragraph (d)(1)(ii) for 3
consecutive years. A continuously
operating FEM PM2.5 monitor satisfies
the requirement in this paragraph
(d)(1)(ii) unless it is identified in the
monitoring agency’s annual monitoring
network plan as not appropriate for
comparison to the NAAQS and the EPA
Regional Administrator has approved
that the data from that monitor may be
excluded from comparison to the
NAAQS.
(iii) Required SLAMS stations whose
measurements determine the 24-hour
design value for their area and whose
data are within plus or minus 5 percent
of the level of the 24-hour PM2.5 NAAQS
must have an FRM or FEM operate on
a daily schedule if that area’s design
value for the annual NAAQS is less than
the level of the annual PM2.5 standard.
A continuously operating FEM or PM2.5
monitor satisfies the requirement in this
paragraph (d)(1)(iii) unless it is
identified in the monitoring agency’s
annual monitoring network plan as not
appropriate for comparison to the
NAAQS and the EPA Regional
Administrator has approved that the
data from that monitor may be excluded
from comparison to the NAAQS. The
daily schedule must be maintained until
the referenced design values no longer
meets the criteria in this paragraph
(d)(1)(iii) for 3 consecutive years.
(iv) Changes in sampling frequency
attributable to changes in design values
shall be implemented no later than
January 1 of the calendar year following
the certification of such data as
described in § 58.15.
*
*
*
*
*
■ 24. Revise § 58.15 to read as follows:
§ 58.15 Annual air monitoring data
certification.
(a) The State, or where appropriate
local, agency shall submit to the EPA
Regional Administrator an annual air
monitoring data certification letter to
certify data collected by FRM and FEM
monitors at SLAMS and SPM sites that
meet criteria in appendix A to this part
from January 1 to December 31 of the
previous year. The head official in each
monitoring agency, or his or her
designee, shall certify that the previous
year of ambient concentration and
quality assurance data are completely
submitted to AQS and that the ambient
concentration data are accurate to the
best of her or his knowledge, taking into
consideration the quality assurance
findings. The annual data certification
letter is due by May 1 of each year.
(b) Along with each certification
letter, the State shall submit to the
Regional Administrator an annual
E:\FR\FM\06MRR3.SGM
06MRR3
16390
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
summary report of all the ambient air
quality data collected by FRM and FEM
monitors at SLAMS and SPM sites. The
annual report(s) shall be submitted for
data collected from January 1 to
December 31 of the previous year. The
annual summary serves as the record of
the specific data that is the object of the
certification letter.
(c) Along with each certification
letter, the State shall submit to the
Regional Administrator a summary of
the precision and accuracy data for all
ambient air quality data collected by
FRM and FEM monitors at SLAMS and
SPM sites. The summary of precision
and accuracy shall be submitted for data
collected from January 1 to December 31
of the previous year.
Subpart C—Special Purpose Monitors
25. Amend § 58.20 by revising
paragraphs (b) through (e) to read as
follows:
■
§ 58.20
Special purpose monitors (SPM).
*
*
*
*
(b) Any SPM data collected by an air
monitoring agency using a Federal
reference method (FRM) or Federal
equivalent method (FEM) must meet the
requirements of §§ 58.11 and 58.12 and
appendix A to this part or an approved
alternative to appendix A. Compliance
with appendix E to this part is optional
but encouraged except when the
monitoring agency’s data objectives are
inconsistent with the requirements in
appendix E. Data collected at an SPM
using a FRM or FEM meeting the
requirements of appendix A must be
submitted to AQS according to the
requirements of § 58.16. Data collected
by other SPMs may be submitted. The
monitoring agency must also submit to
AQS an indication of whether each SPM
reporting data to AQS monitor meets the
requirements of appendices A and E.
(c) All data from an SPM using an
FRM or FEM which has operated for
more than 24 months are eligible for
comparison to the relevant NAAQS,
subject to the conditions of §§ 58.11(e)
and 58.30, unless the air monitoring
agency demonstrates that the data came
from a particular period during which
the requirements of appendix A,
appendix C, or appendix E to this part
were not met, subject to review and EPA
Regional Office approval as part of the
ddrumheller on DSK120RN23PROD with RULES3
*
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
annual monitoring network plan
described in § 58.10.
(d) If an SPM using an FRM or FEM
is discontinued within 24 months of
start-up, the Administrator will not base
a NAAQS violation determination for
the PM2.5 or ozone NAAQS solely on
data from the SPM.
(e) If an SPM using an FRM or FEM
is discontinued within 24 months of
start-up, the Administrator will not
designate an area as nonattainment for
the CO, SO2, NO2, or 24-hour PM10
NAAQS solely on the basis of data from
the SPM. Such data are eligible for use
in determinations of whether a
nonattainment area has attained one of
these NAAQS.
*
*
*
*
*
■ 26. Amend appendix A to part 58 by:
■ a. Revising section 2.6.1 and adding
sections 2.6.1.1 and 2.6.1.2;
■ b. Removing section 3.1.2.2 and
redesignating sections 3.1.2.3, 3.1.2.4,
3.1.2.5, and 3.1.2.6 as sections 3.1.2.2,
3.1.2.3, 3.1.2.4, and 3.1.2.5, respectively;
■ c. Revising sections 3.1.3.3, 3.2.4,
4.2.1, and 4.2.5; and
■ d. In section 6 revising References (1),
(4), (6), (7), (9), (10), and (11) and table
A–1.
The revisions and additions read as
follows:
Appendix A to Part 58—Quality
Assurance Requirements for Monitors
used in Evaluations of National
Ambient Air Quality Standards
*
*
*
*
*
2.6.1 Gaseous pollutant concentration
standards (permeation devices or cylinders of
compressed gas) used to obtain test
concentrations for CO, SO2, NO, and NO2
must be EPA Protocol Gases certified in
accordance with one of the procedures given
in Reference 4 of this appendix.
2.6.1.1 The concentrations of EPA
Protocol Gas standards used for ambient air
monitoring must be certified with a 95percent confidence interval to have an
analytical uncertainty of no more than ±2.0
percent (inclusive) of the certified
concentration (tag value) of the gas mixture.
The uncertainty must be calculated in
accordance with the statistical procedures
defined in Reference 4 of this appendix.
2.6.1.2 Specialty gas producers
advertising certification with the procedures
provided in Reference 4 of this appendix and
distributing gases as ‘‘EPA Protocol Gas’’ for
ambient air monitoring purposes must adhere
to the regulatory requirements specified in 40
PO 00000
Frm 00190
Fmt 4701
Sfmt 4700
CFR 75.21(g) or not use ‘‘EPA’’ in any form
of advertising. Monitoring organizations must
provide information to the EPA on the
specialty gas producers they use on an
annual basis. PQAOs, when requested by the
EPA, must participate in the EPA Ambient
Air Protocol Gas Verification Program at least
once every 5 years by sending a new unused
standard to a designated verification
laboratory.
*
*
*
*
*
3.1.3.3 Using audit gases that are verified
against the NIST standard reference methods
or special review procedures and validated
per the certification periods specified in
Reference 4 of this appendix (EPA
Traceability Protocol for Assay and
Certification of Gaseous Calibration
Standards) for CO, SO2, and NO2 and using
O3 analyzers that are verified quarterly
against a standard reference photometer.
*
*
*
*
*
3.2.4 PM2.5 Performance Evaluation
Program (PEP) Procedures. The PEP is an
independent assessment used to estimate
total measurement system bias. These
evaluations will be performed under the
national performance evaluation program
(NPEP) as described in section 2.4 of this
appendix or a comparable program. A
prescribed number of Performance evaluation
sampling events will be performed annually
within each PQAO. For PQAOs with less
than or equal to five monitoring sites, five
valid performance evaluation audits must be
collected and reported each year. For PQAOs
with greater than five monitoring sites, eight
valid performance evaluation audits must be
collected and reported each year. A valid
performance evaluation audit means that
both the primary monitor and PEP audit
concentrations are valid and equal to or
greater than 2 mg/m3. Siting of the PEP
monitor must be consistent with section
3.2.3.4(c) of this appendix. However, any
horizontal distance greater than 4 meters and
any vertical distance greater than one meter
must be reported to the EPA regional PEP
coordinator. Additionally for every monitor
designated as a primary monitor, a primary
quality assurance organization must:
*
*
*
*
*
4.2.1 Collocated Quality Control Sampler
Precision Estimate for PM10, PM2.5, and Pb.
Precision is estimated via duplicate
measurements from collocated samplers. It is
recommended that the precision be
aggregated at the PQAO level quarterly,
annually, and at the 3-year level. The data
pair would only be considered valid if both
concentrations are greater than or equal to
the minimum values specified in section 4(c)
of this appendix. For each collocated data
pair, calculate ti, using equation 6 to this
appendix:
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
16391
Equation 6 to Section 4.2.1 of Appendix A
Where Xi is the concentration from the
primary sampler and Yi is the concentration
value from the audit sampler. The coefficient
of variation upper bound is calculated using
equation 7 to this appendix:
Equation 7 to Section 4.2.1 of Appendix A
ICV90NAAQS
k x 1:r=l tf - c1:r=1 ti) 2
2k(k - 1)
x
= 100 *
Where k is the number of valid data pairs
being aggregated, and X20.1,k-1 is the 10th
percentile of a chi-squared distribution with
k-1 degrees of freedom. The factor of 2 in the
k- 1
NAAQS Concentration* Xli,k-l
calculated using the PEP audits described in
section 3.2.4. of this appendix. The bias
estimator is based on, si, the absolute
difference in concentrations divided by the
square root of the PEP concentration.
denominator adjusts for the fact that each ti
is calculated from two values with error.
*
*
*
*
*
4.2.5 Performance Evaluation Programs
Bias Estimate for PM2.5. The bias estimate is
Equation 8 to Section 4.2.5 of Appendix A
*
*
*
t=l
n,JNAAQS concentration
*
6. References
(1) American National Standard Institute—
Quality Management Systems For
Environmental Information And
Technology Programs—Requirements
With Guidance For Use. ASQ/ANSI E4–
2014. February 2014. Available from
ANSI Webstore https://webstore.ansi.
org/.
*
*
*
*
*
(4) EPA Traceability Protocol for Assay and
Certification of Gaseous Calibration
Standards. EPA–600/R–12/531. May,
2012. Available from U.S. Environmental
Protection Agency, National Risk
Management Research Laboratory,
Research Triangle Park NC 27711.
https://www.epa.gov/nscep.
*
*
*
*
t
*
where
St
meas - audit
.../audit
= -----
(6) List of Designated Reference and
Equivalent Methods. Available from U.S.
Environmental Protection Agency,
Center for Environmental Measurements
and Modeling, Air Methods and
Characterization Division, MD–D205–03,
Research Triangle Park, NC 27711.
https://www.epa.gov/amtic/airmonitoring-methods-criteria-pollutants.
(7) Transfer Standards for the Calibration of
Ambient Air Monitoring Analyzers for
Ozone. EPA–454/B–13–004 U.S.
Environmental Protection Agency,
Research Triangle Park, NC 27711,
October, 2013. https://www.epa.gov/
sites/default/files/2020-09/documents/
ozonetransferstandardguidance.pdf.
*
*
*
*
*
(9) Quality Assurance Handbook for Air
Pollution Measurement Systems, Volume
1—A Field Guide to Environmental
Quality Assurance. EPA–600/R–94/038a.
April 1994. Available from U.S.
Environmental Protection Agency, ORD
Publications Office, Center for
Environmental Research Information
(CERI), 26 W. Martin Luther King Drive,
Cincinnati, OH 45268. https://
www.epa.gov/amtic/ambient-airmonitoring-qualityassurance#documents.
(10) Quality Assurance Handbook for Air
Pollution Measurement Systems, Volume
II: Ambient Air Quality Monitoring
Program Quality System Development.
EPA–454/B–13–003. https://
www.epa.gov/amtic/ambient-airmonitoring-qualityassurance#documents.
(11) National Performance Evaluation
Program Standard Operating Procedures.
https://www.epa.gov/amtic/ambient-airmonitoring-quality-assurance#npep.
Method
Assessment method
Coverage
Minimum
frequency
Parameters
reported
Gaseous Methods
(CO, NO2, SO2, O3):
VerDate Sep<11>2014
19:03 Mar 05, 2024
AQS assessment
type
ER06MR24.042
ddrumheller on DSK120RN23PROD with RULES3
TABLE A–1 TO SECTION 6 OF APPENDIX A—MINIMUM DATA ASSESSMENT REQUIREMENTS FOR NAAQS RELATED
CRITERIA POLLUTANT MONITORS
Jkt 262001
PO 00000
Frm 00191
Fmt 4701
Sfmt 4700
E:\FR\FM\06MRR3.SGM
06MRR3
ER06MR24.041
*
In s
ER06MR24.043
100 x
16392
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
TABLE A–1 TO SECTION 6 OF APPENDIX A—MINIMUM DATA ASSESSMENT REQUIREMENTS FOR NAAQS RELATED
CRITERIA POLLUTANT MONITORS—Continued
Assessment method
Coverage
Minimum
frequency
Parameters
reported
Response check at
concentration
0.005–0.08 ppm
SO2, NO2, O3, and.
0.5 and 5 ppm CO ....
See section 3.1.2 of
this appendix.
Each analyzer ...........
Once per 2 weeks 5 ..
Audit concentration 1
and measured concentration.2.
One-Point QC.
Each analyzer ...........
Once per year ...........
Annual PE.
Independent Audit .....
20% of sites each
year.
Once per year ...........
Audit concentration 1
and measured concentration 2 for
each level.
Audit concentration1
and measured concentration 2 for
each level.
Collocated samplers
15% ...........................
1-in-12 days ..............
Primary sampler concentration and duplicate sampler
concentration.3.
No Transaction reported as raw data.
Collocated samplers
15% ...........................
1-in-12 days ..............
Primary sampler concentration and duplicate sampler
concentration.3.
No Transaction reported as raw data.
Check of sampler
flow rate.
Each sampler ............
Once every month 5 ..
Flow Rate
Verification.
Check of sampler
flow rate.
Each sampler ............
Once every quarter 5
Check of sampler
flow rate using
independent standard.
Each sampler ............
Once every 6
months 5.
Audit flow rate and
measured flow rate
indicated by the
sampler.
Audit flow rate and
measured flow rate
indicated by the
sampler.
Audit flow rate and
measured flow rate
indicated by the
sampler.
Check of analytical
system with Pb
audit strips/filters.
Analytical ...................
Once each quarter 5 ..
Performance Evaluation Program
PM2.5.
Collocated samplers
Distributed over all 4
quarters 5.
Performance Evaluation Program
Pb-TSP, PbPM10.
Collocated samplers
(1) 5 valid audits for
primary QA orgs,
with ≤5 sites.
(2) 8 valid audits for
primary QA orgs,
with >5 sites.
(3) All samplers in 6
years.
(1) 1 valid audit and 4
collocated samples
for primary QA
orgs, with ≤5 sites.
(2) 2 valid audits and
6 collocated samples for primary QA
orgs with >5 sites.
Method
One-Point QC for
SO2, NO2, O3,
CO.
Annual performance
evaluation for SO2,
NO2, O3, CO.
NPAP for SO2, NO2,
O3, CO.
Particulate Methods:
Continuous 4
method—collocated quality
control sampling
PM2.5.
Manual method—
collocated quality control sampling PM10,
PM2.5, Pb-TSP,
Pb-PM10.
Flow rate
verification
PM10 (low Vol)
PM2.5, Pb-PM10.
Flow rate
verification
PM10 (HighVol), Pb-TSP.
Semi-annual flow
rate audit PM10,
TSP, PM10–2.5,
PM2.5, Pb-TSP,
Pb-PM10.
Pb analysis audits
Pb-TSP, PbPM10.
Distributed over all 4
quarters 5.
Measured value and
audit value (ug Pb/
filter) using AQS
unit code 077.
Primary sampler concentration and performance evaluation sampler concentration.
Primary sampler concentration and performance evaluation sampler concentration. Primary
sampler concentration and duplicate
sampler concentration.
AQS assessment
type
NPAP.
Flow Rate
Verification.
Semi Annual Flow
Rate Audit.
Pb Analysis Audits.
PEP.
PEP.
ddrumheller on DSK120RN23PROD with RULES3
1
Effective concentration for open path analyzers.
Corrected concentration, if applicable for open path analyzers.
Both primary and collocated sampler values are reported as raw data.
4
PM2.5 is the only particulate criteria pollutant requiring collocation of continuous and manual primary monitors.
5
EPA’s recommended maximum number of days that should exist between checks to ensure that the checks are routinely conducted over
time and to limit data impacts resulting from a failed check.
2
3
*
■
*
*
*
*
27. Amend appendix B to part 58 by:
VerDate Sep<11>2014
20:35 Mar 05, 2024
Jkt 262001
a. Revising section 2.6.1 and adding
sections 2.6.1.1 and 2.6.1.2;
■
PO 00000
Frm 00192
Fmt 4701
Sfmt 4700
b. Removing and reserving section
3.1.2.2;
■
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
form of advertising. The PSD PQAOs
must provide information to the PSD
reviewing authority on the specialty gas
producers they use (or will use) for the
duration of the PSD monitoring project.
This information can be provided in the
QAPP or monitoring plan but must be
updated if there is a change in the
specialty gas producers used.
*
*
*
*
*
3.1.3.3 Using audit gases that are
verified against the NIST standard
reference methods or special review
procedures and validated per the
certification periods specified in
Reference 4 of this appendix (EPA
Traceability Protocol for Assay and
Certification of Gaseous Calibration
Standards) for CO, SO2, and NO2 and
using O3 analyzers that are verified
quarterly against a standard reference
photometer.
*
*
*
*
*
3.2.4 PM2.5 Performance Evaluation
Program (PEP) Procedures. The PEP is
an independent assessment used to
estimate total measurement system bias.
These evaluations will be performed
under the NPEP as described in section
2.4 of this appendix or a comparable
program. Performance evaluations will
be performed annually within each
PQAO. For PQAOs with less than or
equal to five monitoring sites, five valid
performance evaluation audits must be
collected and reported each year. For
PQAOs with greater than five
monitoring sites, eight valid
performance evaluation audits must be
collected and reported each year. A
valid performance evaluation audit
means that both the primary monitor
c. Revising sections 3.1.3.3 and 3.2.4;
d. Adding sections 3.2.4.1 through
3.2.4.3;
■ e. Revising sections 4.2.1, and 4.2.5;
and
■ f. In section 6 revising References (1),
(4), (6), (7), (9), (10), and (11) and table
B–1.
The revisions and additions read as
follows:
■
■
Appendix B to Part 58—Quality
Assurance Requirements for Prevention
of Significant Deterioration (PSD) Air
Monitoring
*
*
*
*
*
2.6.1 Gaseous pollutant
concentration standards (permeation
devices or cylinders of compressed gas)
used to obtain test concentrations for
CO, SO2, NO, and NO2 must be EPA
Protocol Gases certified in accordance
with one of the procedures given in
Reference 4 of this appendix.
2.6.1.1 The concentrations of EPA
Protocol Gas standards used for ambient
air monitoring must be certified with a
95-percent confidence interval to have
an analytical uncertainty of no more
than ±2.0 percent (inclusive) of the
certified concentration (tag value) of the
gas mixture. The uncertainty must be
calculated in accordance with the
statistical procedures defined in
Reference 4 of this appendix.
2.6.1.2 Specialty gas producers
advertising certification with the
procedures provided in Reference 4 of
this appendix and distributing gases as
‘‘EPA Protocol Gas’’ for ambient air
monitoring purposes must adhere to the
regulatory requirements specified in 40
CFR 75.21(g) or not use ‘‘EPA’’ in any
16393
and PEP audit concentrations are valid
and equal to or greater than 2 mg/m3.
Siting of the PEP monitor must be
consistent with section 3.2.3.4(c) of this
appendix. However, any horizontal
distance greater than 4 meters and any
vertical distance greater than one meter
must be reported to the EPA regional
PEP coordinator. Additionally for every
monitor designated as a primary
monitor, a primary quality assurance
organization must:
3.2.4.1 Have each method
designation evaluated each year; and,
3.2.4.2 Have all FRM and FEM
samplers subject to a PEP audit at least
once every 6 years, which equates to
approximately 15 percent of the
monitoring sites audited each year.
3.2.4.3 Additional information
concerning the PEP is contained in
Reference 10 of this appendix. The
calculations for evaluating bias between
the primary monitor and the
performance evaluation monitor for
PM2.5 are described in section 4.2.5 of
this appendix.
*
*
*
*
*
4.2.1 Collocated Quality Control
Sampler Precision Estimate for PM10,
PM2.5, and Pb. Precision is estimated via
duplicate measurements from collocated
samplers. It is recommended that the
precision be aggregated at the PQAO
level quarterly, annually, and at the 3year level. The data pair would only be
considered valid if both concentrations
are greater than or equal to the
minimum values specified in section
4(c) of this appendix. For each
collocated data pair, calculate ti, using
equation 6 to this appendix:
Equation 6 to Section 4.2.1 of Appendix B
Where Xi is the concentration from
the primary sampler and Yi is the
concentration value from the audit
sampler. The coefficient of variation
upper bound is calculated using
equation 7 to this appendix:
CV90NAAQS
= 100 *
Where k is the number of valid data
pairs being aggregated, and X20.1,k-1 is
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
k
X
If=t ff - (If=t ti) 2
2k(k -1)
k- 1
X
2
NAAQS Concentration* Xo.t,k-t
the 10th percentile of a chi-squared
distribution with k-1 degrees of
PO 00000
Frm 00193
Fmt 4701
Sfmt 4700
freedom. The factor of 2 in the
denominator adjusts for the fact that
E:\FR\FM\06MRR3.SGM
06MRR3
ER06MR24.045
I
ER06MR24.044
ddrumheller on DSK120RN23PROD with RULES3
Equation 7 to Section 4.2.1 of Appendix B
16394
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
each ti is calculated from two values
with error.
*
*
*
*
*
4.2.5 Performance Evaluation
Programs Bias Estimate for PM2.5. The
bias estimate is calculated using the PEP
audits described in section 3.2.4. of this
appendix. The bias estimator is based
on, si, the absolute difference in
concentrations divided by the square
root of the PEP concentration.
Equation 8 to Section 4.2.5 of Appendix B
100 x
*
*
*
*
,1:n s
1=1 '
n,/NAAQS concentration
*
6. References
(1) American National Standard Institute—
Quality Management Systems For
Environmental Information And
Technology Programs—Requirements
With Guidance For Use. ASQ/ANSI E4–
2014. February 2014. Available from
ANSI Webstore https://webstore.ansi.
org/.
*
*
*
*
*
(4) EPA Traceability Protocol for Assay and
Certification of Gaseous Calibration
Standards. EPA–600/R–12/531. May,
2012. Available from U.S. Environmental
Protection Agency, National Risk
Management Research Laboratory,
Research Triangle Park NC 27711.
https://www.epa.gov/nscep.
*
*
*
*
*
meas - audit
where s - - - - - ' -
(6) List of Designated Reference and
Equivalent Methods. Available from U.S.
Environmental Protection Agency,
Center for Environmental Measurements
and Modeling, Air Methods and
Characterization Division, MD–D205–03,
Research Triangle Park, NC 27711.
https://www.epa.gov/amtic/airmonitoring-methods-criteria-pollutants.
(7) Transfer Standards for the Calibration of
Ambient Air Monitoring Analyzers for
Ozone. EPA–454/B–13–004 U.S.
Environmental Protection Agency,
Research Triangle Park, NC 27711,
October, 2013. https://www.epa.gov/
sites/default/files/2020-09/documents/
ozonetransferstandardguidance.pdf.
*
*
*
*
*
(9) Quality Assurance Handbook for Air
Pollution Measurement Systems, Volume
1—A Field Guide to Environmental
Quality Assurance. EPA–600/R–94/038a.
-../audit
April 1994. Available from U.S.
Environmental Protection Agency, ORD
Publications Office, Center for
Environmental Research Information
(CERI), 26 W. Martin Luther King Drive,
Cincinnati, OH 45268. https://
www.epa.gov/amtic/ambient-airmonitoring-qualityassurance#documents.
(10) Quality Assurance Handbook for Air
Pollution Measurement Systems, Volume
II: Ambient Air Quality Monitoring
Program Quality System Development.
EPA–454/B–13–003. https://
www.epa.gov/amtic/ambient-airmonitoring-qualityassurance#documents.
(11) National Performance Evaluation
Program Standard Operating Procedures.
https://www.epa.gov/amtic/ambient-airmonitoring-quality-assurance#npep.
TABLE B–1 TO SECTION 6 OF APPENDIX B- MINIMUM DATA ASSESSMENT REQUIREMENTS FOR NAAQS RELATED CRITERIA
POLLUTANT PSD MONITORS
Gaseous Methods
(CO, NO2, SO2, O3):
One-Point QC for
SO2, NO2, O3,
CO.
Quarterly performance evaluation
for SO2, NO2,
O3, CO.
NPAP for SO2,
NO2, O3, CO3.
ddrumheller on DSK120RN23PROD with RULES3
Particulate Methods:
Collocated sampling PM10,
PM2.5, Pb.
Flow rate
verification
PM10, PM2.5,
Pb.
Semi-annual flow
rate audit PM10,
PM2.5, Pb.
VerDate Sep<11>2014
Assessment
method
Coverage
Minimum
frequency
Parameters
reported
Response check at
concentration
0.005–0.08 ppm
SO2, NO2, O3, &
0.5 and 5 ppm CO.
See section 3.1.2 of
this appendix.
Each analyzer ...........
Once per 2 weeks5 ...
Audit concentration1
and measured concentration2.
One-Point QC.
Each analyzer ...........
Once per quarter5 .....
Annual PE.
Independent Audit .....
Each primary monitor
Once per year ...........
Audit concentration1
and measured concentration2 for each
level.
Audit concentration1
and measured concentration2 for each
level.
Collocated samplers
1 per PSD Network
per pollutant.
Every 6 days or every
3 days if daily monitoring required.
No Transaction reported as raw data.
Check of sampler
flow rate.
Each sampler ............
Once every month5 ...
Check of sampler
flow rate using
independent standard.
Each sampler ............
Once every 6 months
or beginning, middle and end of
monitoring5.
Primary sampler concentration and duplicate sampler
concentration4.
Audit flow rate and
measured flow rate
indicated by the
sampler.
Audit flow rate and
measured flow rate
indicated by the
sampler.
19:03 Mar 05, 2024
Jkt 262001
PO 00000
Frm 00194
Fmt 4701
Sfmt 4700
E:\FR\FM\06MRR3.SGM
06MRR3
AQS
Assessment type
NPAP.
Flow Rate
Verification.
Semi Annual Flow
Rate Audit.
ER06MR24.046
Method
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
16395
TABLE B–1 TO SECTION 6 OF APPENDIX B- MINIMUM DATA ASSESSMENT REQUIREMENTS FOR NAAQS RELATED CRITERIA
POLLUTANT PSD MONITORS—Continued
Assessment
method
Method
Coverage
Minimum
frequency
Parameters
reported
Measured value and
audit value (ug Pb/
filter) using AQS
unit code 077 for
parameters:.
14129—Pb (TSP) LC
FRM/FEM.
85129—Pb (TSP) LC
Non-FRM/FEM..
Primary sampler concentration and performance evaluation sampler concentration.
Pb analysis audits
Pb-TSP, PbPM10.
Check of analytical
system with Pb
audit strips/filters.
Analytical ...................
Each quarter5 ............
Performance Evaluation Program
PM2.53.
Collocated samplers
Over all 4 quarters5 ..
Performance Evaluation Program
Pb 3.
Collocated samplers
(1) 5 valid audits for
PQAOs with <= 5
sites..
(2) 8 valid audits for
PQAOs with > 5
sites..
(3) All samplers in 6
years.
(1) 1 valid audit and 4
collocated samples
for PQAOs, with
<=5 sites..
(2) 2 valid audits and
6 collocated samples for PQAOs
with >5 sites..
Over all 4 quarters5 ..
Primary sampler concentration and performance evaluation sampler concentration. Primary
sampler concentration and duplicate
sampler concentration.
AQS
Assessment type
Pb Analysis Audits.
PEP.
PEP.
1
Effective concentration for open path analyzers.
Corrected concentration, if applicable for open path analyzers.
3
NPAP, PM2.5, PEP, and Pb-PEP must be implemented if data is used for NAAQS decisions otherwise implementation is at PSD reviewing
authority discretion.
4
Both primary and collocated sampler values are reported as raw data
5 A maximum number of days should be between these checks to ensure the checks are routinely conducted over time and to limit data impacts resulting from a failed check.
2
28. Amend appendix C to part 58 by:
a. Adding sections 2.2 and 2.2.1
through 2.2.19;
■ b. Removing and reserving sections
2.4, 2.4.1;
■ c. Removing sections 2.4.1.1 through
2.4.1.7; and
■ d. Revising section 2.7.1.
The additions and revision reads as
follows:
■
■
Appendix C to Part 58—Ambient Air
Quality Monitoring Methodology
ddrumheller on DSK120RN23PROD with RULES3
*
*
*
*
*
2.2 PM10, PM2.5, or PM10–2.5 continuous
FEMs with existing valid designations may
be calibrated using network data from
collocated FRM and continuous FEM data
under the following provisions:
2.2.1 Data to demonstrate a calibration
may include valid data from State, local, or
Tribal air agencies or data collected by
instrument manufacturers in accordance with
40 CFR 53.35 or other data approved by the
Administrator.
2.2.2 A request to update a designated
methods calibration may be initiated by the
instrument manufacturer of record or the
EPA Administrator. State, local, Tribal, and
multijusistincional organizations of these
entities may work with an instrument
manufacture to update a designated method
calibration.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
2.2.3 Requests for approval of an updated
PM10, PM2.5, or PM10–2.5 continuous FEM
calibration must meet the general submittal
requirements of section 2.7 of this appendix.
2.2.4 Data included in the request should
represent a subset of representative locations
where the method is operational. For cases
with a small number of collocated FRMs and
continuous FEMs sites, an updated candidate
calibration may be limited to the sites where
both methods are in use.
2.2.5 Data included in a candidate
method updated calibration may include a
subset of sites where there is a large grouping
of sites in one part of the country such that
the updated calibration would be
representative of the country as a whole.
2.2.6 Improvements should be national in
scope and ideally implemented through a
firmware change.
2.2.7 The goal of a change to a methods
calibration is to increase the number of sites
meeting measurements quality objectives of
the method as identified in section 2.3.1.1 of
appendix A to this part.
2.2.8 For meeting measurement quality
objectives (MQOs), the primary objective is to
meet the bias goal as this statistic will likely
have the most influence on improving the
resultant data collected.
2.2.9 Precision data are to be included,
but so long as precision data are at least as
good as existing network data or meet the
MQO referenced in section 2.2.8 of this
PO 00000
Frm 00195
Fmt 4701
Sfmt 4700
appendix, no further work is necessary with
precision.
2.2.10 Data available to use may include
routine primary and collocated data.
2.2.11 Audit data may be useful to
confirm the performance of a candidate
updated calibration but should not be used
as the basis of the calibration to keep the
independence of the audit data.
2.2.12 Data utilized as the basis of the
updated calibration may be obtained by
accessing EPA’s AQS database or future
analogous EPA database.
2.2.13 Years of data to use in a candidate
method calibration should include two
recent years where we are past the
certification period for the previous year’s
data, which is May 1 of each year.
2.2.14 Data from additional years is to be
used to test an updated calibration such that
the calibration is independent of the test
years of interest. Data from these additional
years need to minimally demonstrate that a
larger number of sites are expected to meet
bias MQO especially at sites near the level of
the NAAQS for the PM indicator of interest.
2.2.15 Outliers may be excluded using
routine outlier tests.
2.2.16 The range of data used in a
calibration may include all data available or
alternatively use data in the range from the
lowest measured data available up to 125%
of the 24-hour NAAQS for the PM indicator
of interest.
E:\FR\FM\06MRR3.SGM
06MRR3
16396
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
2.2.17 Other improvements to a PM
continuous method may be included as part
of a recommended update so long as
appropriate testing is conducted with input
from EPA’s Office of Research and
Development (ORD) Reference and
Equivalent (R&E) Methods Designation
program.
2.2.18 EPA encourages early
communication by instrument manufacturers
considering an update to a PM method.
Instrument companies should initiate such
dialogue by contacting EPA’s ORD R&E
Methods Designation program. The contact
information for this can be found at 40 CFR
53.4.
2.2.19 Manufacturers interested in
improving instrument’s performance through
an updated factory calibration must submit a
written modification request to EPA with
supporting rationale. Because the testing
requirements and acceptance criteria of any
field and/or lab tests can depend upon the
nature and extent of the intended
modification, applicants should contact
EPA’s R&E Methods Designation program for
guidance prior to development of the
modification request.
*
*
*
*
*
2.7.1 Requests for approval under
sections 2.2, 2.4, 2.6.2, or 2.8 of this
appendix must be submitted to: Director,
Center for Environmental Measurement and
Modeling, Reference and Equivalent Methods
Designation Program (MD–D205–03), U.S.
Environmental Protection Agency, P.O. Box
12055, Research Triangle Park, North
Carolina 27711.
29. Amend appendix D to part 58 by
revising sections 1 and 1.1(b), the
introductory text before the table in
section 4.7.1(a), and sections 4.7.1(b)(3)
and 4.7.2 to read as follows:
■
Appendix D to Part 58—Network
Design Criteria for Ambient Air Quality
Monitoring
ddrumheller on DSK120RN23PROD with RULES3
*
*
*
*
*
1. Monitoring Objectives and Spatial Scales
The purpose of this appendix is to describe
monitoring objectives and general criteria to
be applied in establishing the required
SLAMS ambient air quality monitoring
stations and for choosing general locations
for additional monitoring sites. This
appendix also describes specific
requirements for the number and location of
FRM and FEM sites for specific pollutants,
NCore multipollutant sites, PM10 mass sites,
PM2.5 mass sites, chemically-speciated PM2.5
sites, and O3 precursor measurements sites
(PAMS). These criteria will be used by EPA
in evaluating the adequacy of the air
pollutant monitoring networks.
1.1 * * *
(b) Support compliance with ambient air
quality standards and emissions strategy
development. Data from FRM and FEM
monitors for NAAQS pollutants will be used
for comparing an area’s air pollution levels
against the NAAQS. Data from monitors of
various types can be used in the development
of attainment and maintenance plans.
SLAMS, and especially NCore station data,
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
will be used to evaluate the regional air
quality models used in developing emission
strategies, and to track trends in air pollution
abatement control measures’ impact on
improving air quality. In monitoring
locations near major air pollution sources,
source-oriented monitoring data can provide
insight into how well industrial sources are
controlling their pollutant emissions.
*
*
*
*
*
4.7.1 * * *
(a) State and where applicable, local,
agencies must operate the minimum number
of required PM2.5 SLAMS sites listed in table
D–5 to this appendix. The NCore sites are
expected to complement the PM2.5 data
collection that takes place at non-NCore
SLAMS sites, and both types of sites can be
used to meet the minimum PM2.5 network
requirements. For many State and local
networks, the total number of PM2.5 sites
needed to support the basic monitoring
objectives of providing air pollution data to
the general public in a timely manner,
support compliance with ambient air quality
standards and emission strategy
development, and support for air pollution
research studies will include more sites than
the minimum numbers required in table D–
5 to this appendix. Deviations from these
PM2.5 monitoring requirements must be
approved by the EPA Regional
Administrator.
*
*
*
*
*
(b) * * *
(3) For areas with additional required
SLAMS, a monitoring station is to be sited in
an at-risk community with poor air quality,
particularly where there are anticipated
effects from sources in the area (e.g., a major
industrial area, point source(s), port, rail
yard, airport, or other transportation facility
or corridor).
*
*
*
*
*
4.7.2 Requirement for Continuous PM2.5
Monitoring. The State, or where appropriate,
local agencies must operate continuous PM2.5
analyzers equal to at least one-half (round
up) the minimum required sites listed in
table D–5 to this appendix. At least one
required continuous analyzer in each MSA
must be collocated with one of the required
FRM/FEM monitors, unless at least one of the
required FRM/FEM monitors is itself a
continuous FEM monitor in which case no
collocation requirement applies. State and
local air monitoring agencies must use
methodologies and quality assurance/quality
control (QA/QC) procedures approved by the
EPA Regional Administrator for these
required continuous analyzers.
*
*
*
*
*
30. Revise appendix E to part 58 to
read as follows:
■
Appendix E to Part 58—Probe and
Monitoring Path Siting Criteria for
Ambient Air Quality Monitoring
1. Introduction
2. Monitors and Samplers with Probe Inlets
3. Open Path Analyzers
4. Waiver Provisions
5. References
PO 00000
Frm 00196
Fmt 4701
Sfmt 4700
1. Introduction
1.1 Applicability
(a) This appendix contains specific
location criteria applicable to ambient air
quality monitoring probes, inlets, and optical
paths of SLAMS, NCore, PAMS, and other
monitor types whose data are intended to be
used to determine compliance with the
NAAQS. These specific location criteria are
relevant after the general location has been
selected based on the monitoring objectives
and spatial scale of representation discussed
in appendix D to this part. Monitor probe
material and sample residence time
requirements are also included in this
appendix. Adherence to these siting criteria
is necessary to ensure the uniform collection
of compatible and comparable air quality
data.
(b) The probe and monitoring path siting
criteria discussed in this appendix must be
followed to the maximum extent possible. It
is recognized that there may be situations
where some deviation from the siting criteria
may be necessary. In any such case, the
reasons must be thoroughly documented in a
written request for a waiver that describes
whether the resulting monitoring data will be
representative of the monitoring area and
how and why the proposed or existing siting
must deviate from the criteria. This
documentation should help to avoid later
questions about the validity of the resulting
monitoring data. Conditions under which the
EPA would consider an application for
waiver from these siting criteria are
discussed in section 4 of this appendix.
(c) The pollutant-specific probe and
monitoring path siting criteria generally
apply to all spatial scales except where noted
otherwise. Specific siting criteria that are
phrased with ‘‘shall’’ or ‘‘must’’ are defined
as requirements and exceptions must be
granted through the waiver provisions.
However, siting criteria that are phrased with
‘‘should’’ are defined as goals to meet for
consistency but are not requirements.
2. Monitors and Samplers with Probe Inlets
2.1 Horizontal and Vertical Placement
(a) For O3 and SO2 monitoring, and for
neighborhood or larger spatial scale Pb, PM10,
PM10–2.5, PM2.5, NO2, and CO sites, the probe
must be located greater than or equal to 2.0
meters and less than or equal to 15 meters
above ground level.
(b) Middle scale CO and NO2 monitors
must have sampler inlets greater than or
equal to 2.0 meters and less than or equal to
15 meters above ground level.
(c) Middle scale PM10–2.5 sites are required
to have sampler inlets greater than or equal
to 2.0 meters and less than or equal to 7.0
meters above ground level.
(d) Microscale Pb, PM10, PM10–2.5, and
PM2.5 sites are required to have sampler
inlets greater than or equal to 2.0 meters and
less than or equal to 7.0 meters above ground
level.
(e) Microscale near-road NO2 monitoring
sites are required to have sampler inlets
greater than or equal to 2.0 meters and less
than or equal to 7.0 meters above ground
level.
(f) The probe inlets for microscale carbon
monoxide monitors that are being used to
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
measure concentrations near roadways must
be greater than or equal to 2.0 meters and less
than or equal to 7.0 meters above ground
level. Those probe inlets for microscale
carbon monoxide monitors measuring
concentrations near roadways in downtown
areas or urban street canyons must be greater
than or equal to 2.5 meters and less than or
equal to 3.5 meters above ground level. The
probe must be at least 1.0 meter vertically or
horizontally away from any supporting
structure, walls, parapets, penthouses, etc.,
and away from dusty or dirty areas. If the
probe is located near the side of a building
or wall, then it should be located on the
windward side of the building relative to the
prevailing wind direction during the season
of highest concentration potential for the
pollutant being measured.
ddrumheller on DSK120RN23PROD with RULES3
2.2 Spacing From Minor Sources
(a) It is important to understand the
monitoring objective for a particular site in
order to interpret this requirement. Local
minor sources of a primary pollutant, such as
SO2, lead, or particles, can cause high
concentrations of that particular pollutant at
a monitoring site. If the objective for that
monitoring site is to investigate these local
primary pollutant emissions, then the site
will likely be properly located nearby. This
type of monitoring site would, in all
likelihood, be a microscale-type of
monitoring site. If a monitoring site is to be
used to determine air quality over a much
larger area, such as a neighborhood or city,
a monitoring agency should avoid placing a
monitor probe inlet near local, minor
sources, because a plume from a local minor
source should not be allowed to
inappropriately impact the air quality data
collected at a site. Particulate matter sites
should not be located in an unpaved area
unless there is vegetative ground cover yearround, so that the impact of windblown dusts
will be kept to a minimum.
(b) Similarly, local sources of nitric oxide
(NO) and ozone-reactive hydrocarbons can
have a scavenging effect causing
unrepresentatively low concentrations of O3
in the vicinity of probes for O3. To minimize
these potential interferences from nearby
minor sources, the probe inlet should be
placed at a distance from furnace or
incineration flues or other minor sources of
SO2 or NO. The separation distance should
take into account the heights of the flues,
type of waste or fuel burned, and the sulfur
content of the fuel.
2.3 Spacing From Obstructions
(a) Obstacles may scavenge SO2, O3, or
NO2, and can act to restrict airflow for any
pollutant. To avoid this interference, the
probe inlet must have unrestricted airflow
pursuant to paragraph (b) of this section and
should be located at a distance from
obstacles. The horizontal distance from the
obstacle to the probe inlet must be at least
twice the height that the obstacle protrudes
above the probe inlet. An obstacle that does
not meet the minimum distance requirement
is considered an obstruction that restricts
airflow to the probe inlet. The EPA does not
generally consider objects or obstacles such
as flag poles or site towers used for NOy
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
convertors and meteorological sensors, etc. to
be deemed obstructions.
(b) A probe inlet located near or along a
vertical wall is undesirable because air
moving along the wall may be subject to
removal mechanisms. A probe inlet must
have unrestricted airflow with no
obstructions (as defined in paragraph (a) of
this section) in a continuous arc of at least
270 degrees. An unobstructed continuous arc
of 180 degrees is allowable when the
applicable network design criteria specified
in appendix D of this part require monitoring
in street canyons and the probe is located on
the side of a building. This arc must include
the predominant wind direction for the
season of greatest pollutant concentration
potential. For particle sampling, there must
be a minimum of 2.0 meters of horizontal
separation from walls, parapets, and
structures for rooftop site placement.
(c) A sampling station with a probe inlet
located closer to an obstacle than required by
the criteria in this section should be
classified as middle scale or microscale,
rather than neighborhood or urban scale,
since the measurements from such a station
would more closely represent these smaller
scales.
(d) For near-road monitoring stations, the
monitor probe shall have an unobstructed air
flow, where no obstacles exist at or above the
height of the monitor probe, between the
monitor probe and the outside nearest edge
of the traffic lanes of the target road segment.
2.4
Spacing From Trees
(a) Trees can provide surfaces for SO2, O3,
or NO2 adsorption or reactions and surfaces
for particle deposition. Trees can also act as
obstructions in locations where the trees are
between the air pollutant sources or source
areas and the monitoring site and where the
trees are of a sufficient height and leaf
canopy density to interfere with the normal
airflow around the probe inlet. To reduce this
possible interference/obstruction, the probe
inlet should be 20 meters or more from the
drip line of trees and must be at least 10
meters from the drip line of trees. If a tree
or group of trees is an obstacle, the probe
inlet must meet the distance requirements of
section 2.3 of this appendix.
(b) The scavenging effect of trees is greater
for O3 than for other criteria pollutants.
Monitoring agencies must take steps to
consider the impact of trees on ozone
monitoring sites and take steps to avoid this
problem.
(c) Beginning January 1, 2024, microscale
sites of any air pollutant shall have no trees
or shrubs located at or above the line-of-sight
fetch between the probe and the source under
investigation, e.g., a roadway or a stationary
source.
PO 00000
Frm 00197
Fmt 4701
Sfmt 4700
2.5
16397
Spacing From Roadways
TABLE E–1 TO SECTION 2.5 OF APPENDIX E—MINIMUM SEPARATION
DISTANCE BETWEEN ROADWAYS AND
PROBES FOR MONITORING NEIGHBORHOOD
AND
URBAN SCALE
OZONE (O3) AND OXIDES OF NITROGEN (NO, NO2, NOX, NOy)
Roadway
average daily
traffic,
vehicles per day
≤1,000 ...............
10,000 ...............
15,000 ...............
20,000 ...............
40,000 ...............
70,000 ...............
≥110,000 ...........
Minimum
distance1 3
(meters)
10
10
20
30
50
100
250
Minimum
distance1 2 3
(meters)
10
20
30
40
60
100
250
1 Distance from the edge of the nearest traffic lane. The distance for intermediate traffic
counts should be interpolated from the table
values based on the actual traffic count./
TNOTE≤
2 Applicable
for ozone monitors whose
placement was not approved as of December
18, 2006.
3 All distances listed are expressed as having 2 significant figures. When rounding is performed to assess compliance with these siting
requirements, the distance measurements will
be rounded such as to retain at least two significant figures.
2.5.1 Spacing for Ozone Probes
In siting an O3 monitor, it is important to
minimize destructive interferences from
sources of NO, since NO readily reacts with
O3. Table E–1 of this appendix provides the
required minimum separation distances
between a roadway and a probe inlet for
various ranges of daily roadway traffic. A
sampling site with a monitor probe located
closer to a roadway than allowed by the
Table E–1 requirements should be classified
as middle scale or microscale, rather than
neighborhood or urban scale, since the
measurements from such a site would more
closely represent these smaller scales.
2.5.2 Spacing for Carbon Monoxide Probes
(a) Near-road microscale CO monitoring
sites, including those located in downtown
areas, urban street canyons, and other nearroad locations such as those adjacent to
highly trafficked roads, are intended to
provide a measurement of the influence of
the immediate source on the pollution
exposure on the adjacent area.
(b) Microscale CO monitor probe inlets in
downtown areas or urban street canyon
locations shall be located a minimum
distance of 2.0 meters and a maximum
distance of 10 meters from the edge of the
nearest traffic lane.
(c) Microscale CO monitor probe inlets in
downtown areas or urban street canyon
locations shall be located at least 10 meters
from an intersection, preferably at a midblock
location. Midblock locations are preferable to
intersection locations because intersections
represent a much smaller portion of
downtown space than do the streets between
E:\FR\FM\06MRR3.SGM
06MRR3
16398
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
them. Pedestrian exposure is probably also
greater in street canyon/corridors than at
intersections.
(d) Neighborhood scale CO monitor probe
inlets in downtown areas or urban street
canyon locations shall be located according
to the requirements in Table E–2 of this
appendix.
2 All distances listed are expressed as having 2 significant figures. When rounding is performed to assess compliance with these siting
requirements, the distance measurements will
be rounded such as to retain at least two significant figures.
TABLE E–2 TO SECTION 2.5.2 OF APPENDIX E—MINIMUM SEPARATION
DISTANCE BETWEEN ROADWAYS AND
PROBES FOR MONITORING NEIGHBORHOOD SCALE CARBON MONOXIDE
Roadway average
daily traffic,
vehicles per day
Minimum distance 1
(meters)
≤10,000 .................
15,000 ...................
20,000 ...................
30,000 ...................
40,000 ...................
50,000 ...................
≥60,000 .................
2
10
25
45
80
115
135
150
1 Distance from the edge of the nearest traffic lane. The distance for intermediate traffic
counts should be interpolated from the table
values based on the actual traffic count.
2.5.3 Spacing for Particulate Matter (PM2.5,
PM2.5–10, PM10, Pb) Inlets
(a) Since emissions associated with the
operation of motor vehicles contribute to
urban area particulate matter ambient levels,
spacing from roadway criteria are necessary
for ensuring national consistency in PM
sampler siting.
(b) The intent is to locate localized hot-spot
sites in areas of highest concentrations,
whether it be caused by mobile or multiple
stationary sources. If the area is primarily
affected by mobile sources and the maximum
concentration area(s) is judged to be a traffic
corridor or street canyon location, then the
monitors should be located near roadways
with the highest traffic volume and at
separation distances most likely to produce
the highest concentrations. For microscale
traffic corridor sites, the location must be
greater than or equal 5.0 meters and less than
or equal to 15 meters from the major
roadway. For the microscale street canyon
site, the location must be greater than or
equal 2.0 meters and less than or equal to 10
meters from the roadway. For the middle
scale site, a range of acceptable distances
from the roadway is shown in Figure E–1 of
this appendix. This figure also includes
separation distances between a roadway and
neighborhood or larger scale sites by default.
Any PM probe inlet at a site, 2.0 to 15 meters
high, and further back than the middle scale
requirements will generally be neighborhood,
urban or regional scale. For example,
according to Figure E–1 of this appendix, if
a PM sampler is primarily influenced by
roadway emissions and that sampler is set
back 10 meters from a 30,000 ADT (average
daily traffic) road, the site should be
classified as microscale, if the sampler’s inlet
height is between 2.0 and 7.0 meters. If the
sampler’s inlet height is between 7.0 and 15
meters, the site should be classified as
middle scale. If the sampler is 20 meters from
the same road, it will be classified as middle
scale; if 40 meters, neighborhood scale; and
if 110 meters, an urban scale.
100
0
0
0
...
1
f
)(
II)
'0
cu
0
0::
C>
C
ti
:!
<
JI!
s
.5
E
J
60
I
Cl)~
CD
Neighborhood Scale
~
t i~
C
::,
0
Cl)
.!/!
r--. £
oO
;'t .... Cl)
;; N C
40
Middle Scale
.c:
80
Cl)
(f)
Cl)
"O
"O
-
I-
Urban Scale
C
<
15
0
15
0 5
70
35
140
105
Figure E-1. Distance of PM Samplers to nearest traffic lane (meters)
2.5.4 Spacing for Nitrogen Dioxide (NO2)
Probes
(a) In siting near-road NO2 monitors as
required in section 4.3.2 of appendix D of
this part, the monitor probe shall be as near
as practicable to the outside nearest edge of
the traffic lanes of the target road segment but
shall not be located at a distance greater than
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
50 meters, in the horizontal, from the outside
nearest edge of the traffic lanes of the target
road segment. Where possible, the near-road
NO2 monitor probe should be within 20
meters of the target road segment.
(b) In siting NO2 monitors for
neighborhood and larger scale monitoring, it
is important to minimize near-road
PO 00000
Frm 00198
Fmt 4701
Sfmt 4700
influences. Table E–1 of this appendix
provides the required minimum separation
distances between a roadway and a probe
inlet for various ranges of daily roadway
traffic. A site with a monitor probe located
closer to a roadway than allowed by the
Table E–1 requirements should be classified
E:\FR\FM\06MRR3.SGM
06MRR3
ER06MR24.047
ddrumheller on DSK120RN23PROD with RULES3
Notes: Microscale street canyon sites must reside between 2 and 10 meters from the roadway.
Near-Road sites must be within 50 meters of the roadway.
The slopes of the lines between monitoring scales are one to one.
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
as microscale or middle scale rather than
neighborhood or urban scale.
2.6 Probe Material and Pollutant Sampler
Residence Time
(a) For the reactive gases (SO2, NO2, and
O3), approved probe materials must be used
for monitors. Studies25 34 have been
conducted to determine the suitability of
materials such as polypropylene,
polyethylene, polyvinyl chloride, Tygon®,
aluminum, brass, stainless steel, copper,
borosilicate glass, polyvinylidene fluoride
(PVDF), polytetrafluoroethylene (PTFE),
perfluoroalkoxy (PFA), and fluorinated
ethylene propylene (FEP) for use as intake
sampling lines. Of the above materials, only
borosilicate glass, PVDF, PTFE, PFA, and
FEP have been found to be acceptable for use
as intake sampling lines for all the reactive
gaseous pollutants. Furthermore, the EPA 25
has specified borosilicate glass, FEP Teflon®,
or their equivalents as the only acceptable
probe materials for delivering test
atmospheres in the determination of
reference or equivalent methods. Therefore,
borosilicate glass, PVDF, PTFE, PFA, FEP, or
their equivalents must be the only material
in the sampling train (from probe inlet to the
back of the monitor) that can be in contact
with the ambient air sample for reactive gas
monitors. NafionTM, which is composed
primarily of PTFE, can be considered
equivalent to PTFE; it has been shown in
tests to exhibit virtually no loss of ozone at
20-second residence times.35
(b) For volatile organic compound (VOC)
monitoring at PAMS, FEP Teflon® is
unacceptable as the probe material because of
VOC adsorption and desorption reactions on
the FEP Teflon®. Borosilicate glass, stainless
steel, or their equivalents are the acceptable
probe materials for VOC and carbonyl
sampling. Care must be taken to ensure that
the sample residence time is kept to 20
seconds or less.
(c) No matter how nonreactive the
sampling probe material is initially, after a
period of use, reactive particulate matter is
deposited on the probe walls. Therefore, the
time it takes the gas to transfer from the
probe inlet to the sampling device is critical.
Ozone in the presence of nitrogen oxide (NO)
will show significant losses, even in the most
inert probe material, when the residence time
exceeds 20 seconds.26 Other
16399
studies 27 28indicate that a 10-second or less
residence time is easily achievable.
Therefore, sampling probes for all reactive
gas monitors for SO2, NO2, and O3 must have
a sample residence time less than 20 seconds.
2.7
Summary
Table E–3 of this appendix presents a
summary of the general requirements for
probe siting criteria with respect to distances
and heights. Table E–3 requires different
elevation distances above the ground for the
various pollutants. The discussion in this
appendix for each of the pollutants describes
reasons for elevating the monitor or probe
inlet. The differences in the specified range
of heights are based on the vertical
concentration gradients. For source oriented
and near-road monitors, the gradients in the
vertical direction are very large for the
microscale, so a small range of heights are
used. The upper limit of 15 meters is
specified for the consistency between
pollutants and to allow the use of a single
manifold for monitoring more than one
pollutant.
TABLE E–3 TO SECTION 2.7 OF APPENDIX E—SUMMARY OF PROBE SITING CRITERIA
Pollutant
Scale 9
SO22 3 4 5 .............................................
Middle, Neighborhood, Urban, and
Regional.
Micro [downtown or street canyon
sites].
Micro [Near-Road sites] ......................
Middle and Neighborhood ..................
CO3 4 6 ................................................
O32 3 4 ..................................................
NO22 3 4 ...............................................
PAMS2 3 4 Ozone precursors ..............
PM, Pb 2 3 4 7 .......................................
Height
from
ground to
probe 8
(meters)
Horizontal or
vertical distance
from supporting
structures 1 8 to
probe inlet (meters)
Distance
from drip
line of
trees to
probe 8
(meters)
2.0–15
≥1.0
≥10
Middle, Neighborhood, Urban, and
Regional.
Micro ...................................................
Middle, Neighborhood, Urban, and
Regional.
Neighborhood and Urban ...................
Micro ...................................................
Middle, Neighborhood, Urban and
Regional.
2.5–3.5
Distance from roadways to probe 8
(meters)
N/A.
2.0–7.0
2.0–15
≥1.0
≥10
2.0–15
≥1.0
≥10
2.0–10 for downtown areas or street
canyon microscale.
≤50 for near-road microscale.
See Table E–2 of this appendix for
middle and neighborhood scales.
See Table E–1.
2.0–7.0
2.0–15
≥1.0
≥10
≤50 for near-road micro-scale.
See Table E–1.
≥1.0
≥10
See Table E–1.
≥2.0 (horizontal
distance only)
≥10
See Figure E–1.
2.0–15
2.0–7.0
2.0–15
N/A—Not applicable.
1 When a probe is located on a rooftop, this separation distance is in reference to walls, parapets, or penthouses located on the roof.
2 Should be greater than 20 meters from the dripline of tree(s) and must be 10 meters from the dripline.
3 Distance from sampler or probe inlet to obstacle, such as a building, must be at least twice the height the obstacle protrudes above the sampler or probe inlet.
Sites not meeting this criterion may be classified as microscale or middle scale (see paragraphs 2.3(a) and 2.3(c)).
4 Must have unrestricted airflow in a continuous arc of at least 270 degrees around the probe or sampler; 180 degrees if the probe is on the side of a building or a
wall for street canyon monitoring.
5 The probe or sampler should be away from minor sources, such as furnace or incineration flues. The separation distance is dependent on the height of the minor
source emission point(s), the type of fuel or waste burned, and the quality of the fuel (sulfur, ash, or lead content). This criterion is designed to avoid undue influences
from minor sources.
6 For microscale CO monitoring sites, the probe must be ≥10 meters from a street intersection and preferably at a midblock location.
7 Collocated monitor inlets must be within 4.0 meters of each other and at least 2.0 meters apart for flow rates greater than 200 liters/min or at least 1.0 meter apart
for samplers having flow rates less than 200 liters/min to preclude airflow interference, unless a waiver has been granted by the Regional Administrator pursuant to
paragraph 3.3.4.2(c) of appendix A of part 58. For PM2.5, collocated monitor inlet heights should be within 1.0 meter of each other vertically.
8 All distances listed are expressed as having 2 significant figures. When rounding is performed to assess compliance with these siting requirements, the distance
measurements will be rounded such as to retain at least two significant figures.
9 See section 1.2 of appendix D for definitions of monitoring scales.
ddrumheller on DSK120RN23PROD with RULES3
3.
3.1
Open Path Analyzers
Horizontal and Vertical Placement
(a) For all O3 and SO2 monitoring sites and
for neighborhood or larger spatial scale NO2,
and CO sites, at least 80 percent of the
monitoring path must be located greater than
or equal 2.0 meters and less than or equal to
15 meters above ground level.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
(b) Middle scale CO and NO2 sites must
have monitoring paths greater than or equal
2.0 meters and less than or equal to 15 meters
above ground level.
(c) Microscale near-road monitoring sites
are required to have monitoring paths greater
than or equal 2.0 meters and less than or
equal to 7.0 meters above ground level.
(d) For microscale carbon monoxide
monitors that are being used to measure
PO 00000
Frm 00199
Fmt 4701
Sfmt 4700
concentrations near roadways, the
monitoring path must be greater than or
equal 2.0 meters and less than or equal to 7.0
meters above ground level. If the microscale
carbon monoxide monitors measuring
concentrations near roadways are in
downtown areas or urban street canyons, the
monitoring path must be greater than or
equal 2.5 meters and less than or equal to 3.5
meters above ground level and at least 90
E:\FR\FM\06MRR3.SGM
06MRR3
16400
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
percent of the monitoring path must be at
least 1.0 meter vertically or horizontally
away from any supporting structure, walls,
parapets, penthouses, etc., and away from
dusty or dirty areas. If a significant portion
of the monitoring path is located near the
side of a building or wall, then it should be
located on the windward side of the building
relative to the prevailing wind direction
during the season of highest concentration
potential for the pollutant being measured.
ddrumheller on DSK120RN23PROD with RULES3
3.2 Spacing From Minor Sources
(a) It is important to understand the
monitoring objective for a particular site in
order to interpret this requirement. Local
minor sources of a primary pollutant, such as
SO2 can cause high concentrations of that
particular pollutant at a monitoring site. If
the objective for that monitoring site is to
investigate these local primary pollutant
emissions, then the site will likely be
properly located nearby. This type of
monitoring site would, in all likelihood, be
a microscale type of monitoring site. If a
monitoring site is to be used to determine air
quality over a much larger area, such as a
neighborhood or city, a monitoring agency
should avoid placing a monitoring path near
local, minor sources, because a plume from
a local minor source should not be allowed
to inappropriately impact the air quality data
collected at a site.
(b) Similarly, local sources of nitric oxide
(NO) and ozone-reactive hydrocarbons can
have a scavenging effect causing
unrepresentatively low concentrations of O3
in the vicinity of monitoring paths for O3. To
minimize these potential interferences from
nearby minor sources, at least 90 percent of
the monitoring path should be at a distance
from furnace or incineration flues or other
minor sources of SO2 or NO. The separation
distance should take into account the heights
of the flues, type of waste or fuel burned, and
the sulfur content of the fuel.
3.3 Spacing From Obstructions
(a) Obstacles may scavenge SO2, O3, or
NO2, and can act to restrict airflow for any
pollutant. To avoid this interference, at least
90 percent of the monitoring path must have
unrestricted airflow and should be located at
a distance from obstacles. The horizontal
distance from the obstacle to the monitoring
path must be at least twice the height that the
obstacle protrudes above the monitoring
path. An obstacle that does not meet the
minimum distance requirement is considered
an obstruction that restricts airflow to the
monitoring path. The EPA does not generally
consider objects or obstacles such as flag
poles or site towers used for NOy convertors
and meteorological sensors, etc. to be deemed
obstructions.
(b) A monitoring path located near or along
a vertical wall is undesirable because air
moving along the wall may be subject to
removal mechanisms. At least 90 percent of
the monitoring path for open path analyzers
must have unrestricted airflow with no
obstructions (as defined in paragraph (a) of
this section) in a continuous arc of at least
270 degrees. An unobstructed continuous arc
of 180 degrees is allowable when the
applicable network design criteria specified
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
in appendix D of this part require monitoring
in street canyons and the monitoring path is
located on the side of a building. This arc
must include the predominant wind
direction for the season of greatest pollutant
concentration potential.
(c) Special consideration must be given to
the use of open path analyzers given their
inherent potential sensitivity to certain types
of interferences and optical obstructions. A
monitoring path must be clear of all trees,
brush, buildings, plumes, dust, or other
optical obstructions, including potential
obstructions that may move due to wind,
human activity, growth of vegetation, etc.
Temporary optical obstructions, such as rain,
particles, fog, or snow, should be considered
when siting an open path analyzer. Any of
these temporary obstructions that are of
sufficient density to obscure the light beam
will negatively affect the ability of the open
path analyzer to continuously measure
pollutant concentrations. Transient, but
significant obscuration of especially longer
measurement paths, could occur as a result
of certain meteorological conditions (e.g.,
heavy fog, rain, snow) and/or aerosol levels
that are of a sufficient density to prevent the
open path analyzer’s light transmission. If
certain compensating measures are not
otherwise implemented at the onset of
monitoring (e.g., shorter path lengths, higher
light source intensity), data recovery during
periods of greatest primary pollutant
potential could be compromised. For
instance, if heavy fog or high particulate
levels are coincident with periods of
projected NAAQS-threatening pollutant
potential, the representativeness of the
resulting data record in reflecting maximum
pollution concentrations may be
substantially impaired despite the fact that
the site may otherwise exhibit an acceptable,
even exceedingly high, overall valid data
capture rate.
(d) A sampling station with a monitoring
path located closer to an obstacle than
required by the criteria in this section should
be classified as middle scale or microscale,
rather than neighborhood or urban scale,
since the measurements from such a station
would more closely represent these smaller
scales.
(e) For near-road monitoring stations, the
monitoring path shall have an unobstructed
air flow, where no obstacles exist at or above
the height of the monitoring path, between
the monitoring path and the outside nearest
edge of the traffic lanes of the target road
segment.
3.4 Spacing From Trees
(a) Trees can provide surfaces for SO2, O3,
or NO2 adsorption or reactions. Trees can
also act as obstructions in locations where
the trees are located between the air pollutant
sources or source areas and the monitoring
site, and where the trees are of a sufficient
height and leaf canopy density to interfere
with the normal airflow around the
monitoring path. To reduce this possible
interference/obstruction, at least 90 percent
of the monitoring path should be 20 meters
or more from the drip line of trees and must
be at least 10 meters from the drip line of
trees. If a tree or group of trees could be
PO 00000
Frm 00200
Fmt 4701
Sfmt 4700
considered an obstacle, the monitoring path
must meet the distance requirements of
section 3.3 of this appendix.
(b) The scavenging effect of trees is greater
for O3 than for other criteria pollutants.
Monitoring agencies must take steps to
consider the impact of trees on ozone
monitoring sites and take steps to avoid this
problem.
(c) Beginning January 1, 2024, microscale
sites of any air pollutant shall have no trees
or shrubs located at or above the line-of-sight
fetch between the monitoring path and the
source under investigation, e.g., a roadway or
a stationary source.
3.5 Spacing from Roadways
TABLE E–4 OF SECTION 3.5 OF APPENDIX E—MINIMUM SEPARATION
DISTANCE BETWEEN ROADWAYS AND
MONITORING PATHS FOR MONITORING NEIGHBORHOOD AND URBAN
SCALE OZONE (O3) AND OXIDES OF
NITROGEN (NO, NO2, NOX, NOy)
Roadway
average daily traffic,
vehicles per day
≤1,000 .......................
10,000 .......................
15,000 .......................
20,000 .......................
40,000 .......................
70,000 .......................
≥110,000 ...................
Minimum
distance 1 3
(meters)
Minimum
distance 1 2 3
(meters)
10
10
20
30
50
100
250
10
20
30
40
60
100
250
1 Distance from the edge of the nearest traffic lane. The distance for intermediate traffic
counts should be interpolated from the table
values based on the actual traffic count.
2 Applicable for ozone open path monitors
whose placement was not approved as of December 18, 2006.
3 All distances listed are expressed as having 2 significant figures. When rounding is performed to assess compliance with these siting
requirements, the distance measurements will
be rounded such as to retain at least two significant figures.
3.5.1 Spacing for Ozone Monitoring Paths
In siting an O3 open path analyzer, it is
important to minimize destructive
interferences form sources of NO, since NO
readily reacts with O3. Table E–4 of this
appendix provides the required minimum
separation distances between a roadway and
at least 90 percent of a monitoring path for
various ranges of daily roadway traffic. A
monitoring site with a monitoring path
located closer to a roadway than allowed by
the Table E–4 requirements should be
classified as microscale or middle scale,
rather than neighborhood or urban scale,
since the measurements from such a site
would more closely represent these smaller
scales. The monitoring path(s) must not cross
over a roadway with an average daily traffic
count of 10,000 vehicles per day or more. For
locations where a monitoring path crosses a
roadway with fewer than 10,000 vehicles per
day, monitoring agencies must consider the
entire segment of the monitoring path in the
area of potential atmospheric interference
from automobile emissions. Therefore, this
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
calculation must include the length of the
monitoring path over the roadway plus any
segments of the monitoring path that lie in
the area between the roadway and minimum
separation distance, as determined from
Table E–4 of this appendix. The sum of these
distances must not be greater than 10 percent
of the total monitoring path length.
3.5.2 Spacing for Carbon Monoxide
Monitoring Paths
(a) Near-road microscale CO monitoring
sites, including those located in downtown
areas, urban street canyons, and other nearroad locations such as those adjacent to
highly trafficked roads, are intended to
provide a measurement of the influence of
the immediate source on the pollution
exposure on the adjacent area.
(b) Microscale CO monitoring paths in
downtown areas or urban street canyon
locations shall be located a minimum
distance of 2.0 meters and a maximum
distance of 10 meters from the edge of the
nearest traffic lane.
(c) Microscale CO monitoring paths in
downtown areas or urban street canyon
locations shall be located at least 10 meters
from an intersection, preferably at a midblock
location. Midblock locations are preferable to
intersection locations because intersections
represent a much smaller portion of
downtown space than do the streets between
them. Pedestrian exposure is probably also
greater in street canyon/corridors than at
intersections.
(d) Neighborhood scale CO monitoring
paths in downtown areas or urban street
canyon locations shall be located according
to the requirements in Table E–5 of this
appendix.
TABLE E–5 SECTION 3.5.2 OF APPENDIX E—MINIMUM SEPARATION DISTANCE BETWEEN ROADWAYS AND
MONITORING PATHS FOR MONITORING
NEIGHBORHOOD
SCALE
CARBON MONOXIDE
Roadway average
daily traffic,
vehicles per day
Minimum
distance 1 2
(meters)
≤10,000 .................................
15,000 ...................................
20,000 ...................................
30,000 ...................................
40,000 ...................................
10
25
45
80
115
TABLE E–5 SECTION 3.5.2 OF APPENDIX E—MINIMUM SEPARATION DISTANCE BETWEEN ROADWAYS AND
MONITORING PATHS FOR MONITORING
NEIGHBORHOOD
SCALE
CARBON MONOXIDE—Continued
Roadway average
daily traffic,
vehicles per day
Minimum
distance 1 2
(meters)
50,000 ...................................
≥60,000 .................................
135
150
1 Distance from the edge of the nearest traffic lane. The distance for intermediate traffic
counts should be interpolated from the table
values based on the actual traffic count.
2 All distances listed are expressed as having 2 significant figures. When rounding is performed to assess compliance with these siting
requirements, the distance measurements will
be rounded such as to retain at least two significant figures.
3.5.3 Spacing for Nitrogen Dioxide (NO2)
Monitoring Paths
(a) In siting near-road NO2 monitors as
required in section 4.3.2 of appendix D of
this part, the monitoring path shall be as near
as practicable to the outside nearest edge of
the traffic lanes of the target road segment but
shall not be located at a distance greater than
50 meters, in the horizontal, from the outside
nearest edge of the traffic lanes of the target
road segment.
(b) In siting NO2 open path monitors for
neighborhood and larger scale monitoring, it
is important to minimize near-road
influences. Table E–5 of this appendix
provides the required minimum separation
distances between a roadway and at least 90
percent of a monitoring path for various
ranges of daily roadway traffic. A site with
a monitoring path located closer to a roadway
than allowed by the Table E–4 requirements
should be classified as microscale or middle
scale rather than neighborhood or urban
scale. The monitoring path(s) must not cross
over a roadway with an average daily traffic
count of 10,000 vehicles per day or more. For
locations where a monitoring path crosses a
roadway with fewer than 10,000 vehicles per
day, monitoring agencies must consider the
entire segment of the monitoring path in the
area of potential atmospheric interference
form automobile emissions. Therefore, this
calculation must include the length of the
monitoring path over the roadway plus any
16401
segments of the monitoring path that lie in
the area between the roadway and minimum
separation distance, as determined from
Table E–5 of this appendix. The sum of these
distances must not be greater than 10 percent
of the total monitoring path length.
3.6 Cumulative Interferences on a
Monitoring Path
The cumulative length or portion of a
monitoring path that is affected by minor
sources, trees, or roadways must not exceed
10 percent of the total monitoring path
length.
3.7 Maximum Monitoring Path Length
The monitoring path length must not
exceed 1.0 kilometer for open path analyzers
in neighborhood, urban, or regional scale. For
middle scale monitoring sites, the monitoring
path length must not exceed 300 meters. In
areas subject to frequent periods of dust, fog,
rain, or snow, consideration should be given
to a shortened monitoring path length to
minimize loss of monitoring data due to
these temporary optical obstructions. For
certain ambient air monitoring scenarios
using open path analyzers, shorter path
lengths may be needed in order to ensure that
the monitoring site meets the objectives and
spatial scales defined in appendix D to this
part. The Regional Administrator may require
shorter path lengths, as needed on an
individual basis, to ensure that the SLAMS
sites meet the appendix D requirements.
Likewise, the Administrator may specify the
maximum path length used at NCore
monitoring sites.
3.8 Summary
Table E–6 of this appendix presents a
summary of the general requirements for
monitoring path siting criteria with respect to
distances and heights. Table E–6 requires
different elevation distances above the
ground for the various pollutants. The
discussion in this appendix for each of the
pollutants describes reasons for elevating the
monitoring path. The differences in the
specified range of heights are based on the
vertical concentration gradients. For source
oriented and near-road monitors, the
gradients in the vertical direction are very
large for the microscale, so a small range of
heights are used. The upper limit of 15
meters is specified for the consistency
between pollutants and to allow the use of
a monitoring path for monitoring more than
one pollutant.
TABLE E–6 SECTION 3.8 OF APPENDIX E—SUMMARY OF MONITORING PATH SITING CRITERIA
ddrumheller on DSK120RN23PROD with RULES3
Pollutant
SO2 3 4 5 6 ..........................................
CO4 5 7 ..............................................
VerDate Sep<11>2014
19:03 Mar 05, 2024
Maximum monitoring path
length 9 10
Height from
ground to 80%
of monitoring
path 1 8
(meters)
Horizontal or
vertical distance from
supporting
structures 2 to
90% of monitoring path 1 8
(meters)
Distance from
trees to 90%
of monitoring
path 1 8
(meters)
2.0–15
≥1.0
≥10
N/A.
2.5–3.5
≥1.0
≥10
2.0–10 for downtown areas or street
canyon microscale.
≤50 for near-road microscale.
E:\FR\FM\06MRR3.SGM
06MRR3
<= 300 m for Middle ........................
<= 1.0 km for Neighborhood, Urban,
and Regional
<= 300 m for Micro [downtown or
street canyon sites].
<= 300 m for Micro [Near-Road
sites].
Jkt 262001
PO 00000
Frm 00201
Fmt 4701
2.0–7.0
Sfmt 4700
Distance from roadways to monitoring path 1 8
(meters)
16402
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
TABLE E–6 SECTION 3.8 OF APPENDIX E—SUMMARY OF MONITORING PATH SITING CRITERIA—Continued
Pollutant
O33
4 5
..............................................
3 4 5
NO2
...........................................
PAMS3 4 5
Ozone precursors ..........
length 9 10
Height from
ground to 80%
of monitoring
path 1 8
(meters)
<= 300 m for Middle ........................
2.0–15
Maximum monitoring path
<= 1.0 km for Neighborhood.
<= 300 m for Middle.
<= 1.0 km for Neighborhood, Urban,
and Regional.
Between 50 m–300 m for Micro
(Near-Road).
<= 300 m for Middle ........................
<= 1.0 km for Neighborhood, Urban,
and Regional.
<= 1.0 km for Neighborhood and
Urban.
2.0–15
Horizontal or
vertical distance from
supporting
structures 2 to
90% of monitoring path 1 8
(meters)
Distance from
trees to 90%
of monitoring
path 1 8
(meters)
See Table E–5 of this appendix for
middle and neighborhood scales.
≥1.0
≥10
See Table E–4.
≤50 for near-road micro-scale.
2.0–7.0
≥1.0
≥10
2.0–15
2.0–15
Distance from roadways to monitoring path 1 8
(meters)
See Table E–4.
≥1.0
≥10
See Table E–4.
ddrumheller on DSK120RN23PROD with RULES3
N/A—Not applicable.
1 Monitoring path for open path analyzers is applicable only to middle or neighborhood scale CO monitoring, middle, neighborhood, urban, and regional scale NO
2
monitoring, and all applicable scales for monitoring SO2, O3, and O3 precursors.
2 When the monitoring path is located on a rooftop, this separation distance is in reference to walls, parapets, or penthouses located on roof.
3 At least 90 percent of the monitoring path should be greater than 20 meters from the dripline of tree(s) and must be 10-meters from the dripline.
4 Distance from 90 percent of monitoring path to obstacle, such as a building, must be at least twice the height the obstacle protrudes above the monitoring path.
Sites not meeting this criterion may be classified as microscale or middle scale (see text).
5 Must have unrestricted airflow 270 degrees around at least 90 percent of the monitoring path; 180 degrees if the monitoring path is adjacent to the side of a building or a wall for street canyon monitoring.
6 The monitoring path should be away from minor sources, such as furnace or incineration flues. The separation distance is dependent on the height of the minor
source’s emission point (such as a flue), the type of fuel or waste burned, and the quality of the fuel (sulfur, ash, or lead content). This criterion is designed to avoid
undue influences from minor sources.
7 For microscale CO monitoring sites, the monitoring path must be ≥10. meters from a street intersection and preferably at a midblock location.
8 All distances listed are expressed as having 2 significant figures. When rounding is performed to assess compliance with these siting requirements, the distance
measurements will be rounded such as to retain at least two significant figures.
9 See section 1.2 of appendix D for definitions of monitoring scales.
10 See section 3.7 of this appendix.
4. Waiver Provisions
Most sampling probes or monitors can be
located so that they meet the requirements of
this appendix. New sites, with rare
exceptions, can be located within the limits
of this appendix. However, some existing
sites may not meet these requirements and
may still produce useful data for some
purposes. The EPA will consider a written
request from the State, or where applicable
local, agency to waive one or more siting
criteria for some monitoring sites providing
that the State or their designee can
adequately demonstrate the need (purpose)
for monitoring or establishing a monitoring
site at that location.
4.1 For a proposed new site, a waiver
may be granted only if both the following
criteria are met:
4.1.1 The proposed new site can be
demonstrated to be as representative of the
monitoring area as it would be if the siting
criteria were being met.
4.1.2 The monitor or probe cannot
reasonably be located so as to meet the siting
criteria because of physical constraints (e.g.,
inability to locate the required type of site the
necessary distance from roadways or
obstructions).
4.2 For an existing site, a waiver may be
granted if either the criterion in section 4.1.1
or the criterion in 4.1.2 of this appendix is
met.
4.3 Cost benefits, historical trends, and
other factors may be used to add support to
the criteria in sections 4.1.1 and 4.1.2 of this
appendix; however, by themselves, they will
not be acceptable reasons for the EPA to grant
a waiver. Written requests for waivers must
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
be submitted to the Regional Administrator.
Granted waivers must be renewed minimally
every 5 years and ideally as part of the
network assessment as defined in § 58.10(d).
The approval date of the waiver must be
documented in the annual monitoring
network plan to support the requirements of
§ 58.10(a)(1) and 58.10(b)(10).
5. References
1. Bryan, R.J., R.J. Gordon, and H. Menck.
Comparison of High Volume Air Filter
Samples at Varying Distances from Los
Angeles Freeway. University of Southern
California, School of Medicine, Los Angeles,
CA. (Presented at 66th Annual Meeting of Air
Pollution Control Association. Chicago, IL.
June 24–28, 1973. APCA 73–158.)
2. Teer, E.H. Atmospheric Lead
Concentration Above an Urban Street. Master
of Science Thesis, Washington University, St.
Louis, MO. January 1971.
3. Bradway, R.M., F.A. Record, and W.E.
Belanger. Monitoring and Modeling of
Resuspended Roadway Dust Near Urban
Arterials. GCA Technology Division,
Bedford, MA. (Presented at 1978 Annual
Meeting of Transportation Research Board,
Washington, DC. January 1978.)
4. Pace, T.G., W.P. Freas, and E.M. Afify.
Quantification of Relationship Between
Monitor Height and Measured Particulate
Levels in Seven U.S. Urban Areas. U.S.
Environmental Protection Agency, Research
Triangle Park, NC. (Presented at 70th Annual
Meeting of Air Pollution Control Association,
Toronto, Canada. June 20–24, 1977. APCA
77–13.4.)
5. Harrison, P.R. Considerations for Siting
Air Quality Monitors in Urban Areas. City of
PO 00000
Frm 00202
Fmt 4701
Sfmt 4700
Chicago, Department of Environmental
Control, Chicago, IL. (Presented at 66th
Annual Meeting of Air Pollution Control
Association, Chicago, IL. June 24–28, 1973.
APCA 73–161.)
6. Study of Suspended Particulate
Measurements at Varying Heights Above
Ground. Texas State Department of Health,
Air Control Section, Austin, TX. 1970. p.7.
7. Rodes, C.E. and G.F. Evans. Summary of
LACS Integrated Pollutant Data. In: Los
Angeles Catalyst Study Symposium. U.S.
Environmental Protection Agency, Research
Triangle Park, NC. EPA Publication No. EPA–
600/4–77–034. June 1977.
8. Lynn, D.A. et al. National Assessment of
the Urban Particulate Problem: Volume 1,
National Assessment. GCA Technology
Division, Bedford, MA. U.S. Environmental
Protection Agency, Research Triangle Park,
NC. EPA Publication No. EPA–450/3–75–
024. June 1976.
9. Pace, T.G. Impact of Vehicle-Related
Particulates on TSP Concentrations and
Rationale for Siting Hi-Vols in the Vicinity of
Roadways. OAQPS, U.S. Environmental
Protection Agency, Research Triangle Park,
NC. April 1978.
10. Ludwig, F.L., J.H. Kealoha, and E.
Shelar. Selecting Sites for Monitoring Total
Suspended Particulates. Stanford Research
Institute, Menlo Park, CA. Prepared for U.S.
Environmental Protection Agency, Research
Triangle Park, NC. EPA Publication No. EPA–
450/3–77–018. June 1977, revised December
1977.
11. Ball, R.J. and G.E. Anderson. Optimum
Site Exposure Criteria for SO2 Monitoring.
The Center for the Environment and Man,
E:\FR\FM\06MRR3.SGM
06MRR3
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
Inc., Hartford, CT. Prepared for U.S.
Environmental Protection Agency, Research
Triangle Park, NC. EPA Publication No. EPA–
450/3–77–013. April 1977.
12. Ludwig, F.L. and J.H.S. Kealoha.
Selecting Sites for Carbon Monoxide
Monitoring. Stanford Research Institute,
Menlo Park, CA. Prepared for U.S.
Environmental Protection Agency, Research
Triangle Park, NC. EPA Publication No. EPA–
450/3–75–077. September 1975.
13. Ludwig, F.L. and E. Shelar. Site
Selection for the Monitoring of
Photochemical Air Pollutants. Stanford
Research Institute, Menlo Park, CA. Prepared
for U.S. Environmental Protection Agency,
Research Triangle Park, NC. EPA Publication
No. EPA–450/3–78–013. April 1978.
14. Lead Analysis for Kansas City and
Cincinnati, PEDCo Environmental, Inc.,
Cincinnati, OH. Prepared for U.S.
Environmental Protection Agency, Research
Triangle Park, NC. EPA Contract No. 66–02–
2515, June 1977.
15. Barltrap, D. and C.D. Strelow. Westway
Nursery Testing Project. Report to the Greater
London Council. August 1976.
16. Daines, R. H., H. Moto, and D. M.
Chilko. Atmospheric Lead: Its Relationship to
Traffic Volume and Proximity to Highways.
Environ. Sci. and Technol., 4:318, 1970.
17. Johnson, D. E., et al. Epidemiologic
Study of the Effects of Automobile Traffic on
Blood Lead Levels, Southwest Research
Institute, Houston, TX. Prepared for U.S.
Environmental Protection Agency, Research
Triangle Park, NC. EPA–600/1–78–055,
August 1978.
18. Air Quality Criteria for Lead. Office of
Research and Development, U.S.
Environmental Protection Agency,
Washington, DC EPA–600/8–83–028 aF–dF,
1986, and supplements EPA–600/8–89/049F,
August 1990. (NTIS document numbers
PB87–142378 and PB91–138420.)
19. Lyman, D. R. The Atmospheric
Diffusion of Carbon Monoxide and Lead from
an Expressway, Ph.D. Dissertation,
University of Cincinnati, Cincinnati, OH.
1972.
20. Wechter, S.G. Preparation of Stable
Pollutant Gas Standards Using Treated
Aluminum Cylinders. ASTM STP. 598:40–
54, 1976.
21. Wohlers, H.C., H. Newstein and D.
Daunis. Carbon Monoxide and Sulfur
Dioxide Adsorption On and Description
From Glass, Plastic and Metal Tubings. J. Air
Poll. Con. Assoc. 17:753, 1976.
22. Elfers, L.A. Field Operating Guide for
Automated Air Monitoring Equipment. U.S.
NTIS. p. 202, 249, 1971.
23. Hughes, E.E. Development of Standard
Reference Material for Air Quality
Measurement. ISA Transactions, 14:281–291,
1975.
24. Altshuller, A.D. and A.G. Wartburg.
The Interaction of Ozone with Plastic and
Metallic Materials in a Dynamic Flow
System. Intern. Jour. Air and Water Poll.,
4:70–78, 1961.
25. Code of Federal Regulations. 40 CFR
53.22, July 1976.
26. Butcher, S.S. and R.E. Ruff. Effect of
Inlet Residence Time on Analysis of
Atmospheric Nitrogen Oxides and Ozone,
Anal. Chem., 43:1890, 1971.
27. Slowik, A.A. and E.B. Sansone.
Diffusion Losses of Sulfur Dioxide in
Sampling Manifolds. J. Air. Poll. Con. Assoc.,
24:245, 1974.
28. Yamada, V.M. and R.J. Charlson. Proper
Sizing of the Sampling Inlet Line for a
Continuous Air Monitoring Station. Environ.
Sci. and Technol., 3:483, 1969.
29. Koch, R.C. and H.E. Rector. Optimum
Network Design and Site Exposure Criteria
for Particulate Matter, GEOMET
Technologies, Inc., Rockville, MD. Prepared
for U.S. Environmental Protection Agency,
Research Triangle Park, NC. EPA Contract
No. 68–02–3584. EPA 450/4–87–009. May
1987.
30. Burton, R.M. and J.C. Suggs.
Philadelphia Roadway Study. Environmental
Monitoring Systems Laboratory, U.S.
Environmental Protection Agency, Research
Triangle Park, N.C. EPA–600/4–84–070
September 1984.
31. Technical Assistance Document for
Sampling and Analysis of Ozone Precursors.
Atmospheric Research and Exposure
Assessment Laboratory, U.S. Environmental
Protection Agency, Research Triangle Park,
NC 27711. EPA 600/8–91–215. October 1991.
32. Quality Assurance Handbook for Air
Pollution Measurement Systems: Volume IV.
Meteorological Measurements. Atmospheric
Research and Exposure Assessment
Laboratory, U.S. Environmental Protection
Agency, Research Triangle Park, NC 27711.
EPA 600/4–90–0003. August 1989.
33. On-Site Meteorological Program
Guidance for Regulatory Modeling
Applications. Office of Air Quality Planning
and Standards, U.S. Environmental
Protection Agency, Research Triangle Park,
NC 27711. EPA 450/4–87–013. June 1987F.
34. Johnson, C., A. Whitehill, R. Long, and
R. Vanderpool. Investigation of Gaseous
Criteria Pollutant Transport Efficiency as a
Function of Tubing Material. U.S.
Environmental Protection Agency, Research
16403
Triangle Park, NC 27711. EPA/600/R–22/212.
August 2022.
35. Hannah Halliday, Cortina Johnson, Tad
Kleindienst, Russell Long, Robert
Vanderpool, and Andrew Whitehill.
Recommendations for Nationwide Approval
of NafionTM Dryers Upstream of UVAbsorption Ozone Analyzers. U.S.
Environmental Protection Agency, Research
Triangle Park, NC 27711. EPA/600/R–20/390.
November 2020.
31. Revise appendix G to part 58 to
read as follows:
■
Appendix G to Part 58—Uniform Air
Quality Index (AQI) and Daily
Reporting
1. General Information
2. Reporting Requirements
3. Data Handling
1. General Information
1.1 AQI Overview. The AQI is a tool that
simplifies reporting air quality to the public
in a nationally uniform and easy to
understand manner. The AQI converts
concentrations of pollutants, for which the
EPA has established a national ambient air
quality standard (NAAQS), into a uniform
scale from 0–500. These pollutants are ozone
(O3), particulate matter (PM2.5, PM10), carbon
monoxide (CO), sulfur dioxide (SO2), and
nitrogen dioxide (NO2). The scale of the
index is divided into general categories that
are associated with health messages.
2. Reporting Requirements
2.1 Applicability. The AQI must be
reported daily for a metropolitan statistical
area (MSA) with a population over 350,000.
When it is useful and possible, it is
recommended, but not required for an area to
report a sub-daily AQI as well.
2.2 Contents of AQI Report.
2.2.1 Content of AQI Report
Requirements. An AQI report must contain
the following:
a. The reporting area(s) (the MSA or
subdivision of the MSA).
b. The reporting period (the day for which
the AQI is reported).
c. The main pollutant (the pollutant with
the highest index value).
d. The AQI (the highest index value).
e. The category descriptor and index value
associated with the AQI and, if choosing to
report in a color format, the associated color.
Use only the following descriptors and colors
for the six AQI categories:
ddrumheller on DSK120RN23PROD with RULES3
TABLE 1 TO SECTION 2 OF APPENDIX G—AQI CATEGORIES
And this color 1
For this AQI
Use this descriptor
0 to 50 ................
51 to 100 ............
101 to 150 ..........
151 to 200 ..........
201 to 300 ..........
301 and above ...
‘‘Good’’ .....................................................................................
‘‘Moderate’’ ...............................................................................
‘‘Unhealthy for Sensitive Groups’’ ............................................
‘‘Unhealthy’’ ..............................................................................
‘‘Very Unhealthy’’ .....................................................................
‘‘Hazardous’’ .............................................................................
Green.
Yellow.
Orange.
Red.
Purple.
Maroon1.
1Specific color definitions can be found in the most recent reporting guidance (Technical Assistance Document for the Reporting of Daily Air
Quality), which can be found at https://www.airnow.gov/publications/air-quality-index/technical-assistance-document-for-reporting-the-daily-aqi/.
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
PO 00000
Frm 00203
Fmt 4701
Sfmt 4700
E:\FR\FM\06MRR3.SGM
06MRR3
16404
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
f. The pollutant specific sensitive groups
for any reported index value greater than 100.
The sensitive groups for each pollutant are
identified as part of the periodic review of
the air quality criteria and the NAAQS. For
convenience, the EPA lists the relevant
groups for each pollutant in the most recent
reporting guidance (Technical Assistance
Document for the Reporting of Daily Air
Quality), which can be found at https://www.
airnow.gov/publications/air-quality-index/
technical-assistance-document-for-reportingthe-daily-aqi/.
2.2.2 Contents of AQI Report When
Applicable. When appropriate, the AQI
report may also contain the following, but
such information is not required:
a. Appropriate health and cautionary
statements.
b. The name and index value for other
pollutants, particularly those with an index
value greater than 100.
c. The index values for sub-areas of your
MSA.
d. Causes for unusually high AQI values.
e. Pollutant concentrations.
f. Generally, the AQI report applies to an
area’s MSA only. However, if a significant air
quality problem exists (AQI greater than 100)
in areas significantly impacted by the MSA
but not in it (for example, O3 concentrations
are often highest downwind and outside an
urban area), the report should identify these
areas and report the AQI for these areas as
well.
2.3. Communication, Timing, and
Frequency of AQI Report. The daily AQI
must be reported 7 days per week and made
available via website or other means of
public access. The daily AQI report
represents the air quality for the previous
day. Exceptions to this requirement are in
section 2.4 of this appendix.
a. Reporting the AQI sub-daily is
recommended, but not required, to provide
more timely air quality information to the
public for making health-protective
decisions.
b. Submitting hourly data in real-time to
the EPA’s AirNow (or future analogous)
system is recommended, but not required,
and assists the EPA in providing timely air
quality information to the public for making
health-protective decisions.
c. Submitting hourly data for appropriate
monitors (referenced in section 3.2 of this
appendix) satisfies the daily AQI reporting
requirement because the AirNow system
makes daily and sub-daily AQI reports
widely available through its website and
other communication tools.
d. Forecasting the daily AQI provides
timely air quality information to the public
and is recommended but not required. Subdaily forecasts are also recommended,
especially when air quality is expected to
vary substantially throughout the day, like
during wildfires. Long-term (multi-day)
forecasts can also be made available when
useful.
2.4. Exceptions to Reporting
Requirements.
a. If the index value for a particular
pollutant remains below 50 for a season or
year, then it may be excluded from the
calculation of the AQI in section 3 of this
appendix.
b. If all index values remain below 50 for
a year, then the AQI may be reported at the
discretion of the reporting agency. In
subsequent years, if pollutant levels rise to
where the AQI would be above 50, then the
AQI must be reported as required in section
2 of this appendix.
c. As previously mentioned in section 2.3
of this appendix, submitting hourly data in
real-time from appropriate monitors
(referenced in section 3.2 of this appendix)
to the EPA’s AirNow (or future analogous)
system satisfies the daily AQI reporting
requirement.
3. Data Handling.
3.1 Relationship of AQI and pollutant
concentrations. For each pollutant, the AQI
transforms ambient concentrations to a scale
from 0 to 500. As appropriate, the AQI is
associated with the NAAQS for each
pollutant. In most cases, the index value of
100 is associated with the numerical level of
the short-term standard (i.e., averaging time
of 24-hours or less) for each pollutant. The
index value of 50 is associated with the
numerical level of the annual standard for a
pollutant, if there is one, at one-half the level
of the short-term standard for the pollutant
or at the level at which it is appropriate to
begin to provide guidance on cautionary
language. Higher categories of the index are
based on the potential for increasingly
serious health effects to occur following
exposure and increasing proportions of the
population that are likely to be affected. The
reported AQI corresponds to the pollutant
with the highest calculated AQI. For the
purposes of reporting the AQI, the subindexes for PM10 and PM2.5 are to be
considered separately. The pollutant
responsible for the highest index value (the
reported AQI) is called the ‘‘main’’ pollutant
for that day.
3.2 Monitors Used for AQI Reporting.
Concentration data from State/Local Air
Monitoring Station (SLAMS) or parts of the
SLAMS required by 40 CFR 58.10 must be
used for each pollutant except PM. For PM,
calculate and report the AQI on days for
which air quality data has been measured
(e.g., from continuous PM2.5 monitors
required in appendix D to this part). PM
measurements may be used from monitors
that are not reference or equivalent methods
(for example, continuous PM10 or PM2.5
monitors). Detailed guidance for relating nonapproved measurements to approved
methods by statistical linear regression is
referenced here:
Reference for relating non-approved PM
measurements to approved methods (Eberly,
S., T. Fitz-Simons, T. Hanley, L. Weinstock.,
T. Tamanini, G. Denniston, B. Lambeth, E.
Michel, S. Bortnick. Data Quality Objectives
(DQOs) For Relating Federal Reference
Method (FRM) and Continuous PM2.5
Measurements to Report an Air Quality Index
(AQI). U.S. Environmental Protection
Agency, Research Triangle Park, NC. EPA–
454/B–02–002, November 2002).
3.3 AQI Forecast. The AQI can be
forecasted at least 24-hours in advance using
the most accurate and reasonable procedures
considering meteorology, topography,
availability of data, and forecasting expertise.
The guidance document, ‘‘Guidelines for
Developing an Air Quality (Ozone and PM2.5)
Forecasting Program,’’ can be found at
https://www.airnow.gov/publications/
weathercasters/guidelines-developing-airquality-forecasting-program/.
3.4 Calculation and Equations.
a. The AQI is the highest value calculated
for each pollutant as follows:
i. Identify the highest concentration among
all of the monitors within each reporting area
and truncate as follows:
(A) Ozone—truncate to 3 decimal places
PM2.5—truncate to 1 decimal place
PM10—truncate to integer
CO—truncate to 1 decimal place
SO2—truncate to integer
NO2—truncate to integer
(B) [Reserved]
ii. Using table 2 to this appendix, find the
two breakpoints that contain the
concentration.
iii. Using equation 1 to this appendix,
calculate the index.
iv. Round the index to the nearest integer.
TABLE 2 TO SECTION 3.4 OF APPENDIX G—BREAKPOINTS FOR THE AQI
ddrumheller on DSK120RN23PROD with RULES3
These breakpoints
PM2.5 (μg/m3)
24-hour
PM10 (μg/m3)
24-hour
Equal these AQI’s
SO2
(ppb)
1-hour
O3 (ppm) 1hour1
0.000–0.054 .....
0.055–0.070 .....
0.071–0.085 .....
........................
........................
0.125–0.164
0.0–9.0
9.1–35.4
35.5–55.4
0–54
55–154
155–254
0.0–4.4
4.5–9.4
9.5–12.4
0–35
36–75
76–185
0–53
54–100
101–360
0–50
51–100
101–150
0.086–0.105 .....
0.106–0.200 .....
0.165–0.204
0.205–0.404
55.5–125.4
125.5—225.4
255–354
355–424
12.5–15.4
15.5–30.4
3 186–304
361–649
650–1249
151–200
201–300
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
PO 00000
Frm 00204
CO
(ppm) 8-hour
NO2
(ppb)
1-hour
O3 (ppm) 8-hour
Fmt 4701
3 305–604
Sfmt 4700
E:\FR\FM\06MRR3.SGM
AQI
06MRR3
Category
Good.
Moderate.
Unhealthy for Sensitive
Groups.
Unhealthy.
Very Unhealthy.
16405
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
TABLE 2 TO SECTION 3.4 OF APPENDIX G—BREAKPOINTS FOR THE AQI—Continued
These breakpoints
O3 (ppm) 8-hour
0.201¥(2) ........
O3 (ppm) 1hour1
0.405+
PM2.5 (μg/m3)
24-hour
PM10 (μg/m3)
24-hour
225.5+
425+
Equal these AQI’s
NO2
(ppb)
1-hour
SO2
(ppb)
1-hour
CO
(ppm) 8-hour
30.5+
3 605+
AQI
1250+
Category
301+
4
Hazardous.
1 Areas
are generally required to report the AQI based on 8-hour ozone values. However, there are a small number of areas where an AQI based on 1-hour ozone
values would be more precautionary. In these cases, in addition to calculating the 8-hour ozone index value, the 1-hour ozone index value may be calculated, and the
maximum of the two values reported.
2 8-hour O concentrations do not define higher AQI values (≤301). AQI values > 301 are calculated with 1-hour O concentrations.
3
3
3 1-hr SO concentrations do not define higher AQI values (≥200). AQI values of 200 or greater are calculated with 24-hour SO concentration.
2
2
4 AQI values between breakpoints are calculated using equation 1 to this appendix. For AQI values in the hazardous category, AQI values greater than 500 should
be calculated using equation 1 and the concentration specified for the AQI value of 500. The AQI value of 500 are as follows: O3 1-hour—0.604 ppm; PM2.5 24hour—325.4 μg/m3; PM10 24-hour—604 μg/m3; CO ppm—50.4 ppm; SO2 1-hour—1004 ppb; and NO2 1-hour—2049 ppb.
b. If the concentration is equal to a
breakpoint, then the index is equal to the
corresponding index value in table 2 to this
appendix. However, equation 1 to this
appendix can still be used. The results will
be equal. If the concentration is between two
breakpoints, then calculate the index of that
pollutant with equation 1. It should also be
noted that in some areas, the AQI based on
1-hour O3 will be more precautionary than
using 8-hour values (see footnote 1 to table
2). In these cases, the 1-hour values as well
as 8-hour values may be used to calculate
index values and then use the maximum
index value as the AQI for O3.
Equation 1 to Appendix G to Part 58
Where:
Ip = the index value for pollutantp.
Cp = the truncated concentration of
pollutantp.
BPHi = the breakpoint that is greater than or
equal to Cp.
BPLo = the breakpoint that is less than or
equal to Cp.
IHi = the AQI value corresponding to BPHi.
Ilo = the AQI value corresponding to BPLo.
c. If the concentration is larger than the
highest breakpoint in table 2 to this appendix
then the last two breakpoints in table 2 may
be used when equation 1 to this appendix is
applied.
Example:
d. Using table 2 and equation 1 to this
appendix, calculate the index value for each
of the pollutants measured and select the one
that produces the highest index value for the
AQI. For example, if a PM10 value of 210 mg/
m3 is observed, a 1-hour O3 value of 0.156
ppm, and an 8-hour O3 value of 0.130 ppm,
then do this:
i. Find the breakpoints for PM10 at 210 mg/
m3 as 155 mg/m3 and 254 mg/m3,
corresponding to index values 101 and 150;
ii. Find the breakpoints for 1-hour O3 at
0.156 ppm as 0.125 ppm and 0.164 ppm,
corresponding to index values 101 and 150;
iii. Find the breakpoints for 8-hour O3 at
0.130 ppm as 0.116 ppm and 0.374 ppm,
corresponding to index values 201 and 300;
iv. Apply equation 21 to this appendix for
210 mg/m3, PM10:
Equation 2 to Appendix G to Part 58
150 - 101
254 _ 155 (210 - 155) + 101
= 128
v. Apply equation 3 to this appendix for
0.156 ppm, 1-hour O3:
ER06MR24.050
= 140
ER06MR24.049
vi. Apply equation 4 to this appendix for
0.130 ppm, 8-hour O3:
Equation 4 to Appendix G to Part 58
300 - 201
0.374 _ 0.116 (0.130 - 0.116) + 201
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
PO 00000
Frm 00205
Fmt 4701
Sfmt 4725
= 206
E:\FR\FM\06MRR3.SGM
06MRR3
ER06MR24.048
ddrumheller on DSK120RN23PROD with RULES3
150 - 101
0.164 _ 0.125 (0.156 - 0.125) + 101
ER06MR24.051
Equation 3 to Appendix G to Part 58
16406
Federal Register / Vol. 89, No. 45 / Wednesday, March 6, 2024 / Rules and Regulations
vii. Find the maximum, 206. This is the
AQI. A minimal AQI report could read:
‘‘Today, the AQI for my city is 206, which
is Very Unhealthy, due to ozone.’’ It would
then reference the associated sensitive
groups.
[FR Doc. 2024–02637 Filed 3–5–24; 8:45 am]
ddrumheller on DSK120RN23PROD with RULES3
BILLING CODE 6560–50–P
VerDate Sep<11>2014
19:03 Mar 05, 2024
Jkt 262001
PO 00000
Frm 00206
Fmt 4701
Sfmt 9990
E:\FR\FM\06MRR3.SGM
06MRR3
Agencies
[Federal Register Volume 89, Number 45 (Wednesday, March 6, 2024)]
[Rules and Regulations]
[Pages 16202-16406]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 2024-02637]
[[Page 16201]]
Vol. 89
Wednesday,
No. 45
March 6, 2024
Part III
Environmental Protection Agency
-----------------------------------------------------------------------
40 CFR Parts 50, 53, and 58
Reconsideration of the National Ambient Air Quality Standards for
Particulate Matter; Final Rule
Federal Register / Vol. 89 , No. 45 / Wednesday, March 6, 2024 /
Rules and Regulations
[[Page 16202]]
-----------------------------------------------------------------------
ENVIRONMENTAL PROTECTION AGENCY
40 CFR Parts 50, 53, and 58
[EPA-HQ-OAR-2015-0072; FRL-8635-02-OAR]
RIN 2060-AV52
Reconsideration of the National Ambient Air Quality Standards for
Particulate Matter
AGENCY: Environmental Protection Agency (EPA).
ACTION: Final rule.
-----------------------------------------------------------------------
SUMMARY: Based on the Environmental Protection Agency's (EPA's)
reconsideration of the air quality criteria and the national ambient
air quality standards (NAAQS) for particulate matter (PM), the EPA is
revising the primary annual PM2.5 standard by lowering the
level from 12.0 [micro]g/m\3\ to 9.0 [micro]g/m\3\. The Agency is
retaining the current primary 24-hour PM2.5 standard and the
primary 24-hour PM10 standard. The Agency also is not
changing the secondary 24-hour PM2.5 standard, secondary
annual PM2.5 standard, and secondary 24-hour PM10
standard at this time. The EPA is also finalizing revisions to other
key aspects related to the PM NAAQS, including revisions to the Air
Quality Index (AQI) and monitoring requirements for the PM NAAQS.
DATES: This final rule is effective May 6, 2024.
ADDRESSES: The EPA has established a docket for this action under
Docket ID No. EPA-HQ-OAR-2015-0072. All documents in the docket are
listed on the https://www.regulations.gov website. Although listed in
the index, some information is not publicly available, e.g., CBI or
other information whose disclosure is restricted by statute. Certain
other material, such as copyrighted material, is not placed on the
internet and will be publicly available only in hard copy form.
Publicly available docket materials are available electronically
through https://www.regulations.gov.
FOR FURTHER INFORMATION CONTACT: Dr. Lars Perlmutt, Health and
Environmental Impacts Division, Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency, Mail Code C539-04,
Research Triangle Park, NC 27711; telephone: (919) 541-3037; fax: (919)
541-5315; email: [email protected].
SUPPLEMENTARY INFORMATION:
Table of Contents
The following topics are discussed in this preamble:
Executive Summary
I. Background
A. Legislative Requirements
B. Related PM Control Programs
C. Review of the Air Quality Criteria and Standards for
Particulate Matter
1. Reviews Completed in 1971 and 1987
2. Review Completed in 1997
3. Review Completed in 2006
4. Review Completed in 2012
5. Review Initiated in 2014
a. 2020 Proposed and Final Decisions
b. Reconsideration of the 2020 PM NAAQS Final Action
D. Air Quality Information
1. Distribution of Particle Size in Ambient Air
2. Sources and Emissions Contributing to PM in the Ambient Air
3. Monitoring of Ambient PM
4. Ambient Concentrations and Trends
a. PM2.5 Mass
b. PM2.5 Components
c. PM10
d. PM10-2.5
e. UFP
5. Characterizing Ambient PM2.5 Concentrations for
Exposure
a. Predicted Ambient PM2.5 and Exposure Based on
Monitored Data
b. Comparison of PM2.5 Fields in Estimating Exposure
and Relative to Design Values
6. Background PM
II. Rationale for Decisions on the Primary PM2.5
Standards
A. Introduction
1. Background on the Current Standards
2. Overview of the Health Effects Evidence
a. Nature of Effects
i. Mortality
ii. Cardiovascular Effects
iii. Respiratory Effects
iv. Cancer
v. Nervous System Effects
vi. Other Effects
b. Public Health Implications and At-Risk Populations
c. PM2.5 Concentrations in Key Studies Reporting
Health Effects
i. PM2.5 Exposure Concentrations Evaluated in
Experimental Studies
ii. Ambient PM2.5 Concentrations in Locations of
Epidemiologic Studies
d. Uncertainties in the Health Effects Evidence
3. Summary of Exposure and Risk Estimates
a. Key Design Aspects
b. Key Limitations and Uncertainties
c. Summary of Risk Estimates
B. Conclusions on the Primary PM2.5 Standards
1. CASAC Advice
2. Basis for the Proposed Decision
3. Comments on the Proposed Decision
4. Administrator's Conclusions
C. Decisions on the Primary PM2.5 Standards
III. Rationale for Decisions on the Primary PM10 Standard
A. Introduction
1. Background on the Current Standard
2. Overview of Health Effects Evidence
a. Nature of Effects
i. Mortality
ii. Cardiovascular Effects
iii. Respiratory Effects
iv. Cancer
v. Metabolic Effects
vi. Nervous System Effects
B. Conclusions on the Primary PM10 Standard
1. CASAC Advice
2. Basis for the Proposed Decision
3. Comments on the Proposed Decision
4. Administrator's Conclusions
C. Decisions on the Primary PM10 Standard
IV. Communication of Public Health
A. Air Quality Index Overview
B. Air Quality Index Category Breakpoints for PM2.5
1. Summary of Proposed Revisions
a. Air Quality Index Values of 50, 100, and 150
b. Air Quality Index Values of 200 and Above
2. Summary of Significant Comments on Proposed Revisions
a. Air Quality Index Values of 50, 100, and 150
b. Air Quality Index Values of 200 and Above
c. Other Comments
3. Summary of Final Revisions
C. Air Quality Index Category Breakpoints for PM10
D. Air Quality Index Reporting
1. Summary of Proposed Revisions
2. Summary of Significant Comments on Proposed Revisions
3. Summary of Final Revisions
V. Rationale for Decisions on the Secondary PM Standards
A. Introduction
1. Background on the Current Standards
a. Non-Visibility Effects
b. Visibility Effects
2. Overview of Welfare Effects Evidence
a. Nature of Effects
i. Visibility
ii. Climate
iii. Materials
3. Summary of Air Quality and Quantitative Information
a. Visibility Effects
i. Target Level of Protection in Terms of a PM2.5
Visibility Index
ii. Relationship Between the PM2.5 Visibility Index
and the Current Secondary 24-Hour PM2.5 Standard
b. Non-Visibility Effects
B. Conclusions on the Secondary PM Standards
1. CASAC Advice
2. Basis for the Proposed Decision
3. Comments on the Proposed Decision
4. Administrator's Conclusions
C. Decisions on the Secondary PM Standards
VI. Interpretation of the NAAQS for PM
A. Amendments to Appendix K: Interpretation of the NAAQS for
Particulate Matter
B. Amendments to Appendix N: Interpretation of the NAAQS for
PM2.5
VII. Amendments to Ambient Monitoring and Quality Assurance
Requirements
[[Page 16203]]
A. Amendment to 40 CFR Part 50 (Appendix L): Reference Method
for the Determination of Fine Particulate Matter as PM2.5
in the Atmosphere--Addition of the Tisch Cyclone as an Approved
Second Stage Separator
B. Issues Related to 40 CFR Part 53 (Reference and Equivalent
Methods)
C. Changes to 40 CFR Part 58 (Ambient Air Quality Surveillance)
D. Incorporating Data From Next-Generation Technologies
VIII. Clean Air Act Implementation Requirements for the Revised
Primary Annual PM2.5 NAAQS
A. Designation of Areas
B. Section 110(a)(1) and (2) Infrastructure SIP Requirements
C. Implementing the Revised Primary Annual PM2.5
NAAQS in Nonattainment Areas
D. Implementing the Primary and Secondary PM10 NAAQS
E. Prevention of Significant Deterioration and Nonattainment New
Source Review Programs for the Revised Primary Annual
PM2.5 NAAQS
F. Transportation Conformity Program
G. General Conformity Program
IX. Statutory and Executive Order Reviews
A. Executive Order 12866: Regulatory Planning and Review and
Executive Order 14094: Modernizing Regulatory Review
B. Paperwork Reduction Act (PRA)
C. Regulatory Flexibility Act (RFA)
D. Unfunded Mandates Reform Act (UMRA)
E. Executive Order 13132: Federalism
F. Executive Order 13175: Consultation and Coordination With
Indian Tribal Governments
G. Executive Order 13045: Protection of Children From
Environmental Health and Safety Risks
H. Executive Order 13211: Actions Concerning Regulations That
Significantly Affect Energy Supply, Distribution or Use
I. National Technology Transfer and Advancement Act (NTTAA)
J. Executive Order 12898: Federal Actions To Address
Environmental Justice in Minority Populations and Low-Income
Populations and Executive Order 14096: Revitalizing Our Nation's
Commitment to Environmental Justice for All
K. Congressional Review Act (CRA)
References
Executive Summary
This document presents the Administrator's final decisions for the
reconsideration of the 2020 final decision on the primary (health-
based) and secondary (welfare-based) National Ambient Air Quality
Standards (NAAQS) for Particulate Matter (PM). More specifically, this
document summarizes the background and rationale for the
Administrator's final decisions to revise the primary annual
PM2.5 standard by lowering the level from 12.0 [micro]g/m\3\
to 9.0 [micro]g/m\3\; to retain the current primary 24-hour
PM2.5 standard (at a level of 35 [micro]g/m\3\); to retain
the primary 24-hour PM10 standard; and, not to change the
secondary PM standards at this time. In reaching his final decisions,
the Administrator considered the currently available scientific
evidence in the 2019 Integrated Science Assessment (2019 ISA) and the
Supplement to the 2019 ISA (ISA Supplement), quantitative and policy
analyses presented in the 2022 Policy Assessment (2022 PA), advice from
the Clean Air Scientific Advisory Committee (CASAC), and public
comments on the proposal. The EPA has established primary and secondary
standards for PM2.5, which includes particles with diameters
generally less than or equal to 2.5 [micro]m, and PM10,
which includes particles with diameters generally less than or equal to
10 [micro]m. The standards include two primary PM2.5
standards: an annual average standard, averaged over three years, with
a level of 12.0 [micro]g/m\3\, and a 24-hour standard with a 98th
percentile form, averaged over three years, and a level of 35 [micro]g/
m\3\. It also includes a primary PM10 standard with a 24-
hour averaging time, and a level of 150 [micro]g/m\3\, not to be
exceeded more than once per year on average over three years. Secondary
PM standards are set equal to the primary standards, except that the
level of the secondary annual PM2.5 standard is 15.0
[micro]g/m\3\.
The most recent of the PM NAAQS was completed in December 2020. In
that review, the EPA retained the primary and secondary NAAQS, without
revision (85 FR 82684, December 18, 2020). Following publication of the
2020 final action, several parties filed petitions for review and
petitions for reconsideration of the EPA's final decision.
In June 2021, the Agency announced its decision to reconsider the
2020 PM NAAQS final action.\1\ The EPA decided to reconsider the
December 2020 decision because the available scientific evidence and
technical information indicated that the current standards may not be
adequate to protect public health and welfare, as required by the Clean
Air Act. The EPA noted that the 2020 PA concluded that the scientific
evidence and information called into question the adequacy of the
primary PM2.5 standards and supported consideration of
revising the level of the primary annual PM2.5 standard to
below the current level of 12.0 [micro]g/m\3\ while retaining the
primary 24-hour PM2.5 standard (U.S. EPA, 2020b). The EPA
also noted that the 2020 PA concluded that the available scientific
evidence and information did not call into question the adequacy of the
primary PM10 or secondary PM standards and supported
consideration of retaining the primary PM10 standard and
secondary PM standards without revision (U.S. EPA, 2020b).
---------------------------------------------------------------------------
\1\ The press release for this announcement is available at:
https://www.epa.gov/newsreleases/epa-reexamine-health-standards-harmful-soot-previous-administration-left-unchanged.
---------------------------------------------------------------------------
The final decisions presented in this document on the primary
PM2.5 standards have been informed by key aspects of the
available health effects evidence and conclusions contained in the 2019
ISA and ISA Supplement, quantitative exposure/risk analyses and policy
evaluations presented in the 2022 PA, advice from the CASAC \2\ and
public comment received as part of this reconsideration.\3\ The health
effects evidence newly available in this reconsideration, in
conjunction with the full body of evidence critically evaluated in the
2019 ISA, supports a causal relationship between long- and short-term
exposures and mortality and cardiovascular effects, and the evidence
supports a likely to be a causal relationship between long-term
exposures and respiratory effects, nervous system effects, and cancer.
The longstanding evidence base, including animal toxicological studies,
controlled human exposure studies, and epidemiologic studies,
reaffirms, and in some cases strengthens, the conclusions from past
reviews regarding the health effects of PM2.5 exposures.
Epidemiologic studies available in this reconsideration demonstrate
generally positive, and often statistically significant,
PM2.5 health effect associations. Such studies report
associations between estimated PM2.5 exposures and non-
accidental, cardiovascular, or respiratory mortality; cardiovascular or
respiratory hospitalizations or emergency room visits; and other
mortality/morbidity outcomes (e.g., lung cancer mortality or incidence,
asthma development). The scientific evidence available in this
reconsideration, as evaluated in the 2019 ISA and ISA Supplement,
includes
[[Page 16204]]
a number of epidemiologic studies that use various methods to
characterize exposure to PM2.5 (e.g., ground-based monitors
and hybrid modeling approaches) and to evaluate associations between
health effects and lower ambient PM2.5 concentrations. There
are a number of recent epidemiologic studies that use varying study
designs that reduce uncertainties related to confounding and exposure
measurement error. The results of these analyses provide further
support for the robustness of associations between PM2.5
exposures and mortality and morbidity. Moreover, the Administrator
notes that recent epidemiologic studies strengthen support for health
effect associations at lower PM2.5 concentrations, with
these new studies finding positive and significant associations when
assessing exposure in locations and time periods with lower annual mean
and 25th percentile concentrations than those evaluated in
epidemiologic studies available at the time of previous reviews.
Additionally, the experimental evidence (i.e., animal toxicological and
controlled human exposure studies) strengthens the coherence of effects
across scientific disciplines and provides additional support for
potential biological pathways through which PM2.5 exposures
could lead to the overt population-level outcomes reported in
epidemiologic studies for the health effect categories for which a
causal relationship (i.e., short- and long-term PM2.5
exposure and mortality and cardiovascular effects) or likely to be
causal relationship (i.e., short- and long-term PM2.5
exposure and respiratory effects; and long-term PM2.5
exposure and nervous system effects and cancer) was concluded.
---------------------------------------------------------------------------
\2\ In 2021, the Administrator announced his decision to
reestablish the membership of the CASAC. The Administrator selected
seven members to serve on the chartered CASAC, and appointed a PM
CASAC panel to support the chartered CASAC's review of the draft ISA
Supplement and the draft PA as a part of this reconsideration (see
section I.C.6.b below for more information).
\3\ More information regarding the CASAC review of the draft ISA
Supplement and the draft PA, including opportunities for public
comment, can be found in the following Federal Register notices: 86
FR 54186, September 30, 2021; 86 FR 52673, September 22, 2021; 86 FR
56263, October 8, 2021; 87 FR 958, January 7, 2022.
---------------------------------------------------------------------------
The available evidence in the 2019 ISA continues to provide support
for factors that may contribute to increased risk of PM2.5-
related health effects including lifestage (children and older adults),
pre-existing diseases (cardiovascular disease and respiratory disease),
race/ethnicity, and socioeconomic status. For example, the 2019 ISA and
ISA Supplement conclude that there is strong evidence that Black and
Hispanic populations, on average, experience higher PM2.5
exposures and PM2.5-related health risks than non-Hispanic
White populations. In addition, studies evaluated in the 2019 ISA and
ISA Supplement also provide evidence indicating that communities with
lower socioeconomic status (SES), as assessed in epidemiologic studies
using indicators of SES including income and educational attainment
are, on average, exposed to higher concentrations of PM2.5
compared to higher SES communities.
The quantitative risk assessment, as well as policy considerations
in the 2022 PA, also inform the final decisions on the primary
PM2.5 standards. The risk assessment in this reconsideration
focuses on all-cause or nonaccidental mortality associated with long-
and short-term PM2.5 exposures. The primary analyses focus
on exposure and risk associated with air quality that might occur in an
area under air quality conditions that just meet the current and
potential alternative standards. The risk assessment estimates that the
current primary PM2.5 standards could allow a substantial
number of PM2.5-associated premature deaths in the United
States, and that public health improvements would be associated with
just meeting all of the alternative (more stringent) annual and 24-hour
standard levels modeled. Additionally, the results of the risk
assessment suggest that for most of the U.S., the annual standard is
the controlling standard and that revision to that standard has the
most potential to reduce PM2.5 exposure-related risk. The
analyses are summarized in this document and in the proposal and are
described in detail in the 2022 PA.
In its advice to the Administrator, in its review of the 2021 draft
PA, the CASAC concurred that the currently available health effects
evidence calls into question the adequacy of the primary annual
PM2.5 standard. With regard to the primary annual
PM2.5 standard, the majority of the CASAC concluded that the
level of the standard should be revised within the range of 8.0 to 10.0
[micro]g/m\3\, while the minority of the CASAC concluded that the
primary annual PM2.5 standard should be revised to a level
of 10.0 to 11.0 [micro]g/m\3\. With regard to the primary 24-hour
PM2.5 standard, the CASAC did not reach consensus on the
adequacy of the current standard. The majority of the CASAC concluded
that the primary 24-hour PM2.5 was not adequate and that the
level of the standard should be revised to within the range of 25 to 30
[micro]g/m\3\, while the minority of the CASAC concluded that the
standard was adequate and should be retained, without revision.
Additionally, in their review of the 2019 draft PA, the CASAC did not
reach consensus on the adequacy of the primary annual PM2.5
standard, with the minority recommending revision and the majority
recommending the standard be retained. In their review of the 2019
draft PA, the CASAC reached consensus regarding the adequacy of the
primary 24-hour PM2.5 standard, concluding that the standard
should be retained.
In considering how to revise the suite of primary PM2.5
standards to provide the requisite degree of protection, the
Administrator recognizes that the current annual standard and 24-hour
standard, together, are intended to provide public health protection
against the full distribution of short- and long-term PM2.5
exposures. Further, he recognizes that changes in PM2.5 air
quality designed to meet either the annual or the 24-hour standard
would likely result in changes to both long-term average and short-term
peak PM2.5 concentrations.
As in 2012, the Administrator concludes that the most effective way
to reduce total population risk associated with both long- and short-
term PM2.5 exposures is to set a generally controlling
annual standard, and to provide supplemental protection against the
occurrence of peak 24-hour PM2.5 concentrations by means of
a 24-hour standard set at the appropriate level. Based on the current
evidence and quantitative information, as well as consideration of
CASAC advice and public comments, the Administrator concludes that the
current primary annual PM2.5 standard is not adequate to
protect public health with an adequate margin of safety. The
Administrator notes that the CASAC was unanimous in its advice on the
2021 draft PA regarding the need to revise the annual standard. In
considering the appropriate level for a revised annual standard, the
Administrator concludes that a standard set at a level of 9.0 [micro]g/
m\3\ reflects his judgment about placing the most weight on the
strongest available evidence while appropriately weighing the
uncertainties.
With regard to the primary 24-hour PM2.5 standard, the
Administrator finds the available scientific evidence and quantitative
information to be insufficient to call into question the adequacy of
the public health protection afforded by the current 24-hour standard.
He further notes that a more stringent annual standard set at a level
of 9.0 [micro]g/m\3\ is expected to reduce both average (annual)
concentrations and peak (daily) concentrations. The Administrator also
notes that, in their review of the 2021 draft PA, the CASAC did not
reach consensus on whether revisions to the primary 24-hour
PM2.5 standard are warranted at this time. He also notes
that, in their review of the 2019 draft PA, the CASAC did reach
consensus that the primary 24-hour PM2.5 standard should be
retained. The Administrator concludes that the 24-hour standard should
be retained to
[[Page 16205]]
continue to provide requisite protection against short-term peak
PM2.5 concentrations, particularly when considered in
conjunction with the protection provided by the suite of standards and
the decision to revise the annual standard to a level of 9.0 [micro]g/
m\3\.
The primary PM10 standard is intended to provide public
health protection against health effects related to exposures to
PM10-2.5, which are particles with a diameter between 10
[micro]m and 2.5 [micro]m. The final decision to retain the current 24-
hour PM10 standard has been informed by key aspects of the
available health effects evidence and conclusions contained in the 2019
ISA, the policy evaluations presented in the 2022 PA, advice from the
CASAC and public comments. Specifically, the health effects evidence
for PM10-2.5 exposures is somewhat strengthened since past
reviews, although the strongest evidence still only provides support
for a suggestive of, but not sufficient to infer, causal relationship
with long- and short-term exposures and mortality and cardiovascular
effects, short-term exposures and respiratory effects, and long-term
exposures and cancer, nervous system effects, and metabolic effects. In
reaching his final decision on the primary PM10 standard,
the Administrator recognizes that, while the available health effects
evidence has expanded, recent studies are subject to the same types of
uncertainties that were judged to be important in previous reviews. He
also recognizes that, in their review of the 2019 draft PA and the 2021
draft PA, the CASAC generally agreed that it was reasonable to retain
the primary 24-hour PM10 standard given the available
scientific evidence, including retaining PM10 as the
indicator. He concludes that the newly available evidence does not call
into question the adequacy of the current primary PM10
standard, and retains that standard, without revision.
With respect to the secondary PM standards, this reconsideration
focuses on visibility, climate, and materials effects.\4\ The
Administrator's final decision to not change the current secondary
standards at this time has been informed by key aspects of the
currently available welfare effects evidence as well as the conclusions
contained in the 2019 ISA and ISA Supplement; quantitative analyses of
visibility impairment; policy evaluations presented in the 2022 PA;
advice from the CASAC; and public comments. Specifically, the welfare
effects evidence available in this reconsideration is consistent with
the evidence available in previous reviews and supports a causal
relationship between PM and visibility, climate, and materials effects.
With regard to visibility effects, the Administrator notes that he
judges that the evidence supports a target level of protection of 27
dv. He further notes that the results of quantitative analyses of
visibility impairment suggest that in areas that meet the current
secondary 24-hour PM2.5 standard that estimated light
extinction in terms of a 3-year visibility metric would be at or well
below the target level of protection. With regard to climate and
materials effects, while the evidence has expanded since previous
reviews, significant limitations and uncertainties remain in the
evidence. While the evidence has expanded since previous reviews, the
available scientific evidence remains insufficient to allow the
Administrator to make a reasoned judgment about what specific
standard(s) would be requisite to protect against known or anticipated
adverse effects to public welfare from PM's effects on materials damage
or climate.-In their review of the 2019 draft PA and the 2021 draft PA,
the CASAC did not recommend revising the secondary PM standards. In
considering the available evidence and quantitative information, with
its inherent uncertainties and limitations, the Administrator judges
that it is appropriate not to change the secondary PM standards at this
time.
---------------------------------------------------------------------------
\4\ Consistent with the 2016 Integrated Review Plan (U.S. EPA,
2016), other welfare effects of PM, such as ecological effects, are
being considered in the separate, on-going review of the secondary
NAAQS for oxides of nitrogen, oxides of sulfur and PM. Accordingly,
the public welfare protection provided by the secondary PM standards
against ecological effects such as those related to deposition of
nitrogen- and sulfur-containing compounds in vulnerable ecosystems
is being considered in that separate review. Thus, the
Administrator's conclusion in this reconsideration of the 2020 final
decision is focused only and specifically on the adequacy of public
welfare protection provided by the secondary PM standards from
effects related to visibility, climate, and materials and hereafter
``welfare effects'' refers to those welfare effects.
---------------------------------------------------------------------------
The final revisions to the primary annual PM2.5 NAAQS
trigger a process under which States (and Tribes, if they choose) make
recommendations to the Administrator regarding designations,
identifying areas of the country that either meet or do not meet the
new or revised PM NAAQS. Those areas that do not meet the revised PM
NAAQS will need to develop plans that demonstrate how they will meet
the standards. As part of these plans, states have the opportunity to
advance environmental justice, in this case for overburdened
communities in areas with high PM concentrations above the NAAQS, by
using the tools described in the current PM NAAQS implementation
guidance (80 FR 58010, 58136, August 25, 2016). The EPA is not making
changes to any of the current PM NAAQS implementation programs in this
final rulemaking.
On other topics, the EPA is finalizing two sets of changes to the
PM2.5 sub-index of the Air Quality Index (AQI). First, the
EPA is continuing to use the approach used in the revisions to the AQI
in 2012 (77 FR 38890, June 29, 2012) of setting the lower breakpoints
(50, 100 and 150) based on the levels of the primary annual and 24-hour
PM2.5 standards. In so doing, the EPA is revising the AQI
value of 50 to 9.0 [micro]g/m\3\ and is retaining the AQI values of 100
and 150 at 35.4 [micro]g/m\3\ and 55.4 [micro]g/m\3\, respectively.
Second, the EPA is revising the upper AQI breakpoints (200 and above),
and replacing the linear-relationship approach used in 1999 (64 FR
42530, August 4, 1999) to set these breakpoints, with an approach that
more fully considers the PM2.5 health effects evidence from
controlled human exposure and epidemiologic studies that has become
available in the last 20 years. The EPA is also revising the AQI values
of 200, 300 and 500 to 125.4 [micro]g/m\3\, 225.4 [micro]g/m\3\, and
325.4 [micro]g/m\3\, respectively. In addition, this final rule revises
the daily reporting requirement from 5 days per week to 7 days per
week, while also reformatting appendix G and providing clarifications.
With regard to monitoring-related activities, the EPA finalizes
revisions to data calculations and ambient air monitoring requirements
for PM to improve the usefulness and appropriateness of data used in
regulatory decision making and to better characterize air quality in
communities that are at increased risk of PM2.5 exposure and
health risk. These changes are found in 40 CFR part 50 (appendices K,
L, and N), part 53, and part 58 with associated appendices (A, B, C, D,
and E). These changes include addressing updates in data calculations,
approval of reference and equivalent methods, updates in quality
assurance statistical calculations to account for lower concentration
measurements, updates to support improvements in PM methods, a revision
to the PM2.5 network design to account for at-risk
populations, and updates to the Probe and Monitoring Path Siting
Criteria for NAAQS pollutants.
In setting the NAAQS, the EPA may not consider the costs of
implementing the standards. This was confirmed by the Supreme Court in
Whitman v. American Trucking Associations, 531 U.S. 457, 465-472, 475-
76 (2001), as discussed in section II.A of this document. As has
traditionally been
[[Page 16206]]
done in NAAQS rulemaking, the EPA prepared a Regulatory Impact Analysis
(RIA) to provide the public with information on the potential costs and
benefits of attaining several alternative PM2.5 standard
levels. In NAAQS rulemaking, the RIA is done for informational purposes
only, and the final decisions on the NAAQS in this rulemaking are not
based on consideration of the information or analyses in the RIA. The
RIA fulfills the requirements of Executive Orders 14094, 13563, and
12866. The RIA estimates the costs and monetized human health benefits
of attaining the revised and two alternative annual PM2.5
standard levels and one alternative 24-hour PM2.5 standard
level. Specifically, the RIA examines the revised annual standard level
of 9.0 [micro]g/m\3\ in combination with the current 24-hour standard
of 35 [micro]g/m\3\ (i.e., 9.0/35 [micro]g/m\3\), as well as the
following less and more stringent alternative standard levels: (1) An
alternative annual standard level of 10.0 [micro]g/m\3\ in combination
with the current 24-hour standard (i.e., 10.0/35 [micro]g/m\3\), (2) an
alternative annual standard level of 8.0 [micro]g/m\3\ in combination
with the current 24-hour standard (i.e., 8.0/35 [micro]g/m\3\), and (3)
an alternative 24-hour standard level of 30 [micro]g/m\3\ in
combination with an alternative annual standard level of 10 [micro]g/
m\3\ (i.e., 10.0/30 [micro]g/m\3\). The RIA presents estimates of the
costs and benefits of applying illustrative national control strategies
in 2032 after implementing existing and expected regulations and
assessing emissions reductions to meet the current annual and 24-hour
particulate matter NAAQS (12.0/35 [micro]g/m\3\).
I. Background
A. Legislative Requirements
Two sections of the Clean Air Act (CAA) govern the establishment
and revision of the NAAQS. Section 108 (42 U.S.C. 7408) directs the
Administrator to identify and list certain air pollutants and then to
issue air quality criteria for those pollutants. The Administrator is
to list those pollutants ``emissions of which, in his judgment, cause
or contribute to air pollution which may reasonably be anticipated to
endanger public health or welfare''; ``the presence of which in the
ambient air results from numerous or diverse mobile or stationary
sources''; and for which he ``plans to issue air quality criteria. . .
.'' (42 U.S.C. 7408(a)(1)). Air quality criteria are intended to
``accurately reflect the latest scientific knowledge useful in
indicating the kind and extent of all identifiable effects on public
health or welfare which may be expected from the presence of [a]
pollutant in the ambient air. . . .'' (42 U.S.C. 7408(a)(2)).
Section 109 [42 U.S.C. 7409] directs the Administrator to propose
and promulgate ``primary'' and ``secondary'' NAAQS for pollutants for
which air quality criteria are issued [42 U.S.C. 7409(a)]. Section
109(b)(1) defines primary standards as ones ``the attainment and
maintenance of which in the judgment of the Administrator, based on
such criteria and allowing an adequate margin of safety, are requisite
to protect the public health.'' \5\ Under section 109(b)(2), a
secondary standard must ``specify a level of air quality the attainment
and maintenance of which, in the judgment of the Administrator, based
on such criteria, is requisite to protect the public welfare from any
known or anticipated adverse effects associated with the presence of
[the] pollutant in the ambient air.'' \6\
---------------------------------------------------------------------------
\5\ The legislative history of section 109 indicates that a
primary standard is to be set at ``the maximum permissible ambient
air level . . . which will protect the health of any [sensitive]
group of the population,'' and that for this purpose ``reference
should be made to a representative sample of persons comprising the
sensitive group rather than to a single person in such a group.'' S.
Rep. No. 91-1196, 91st Cong., 2d Sess. 10 (1970).
\6\ Under CAA section 302(h) (42 U.S.C. 7602(h)), effects on
welfare include, but are not limited to, ``effects on soils, water,
crops, vegetation, manmade materials, animals, wildlife, weather,
visibility, and climate, damage to and deterioration of property,
and hazards to transportation, as well as effects on economic values
and on personal comfort and well-being.''
---------------------------------------------------------------------------
In setting primary and secondary standards that are ``requisite''
to protect public health and welfare, respectively, as provided in
section 109(b), the EPA's task is to establish standards that are
neither more nor less stringent than necessary. In so doing, the EPA
may not consider the costs of implementing the standards. See generally
Whitman v. American Trucking Associations, 531 U.S. 457, 465-472, 475-
76 (2001). Likewise, ``[a]ttainability and technological feasibility
are not relevant considerations in the promulgation of national ambient
air quality standards.'' American Petroleum Institute v. Costle, 665
F.2d 1176, 1185 (D.C. Cir. 1981); accord Murray Energy Corporation v.
EPA, 936 F.3d 597, 623-24 (D.C. Cir. 2019).
The requirement that primary standards provide an adequate margin
of safety was intended to address uncertainties associated with
inconclusive scientific and technical information available at the time
of standard setting. It was also intended to provide a reasonable
degree of protection against hazards that research has not yet
identified. See Lead Industries Association v. EPA, 647 F.2d 1130, 1154
(D.C. Cir. 1980); American Petroleum Institute v. Costle, 665 F.2d at
1186; Coalition of Battery Recyclers Ass'n v. EPA, 604 F.3d 613, 617-18
(D.C. Cir. 2010); Mississippi v. EPA, 744 F.3d 1334, 1353 (D.C. Cir.
2013). Both kinds of uncertainties are components of the risk
associated with pollution at levels below those at which human health
effects can be said to occur with reasonable scientific certainty.
Thus, in selecting primary standards that include an adequate margin of
safety, the Administrator is seeking not only to prevent pollution
levels that have been demonstrated to be harmful but also to prevent
lower pollutant levels that may pose an unacceptable risk of harm, even
if the risk is not precisely identified as to nature or degree. The CAA
does not require the Administrator to establish a primary NAAQS at a
zero-risk level or at background concentration levels, see Lead
Industries Ass'n v. EPA, 647 F.2d at 1156 n.51, Mississippi v. EPA, 744
F.3d at 1351, but rather at a level that reduces risk sufficiently so
as to protect public health with an adequate margin of safety.
In addressing the requirement for an adequate margin of safety, the
EPA considers such factors as the nature and severity of the health
effects involved, the size of the sensitive population(s), and the kind
and degree of uncertainties. The selection of any particular approach
to providing an adequate margin of safety is a policy choice left
specifically to the Administrator's judgment. See Lead Industries Ass'n
v. EPA, 647 F.2d at 1161-62; Mississippi v. EPA, 744 F.3d at 1353.
Section 109(d)(1) of the Act requires the review every five years
of existing air quality criteria and, if appropriate, the revision of
those criteria to reflect advances in scientific knowledge on the
effects of the pollutant on public health and welfare. Under the same
provision, the EPA is also to review every five years and, if
appropriate, revise the NAAQS, based on the revised air quality
criteria. Section 109(d)(1) also provides that the Administrator may
review and revise criteria or promulgate new standards earlier or more
frequently.
Section 109(d)(2) addresses the appointment and advisory functions
of an independent scientific review committee. Section 109(d)(2)(A)
requires the Administrator to appoint this committee, which is to be
composed of ``seven members including at least one member of the
National Academy of Sciences, one physician, and one person
representing State air
[[Page 16207]]
pollution control agencies.'' Section 109(d)(2)(B) provides that the
independent scientific review committee ``shall complete a review of
the criteria . . . and the national primary and secondary ambient air
quality standards . . . and shall recommend to the Administrator any
new . . . standards and revisions of existing criteria and standards as
may be appropriate. . . .'' Since the early 1980s, this independent
review function has been performed by the Clean Air Scientific Advisory
Committee (CASAC) of the EPA's Science Advisory Board.
As previously noted, the Supreme Court has held that section 109(b)
``unambiguously bars cost considerations from the NAAQS-setting
process.'' Whitman v. Am. Trucking Associations, 531 U.S. 457, 471
(2001). Accordingly, while some of these issues regarding which
Congress has directed the CASAC to advise the Administrator are ones
that are relevant to the standard setting process, others are not.
Issues that are not relevant to standard setting may be relevant to
implementation of the NAAQS once they are established.
B. Related PM Control Programs
States are primarily responsible for ensuring attainment and
maintenance of ambient air quality standards once the EPA has
established them. Under section 110, Part C, and Part D, Subparts 1 and
4 of the CAA, and related provisions and regulations, States are to
submit, for the EPA's approval, State implementation plans (SIPs) that
provide for the attainment and maintenance of the NAAQS for PM through
control programs directed to sources of the pollutants involved. The
States, in conjunction with the EPA, also administer the prevention of
significant deterioration of air quality program that covers these
pollutants (see 42 U.S.C. 7470-7479). In addition, Federal programs
provide for or result in nationwide reductions in emissions of PM and
its precursors under Title II of the Act, 42 U.S.C. 7521-7574, which
involves controls for motor vehicles and nonroad engines and equipment;
the new source performance standards under section 111 of the Act, 42
U.S.C. 7411; and the national emissions standards for hazardous
pollutants under section 112 of the Act, 42 U.S.C. 7412.
C. Review of the Air Quality Criteria and Standards for Particulate
Matter
1. Reviews Completed in 1971 and 1987
The EPA first established NAAQS for PM in 1971 (36 FR 8186, April
30, 1971), based on the original Air Quality Criteria Document (AQCD)
(DHEW, 1969).\7\ The Federal reference method (FRM) specified for
determining attainment of the original standards was the high-volume
sampler, which collects PM up to a nominal size of 25 to 45 [micro]m
(referred to as total suspended particulates or TSP). The primary
standards were set at 260 [micro]g/m\3\, 24-hour average, not to be
exceeded more than once per year, and 75 [micro]g/m\3\, annual
geometric mean. The secondary standards were set at 150 [micro]g/m\3\,
24-hour average, not to be exceeded more than once per year, and 60
[micro]g/m\3\, annual geometric mean.
---------------------------------------------------------------------------
\7\ Prior to the review initiated in 2007 (see below), the AQCD
provided the scientific foundation (i.e., the air quality criteria)
for the NAAQS. Beginning in that review, the Integrated Science
Assessment (ISA) has replaced the AQCD.
---------------------------------------------------------------------------
In October 1979 (44 FR 56730, October 2, 1979), the EPA announced
the first periodic review of the air quality criteria and NAAQS for PM.
Revised primary and secondary standards were promulgated in 1987 (52 FR
24634, July 1, 1987). In the 1987 decision, the EPA changed the
indicator for particles from TSP to PM10, in order to focus
on the subset of inhalable particles small enough to penetrate to the
thoracic region of the respiratory tract (including the
tracheobronchial and alveolar regions), referred to as thoracic
particles.\8\ The level of the 24-hour standards (primary and
secondary) was set at 150 [micro]g/m\3\, and the form was one expected
exceedance per year, on average over three years. The level of the
annual standards (primary and secondary) was set at 50 [micro]g/m\3\,
and the form was the annual arithmetic mean, averaged over three years.
---------------------------------------------------------------------------
\8\ PM10 refers to particles with a nominal mean
aerodynamic diameter less than or equal to 10 [micro]m. More
specifically, 10 [micro]m is the aerodynamic diameter for which the
efficiency of particle collection is 50 percent.
---------------------------------------------------------------------------
2. Review Completed in 1997
In April 1994, the EPA announced its plans for the second periodic
review of the air quality criteria and NAAQS for PM, and in 1997 the
EPA promulgated revisions to the NAAQS (62 FR 38652, July 18, 1997). In
the 1997 decision, the EPA determined that the fine and coarse
fractions of PM10 should be considered separately. This
determination was based on evidence that serious health effects were
associated with short- and long-term exposures to fine particles in
areas that met the existing PM10 standards. The EPA added
new standards, using PM2.5 as the indicator for fine
particles (with PM2.5 referring to particles with a nominal
mean aerodynamic diameter less than or equal to 2.5 [micro]m). The new
primary standards were as follows: (1) An annual standard with a level
of 15.0 [micro]g/m\3\, based on the 3-year average of annual arithmetic
mean PM2.5 concentrations from single or multiple community-
oriented monitors; \9\ and (2) a 24-hour standard with a level of 65
[micro]g/m\3\, based on the 3-year average of the 98th percentile of
24-hour PM2.5 concentrations at each monitor within an area.
Also, the EPA established a new reference method for the measurement of
PM2.5 in the ambient air and adopted rules for determining
attainment of the new standards. To continue to address the health
effects of the coarse fraction of PM10 (referred to as
thoracic coarse particles or PM10-2.5, generally including
particles with a nominal mean aerodynamic diameter greater than 2.5
[micro]m and less than or equal to 10 [micro]m), the EPA retained the
primary annual PM10 standard and revised the form of the
primary 24-hour PM10 standard to be based on the 99th
percentile of 24-hour PM10 concentrations at each monitor in
an area. The EPA revised the secondary standards by setting them equal
in all respects to the primary standards.
---------------------------------------------------------------------------
\9\ The 1997 annual PM2.5 standard was compared with
measurements made at the community-oriented monitoring site
recording the highest concentration or, if specific constraints were
met, measurements from multiple community-oriented monitoring sites
could be averaged (i.e., ``spatial averaging''). In the last review
(completed in 2012) the EPA replaced the term ``community-oriented''
monitor with the term ``area-wide'' monitor. Area-wide monitors are
those sited at the neighborhood scale or larger, as well as those
monitors sited at micro- or middle-scales that are representative of
many such locations in the same core-based statistical area (CBSA)
(78 FR 3236, January 15, 2013).
---------------------------------------------------------------------------
Following promulgation of the 1997 PM NAAQS, petitions for review
were filed by several parties, addressing a broad range of issues. In
May 1999, the U.S. Court of Appeals for the District of Columbia
Circuit (D.C. Circuit) upheld the EPA's decision to establish fine
particle standards and to regulate coarse particle pollution, but
vacated the 1997 PM10 standards, concluding that the EPA had
not provided a reasonable explanation justifying use of PM10
as an indicator for coarse particles. American Trucking Associations,
Inc. v. EPA, 175 F. 3d 1027 (D.C. Cir. 1999). Pursuant to the D.C.
Circuit's decision, the EPA removed the vacated 1997 PM10
standards, and the pre-existing 1987 PM10 standards remained
in place (65 FR 80776, December 22, 2000). The D.C. Circuit also upheld
the EPA's determination not to establish more stringent secondary
standards for fine particles to address effects on visibility. American
Trucking Associations v. EPA, 175 F. 3d at 1027.
[[Page 16208]]
The D.C. Circuit also addressed more general issues related to the
NAAQS, including issues related to the consideration of costs in
setting NAAQS and the EPA's approach to establishing the levels of
NAAQS. Regarding the cost issue, the court reaffirmed prior rulings
holding that in setting NAAQS the EPA is ``not permitted to consider
the cost of implementing those standards.'' American Trucking
Associations v. EPA, 175 F. 3d at 1040-41. Regarding the levels of
NAAQS, the court held that the EPA's approach to establishing the level
of the standards in 1997 (i.e., both for PM and for the ozone NAAQS
promulgated on the same day) effected ``an unconstitutional delegation
of legislative authority.'' American Trucking Associations v. EPA, 175
F. 3d at 1034-40. Although the court stated that ``the factors EPA uses
in determining the degree of public health concern associated with
different levels of ozone and PM are reasonable,'' it remanded the rule
to the EPA, stating that when the EPA considers these factors for
potential non-threshold pollutants ``what EPA lacks is any determinate
criterion for drawing lines'' to determine where the standards should
be set.
The D.C. Circuit's holding on the cost and constitutional issues
were appealed to the United States Supreme Court. In February 2001, the
Supreme Court issued a unanimous decision upholding the EPA's position
on both the cost and constitutional issues. Whitman v. American
Trucking Associations, 531 U.S. 457, 464, 475-76. On the constitutional
issue, the Court held that the statutory requirement that NAAQS be
``requisite'' to protect public health with an adequate margin of
safety sufficiently guided the EPA's discretion, affirming the EPA's
approach of setting standards that are neither more nor less stringent
than necessary.
The Supreme Court remanded the case to the D.C. Circuit for
resolution of any remaining issues that had not been addressed in that
court's earlier rulings. Id. at 475-76. In a March 2002 decision, the
D.C. Circuit rejected all remaining challenges to the standards,
holding that the EPA's PM2.5 standards were reasonably
supported by the administrative record and were not ``arbitrary and
capricious.'' American Trucking Associations v. EPA, 283 F. 3d 355,
369-72 (D.C. Cir. 2002).
3. Review Completed in 2006
In October 1997, the EPA published its plans for the third periodic
review of the air quality criteria and NAAQS for PM (62 FR 55201,
October 23, 1997). After the CASAC and public review of several drafts,
the EPA's National Center for Environmental Assessment (NCEA) finalized
the AQCD in October 2004 (U.S. EPA, 2004a). The EPA's Office of Air
Quality Planning and Standards (OAQPS) finalized a Risk Assessment and
Staff Paper in December 2005 (Abt Associates, 2005; U.S. EPA,
2005).\10\ On December 20, 2005, the EPA announced its proposed
decision to revise the NAAQS for PM and solicited public comment on a
broad range of options (71 FR 2620, January 17, 2006). On September 21,
2006, the EPA announced its final decisions to revise the primary and
secondary NAAQS for PM to provide increased protection of public health
and welfare, respectively (71 FR 61144, October 17, 2006). With regard
to the primary and secondary standards for fine particles, the EPA
revised the level of the 24-hour PM2.5 standards to 35
[micro]g/m\3\, retained the level of the annual PM2.5
standards at 15.0 [micro]g/m\3\, and revised the form of the annual
PM2.5 standards by narrowing the constraints on the optional
use of spatial averaging. With regard to the primary and secondary
standards for PM10, the EPA retained the 24-hour standards,
with levels at 150 [micro]g/m\3\, and revoked the annual standards. The
then-Administrator judged that the available evidence generally did not
suggest a link between long-term exposure to existing ambient levels of
coarse particles and health or welfare effects. In addition, a new
reference method was added for the measurement of PM10-2.5
in the ambient air in order to provide a basis for approving Federal
Equivalent Methods (FEMs) and to promote the gathering of scientific
data to support future reviews of the PM NAAQS.
---------------------------------------------------------------------------
\10\ Prior to the review initiated in 2007, the Staff Paper
presented the EPA staff's considerations and conclusions regarding
the adequacy of existing NAAQS and, when appropriate, the potential
alternative standards that could be supported by the evidence and
information. More recent reviews present this information in the
Policy Assessment.
---------------------------------------------------------------------------
Several parties filed petitions for review following promulgation
of the revised PM NAAQS in 2006. On February 24, 2009, the D.C. Circuit
issued its opinion in the case American Farm Bureau Federation v. EPA,
559 F. 3d 512 (D.C. Cir. 2009). The court remanded the primary annual
PM2.5 NAAQS to the EPA because the Agency had failed to
adequately explain why the standards provided the requisite protection
from both short- and long-term exposures to fine particles, including
protection for at-risk populations. Id. at 520-27. With regard to the
standards for PM10, the court upheld the EPA's decisions to
retain the 24-hour PM10 standard to provide protection from
thoracic coarse particle exposures and to revoke the annual
PM10 standard. Id. at 533-38. With regard to the secondary
PM2.5 standards, the court remanded the standards to the EPA
because the Agency failed to adequately explain why setting the
secondary PM standards identical to the primary standards provided the
required protection for public welfare, including protection from
visibility impairment. Id. at 528-32. The EPA responded to the court's
remands as part of the next review of the PM NAAQS, which was initiated
in 2007 (discussed below).
4. Review Completed in 2012
In June 2007, the EPA initiated the fourth periodic review of the
air quality criteria and the PM NAAQS by issuing a call for information
(72 FR 35462, June 28, 2007). Based on the NAAQS review process, as
revised in 2008 and again in 2009,\11\ the EPA held science/policy
issue workshops on the primary and secondary PM NAAQS (72 FR 34003,
June 20, 2007; 72 FR 34005, June 20, 2007), and prepared and released
the planning and assessment documents that comprise the review process
(i.e., Integrated Review Plan, (IRP; U.S. EPA, 2008), Integrated
Science Assessment (ISA; U.S. EPA, 2009a), Risk and Exposure Assessment
(REA) planning documents for health and welfare (U.S. EPA, 2009b, U.S.
EPA, 2009c), a quantitative health risk assessment (U.S. EPA, 2010a)
and an urban-focused visibility assessment (U.S. EPA, 2010b), and a
Policy Assessment (PA; U.S. EPA, 2011). In June 2012, the EPA announced
its proposed decision to revise the NAAQS for PM (77 FR 38890, June 29,
2012).
---------------------------------------------------------------------------
\11\ The history of the NAAQS review process, including
revisions to the process, is discussed at https://www.epa.gov/naaqs/historical-information-naaqs-review-process.
---------------------------------------------------------------------------
In December 2012, the EPA announced its final decisions to revise
the primary NAAQS for PM to provide increased protection of public
health (78 FR 3086, January 15, 2013). With regard to primary standards
for PM2.5, the EPA revised the level of the annual
PM2.5 standard \12\ to 12.0 [micro]g/m\3\ and retained the
24-hour PM2.5 standard, with its level of 35 [micro]g/m\3\.
For the primary PM10 standard, the EPA retained the 24-hour
standard to continue to provide protection against effects associated
with short-term exposure to thoracic coarse particles (i.e.,
PM10-2.5). With regard to the secondary PM standards, the
EPA generally retained the 24-hour
[[Page 16209]]
and annual PM2.5 standards \13\ and the 24-hour
PM10 standard to address visibility and non-visibility
welfare effects.
---------------------------------------------------------------------------
\12\ The EPA also eliminated the option for spatial averaging.
\13\ Consistent with the primary standard, the EPA eliminated
the option for spatial averaging with the annual standard.
---------------------------------------------------------------------------
As with previous reviews, petitioners challenged the EPA's final
rule. Petitioners argued that the EPA acted unreasonably in revising
the level and form of the annual standard and in amending the
monitoring network provisions. On judicial review, the revised
standards and monitoring requirements were upheld in all respects. NAM
v. EPA, 750 F.3d 921 (D.C. Cir. 2014).
5. Review Initiated in 2014
In December 2014, the EPA announced the initiation of the current
periodic review of the air quality criteria for PM and of the
PM2.5 and PM10 NAAQS and issued a call for
information (79 FR 71764, December 3, 2014). On February 9 to 11, 2015,
the EPA's NCEA and OAQPS held a public workshop to inform the planning
for the review of the PM NAAQS (announced in 79 FR 71764, December 3,
2014). Workshop participants, including a wide range of external
experts as well as the EPA staff representing a variety of areas of
expertise (e.g., epidemiology, human and animal toxicology, risk/
exposure analysis, atmospheric science, visibility impairment, climate
effects), were asked to highlight significant new and emerging PM
research, and to make recommendations to the Agency regarding the
design and scope of the review. This workshop provided for a public
discussion of the key science and policy-relevant issues around which
the EPA structured the review of the PM NAAQS and of the most
meaningful new scientific information that would be available in the
review to inform understanding of these issues.
The input received at the workshop guided the EPA staff in
developing a draft IRP, which was reviewed by the CASAC Particulate
Matter Panel and discussed on public teleconferences held in May 2016
(81 FR 13362, March 14, 2016) and August 2016 (81 FR 39043, June 15,
2016). Advice from the CASAC, supplemented by the Particulate Matter
Panel, and input from the public were considered in developing the
final IRP (U.S. EPA, 2016). The final IRP discusses the approaches to
be taken in developing key scientific, technical, and policy documents
in the review and the key policy-relevant issues that frame the EPA's
consideration of whether the primary and/or secondary NAAQS for PM
should be retained or revised.
In May 2018, the then-Administrator issued a memorandum announcing
the Agency's intention to conduct the review of the PM NAAQS in such a
manner as to ensure that any necessary revisions were finalized by
December 2020 (Pruitt, 2018). Following this memo, on October 10, 2018,
the then-Administrator additionally announced that the role of
reviewing the key assessments developed as part of the ongoing review
of the PM NAAQS (i.e., drafts of the ISA and PA) would be performed by
the seven-member chartered CASAC (i.e., rather than the CASAC
Particulate Matter Panel that reviewed the draft IRP).\14\
---------------------------------------------------------------------------
\14\ Announcement available at: https://www.regulations.gov/document/EPA-HQ-OAR-2015-0072-0223.
---------------------------------------------------------------------------
The EPA released the draft ISA in October 2018 (83 FR 53471,
October 23, 2018). The draft ISA was reviewed by the chartered CASAC at
a public meeting held in Arlington, VA in December 2018 (83 FR 55529,
November 6, 2018) and was discussed on a public teleconference in March
2019 (84 FR 8523, March 8, 2019). The CASAC provided its advice on the
draft ISA in a letter to the then-Administrator dated April 11, 2019
(Cox, 2019a). The EPA addressed these comments in the final ISA, which
was released in December 2019 (U.S. EPA, 2019a).
The EPA released the draft PA in September 2019 (84 FR 47944,
September 11, 2019). The draft PA was reviewed by the chartered CASAC
and discussed in October 2019 at a public meeting held in Cary, NC.
Public comments were received via a separate public teleconference (84
FR 51555, September 30, 2019). A public meeting to discuss the
chartered CASAC letter and response to charge questions on the draft PA
was held in Cary, NC, in October 2019 (84 FR 51555, September 30,
2019), and the CASAC provided its advice on the draft PA, including its
advice on the current primary and secondary PM standards, in a letter
to the then-Administrator dated December 16, 2019 (Cox, 2019b). With
regard to the primary standards, the CASAC recommended retaining the
current 24-hour PM2.5 and PM10 standards but did
not reach consensus on the adequacy of the current annual
PM2.5 standard. Some CASAC members expressed support for
retaining the current primary annual PM2.5 standard while
other members expressed support for revising that standard in order to
increase public health protection (Cox, 2019b, p. 1 of letter). These
views are described in greater detail in the letter to the then-
Administrator (Cox, 2019b) and in the notice of final rulemaking (85 FR
82706-82707, December 18, 2020), as well as below. With regard to the
secondary standards, the CASAC recommended retaining the current
standards. In response to the CASAC's comments, the 2020 final PA
incorporated a number of changes (Cox, 2019b, U.S. EPA, 2020b), as
described in detail in section I.C.5 of the 2020 proposal document (85
FR 24100, April 30, 2020).
a. 2020 Proposed and Final Actions
On April 14, 2020, the EPA proposed to retain all of the primary
and secondary PM standards, without revision. These proposed decisions
were published in the Federal Register on April 30, 2020 (85 FR 24094,
April 30, 2020). The EPA's final decision on the PM NAAQS was published
in the Federal Register on December 18, 2020 (85 FR 82684, December 18,
2020). In the 2020 rulemaking, the EPA retained the primary and
secondary PM2.5 and PM10 standards, without
revision. The then-Administrator's rationale for his decisions is
described in more detail in section II, III, and V below, and is
briefly summarized here.
In reaching his final decision to retain the primary annual and 24-
hour PM2.5 standards, the then-Administrator considered the
available scientific evidence, quantitative information, CASAC advice,
and public comments in his supporting rationale in the 2020 final
action (85 FR 82714, December 18, 2020). In so doing, he concluded that
the available controlled human exposure studies did not provide support
for additional public health protection against exposures to peak
PM2.5 concentrations, beyond the protection provided by the
combination of the current primary annual and 24-hour PM2.5
standards. He also noted that the available epidemiologic studies did
not indicate that associations in those studies are strongly influenced
by exposures to peak concentrations in the air quality distribution and
thus did not indicate the need for additional protection against short-
term exposures to peak PM2.5 concentrations. Accordingly,
and taking into account consensus CASAC advice to retain the current
primary 24-hour PM2.5 standard, the then-Administrator
concluded the primary 24-hour PM2.5 standard should be
retained.
With respect to the annual PM2.5 standard, the then-
Administrator recognized that important uncertainties and limitations
that were present in epidemiologic studies in previous
[[Page 16210]]
reviews remained in the evidence assessed in the 2019 ISA. In
considering the epidemiologic evidence, the then-Administrator noted
that: (1) The reported mean concentration in the majority of the key
U.S. epidemiologic studies using ground-based monitoring data are above
the level of the current annual standard; (2) the mean of the reported
study means (or medians) (i.e., 13.5 [micro]g/m\3\) is above the level
of the current primary annual PM2.5 standard of 12 [micro]g/
m\3\; (3) air quality analyses show the study means to be lower than
their corresponding design by 10-20%; and (4) that these analyses must
be considered in light of uncertainties inherent in the epidemiologic
evidence. The then-Administrator further considered other available
information, including the risk assessment, accountability studies, and
controlled human exposure studies, and found that, in considering all
of the evidence together along with advice from the CASAC, the suite of
primary PM2.5 standards were requisite to protect public
health with an adequate margin of safety, and should be retained,
without revision.
With regard to the primary PM10 standard, the then-
Administrator noted that the expanded body of evidence has broadened
the range of effects that have been linked with PM10-2.5
exposures. In light of that information, as well as continued
uncertainties in the evidence and advice from the CASAC to retain the
standard, the then-Administrator judged it appropriate to retain the
primary PM10 standard to provide the requisite degree of
public health protection against PM10-2.5 exposures,
regardless of location, source of origin, or particle composition (85
FR 82725, December 18, 2020).
With regard to the secondary PM standards, the then-Administrator
concluded that there was insufficient information available to
establish any distinct secondary PM standards to address climate and
materials effects of PM. For visibility effects, he found that in the
absence of a monitoring network for direct measurement of light
extinction, a calculated light extinction indicator that utilizes the
IMPROVE algorithms continued to provide a reasonable basis for defining
a target level of protection against PM-related visibility impairment.
He further found that a visibility index with a 24-hour averaging time
was reasonable based on its stability and suitability for representing
subdaily periods, and a form based on the 3-year average of annual 90th
percentile values was reasonable based on its stability and that it
represents the median of the 20 percent worst visibility days which are
targeted under the Regional Haze program. With regard to the level of a
visibility index, the then-Administrator judged it appropriate to
establish a target level of protection of 30 dv, reflecting the upper
end of the range of visibility impairment judged to be acceptable by at
least 50% of study participants in the available public preference
studies, taking into consideration the variability, limitations and
uncertainties of the public preference studies. The then-Administrator
judged that the secondary 24-hour PM2.5 standard with its
level of 35 [micro]g/m\3\ would provide at least the target level of
protection for visual air quality of 30 dv which he judged appropriate.
Accordingly, taking into consideration the advice of the CASAC to
retain the current secondary PM standards, the then-Administrator found
the current secondary standards provide the requisite degree of
protection and that they should be retained (85 FR 82742, December 18,
2020).
Following publication of the 2020 final action, several parties
filed petitions for review and petitions for reconsideration of the
EPA's final decision. The petitions for review were filed in the D.C.
Circuit and the Court consolidated the cases.\15\ Following EPA's
decision to reconsider the 2020 final decision, the Court ordered the
consolidated cases to be held in abeyance.
---------------------------------------------------------------------------
\15\ See California v. EPA, (D.C. Cir., No. 21-2014 consolidated
with Nos. 21-1027, 21-1054).
---------------------------------------------------------------------------
b. Reconsideration of the 2020 PM NAAQS Final Action
Executive Order 13990 directed review of certain agency actions (86
FR 7037, January 25, 2021).\16\ An accompanying fact sheet provided a
non-exclusive list of agency actions that agency heads should review in
accordance with that order, including the 2020 Particulate Matter NAAQS
Decision.\17\
---------------------------------------------------------------------------
\16\ See https://www.whitehouse.gov/briefing-room/presidential-actions/2021/01/20/executive-order-protecting-public-health-and-environment-and-restoring-science-to-tackle-climate-crisis/.
\17\ See https://www.whitehouse.gov/briefing-room/statements-releases/2021/01/20/fact-sheet-list-of-agency-actions-for-review/.
---------------------------------------------------------------------------
On June 10, 2021, the Agency announced its decision to reconsider
the 2020 PM NAAQS final action because the available scientific
evidence and technical information indicate that the current standards
may not be adequate to protect public health and welfare, as required
by the Clean Air Act.\18\ The Administrator reached this decision in
part based on the fact that the EPA noted that the 2020 PA concluded
that the scientific evidence and information called into question the
adequacy of the primary annual PM2.5 standard and supported
revising the level to below the current level of 12.0 [micro]g/m\3\
while retaining the primary 24-hour PM2.5 standard (U.S.
EPA, 2020b). The EPA also noted that the 2020 PA concluded that the
available scientific evidence and information supported retaining the
primary PM10 standard and secondary PM standards without
revision (U.S. EPA, 2020b).
---------------------------------------------------------------------------
\18\ The press release for this announcement is available at:
https://www.epa.gov/newsreleases/epa-reexamine-health-standards-harmful-soot-previous-administration-left-unchanged.
---------------------------------------------------------------------------
The EPA staff conclusions detailed in the 2020 PA in combination
with the CASAC advice that informed the Administrator's decisions
regarding the 2020 final action, studies highlighted by public comments
on the 2020 proposal, and the numerous studies published since the
literature cutoff date of the 2019 ISA all informed the scope of the
reconsideration.
In its review of the 2019 draft PA, some members of the CASAC had
recommended that greater attention should be given to accountability
studies and epidemiologic studies that employ alternative methods for
confounder control (also referred to as causal inference or causal
modeling studies) in order to ``more fully account for effects of
confounding, measurement and estimation errors, model uncertainty, and
heterogeneity'' in epidemiologic studies (Cox, 2019b, p. 8 of consensus
responses). In addition, public commenters submitted a number of recent
studies published after the literature cutoff date for the 2019 ISA
that would have been considered within the scope of the 2019 ISA. While
the EPA provisionally considered these studies in responding to public
comments,\19\ it was determined that, at the time of the 2020 final
action, these studies were generally consistent with the evidence
assessed in the 2019 ISA (85 FR 82690, December 18, 2020; U.S. EPA,
2020a). As such, and consistent with previous NAAQS reviews, the EPA
concluded that the new studies did not materially change any of the
broad scientific conclusions regarding the health and welfare effects
of PM in ambient air made in the air quality criteria, and therefore,
reopening of the air quality criteria was not warranted (85 FR 82691,
December 18, 2020). However, at that time, the EPA
[[Page 16211]]
recognized that its ``provisional consideration of these studies did
not and could not provide the kind of in-depth critical review'' (85 FR
82690, December 18, 2020) that studies undergo in the development of an
ISA.
---------------------------------------------------------------------------
\19\ The list of provisionally considered studies is included in
Appendix A to the 2020 Response to Comments document (U.S. EPA,
2020a).
---------------------------------------------------------------------------
In preparing to reconsider the 2020 final decision for the PM
NAAQS, the Agency revisited the need to reopen the air quality
criteria, given the amount of time that had passed since the literature
cutoff date of the 2019 ISA (i.e., approximately January 2018) and the
volume of literature that had become available, including those studies
provisionally considered in responding to comments in 2020. In so
doing, the EPA preliminarily concluded that at least some of these
studies were likely to be relevant to its reconsideration of the air
quality criteria and the PM NAAQS and that, in considering public
comments on any proposed decisions for the reconsideration, these
studies were likely to be raised by public commenters and would
potentially warrant a reopening of the air quality criteria. For
example, on February 16, 2021, the EPA received two petitions to
reconsider the PM NAAQS. One petition objected to the EPA's provisional
consideration of studies submitted in public comments on the 2020
proposal and suggested that the provisional consideration was
inadequate because the studies could be important in determining
whether the existing standards are adequately protective. See, Petition
for Reconsideration of National Ambient Air Quality Standards for
Particulate Matter, submitted by American Lung Association, et al,
dated Feb. 16, 2020. The other petition identified a number of new
studies, including one epidemiologic study that was published after the
provisional consideration was completed that could further inform the
concern expressed by the CASAC that associations reported in
epidemiologic studies do not adequately account for ``uncontrolled
confounding and other potential sources of error and bias.'' See
Petition for Reconsideration of ``Review of the National Ambient Air
Quality Standards for Particulate Matter,'' submitted by the State of
California, dated Feb. 16, 2020. This was also an uncertainty noted by
the then-Administrator in the 2020 decision, who also recognized ``that
methodological study designs to address confounding, such as causal
inference methods, are an emerging field of study.'' Thus, the Agency
concluded it was appropriate to reconsider not only the standards but
also the air quality criteria, in light of public comments during the
2020 PM NAAQS proposal and recent studies published since the cutoff
date of the 2019 ISA, as reflected in petitions. In deciding to reopen
the air quality criteria, the Agency concluded it was reasonable to
focus on studies that were most likely to inform decisions on the
appropriate standard, but not to reassess areas which, based on the
assessment of available science published since the cutoff date of the
2019 ISA and through 2021, were judged unlikely to have new information
that would be useful for the Administrator's decision making. The
Agency accordingly announced that, in support of the reconsideration,
it would develop a supplement to the 2019 ISA and a revised PA.
The EPA also explained that the draft ISA Supplement and draft PA
would be reviewed at a public meeting by the CASAC, and the public
would have opportunities to comment on these documents during the CASAC
review process, as well as to provide input during the rulemaking
through the public comment process and public hearings on the proposed
rulemaking.
On March 31, 2021, the Administrator announced his decision to
reestablish the membership of the CASAC to ``ensure the agency received
the best possible scientific insight to support our work to protect
human health and the environment.'' \20\ Consistent with this
memorandum, a call for nominations of candidates to the EPA's chartered
CASAC was published in the Federal Register (86 FR 17146, April 1,
2021). On June 17, 2021, the Administrator announced his selection of
the seven members to serve on the chartered CASAC.21 22
Additionally, a call for nominations of candidates to a PM-specific
panel was published in the Federal Register (86 FR 33703, June 25,
2021). The members of the PM CASAC panel were announced on August 30,
2021.\23\
---------------------------------------------------------------------------
\20\ The press release for this announcement is available at:
https://www.epa.gov/newsreleases/administrator-regan-directs-epa-reset-critical-science-focused-federal-advisory.
\21\ The press release for this announcement is available at:
https://www.epa.gov/newsreleases/epa-announces-selections-charter-members-clean-air-scientific-advisory-committee.
\22\ The list of members of the chartered CASAC and their
biosketches are available at: https://casac.epa.gov/ords/sab/r/sab_apex/casac/mems?p14_committeeon=2021%20CASAC%20PM%20Panel&session=17433386035954
.
\23\ The list of members of the PM CASAC panel and their
biosketches are available at: https://casac.epa.gov/ords/sab/f?p=105:14:9979229564047:::14:P14_COMMITTEEON:2021%20CASAC%20PM%20Panel.
---------------------------------------------------------------------------
The draft ISA Supplement was released in September 2021 (U.S. EPA,
2021a; 86 FR 54186, September 30, 2021), and included a discussion of
the rationale and scope of the Supplement. As explained therein, the
ISA Supplement focuses on a thorough evaluation of some studies that
became available after the literature cutoff date of the 2019 ISA that
could either further inform the adequacy of the current PM NAAQS or
address key scientific topics that have evolved since the literature
cutoff date for the 2019 ISA. In selecting the health effects to
evaluate within the ISA Supplement, the EPA focused on health effects
for which the evidence supported a ``causal relationship'' because
those were the health effects that were most useful in informing
conclusions in the 2020 PA (U.S. EPA, 2022a, section 1.2.1).\24\
Consistent with the rationale for the focus on certain health effects,
in selecting the non-ecological welfare effects to evaluate within the
ISA Supplement, the EPA focused on the non-ecological welfare effects
for which the evidence supported a ``causal relationship'' and for
which quantitative analyses could be supported by the evidence because
those were the welfare effects that were most useful in informing
conclusions in the 2020 PA.\25\ Specifically, for non-ecological
welfare effects, the focus within the ISA Supplement is on visibility
effects. The ISA Supplement also considers recent health effects
evidence that addresses key scientific topics where the literature has
evolved since the 2020 review was completed,
[[Page 16212]]
specifically since the literature cutoff date for the 2019 ISA.\26\
---------------------------------------------------------------------------
\24\ As described in section 1.2.1 of the ISA Supplement: ``In
considering the public health protection provided by the current
primary PM2.5 standards, and the protection that could be
provided by alternatives, [the U.S. EPA, within the 2020 PM PA]
emphasized health outcomes for which the ISA determined that the
evidence supports either a `causal' or a `likely to be causal'
relationship with PM2.5 exposures'' (U.S. EPA, 2020b).
Although the 2020 PA initially focused on this broader set of
evidence, the basis of the discussion on potential alternative
standards primarily focused on health effect categories where the
2019 PM ISA concluded a `causal relationship' (i.e., short- and
long-term PM2.5 exposure and cardiovascular effects and
mortality) as reflected in Figures 3-7 and 3-8 of the 2020 PA (U.S.
EPA, 2020b).''
\25\ As described in section 1.2.1 of the ISA Supplement: ``The
2019 PM ISA concluded a `causal relationship' for each of the
welfare effects categories evaluated (i.e., visibility, climate
effects and materials effects). While the 2020 PA considered the
broader set of evidence for these effects, for climate effects and
material effects, it concluded that there remained `substantial
uncertainties with regard to the quantitative relationships with PM
concentrations and concentration patterns that limit[ed] [the]
ability to quantitatively assess the public welfare protection
provided by the standards from these effects' (U.S. EPA, 2020b).''
\26\ These key scientific topics include experimental studies
conducted at near-ambient concentrations, epidemiologic studies that
employed alternative methods for confounder control or conducted
accountability analyses, studies that assess the relationship
between PM2.5 exposure and severe acute respiratory
syndrome coronavirus 2 (SARS-CoV-2) infection and coronavirus
disease 2019 (COVID-19) death; and in accordance with recent EPA
goals on addressing environmental justice, studies that examine
disparities in PM2.5 exposure and the risk of health
effects by race/ethnicity or socioeconomic status (SES) (U.S. EPA,
2022a, section 1.2.1).
---------------------------------------------------------------------------
Building on the rationale presented in section 1.2.1, the ISA
Supplement considers peer-reviewed studies published from approximately
January 2018 through March 2021 that meet the following criteria:
Health Effects
[cir] U.S. and Canadian epidemiologic studies for health effect
categories where the 2019 ISA concluded a ``causal relationship''
(i.e., short- and long-term PM2.5 exposure and
cardiovascular effects and mortality).
[ssquf] U.S. and Canadian epidemiologic studies that employed
alternative methods for confounder control or conducted accountability
analyses (i.e., examined the effect of a policy on reducing
PM2.5 concentrations).
Welfare Effects
[cir] U.S. and Canadian studies that provide new information on
public preferences for visibility impairment and/or developed
methodologies or conducted quantitative analyses of light extinction.
Key Scientific Topics
[cir] Experimental studies (i.e., controlled human exposure and
animal toxicological) conducted at near-ambient PM2.5
concentrations experienced in the U.S.
[cir] U.S.- and Canadian-based epidemiologic studies that examined
the relationship between PM2.5 exposures and severe acute
respiratory syndrome coronavirus 2 (SARS-CoV-2) infection and
coronavirus disease 2019 (COVID-19) death.
[cir] At-Risk Populations.
[ssquf] U.S.- and Canadian-based epidemiologic or exposure studies
examining potential disparities in either PM2.5 exposures or
the risk of health effects by race/ethnicity or socioeconomic status
(SES).
Given the narrow scope of the ISA Supplement, it is important to
recognize that the evaluation does not encompass the full
multidisciplinary evaluation presented within the 2019 ISA that would
result in weight-of-evidence conclusions on causality (i.e., causality
determinations). The ISA Supplement critically evaluates and provides
key study-specific information for those recent studies deemed to be of
greatest significance for informing preliminary conclusions on the PM
NAAQS in the context of the body of evidence and scientific conclusions
presented in the 2019 ISA.
In developing a revised PA to support the reconsideration, the EPA
considered the available scientific evidence, including the evidence
presented in the 2019 ISA and ISA Supplement. The 2022 PA considered
the quantitative and technical information presented in the 2020 PA, in
addition to new and updated analyses conducted since the 2020 final
decision. For those health and welfare effects for which the ISA
Supplement evaluated recently available studies (i.e.,
PM2.5-related health effects and visibility effects), new
updated quantitative analyses were conducted as a part of the
development of the 2022 PA. The newly available scientific and
technical information presented in the 2022 PA were considered in
reaching conclusions regarding the adequacy of the current standards
and any potential alternative standards. For those health and welfare
effects for which newly available scientific and technical information
were not evaluated (i.e., PM10-2.5-related health effects
and non-visibility welfare effects), the conclusions presented in the
2022 PA rely heavily on the information that supported the conclusions
in the 2020 PA.
The CASAC PM panel met at a virtual public meeting in November 2021
to review the draft ISA Supplement (86 FR 52673, September 22, 2021). A
virtual public meeting was then held in February 2022, and during this
meeting the chartered CASAC considered the CASAC PM panel's draft
letter to the Administrator on the draft ISA Supplement (87 FR 958,
January 7, 2022).
The chartered CASAC provided its advice on the draft ISA Supplement
in a letter to the EPA Administrator dated March 18, 2022 (Sheppard,
2022b). In its review of the draft ISA Supplement, the CASAC noted that
they found ``the Draft ISA Supplement to be a well-written,
comprehensive evaluation of the new scientific information published
since the 2019 PM ISA'' (Sheppard, 2022b, p. 2 of letter). Furthermore,
the CASAC stated that ``the final Integrated Science Assessment (ISA)
Supplement . . . deserve[s] the Administrator's full consideration and
[is] adequate for rulemaking'' (Sheppard, 2022b, p. 2 of letter). The
CASAC generally endorsed EPA's decisions regarding the limited scope of
the draft ISA Supplement, stating that ``this limitation [on scope] is
appropriate for the targeted purpose of the Draft ISA Supplement''
although the CASAC noted it would not be appropriate for ISAs
generally, and recommended that the EPA provide additional
acknowledgment and explanation for the limited scope (Sheppard, 2022b,
p. 2 of letter; see also pp. 2-3 of consensus responses). The EPA
specifically noted in the final ISA Supplement, which was released in
May 2022 (U.S. EPA, 2022a; hereafter referred to as the ISA Supplement
throughout this document) that the ``targeted approach to developing
the Supplement to the 2019 PM ISA for the purpose of reconsidering the
2020 PM NAAQS decision does not reflect a change to EPA's approach for
developing ISAs for NAAQS reviews.'' Thus, the evidence presented
within the 2019 ISA, along with the targeted identification and
evaluation of new scientific information in the ISA Supplement,
provides the scientific basis for the reconsideration of the 2020 PM
NAAQS final decision.
The draft PA was released in October 2021 (86 FR 56263, October 8,
2021). The CASAC PM panel met at a virtual public meeting in December
2021 to review the draft PA (86 FR 52673, September 22, 2021). A
virtual public meeting was then held in February 2022 and March 2022,
and during this meeting the chartered CASAC considered the CASAC PM
panel's draft letter to the Administrator on the draft PA (87 FR 958,
January 7, 2022). The chartered CASAC provided its advice on the draft
PA in a letter to the EPA Administrator dated March 18, 2022 (Sheppard,
2022a). The EPA took steps to address these comments in revising and
finalizing the PA. The 2022 PA considers the scientific evidence
presented in the 2019 ISA and ISA Supplement and considers the
quantitative and technical information presented in the 2020 PA, along
with updated and newly available analyses since the completion of the
2020 review. For those health and welfare effects for which the ISA
Supplement evaluated recently available evidence and for which updated
quantitative analyses were supported (i.e., PM2.5-related
health effects and visibility effects), the 2022 PA includes
consideration of this newly available scientific and technical
information in reaching preliminary conclusions. For those health and
welfare effects for which newly available scientific and technical
[[Page 16213]]
information were not evaluated (i.e., PM10-2.5-related
health effects and non-visibility effects), the conclusions presented
in the 2022 PA rely heavily on the information that supported the
conclusions in the 2020 PA. The final PA was released in May 2022 (U.S.
EPA, 2022b; hereafter referred to as the 2022 PA throughout this
document).
Drawing from his consideration of the scientific evidence assessed
in the 2019 ISA and ISA Supplement and the analyses in the 2022 PA,
including the uncertainties in the evidence and analyses, and from his
consideration of advice from the CASAC, on January 5, 2023, the
Administrator proposed to revise the level of the primary annual
PM2.5 standard and to retain the primary 24-hour
PM2.5 standard, the primary 24-hour PM10
standard, and the secondary PM standards. These proposed decisions were
published in the Federal Register on January 27, 2023 (88 FR 5558,
January 27, 2023). The EPA held a multi-day virtual public hearing on
February 21-23, 2023 (88 FR 6215, January 31, 2023). In total, the EPA
received nearly 700,000 comments on the proposal from members of the
public by the close of the public comment period on March 28, 2023.
Major issues raised in the public comments are discussed throughout the
preamble of this final action. A more detailed summary of all
significant comments, along with the EPA's responses (henceforth
``Response to Comments'' document), can be found in the docket for this
rulemaking (Docket No. EPA-HQ-OAR-2015-0072).
As in prior reviews, the EPA is basing its decision in this
reconsideration on studies and related information in the air quality
criteria, which have undergone CASAC and public review. These studies
assessed in the 2019 ISA \27\ and ISA Supplement \28\ and the 2022 PA,
and the integration of the scientific evidence presented in them, have
undergone extensive critical review by the EPA, the CASAC, and the
public. Decisions on the NAAQS should be based on studies that have
been rigorously assessed in an integrative manner not only by the EPA
but also by the statutorily mandated independent scientific advisory
committee, as well as the public review that accompanies this process.
It is for this reason that the EPA preliminarily concluded that the
scientific evidence available since the completion of the 2019 ISA,
including those raised in public comments on the proposal in 2020,
warranted a partial reopening of the air quality criteria and prepared
an ISA Supplement to enable the EPA, the CASAC, and the public to
consider them further. Some commenters have referred to and discussed
additional individual scientific studies on the health effects of PM
that were not included in the 2019 ISA or ISA Supplement (``new
studies'') and that have not gone through this comprehensive review
process. In considering and responding to comments for which such
``new'' studies were cited in support, the EPA has provisionally
considered the cited studies in the context of the findings of the 2019
ISA and ISA Supplement. The EPA's provisional consideration of these
studies did not and could not provide the kind of in-depth critical
review described above, but rather was focused on determining whether
they warranted further reopening the review of the air quality criteria
to enable the EPA, the CASAC, and the public to consider them further.
---------------------------------------------------------------------------
\27\ In addition to the 2020 review's opening ``call for
information'' (79 FR 71764, December 3, 2014), the 2019 ISA
identified and evaluated studies and reports that have undergone
scientific peer review and were published or accepted for
publication between January 1, 2009, through approximately January
2018 (U.S. EPA, 2019a, p. ES-2). References that are cited in the
2019 ISA, the references that were considered for inclusion but not
cited, and electronic links to bibliographic information and
abstracts can be found at: https://hero.epa.gov/hero/particulate-matter.
\28\ As described above, the ISA Supplement represents an
evaluation of recent studies that are of greatest policy relevance
and utility to the reconsideration of the 2020 final decision on the
PM NAAQS (U.S. EPA, 2022a).
---------------------------------------------------------------------------
This approach, and the decision to rely on the studies and related
information in the air quality criteria, which have undergone CASAC and
public review, is consistent with the EPA's practice in prior NAAQS
reviews and its interpretation of the requirements of the CAA. Since
the 1970 amendments, the EPA has taken the view that NAAQS decisions
are to be based on scientific studies and related information that have
been assessed as a part of the pertinent air quality criteria, and the
EPA has consistently followed this approach. This longstanding
interpretation was strengthened by new legislative requirements enacted
in 1977, which added section 109(d)(2) of the Act concerning CASAC
review of air quality criteria. See 71 FR 6114, 61148 (October 17,
2006, final decision on review of NAAQS for particulate matter) for a
detailed discussion of this issue and the EPA's past practice.
As discussed in the EPA's 1993 decision not to review the
O3 NAAQS, ``new'' studies may sometimes be of such
significance that it is appropriate to delay a decision in a NAAQS
review and to supplement the pertinent air quality criteria so the
studies can be taken into account (58 FR 13013-13014, March 9, 1993).
In the present case, the EPA decided to partially reopen the air
quality criteria and prepared an ISA Supplement as a part of the
reconsideration to facilitate evaluation of these studies by the EPA,
the CASAC, and the public. The narrow scope of the ISA Supplement is
supported by EPA's provisional consideration of ``new'' studies
submitted in response to public comments on the 2020 proposal which
concluded that, taken in context, the ``new'' information and findings
do not materially change any of the broad scientific conclusions
regarding the health and welfare effects of PM in ambient air made in
the air quality criteria. Therefore, a full reopening of the air
quality criteria was not warranted to assess the health and welfare
effects of PM for purposes of the review.
Accordingly, the EPA is basing the final decisions in this
reconsideration on the studies and related information included in the
PM air quality criteria (including the 2019 PM ISA and ISA Supplement)
that have undergone rigorous review by the EPA, the CASAC, and the
public. The EPA will consider these ``new'' studies for inclusion in
the air quality criteria for the next PM NAAQS review, which the EPA
expects to begin soon after the conclusion of this reconsideration and
which will provide the opportunity to fully assess these studies
through a more rigorous review process involving the EPA, the CASAC,
and the public.
D. Air Quality Information
This section provides a summary of basic information related to PM
ambient air quality. It summarizes information on the distribution of
particle size in ambient air (section I.D.1), sources and emissions
contributing to PM in the ambient air (section I.D.2), monitoring
ambient PM in the U.S. (section I.D.3), ambient PM concentrations and
trends in the U.S. (I.D.4), characterizing ambient PM2.5
concentrations for exposure (section I.D.5), and background PM (section
I.D.6). Additional detail on PM air quality can be found in Chapter 2
of the 2022 PA (U.S. EPA, 2022b).
1. Distribution of Particle Size in Ambient Air
In ambient air, PM is a mixture of substances suspended as small
liquid and/or solid particles (U.S. EPA, 2019a, section 2.2) and
distinct health and welfare effects have been linked with exposures to
particles of different sizes. Particles in the atmosphere range in size
from less than 0.01 to more than 10 [mu]m
[[Page 16214]]
in diameter (U.S. EPA, 2019a, section 2.2). The EPA defines
PM2.5, also referred to as fine particles, as particles with
aerodynamic diameters generally less than or equal to 2.5 [mu]m. The
size range for PM10-2.5, also called coarse or thoracic
coarse particles, includes those particles with aerodynamic diameters
generally greater than 2.5 [mu]m and less than or equal to 10 [mu]m.
PM10, which is comprised of both fine and coarse fractions,
includes those particles with aerodynamic diameters generally less than
or equal to 10 [mu]m. In addition, ultrafine particles (UFP) are often
defined as particles with a diameter of less than 0.1 [mu]m based on
physical size, thermal diffusivity or electrical mobility (U.S. EPA,
2019a, section 2.2). Atmospheric lifetimes are generally longest for
PM2.5, which often remains in the atmosphere for days to
weeks (U.S. EPA, 2019a, Table 2-1) before being removed by wet or dry
deposition, while atmospheric lifetimes for UFP and PM10-2.5
are shorter and are generally removed from the atmosphere within hours,
through wet or dry deposition (U.S. EPA, 2019a, Table 2-1; U.S. EPA,
2022b, section 2.1).
2. Sources and Emissions Contributing to PM in the Ambient Air
PM is composed of both primary (directly emitted particles) and
secondary particles. Primary PM is derived from direct particle
emissions from specific PM sources while secondary PM originates from
gas-phase precursor chemical compounds present in the atmosphere that
have participated in new particle formation or condensed onto existing
particles (U.S. EPA, 2019a, section 2.3). As discussed further in the
2019 ISA (U.S. EPA, 2019a, section 2.3.2.1), secondary PM is formed in
the atmosphere by photochemical oxidation reactions of both inorganic
and organic gas-phase precursors. Precursor gases include sulfur
dioxide (SO2), nitrogen oxides (NOX), and
volatile organic compounds (VOC) (U.S. EPA, 2019a, section 2.3.2.1).
Ammonia also plays an important role in the formation of nitrate PM by
neutralizing sulfuric acid and nitric acid. Sources and emissions of PM
are discussed in more detail the 2022 PA (U.S. EPA, 2022b, section
2.1.1). Briefly, anthropogenic sources of PM include both stationary
(e.g., fuel combustion for electricity production and other purposes,
industrial processes, agricultural activities) and mobile (e.g.,
diesel- and gasoline-powered highway vehicles and other engine-driven
sources) sources. Natural sources of PM include dust from the wind
erosion of natural surfaces, sea salt, wildfires, primary biological
aerosol particles (PBAP) such as bacteria and pollen, oxidation of
biogenic hydrocarbons, such as isoprene and terpenes to produce
secondary organic aerosol (SOA), and geogenic sources, such as sulfate
formed from volcanic production of SO2. Wildland fire, which
encompass both wildfire and prescribed fire, accounts for 44% of
emissions of primary PM2.5 emissions (U.S. EPA, 2021b).
Emissions from wildfire comprises 29% of primary PM2.5
emissions.
In recent years, the frequency and magnitude of wildfires have
increased (U.S. EPA, 2019a). The magnitude of the public health impact
of wildfires is substantial both because of the increase in
PM2.5 concentrations as well as the duration of the wildfire
smoke season, which is considered to range from May to November.
Wildfire can make a large contribution to air pollution (including
PM2.5), and wildfire events can threaten public safety and
life. The impacts of wildfire events can be mitigated through
management of wildland vegetation, including through prescribed fire.
Prescribed fire (and some wildfires) can mimic the natural processes
necessary to maintain fire-dependent ecosystems, minimizing
catastrophic wildfires and the risks they pose to safety, property and
air quality (see, e.g., 81 FR 58010, 58038, August 24, 2016). The EPA
views the strategic use of prescribed fire as an important tool for
reducing wildfire risk and the severity of wildfires and wildfire smoke
(88 FR, 54118, 54126, August 9, 2023).\29\ As noted in the PM NAAQS
proposal, agencies have efforts in place to reduce the frequency and
severity of human-caused wildfires (88 FR 5570, January 27, 2023).
---------------------------------------------------------------------------
\29\ See also: https://www.usda.gov/sites/default/files/documents/usda-epa-doi-cdc-mou.pdf.
---------------------------------------------------------------------------
Wildfire events produce high PM emissions that may impact the PM
concentrations in ambient air to the extent that the concentrations
result in an exceedance or violation which may affect the design value
in a given area. The EPA's Exceptional Events Rule (81 FR 68216,
October 3, 2016) describes the process by which air agencies may
request to exclude `event-influenced' data caused by exceptional
events, which can include wildfires and prescribed fires on wildland.
The EPA has issued guidance specifically addressing exceptional events
demonstrations for both wildfires and prescribed fires on wildland.
These documents are available on EPA's Exceptional Events Program
website.\30\ The EPA will develop fire-related exceptional events
implementation tools, including updates as needed to existing guidance
to facilitate more efficient processing of PM2.5-related
exceptional events demonstrations for both the 24-hour and annual
standards.
---------------------------------------------------------------------------
\30\ See: https://www.epa.gov/air-quality-analysis/final-2016-exceptional-events-rule-supporting-guidance-documents-updated-faqs.
---------------------------------------------------------------------------
3. Monitoring of Ambient PM
To promote uniform application of the air quality standards set
forth under the CAA and to achieve the degree of public health and
welfare protection intended for the NAAQS, the EPA establishes PM
Federal Reference Methods (FRMs) for both PM10 and
PM2.5 in appendices J and L to 40 CFR part 50, both of which
were amended following the 2006 and 2012 PM NAAQS reviews. The current
PM monitoring network relies on FRMs and automated continuous Federal
Equivalent Methods (FEMs) approved pursuant to 40 CFR part 53, in part
to support changes necessary for implementation of the revised PM
standards. Additionally, 40 CFR part 58, appendices A through E, detail
the requirements to measure ambient air quality and report ambient air
quality data and related information. More information on PM ambient
monitoring networks is available in section 2.2 of the 2022 PA (U.S.
EPA, 2022b).
The PM2.5 monitoring program is one of the major ambient
air monitoring programs with a robust, nationally consistent network of
ambient air monitoring sites providing mass and/or chemical speciation
measurements. 40 CFR part 58, appendix D, section 4.7 provides the
applicable PM2.5 network design criteria. For most urban
locations, PM2.5 monitors are sited at the neighborhood
scale,\31\ where PM2.5 concentrations are reasonably
homogeneous throughout an entire urban sub-region. In each CBSA with a
monitoring requirement, at least one PM2.5 monitoring
station representing
[[Page 16215]]
area-wide air quality is sited in an area of expected maximum
concentration.\32\ By ensuring the area of expected maximum
concentration in a CBSA has a site compared to both the annual and 24-
hour NAAQS, all other similar locations are thus protected. Sites that
represent relatively unique microscale, localized hot-spot, or unique
middle scale impact sites are only eligible for comparison to the 24-
hour PM2.5 NAAQS.
---------------------------------------------------------------------------
\31\ For PM2.5, neighborhood scale is defined at 40
CFR part 58, appendix D, 4.7.1(c)(3) as follows: Measurements in
this category would represent conditions throughout some reasonably
homogeneous urban sub-region with dimensions of a few kilometers and
of generally more regular shape than the middle scale. Homogeneity
refers to the particulate matter concentrations, as well as the land
use and land surface characteristics. Much of the PM2.5
exposures are expected to be associated with this scale of
measurement. In some cases, a location carefully chosen to provide
neighborhood scale data would represent the immediate neighborhood
as well as neighborhoods of the same type in other parts of the
city. PM2.5 sites of this kind provide good information
about trends and compliance with standards because they often
represent conditions in areas where people commonly live and work
for periods comparable to those specified in the NAAQS. In general,
most PM2.5 monitoring in urban areas should have this
scale.
\32\ 40 CFR part 58, app. D, 4.7.1(b)(2).
---------------------------------------------------------------------------
Under 40 CFR part 50, appendix L, and 40 CFR part 53, and 40 CFR
part 58 appendix D there are three main methods components of the
PM2.5 monitoring program: filter-based FRMs measuring
PM2.5 mass, FEMs measuring PM2.5 mass, and other
samplers used to collect the aerosol used in subsequent laboratory
analysis for measuring PM2.5 chemical speciation. The FRMs
are primarily used for comparison to the NAAQS, but also serve other
important purposes, such as developing trends and evaluating the
performance of FEMs. PM2.5 FEMs are typically continuous
methods used to support forecasting and reporting of the Air Quality
Index (AQI) but are also used for comparison to the NAAQS. Samplers
that are part of the Chemical Speciation Network (CSN) and Interagency
Monitoring of Protected Visual Environments (IMPROVE) network are used
to provide chemical composition of the aerosol and serve a variety of
objectives. More detail on of each of these components of the
PM2.5 monitoring program and of recent changes to
PM2.5 monitoring requirements are described in detail in the
2022 PA (U.S. EPA, 2022b, section 2.2.3).
4. Ambient Concentrations and Trends
This section summarizes available information on recent ambient PM
concentrations in the U.S. and on trends in PM air quality. Sections
I.D.4.a and I.D.4.b summarize information on PM2.5 mass and
components, respectively. Section I.D.4.c summarizes information on
PM10. Sections I.D.4.d and I.D.4.e summarize the more
limited information on PM10-2.5 and UFP, respectively.
Additional detail on PM air quality and trends can be found in the 2022
PA (U.S. EPA, 2022b, section 2.3).
a. PM2.5 mass
At monitoring sites in the U.S., annual PM2.5
concentrations from 2017 to 2019 averaged 8.0 [mu]g/m\3\ (with the 10th
and 90th percentiles at 5.9 and 10.0 [mu]g/m\3\,
respectively) and the 98th percentiles of 24-hour concentrations
averaged 21.3 [mu]g/m\3\ (with the 10th and 90th percentiles at 14.0
and 29.7 [mu]g/m\3\, respectively) (U.S. EPA, 2022b, section 2.3.2.1).
The highest ambient PM2.5 concentrations occur in the
western U.S., particularly in California and the Pacific Northwest
(U.S. EPA, 2022b, Figure 2-15). Much of the eastern U.S. has lower
ambient concentrations, with annual average concentrations generally at
or below 12.0 [mu]g/m\3\ and 98th percentiles of 24-hour concentrations
generally at or below 30 [mu]g/m\3\ (U.S. EPA, 2022b, section 2.3.2.1).
Recent ambient PM2.5 concentrations reflect the
substantial reductions that have occurred across much of the U.S. (U.S.
EPA, 2022b, section 2.3.2.1). From 2000 to 2019, national annual
average PM2.5 concentrations declined from 13.5 [mu]g/m\3\
to 7.6 [mu]g/m\3\, a 43% decrease (U.S. EPA, 2022b, section
2.3.2.1).\33\ These declines have occurred at urban and rural
monitoring sites, although urban PM2.5 concentrations remain
consistently higher than those in rural areas (Chan et al., 2018) due
to the impact of local sources in urban areas. Analyses at individual
monitoring sites indicate that declines in ambient PM2.5
concentrations have been most consistent across the eastern U.S. and in
parts of coastal California, where both annual average and 98th
percentiles of 24-hour concentrations declined significantly (U.S. EPA,
2022b, section 2.3.2.1). In contrast, trends in ambient
PM2.5 concentrations have been less consistent over much of
the western U.S., with no significant changes since 2000 observed at
some sites in the Pacific Northwest, the northern Rockies and plains,
and the Southwest, particularly for 98th percentiles of 24-hour
concentrations (U.S. EPA, 2022b, section 2.3.2.1). As noted below, some
sites in the northwestern U.S. and California, where wildfire have been
relatively common in recent years, have experienced high concentrations
over shorter periods (i.e., 2-hour averages).
---------------------------------------------------------------------------
\33\ See https://www.epa.gov/air-trends/particulate-matter-pm25-trends for up-to-date PM2.5 trends information.
---------------------------------------------------------------------------
The recent deployment of PM2.5 monitors near major roads
in large urban areas provides information on PM2.5
concentrations near an important emissions source. For 2016-2018, Gantt
et al. (2021) reported that 52% and 24% of the time near-road sites
reported the highest annual and 24-hour PM2.5 design value
\34\ in the CBSA, respectively. Of the CBSAs with the highest annual
design values at near-road sites reported by Gantt et al. (2021), those
design values were, on average, 0.8 [micro]g/m\3\ higher than at the
highest measuring non-near-road sites (range is 0.1 to 2.1 [micro]g/
m\3\ higher at near-road sites). Although most near-road monitoring
sites do not have sufficient data to evaluate long-term trends in near-
road PM2.5 concentrations, analyses of the data at one near-
road-like site in Elizabeth, NJ, \35\ show that the annual average
near-road increment has generally decreased between 1999 and 2017 from
about 2.0 [mu]g/m\3\ to about 1.3 [mu]g/m\3\ (U.S. EPA, 2022b, section
2.3.2.1).
---------------------------------------------------------------------------
\34\ A design value is considered valid if it meets the data
handling requirements given in appendix N to 40 CFR part 50.
\35\ The Elizabeth Lab site in Elizabeth, NJ, is situated
approximately 30 meters from travel lanes of the Interchange 13 toll
plaza of the New Jersey Turnpike and within 200 meters of travel
lanes for Interstate 278 and the New Jersey Turnpike.
---------------------------------------------------------------------------
Ambient PM2.5 concentrations can exhibit a diurnal cycle
that varies due to impacts from intermittent emission sources,
meteorology, and atmospheric chemistry. The PM2.5 monitoring
network in the U.S. has an increasing number of continuous FEM monitors
reporting hourly PM2.5 mass concentrations that reflect this
diurnal variation. The 2019 ISA describes a two-peaked diurnal pattern
in urban areas, with morning peaks attributed to rush-hour traffic and
afternoon peaks attributed to a combination of rush hour traffic,
decreasing atmospheric dilution, and nucleation (U.S. EPA, 2019a,
section 2.5.2.3, Figure 2-32). Because a focus on annual average and
24-hour average PM2.5 concentrations could mask subdaily
patterns, and because some health studies examine PM exposure durations
shorter than 24-hours, it is useful to understand the broader
distribution of subdaily PM2.5 concentrations across the
U.S. The 2022 PA presents information on the frequency distribution of
2-hour average PM2.5 mass concentrations from all FEM
PM2.5 monitors in the U.S. for 2017-2019. At sites meeting
the current primary PM2.5 standards, these 2-hour
concentrations generally remain below 10 [mu]g/m\3\, and rarely exceed
30 [mu]g/m\3\. Two-hour concentrations are higher at sites violating
the current standards, generally remaining below 16 [mu]g/m\3\ and
rarely exceeding 80 [mu]g/m\3\ (U.S. EPA, 2022b, section 2.3.2.2.3).
The extreme upper end of the distribution of 2-hour PM2.5
concentrations is shifted higher during the warmer months, generally
corresponding to the period of peak wildfire frequency (April to
September) in the U.S. At sites meeting the current primary standards,
the highest 2-hour concentrations measured rarely occur outside of the
period of peak wildfire frequency. Most of the sites measuring
[[Page 16216]]
these very high concentrations are in the northwestern U.S. and
California, where wildfires have been relatively common in recent years
(see U.S. EPA, 2022b, Appendix A, Figure A-1). When the period of peak
wildfire frequency is excluded from the analysis, the extreme upper end
of the distribution is reduced (U.S. EPA, 2022b, section 2.3.2.2.3).
b. PM2.5 Components
Based on recent air quality data, the major chemical components of
PM2.5 have distinct spatial distributions. Sulfate
concentrations tend to be highest in the eastern U.S., while in the
Ohio Valley, Salt Lake Valley, and California nitrate concentrations
are highest, and relatively high concentrations of organic carbon are
widespread across most of the continental U.S. (U.S. EPA, 2022b,
section 2.3.2.3). Elemental carbon, crustal material, and sea salt are
found to have the highest concentrations in the northeast U.S.,
southwest U.S., and coastal areas, respectively.
An examination of PM2.5 composition trends can provide
insight into the factors contributing to overall reductions in ambient
PM2.5 concentrations. The biggest change in PM2.5
composition that has occurred in recent years is the reduction in
sulfate concentrations due to reductions in SO2 emissions.
Between 2000 and 2015, the nationwide annual average sulfate
concentration decreased by 17% at urban sites and 20% at rural sites.
This change in sulfate concentrations is most evident in the eastern
U.S. and has resulted in organic matter or nitrate now being the
greatest contributor to PM2.5 mass in many locations (U.S.
EPA, 2019a, Figure 2-19). The overall reduction in sulfate
concentrations has contributed substantially to the decrease in
national average PM2.5 concentrations as well as the decline
in the fraction of PM10 mass accounted for by
PM2.5 (U.S. EPA, 2019a, section 2.5.1.1.6; U.S. EPA, 2022b,
section 2.3.1).
c. PM10
At long-term monitoring sites in the U.S., the 2017-2019 average of
2nd highest 24-hour PM10 concentration was 68 [mu]g/m\3\
(with 10th and 90th percentiles at 28 and 124 [mu]g/m\3\, respectively)
(U.S. EPA, 2022b, section 2.3.2.4).\36\ The highest PM10
concentrations tend to occur in the western U.S. Seasonal analyses
indicate that ambient PM10 concentrations are generally
higher in the summer months than at other times of year, though the
most extreme high concentration events are more likely in the spring
(U.S. EPA, 2019a, Table 2-5). This is due to fact that the major
PM10 emission sources, dust and agriculture, are more active
during the warmer and drier periods of the year.
---------------------------------------------------------------------------
\36\ The form of the current 24-hour PM10 standard is
one-expected-exceedance, averaged over three years.
---------------------------------------------------------------------------
Recent ambient PM10 concentrations reflect reductions
that have occurred across much of the U.S. (U.S. EPA, 2022b, section
2.3.2.4). From 2000 to 2019, 2nd highest 24-hour PM10
concentrations have declined by about 46% (U.S. EPA, 2022b, section
2.3.2.4).\37\ Analyses at individual monitoring sites indicate that
annual average PM10 concentrations have generally declined
at most sites across the U.S., with much of the decrease in the eastern
U.S. associated with reductions in PM2.5 concentrations
(U.S. EPA, 2022b, section 2.3.2.4). Annual 2nd highest 24-hour
PM10 concentrations have generally declined in the eastern
U.S., while concentrations in much of the midwest and western U.S. have
remained unchanged or increased since 2000 (U.S. EPA, 2022b, section
2.3.2.4).
---------------------------------------------------------------------------
\37\ For more information, see https://www.epa.gov/air-trends/particulate-matter-pm10-trends#pmnat.
---------------------------------------------------------------------------
Compared to previous reviews, data available from the NCore
monitoring network in the current reconsideration allows a more
comprehensive analysis of the relative contributions of
PM2.5 and PM10-2.5 to PM10 mass.
PM2.5 generally contributes more to annual average
PM10 mass in the eastern U.S. than the western U.S. (U.S.
EPA, 2022b, Figure 2-23). At most sites in the eastern U.S., the
majority of PM10 mass is comprised of PM2.5. As
ambient PM2.5 concentrations have declined in the eastern
U.S. (U.S. EPA, 2022b, section 2.3.2.2), the ratios of PM2.5
to PM10 have also declined. For sites with days having
concurrently very high PM2.5 and PM10
concentrations (U.S. EPA, 2022b, Figure 2-24), the PM2.5/
PM10 ratios are typically higher than the annual average
ratios. This is particularly true in the northwestern U.S. where the
high PM10 concentrations can occur during wildfires with
high PM2.5 (U.S. EPA, 2022b, section 2.3.2.4).
d. PM10-2.5
Since the 2012 review, the availability of PM10-2.5
ambient concentration data has greatly increased because of additions
to the PM10-2.5 monitoring capabilities to the national
monitoring network. As illustrated in the 2022 PA (U.S. EPA, 2022b,
section 2.3.2.5), annual average and 98th percentile
PM10-2.5 concentrations exhibit less distinct differences
between the eastern and western U.S. than for either PM2.5
or PM10.
Due to the short atmospheric lifetime of PM10-2.5
relative to PM2.5, many of the high concentration sites are
isolated and likely near emission sources associated with wind-blown
and fugitive dust. The spatial distributions of annual average and 98th
percentile concentrations of PM10-2.5 are more similar than
that of PM2.5, suggesting that the same dust-related
emission sources are affecting both long-term and episodic
concentrations (U.S. EPA, 2022b, Figure 2-25). The highest
concentrations of PM10-2.5 are in the southwest U.S. where
widespread dry and windy conditions contribute to wind-blown dust
emissions. Additionally, compared to PM2.5 and
PM10, changes in PM10-2.5 concentrations have
been small in magnitude and inconsistent in direction (U.S. EPA, 2022b,
Figure 2-25). The majority of PM10-2.5 sites in the U.S. do
not have a concentration trend from 2000-2019, reflecting the
relatively consistent level of dust emissions across the U.S. during
the same time period (U.S. EPA, 2022b, section 2.3.2.5).\38\
---------------------------------------------------------------------------
\38\ PM from dust emissions in the National Emissions Inventory
(NEI) remain fairly consistent from year-to-year, except when there
are severe weather incursions or there is a dust event that
transports or causes major local dust storms to occur (particularly
in the western U.S.). These dust events and weather incursions
needed to effect dust emissions on a national level are not common
and only seldomly occur. In the emissions trends analysis presented
in the 2022 PA (U.S. EPA, 2022b, section 2.1.1), dust is included in
the NEI sector labeled ``miscellaneous.''
---------------------------------------------------------------------------
e. UFP
Compared to PM2.5 mass, there is relatively little data
on U.S. particle number concentrations, which are dominated by UFP. In
the published literature, annual average particle number concentrations
reaching about 20,000 to 30,000 cm\3\ have been reported in U.S. cities
(U.S. EPA, 2019a). In addition, based on UFP measurements in two urban
areas (New York City, Buffalo) and at a background site (Steuben
County) in New York, there is a pronounced difference in particle
number concentration between different types of locations (U.S. EPA,
2022b, Figure 2-26; U.S. EPA, 2019a, Figure 2-18). Urban particle
number counts were several times higher than at the background site,
and the highest particle number counts in an urban area with multiple
sites (Buffalo) were observed at a near-road location (U.S. EPA, 2022b,
section 2.3.2.6).
Long-term trends in UFP are not routinely available at U.S.
monitoring
[[Page 16217]]
sites. At one background site in Illinois with long-term data
available, the annual average particle number concentration declined
between 2000 and 2019, closely matching the reductions in annual
PM2.5 mass over that same period (U.S. EPA, 2022b, section
2.3.2.6). In addition, a small number of published studies have
examined UFP trends over time. While limited, these studies also
suggest that UFP number concentrations have declined over time along
with decreases in PM2.5 (U.S. EPA, 2022b, section 2.3.2.6).
However, the relationship between changes in ambient PM2.5
and UFPs cannot be comprehensively characterized due to the high
variability and limited monitoring of UFPs (U.S. EPA, 2022b, section
2.3.2.6).
5. Characterizing Ambient PM2.5 Concentrations for Exposure
Epidemiologic studies use various methods to characterize exposure
to ambient PM2.5. The methods used to estimate
PM2.5 concentrations can vary from traditional methods using
monitoring data from ground-based monitors to newer methods using more
complex hybrid modeling approaches. Studies using hybrid modeling
approaches aim to broaden the spatial coverage, as well as estimate
more spatially-resolved ambient PM2.5 concentrations, by
expanding beyond just those areas with monitors and providing estimates
in areas that do not have ground-based monitors (i.e., areas that are
generally less densely populated and tend to have lower
PM2.5 concentrations) and at finer spatial resolutions
(e.g., 1 km x 1 km grid cells). Ground-based PM2.5 monitors
are generally sited in areas of expected maximum concentration. As
such, the hybrid modeling approaches tend to broaden the areas captured
in the exposure assessment, and in doing so, the studies that utilize
these methods tend to report lower mean PM2.5 concentrations
than monitor-based approaches. Further, other aspects of the approaches
applied in the various epidemiologic studies to estimate
PM2.5 exposure and/or to calculate the related study-
reported mean concentration (i.e., population weighting, trim mean
approaches) can affect those data values. More detail related to hybrid
modeling methods, performance of the methods, and how the reported mean
concentrations compare across approaches is provided in section 2.3.3.2
of the 2022 PA (U.S. EPA, 2022b). The subsections below discuss the
characterization of PM2.5 concentrations based on monitoring
data (I.D.5.a) and using hybrid modeling approaches (I.D.5.b).
a. Predicted Ambient PM2.5 and Exposure Based on Monitored
Data
Ambient concentrations of PM2.5 are often characterized
using measurements from national monitoring networks due to the
accuracy and precision of the measurements and the public availability
of data. For applications requiring PM2.5 characterizations
across large areas or provide complete coverage from the site
measurements, data interpolation and averaging techniques (such as
Average Nearest Neighbor tools, and area-wide or population-weighted
averaging of monitors) are sometimes used (U.S. EPA, 2019a, chapter 3).
For an area to meet the NAAQS, all valid design values \39\ in that
area, including the highest annual and 24-hour design values, must be
at or below the levels of the standards. Because the monitoring network
siting requirements are specified to capture the high PM2.5
concentrations (U.S. EPA, 2022b, section 2.2.3), areas meeting an
annual PM2.5 standard with a particular level would be
expected to have long-term average monitored PM2.5
concentrations (i.e., averaged across space and over time in the area)
somewhat below that standard level. This means that the
PM2.5 design value in an area is associated with a
distribution of PM2.5 concentrations in that area, and,
based on monitoring siting requirements, should represent the highest
concentration location applicable to be monitored under the
PM2.5 NAAQS. Analyses in the 2022 PA indicate that, based on
recent air quality in U.S. CBSAs, maximum annual PM2.5
design values are often 10% to 20% higher than annual average
concentrations (i.e., averaged across multiple monitors in the same
CBSA) (U.S. EPA, 2022b, section 2.3.3.1, Figures 2-28 and 2-29). This
difference between the maximum annual design value and the average
concentration in an area can vary, depending on factors such as the
number of monitors, monitor siting characteristics, and the
distribution of ambient PM2.5 concentrations. Given that
higher PM2.5 concentrations have been reported at some near-
road monitoring sites relative to the surrounding area (U.S. EPA,
2022b, section 2.3.2.2.2), recent requirements for PM2.5
monitoring at near-road locations in large urban areas (U.S. EPA,
2022b, section 2.2.3.3) may increase the ratios of maximum design
values to average annual design values in some areas. Such ratios may
also depend on how the averages are calculated (i.e., averaged across
monitors versus across modeled grid cells, as described below in
section I.5.b). Compared to annual design values, the analysis in the
2022 PA indicates a more variable relationship between maximum 24-hour
PM2.5 design values and annual average concentrations (U.S.
EPA, 2022b, section 2.3.3.1, Figure 2-29).
---------------------------------------------------------------------------
\39\ For the annual PM2.5 standard, design values are
calculated as the annual arithmetic mean PM2.5
concentration, averaged over 3 years. For the 24-hour standard,
design values are calculated as the 98th percentile of the annual
distribution of 24-hour PM2.5 concentrations, averaged
over three years (appendix N of 40 CFR part 50).
---------------------------------------------------------------------------
b. Comparison of PM2.5 Hybrid Modeling Approaches in
Estimating Exposure and Relative to Design Values
Two types of hybrid approaches that have been utilized in several
key PM2.5 epidemiologic studies in the 2019 ISA and ISA
Supplement include neural network approaches and a satellite-based
method with regression of residual PM2.5 with land-use and
other variables to improve estimates of PM2.5 concentration
in the U.S. As such, the 2022 PA further compares these two types of
approaches across various scales (e.g., CBSA versus nationwide), taking
into account population weighting approaches utilized in epidemiologic
studies when estimating PM2.5 exposure (U.S. EPA, 2022b,
section 2.3.3.2.4). Additionally, the 2022 PA assesses how average
PM2.5 concentrations computed in epidemiologic studies using
these hybrid surfaces compare to the maximum design values measured at
ground-based monitors. For this assessment, the 2022 PA evaluates the
DI2019 \40\ and HA2020 \41\ hybrid surfaces, surfaces that are used in
several of the key epidemiologic studies in the 2022 PA. This analysis
is intended to help inform how the magnitude of the overall study-
reported mean PM2.5 concentrations in epidemiologic studies
may be
[[Page 16218]]
influenced by the approach used to compute that mean and how that value
might compare to monitor reported concentrations. The PM2.5
standards are expected to achieve a pattern of air quality through the
attainment of a specific design value at each monitor in the monitoring
network. As a result, it is important to be able to assess the
relationship between monitor concentrations and patterns of air quality
evaluated in the epidemiologic studies.
---------------------------------------------------------------------------
\40\ This analysis includes an updated version of the surface
used in Di et al. (2016). Predictions in Di et al. (2016) were for
2000 to 2012 using a neural network model. The Di et al. (2019)
study improved on that effort in several ways. First, a generalized
additive model was used that accounted for geographic variations in
performance to combine predictions from three models (neural
network, random forest, and gradient boosting) to make the final
optimal PM2.5 predictions. Second, the datasets were
updated that were used in model training and included additional
variables such as 12-km CMAQ modeling as predictors. Finally, more
recent years were included in the Di et al. (2019) study.
\41\ The HA2020 field is based on the V4.NA.03 product available
at: https://sites.wustl.edu/acag/datasets/surface-pm2-5/. The name
``HA2020'' comes from the references for this product (Hammer et
al., 2020; van Donkelaar et al., 2019).
---------------------------------------------------------------------------
In estimating exposure, some studies focus on estimating
concentrations in urban areas, while others examine the entire U.S. or
large portions of the country. In general, the areas that are not
included in the CBSA-only analysis tend to be more rural or less
densely populated areas, tend to have lower PM2.5
concentrations, and likely correspond to those locations where
monitoring data availability is limited or nonexistent (U.S. EPA,
2022b, section 2.3.3.2.4, Figure 2-37). To evaluate the differences in
mean PM2.5 concentrations across different spatial scales,
the 2022 PA analysis compares the DI2019 and HA2020 surfaces. At the
national scale, the two surfaces generally produce similar average
annual PM2.5 concentrations, with the DI2019 surface being
slightly higher compared to the HA2020 surface. The average annual
PM2.5 concentrations are also slightly higher using the
DI2019 surface compared to the HA2020 surface when the analyses are
conducted for CBSAs. Also, regardless of which surface is used, the
average annual and 3-year average of the average annual
PM2.5 concentrations for the CBSA-only analyses are somewhat
higher than for the nationwide analyses (4-8% higher) (U.S. EPA, 2022b,
section 2.3.3.2.4, Table 2-5).\42\ Overall, these analyses suggest that
there are only slight differences in the average PM2.5
concentrations depending on the hybrid modeling method employed, though
including other hybrid modeling methods in this comparison could result
in larger differences.
---------------------------------------------------------------------------
\42\ For the national scale, 3-year averages of the average
annual PM2.5 concentrations generally range from about
5.3 [micro]g/m\3\ to 8.1 [micro]g/m\3\, compared to the CBSA scale,
which ranges from 5.7 [micro]g/m\3\ to 8.7 [micro]g/m\3\. (U.S. EPA,
2022b, section 2.3.3.2.4, Table 2-6).
---------------------------------------------------------------------------
The 2022 PA next evaluates how the averages of the hybrid model
surfaces compare to regulatory design values using both the DI2019 and
HA2020 surfaces and how population weighting influences the mean
PM2.5 concentration.\43\ As presented in the 2022 PA, the
results using the DI2019 and HA2020 surfaces are similar for the
average annual PM2.5 concentrations, for each 3-year period.
When population weighting is not applied, the average annual
PM2.5 concentrations generally range from 7.0 to 8.6
[micro]g/m\3\. When population weighting is applied, the average annual
PM2.5 concentrations are slightly higher, ranging from 8.2
to 10.2 [micro]g/m\3\. As with CBSAs versus the national comparison
above, population weighting results in a higher average
PM2.5 concentration than when population weighting is not
applied (U.S. EPA, 2022b, section 2.3.3.2.4, Table 2-7). For the CBSAs
included in the population weighted analyses, the average maximum
annual design values generally range from 9.5 to 11.7 [micro]g/m\3\.
The results are similar for both the DI2019 and HA2020 surfaces and the
maximum annual PM2.5 design values measured at the monitors
are often 40% to 50% higher than average annual PM2.5
concentrations predicted by hybrid modeling methods when population
weighting is not applied. However, when population weighting is
applied, the ratio of the maximum annual PM2.5 design values
to the predicted average annual PM2.5 concentrations are
lower than when population weighting is not applied, with monitored
design values generally 15% to 18% higher than population-weighted
hybrid modeling average annual PM2.5 concentrations (U.S.
EPA, 2022b, section 2.3.3.2.4, Table 2-7).
---------------------------------------------------------------------------
\43\ For this analysis, the 2022 PA includes CBSAs with three or
more valid design values for the 3-year period. The regulatory
design values for the CBSAs were calculated for each 3-year period
for the CBSAs with 3 or more design values in each of the 3-year
periods. Using the maximum design value for each CBSA and by each 3-
year period, the ratio of maximum design values to modeled average
annual PM2.5 concentrations were calculated, for each 3-
year period. More details about the analytical methods used for this
analysis are described in section A.6 of Appendix A in the 2022 PA
(U.S. EPA, 2022b).
---------------------------------------------------------------------------
6. Background PM
In this reconsideration, background PM is defined as all particles
that are formed by sources or processes that cannot be influenced by
actions within the jurisdiction of concern. U.S. background PM is
defined as any PM formed from emissions other than U.S. anthropogenic
(i.e., manmade) emissions. Potential sources of U.S. background PM
include both natural sources (i.e., PM that would exist in the absence
of any anthropogenic emissions of PM or PM precursors) and
transboundary sources originating outside U.S. borders. Background PM
is discussed in more detail in the 2022 PA (U.S. EPA, 2022b, section
2.4). At annual and national scales, estimated background PM
concentrations in the U.S. are small compared to contributions from
domestic anthropogenic sources.\44\ For example, based on zero-out
modeling in the last review of the PM NAAQS, annual background
PM2.5 concentrations were estimated to range from 0.5-3
[micro]g/m\3\ across the sites examined. In addition, speciated
monitoring data from IMPROVE sites can provide some insights into how
contributions from different sources, including sources of background
PM, may have changed over time. Such data suggests the estimates of
background concentrations using speciated monitoring data from IMPROVE
monitors are around 1-3 [micro]g/m\3\ and have not changed
significantly since the 2012 review. Contributions to background PM in
the U.S. result mainly from sources within North America. Contributions
from intercontinental events have also been documented (e.g., transport
from dust storms occurring in deserts in North Africa and Asia), but
these events are less frequent and represent a relatively small
fraction of background PM in most of the U.S. (U.S. EPA, 2022b, section
2.4).
---------------------------------------------------------------------------
\44\ Sources that contribute to natural background PM include
dust from the wind erosion of natural surfaces, sea salt, wildland
fires, primary biological aerosol particles such as bacteria and
pollen, oxidation of biogenic hydrocarbons such as isoprene and
terpenes to produce secondary organic aerosols (SOA), and geogenic
sources such as sulfate formed from volcanic production of
SO2 and oceanic production of dimethyl-sulfide (U.S. EPA,
2022b, section 2.4). While most of these sources release or
contribute predominantly to fine aerosol, some sources including
windblown dust, and sea salt also produce particles in the coarse
size range (U.S. EPA, 2019a, section 2.3.3).
---------------------------------------------------------------------------
II. Rationale for Decisions on the Primary PM2.5 Standards
This section presents the rationale for the Administrator's
decision to revise the primary annual PM2.5 standard down to
a level of 9 [micro]g/m\3\ and retain the primary 24-hour
PM2.5 standard. This rationale is based on a thorough review
of the scientific evidence generally published through January
2018,\45\ as evaluated in the 2019 ISA (U.S. EPA, 2019a), on the human
health effects of PM2.5 associated with long- and short-term
exposures \46\ to PM2.5 in
[[Page 16219]]
the ambient air. Additionally, this rationale is based on a thorough
evaluation of some studies that became available after the literature
cutoff date of the 2019 ISA, as evaluated in the ISA Supplement, that
could either further inform the adequacy of the current PM NAAQS or
address key scientific topics that have evolved since the literature
cutoff date for the 2019 ISA, generally through March 2021 (U.S. EPA,
2022a).\47\ The Administrator's rationale also takes into account: (1)
The 2022 PA evaluation of the policy-relevant information in the 2019
ISA and ISA Supplement and presentation of quantitative analyses of air
quality and health risks; (2) CASAC advice and recommendations; and (3)
public comments received during the development of these documents.
---------------------------------------------------------------------------
\45\ In addition to the 2020 review's opening ``call for
information'' (79 FR 71764, December 3, 2014), the 2019 ISA
identified and evaluated studies and reports that have undergone
scientific peer review and were published or accepted for
publication between January 1, 2009, through approximately January
2018 (U.S. EPA, 2019a, p. ES-2). References that are cited in the
2019 ISA, the references that were considered for inclusion but not
cited, and electronic links to bibliographic information and
abstracts can be found at: https://hero.epa.gov/hero/particulate-matter.
\46\ Short-term exposures are defined as those exposures
occurring over hours up to 1 month, whereas long-term exposures are
defined as those exposures occurring over 1 month to years (U.S.
EPA, 2019a, section P.3.1).
\47\ The ISA Supplement represents an evaluation of recent
studies that are of greatest policy relevance to the reconsideration
of the 2020 final decision on the PM NAAQS. Specifically, the ISA
Supplement focuses on studies of health effects for which the
evidence in the 2019 ISA supported a ``causal relationship'' (i.e.,
short- and long-term PM2.5 exposure and mortality and
cardiovascular effects) because those were the health effects that
were most useful in informing conclusions in the 2020 PA. The ISA
Supplement does not include an evaluation of studies for other
PM2.5-related health effects (U.S. EPA, 2022a).
---------------------------------------------------------------------------
In presenting the rationale for the Administrator's decisions and
its foundations, section II.A provides background on the general
approach for this reconsideration and the basis for the existing
standard, and also presents brief summaries of key aspects of the
currently available health effects and risk information. Section II.B
summarizes the CASAC advice and the basis for the proposed conclusions,
addresses public comments received on the proposal and presents the
Administrator's conclusions on the adequacy of the current standards,
drawing on consideration of the scientific evidence and quantitative
risk information, advice from the CASAC, and comments from the public.
Section II.C summarizes the Administrator's decision on the primary
PM2.5 standards.
A. Introduction
The general approach for this reconsideration of the 2020 final
decision on the primary PM2.5 standards is fundamentally
based on using the EPA's assessment of the current scientific evidence
and associated quantitative analyses to inform the Administrator's
judgment regarding primary PM2.5 standards that protect
public health with an adequate margin of safety. The EPA's assessments
are primarily documented in the 2019 ISA, ISA Supplement, and 2022 PA,
all of which have received CASAC review and public comment (83 FR
53471, October 23, 2018; 83 FR 55529, November 6, 2018; 85 FR 4655,
January 27, 2020; 86 FR 52673, September 22, 2021; 86 FR 54186,
September 30, 2021; 86 FR 56263, October 8, 2021; 87 FR 958, January 7,
2022; 87 FR 22207, April 14, 2022; 87 FR 31965, May 26, 2022). In
bridging the gap between the scientific assessments of the 2019 ISA and
ISA Supplement and the judgments required of the Administrator in
determining whether the current standards provide the requisite public
health protection, the 2022 PA evaluates policy implications of the
evaluation of the current evidence in the 2019 ISA and ISA Supplement,
and the risk information documented in the 2022 PA. In evaluating the
public health protection afforded by the current standards, the four
basic elements of the NAAQS (i.e., indicator, averaging time, level,
and form) are considered collectively.
The final decision on the adequacy of the current primary
PM2.5 standards is a public health policy judgment to be
made by the Administrator. In reaching conclusions with regard to the
standards, the decision will draw on the scientific information and
analyses about health effects and population risks, as well as
judgments about how to consider the range and magnitude of
uncertainties that are inherent in the scientific evidence and
analyses. This approach is based on the recognition that the available
health effects evidence generally reflects a continuum, consisting of
levels at which scientists generally agree that health effects are
likely to occur, through lower levels at which the likelihood and
magnitude of the response become increasingly uncertain. This approach
is consistent with the requirements of the NAAQS provisions of the
Clean Air Act and with how the EPA and the courts have historically
interpreted the Act (summarized in section I.A above). These provisions
require the Administrator to establish primary standards that, in the
judgment of the Administrator, are requisite to protect public health
with an adequate margin of safety. In so doing, the Administrator seeks
to establish standards that are neither more nor less stringent than
necessary for this purpose. The Act does not require that primary
standards be set at a zero-risk level, but rather at a level that
avoids unacceptable risks to public health, including the health of
sensitive (also referred to as ``at-risk'') groups.\48\
---------------------------------------------------------------------------
\48\ As noted in section I.A above, the legislative history
describes such protection for the sensitive group of individuals and
not for a single person in the sensitive group (see S. Rep. No. 91-
1196, 91st Cong, 2d Sess. 10 [1970]); see also Am. Lung Ass'n v.
EPA, 134 F.3d 388, 389 (D.C. Cir. 1998).
---------------------------------------------------------------------------
1. Background on the Current Standards
The current primary PM2.5 standards were retained in
2020 based on the scientific evidence and quantitative risk information
available at that time, as well as the then-Administrator's judgments
regarding the available health effects evidence and the appropriate
degree of public health protection afforded by the existing standards
(85 FR 82718, December 18, 2020). With the 2020 decision, the then-
Administrator retained the primary annual PM2.5 standard
with its level of 12.0 [mu]g/m\3\ and retained the primary 24-hour
PM2.5 standard with its level of 35 [mu]g/m\3\. The key
considerations and the then-Administrator's conclusions regarding the
primary PM2.5 standards in the 2020 review are summarized
below.
The health effects evidence base available in the 2020 review
included extensive evidence from previous reviews as well as the
evidence that had emerged since the prior review had been completed in
2012. This evidence base, spanning several decades, documents the
relationship between short- and long-term PM2.5 exposure and
mortality or serious morbidity effects. The evidence available in the
2019 ISA reaffirmed, and in some cases strengthened, the conclusions
from the 2009 ISA regarding the health effects of PM2.5
exposures (U.S. EPA, 2019a). Much of the evidence came from
epidemiologic studies conducted in North America, Europe, or Asia
examining short-term and long-term exposures that demonstrated
generally positive, and often statistically significant,
PM2.5 health effect associations with a range of outcomes
including non- accidental, cardiovascular, or respiratory mortality;
cardiovascular- or respiratory-related hospitalizations or emergency
department visits; and other mortality/morbidity outcomes (e.g., lung
cancer mortality or incidence, asthma development). Experimental
evidence, as well as evidence from panel studies, strengthened support
for potential biological pathways through which PM2.5
exposures could lead to health effects reported in many population-
based epidemiologic studies, including support for pathways that could
lead to cardiovascular, respiratory, nervous system, and cancer-related
effects.
[[Page 16220]]
Based on this evidence, the 2019 ISA concluded there to be a causal
relationship between long- and short-term PM2.5 exposure and
mortality and cardiovascular effects, as well as likely to be causal
relationships between long- and short-term PM2.5 exposure
and respiratory effects, and between long-term PM2.5
exposure and cancer and nervous system effects (U.S. EPA, 2019a,
section 1.7).
Epidemiologic studies reported PM2.5 health effect
associations with mortality and/or morbidity across multiple U.S.
cities and in diverse populations, including in studies examining
populations and lifestages that may be at increased risk of
experiencing a PM2.5-related health effect (e.g., older
adults, children). The 2019 ISA cited extensive evidence indicating
that ``both the general population as well as specific populations and
lifestages are at risk for PM2.5-related health effects''
(U.S. EPA, 2019a, p. 12-1), including children and older adults, people
with pre-existing respiratory or cardiovascular disease, minority
populations, and low socioeconomic status (SES) populations.
The risk information available in the 2020 review included risk
estimates for air quality conditions just meeting the existing primary
PM2.5 standards, and also for air quality conditions just
meeting potential alternative standards. The general approach to
estimating PM2.5-associated health risks combined
concentration-response (C-R) functions from epidemiologic studies with
model-based PM2.5 air quality surfaces, baseline health
incidence data, and population demographics for 47 urban areas (U.S.
EPA, 2020b, section 3.3, Figure 3-10, Appendix C). The risk assessment
estimated that the existing primary PM2.5 standards could
allow a substantial number of PM2.5-associated deaths in the
U.S. Uncertainty in risk estimates (e.g., in the size of risk
estimates) can result from a number of factors, including assumptions
about the shape of the C-R relationship with mortality at low ambient
PM2.5 concentrations, the potential for confounding and/or
exposure measurement error, and the methods used to adjust
PM2.5 air quality.
Consistent with the general approach routinely employed in NAAQS
reviews, the initial consideration in the 2020 review of the primary
PM2.5 standards was with regard to the adequacy of the
protection provided by the existing standards.
As an initial matter, the then-Administrator considered the range
of scientific evidence evaluating these effects, including studies of
at-risk populations, to inform his review of the primary
PM2.5 standards, placing the greatest weight on evidence of
effects for which the 2019 ISA determined there to be a causal or
likely to be causal relationship with long- and short-term
PM2.5 exposures (85 FR 82714-82715, December 18, 2020).
With regard to indicator, the then-Administrator recognized that,
consistent with the evidence available in prior reviews, the scientific
evidence continued to provide strong support for health effects
following short- and long-term PM2.5 exposures. He noted the
2020 PA conclusions that the information continued to support the
PM2.5 mass-based indicator and remained too limited to
support a distinct standard for any specific PM2.5 component
or group of components, and too limited to support a distinct standard
for the ultrafine fraction. Thus, the then-Administrator concluded that
it was appropriate to retain PM2.5 as the indicator for the
primary standards for fine particles (85 FR 82715, December 18, 2020).
With respect to averaging time and form, the then-Administrator
noted that the scientific evidence continued to provide strong support
for health effects associations with both long-term (e.g., annual or
multi-year) and short-term (e.g., mostly 24-hour) exposures to
PM2.5, consistent with the conclusions in the 2020 PA. In
the 2019 ISA, epidemiologic and controlled human exposure studies
examined a variety of PM2.5 exposure durations.
Epidemiologic studies continued to provide strong support for health
effects associated with short-term PM2.5 exposures based on
24-hour PM2.5 averaging periods, and the EPA noted that
associations with subdaily estimates are less consistent and, in some
cases, smaller in magnitude (U.S. EPA, 2019a, section 1.5.2.1; U.S.
EPA, 2020b, section 3.5.2.2). In addition, controlled human exposure
and panel-based studies of subdaily exposures typically examined
subclinical effects, rather than the more serious population-level
effects that have been reported to be associated with 24-hour exposures
(e.g., mortality, hospitalizations). Taken together, the 2019 ISA
concluded that epidemiologic studies did not indicate that subdaily
averaging periods were more closely associated with health effects than
the 24-hour average exposure metric (U.S. EPA, 2019a, section 1.5.2.1).
Additionally, while controlled human exposure studies provided
consistent evidence for cardiovascular effects following
PM2.5 exposures for less than 24 hours (i.e., <30 minutes to
5 hours), exposure concentrations in the studies were well-above the
ambient concentrations typically measured in locations meeting the
existing standards (U.S. EPA, 2020b, section 3.2.3.1). Thus, these
studies also did not suggest the need for additional protection against
subdaily PM2.5 exposures (U.S. EPA, 2020b, section 3.5.2.2).
Therefore, the then-Administrator judged that the 24-hour averaging
time remained appropriate (85 FR 82715, December 18, 2020).
With regard to the form of the 24-hour standard (98th percentile,
averaged over three years), the then-Administrator noted that
epidemiologic studies continued to provide strong support for health
effect associations with short-term (e.g., mostly 24-hour)
PM2.5 exposures (U.S. EPA, 2020b, section 3.5.2.3) and that
controlled human exposure studies provided evidence for health effects
following single short-term ``peak'' PM2.5 exposures. Thus,
the evidence supported retaining a standard focused on providing
supplemental protection against short-term peak exposures and supported
a 98th percentile form for a 24-hour standard. The then-Administrator
further noted that this form also provided an appropriate balance
between limiting the occurrence of peak 24-hour PM2.5
concentrations and identifying a stable target for risk management
programs (U.S. EPA, 2020b, section 3.5.2.3). As such, the then-
Administrator concluded that the available information supported
retaining the form and averaging time of the current 24-hour standard
(98th percentile, averaged over three years) and annual standard
(annual average, averaged over three years) (85 FR 82715, December 18,
2020).
With regard to the level of the standards, in reaching his final
decision, the then-Administrator considered the large body of evidence
presented and assessed in the 2019 ISA (U.S. EPA, 2019a), the policy-
relevant and risk-based conclusions and rationales as presented in the
2020 PA (U.S. EPA, 2020b), advice from the CASAC, and public comments.
In particular, in considering the 2019 ISA and 2020 PA, he considered
key epidemiologic studies that evaluated associations between
PM2.5 air quality distributions and mortality and morbidity,
including key accountability studies; the availability of experimental
studies to support biological plausibility; controlled human exposure
studies examining effects following short-term PM2.5
exposures; air quality analyses; and the important uncertainties and
limitations associated with the information (85 FR 82715, December 18,
2020).
[[Page 16221]]
As an initial matter, the then-Administrator considered the
protection afforded by both the annual and 24-hour standards together
against long- and short-term PM2.5 exposures and health
effects. The Administrator recognized that the annual standard was most
effective in controlling ``typical'' PM2.5 concentrations
near the middle of the air quality distribution (i.e., around the mean
of the distribution), but also provided some control over short-term
peak PM2.5 concentrations. On the other hand, the 24-hour
standard, with its 98th percentile form, was most effective at limiting
peak 24-hour PM2.5 concentrations, but in doing so also had
an effect on annual average PM2.5 concentrations. Thus,
while either standard could be viewed as providing some measure of
protection against both average exposures and peak exposures, the 24-
hour and annual standards were not expected to be equally effective at
limiting both types of exposures. Thus, consistent with previous
reviews, the then-Administrator's consideration of the public health
protection provided by the existing primary PM2.5 standards
was based on his consideration of the combination of the annual and 24-
hour standards. Specifically, he recognized that the annual standard
was more likely to appropriately limit the ``typical'' daily and annual
exposures that are most strongly associated with the health effects
observed in epidemiologic studies. The then-Administrator concluded
that an annual standard (as the arithmetic mean, averaged over three
years) remained appropriate for targeting protection against the annual
and daily PM2.5 exposures around the middle portion of the
PM2.5 air quality distribution. Further, recognizing that
the 24-hour standard (with its 98th percentile form) was more directly
tied to short-term peak PM2.5 concentrations, and more
likely to appropriately limit exposures to such concentrations, the
then-Administrator concluded that the current 24-hour standard (with
its 98th percentile form, averaged over three years) remained
appropriate to provide a balance between limiting the occurrence of
peak 24-hour PM2.5 concentrations and identifying a stable
target for risk management programs. However, the then-Administrator
recognized that changes in PM2.5 air quality to meet an
annual standard would likely result not only in lower short- and long-
term PM2.5 concentrations near the middle of the air quality
distribution, but also in fewer and lower short-term peak
PM2.5 concentrations. The then-Administrator further
recognized that changes in air quality to meet a 24-hour standard, with
a 98th percentile form, would result not only in fewer and lower peak
24-hour PM2.5 concentrations, but also in lower annual
average PM2.5 concentrations (85 FR 82715-82716, December
18, 2020).
Thus, in considering the adequacy of the 24-hour standard, the
then-Administrator noted the importance of considering whether
additional protection was needed against short-term exposures to peak
PM2.5 concentrations. In examining the scientific evidence,
he noted the limited utility of the animal toxicological studies in
directly informing conclusions on the appropriate level of the standard
given the uncertainty in extrapolating from effects in animals to those
in human populations. The then-Administrator noted that controlled
human exposure studies provided evidence for health effects following
single, short-term PM2.5 exposures that corresponded best to
exposures that might be experienced in the upper end of the
PM2.5 air quality distribution in the U.S. (i.e., ``peak''
concentrations). However, most of these studies examined exposure
concentrations considerably higher than are typically measured in areas
meeting the standards (U.S. EPA, 2020b, section 3.2.3.1). In
particular, controlled human exposure studies often reported
statistically significant effects on one or more indicators of
cardiovascular function following 2-hour exposures to PM2.5
concentrations at and above 120 [mu]g/m\3\ (at and above 149 [mu]g/m\3\
for vascular impairment, the effect shown to be most consistent across
studies). To provide insight into what these studies may indicate
regarding the primary PM2.5 standards, the 2020 PA (U.S.
EPA, 2020b, p. 3-49) noted that 2-hour ambient concentrations of
PM2.5 at monitoring sites meeting the current standards
almost never exceeded 32 [mu]g/m\3\. In fact, even the extreme upper
end of the distribution of 2-hour PM2.5 concentrations at
sites meeting the primary PM2.5 standards remained well-
below the PM2.5 exposure concentrations consistently shown
in controlled human exposure studies to elicit effects (i.e., 99.9th
percentile of 2-hour concentrations at these sites is 68 [mu]g/m\3\
during the warm season). Thus, the experimental evidence did not
indicate the need for additional protection against exposures to peak
PM2.5 concentrations, beyond the protection provided by the
combination of the 24-hour and the annual standards (U.S. EPA, 2020b,
section 3.2.3.1; 85 FR 82716, December 18, 2020).
With respect to the epidemiologic evidence, the then-Administrator
noted that the studies did not indicate that associations in those
studies were strongly influenced by exposures to peak concentrations in
the air quality distribution and thus did not indicate the need for
additional protection against short-term exposures to peak
PM2.5 concentrations (U.S. EPA, 2020b, section 3.5.1). The
then-Administrator noted that this was consistent with CASAC consensus
support for retaining the current 24-hour standard. Thus, the then-
Administrator concluded that the 24-hour standard with its level of 35
[micro]g/m\3\ was adequate to provide supplemental protection (i.e.,
beyond that provided by the annual standard alone) against short-term
exposures to peak PM2.5 concentrations (85 FR 82716,
December 18, 2020).
With regard to the level of the annual standard, the then-
Administrator recognized that the annual standard, with its form based
on the arithmetic mean concentration, was most appropriately meant to
limit the ``typical'' daily and annual exposures that were most
strongly associated with the health effects observed in epidemiologic
studies. However, the then-Administrator also noted that while
epidemiologic studies examined associations between distributions of
PM2.5 air quality and health outcomes, they did not identify
particular PM2.5 exposures that cause effects and thus, they
could not alone identify a specific level at which the standard should
be set, as such a determination necessarily required the then-
Administrator's judgment. Thus, consistent with the approaches in
previous NAAQS reviews, the then-Administrator recognized that any
approach that used epidemiologic information in reaching decisions on
what standards are appropriate necessarily required judgments about how
to translate the information from the epidemiologic studies into a
basis for appropriate standards. This approach included consideration
of the uncertainties in the reported associations between daily or
annual average PM2.5 exposures and mortality or morbidity in
the epidemiologic studies. Such an approach is consistent with setting
standards that are neither more nor less stringent than necessary,
recognizing that a zero-risk standard is not required by the Clean Air
Act (CAA) (85 FR 82716, December 18, 2020).
The then-Administrator emphasized uncertainties and limitations
that were present in epidemiologic studies in previous reviews and
persisted in the 2020 review. These uncertainties
[[Page 16222]]
included exposure measurement error, potential confounding by
copollutants, increasing uncertainty of associations at lower
PM2.5 concentrations, and heterogeneity of effects across
different cities or regions (85 FR 82716, December 18, 2020). The then-
Administrator also noted the advice given by the CASAC on this matter.
As described in section I.C.5 above, the CASAC did not reach consensus
on the adequacy of the primary annual PM2.5 standard. ``Some
CASAC members'' expressed support for retaining the primary annual
PM2.5 standard while ``other members'' expressed support for
revising that standard in order to increase public health protection
(Cox, 2019b, p. 1 of consensus letter). The CASAC members who supported
retaining the annual standard expressed their concerns with the
epidemiologic studies, asserting that these studies did not provide a
sufficient basis for revising the existing standards. They also
identified several key concerns regarding the associations reported in
epidemiologic studies and concluded that ``while the data on
associations should certainly be carefully considered, this data should
not be interpreted more strongly than warranted based on its
methodological limitations'' (Cox, 2019b, p. 8 consensus responses).
Taking into consideration the views expressed by the CASAC members
who supported retaining the annual standard, the then-Administrator
recognized that epidemiologic studies examined associations between
distributions of PM2.5 air quality and health outcomes, and
they did not identify particular PM2.5 exposures that cause
effects (U.S. EPA, 2020b, section 3.1.2). While the Administrator
remained concerned about placing too much weight on epidemiologic
studies to inform conclusions on the adequacy of the primary standards,
he noted the approach to considering such studies in the 2012 review.
In the 2012 review, it was noted that the evidence of an association in
any epidemiologic study was ``strongest at and around the long-term
average where the data in the study are most concentrated'' (78 FR
3140, January 15, 2013). In considering the characterization of
epidemiologic studies, the then-Administrator viewed that when
assessing the mean concentrations of the key short-term and long-term
epidemiologic studies in the U.S. that used ground-based monitoring
(i.e., those studies where the mean is most directly comparable to the
current annual standard), the majority of studies had mean
concentrations at or above the level of the existing annual standard,
with the mean of the study-reported means or medians equal to 13.5
[micro]g/m\3\, a concentration level above the existing level of the
primary annual standard of 12 [micro]g/m\3\. The then-Administrator
further noted his caution in directly comparing the reported study mean
values to the standard level given that study-reported mean
concentrations, by design, are generally lower than the design value of
the highest monitor in an area, which determines compliance. In the
2020 PA, analyses of recent air quality in U.S. CBSAs indicated that
maximum annual PM2.5 design values for a given three-year
period were often 10% to 20% higher than average monitored
concentrations (i.e., averaged across multiple monitors in the same
CBSA) (U.S. EPA, 2020b, Appendix B, section B.7). He further noted his
concern in placing too much weight on any one epidemiologic study but
instead judged that it was more appropriate to focus on the body of
studies together and therefore noted the calculation of the mean of
study-reported means (or medians). Thus, while the then-Administrator
was cautious in placing too much weight on the epidemiologic evidence
alone, he noted that: (1) The reported mean concentration in the
majority of the key U.S. epidemiologic studies using ground-based
monitoring data were above the level of the existing annual standard;
(2) the mean of the reported study means (or medians) (i.e., 13.5
[micro]g/m\3\) was above the level of the current standard; \49\ (3)
air quality analyses showed the study means to be lower than their
corresponding design values by 10-20%; and (4) these analyses must be
considered in light of uncertainties inherent in the epidemiologic
evidence. When taken together, the then-Administrator judged that, even
if it were appropriate to place more weight on the epidemiologic
evidence, this information did not call into question the adequacy of
the current standards (85 FR 82716-17, December 18, 2020).
---------------------------------------------------------------------------
\49\ The median of the study-reported mean (or median)
PM2.5 concentrations is 13.3 [micro]g/m\3\, which was
also above the level of the existing standard.
---------------------------------------------------------------------------
In addition to the evidence, the then-Administrator also considered
the potential implications of the risk assessment. He noted that all
risk assessments have limitations and that he remained concerned about
the uncertainties in the underlying epidemiologic data used in the risk
assessment. The then-Administrator also noted that in previous reviews,
these uncertainties and limitations have often resulted in less weight
being placed on quantitative estimates of risk than on the underlying
scientific evidence itself (e.g., 78 FR 3086, 3098-99, January 15,
2013). These uncertainties and limitations included uncertainty in the
shapes of C-R functions, particularly at low concentrations;
uncertainties in the methods used to adjust air quality; and
uncertainty in estimating risks for populations, locations and air
quality distributions different from those examined in the underlying
epidemiologic study (U.S. EPA, 2020b, section 3.3.2.4). Additionally,
the then-Administrator noted similar concern expressed by some members
of the CASAC who support retaining the existing standards; they
highlighted similar uncertainties and limitations in the risk
assessment (Cox, 2019b). In light of all of this, the then-
Administrator judged it appropriate to place little weight on
quantitative estimates of PM2.5-associated mortality risk in
reaching conclusions about the level of the primary PM2.5
standards (85 FR 82717, December 18, 2020).
The then-Administrator additionally considered an emerging body of
evidence from accountability studies that examined past reductions in
ambient PM2.5 and the degree to which those reductions
resulted in public health improvements. While the then-Administrator
agreed with public commenters that well-designed and conducted
accountability studies can be informative, he viewed the interpretation
of such studies in the context of the primary PM2.5
standards as complicated by the fact that some of the available studies
had not evaluated PM2.5 specifically (e.g., as opposed to
PM10 or total suspended particulates), did not show changes
in PM2.5 air quality, or had not been able to disentangle
health impacts of the interventions from background trends in health
(U.S. EPA, 2020b, section 3.5.1). He further recognized that the small
number of available studies that did report public health improvements
following past declines in ambient PM2.5 had not examined
air quality meeting the existing standards (U.S. EPA, 2020b, Table 3-
3). This included U.S. studies that reported increased life expectancy,
decreased mortality, and decreased respiratory effects following past
declines in ambient PM2.5 concentrations. Such studies
examined ``starting'' annual average PM2.5 concentrations
(i.e., prior to the reductions being evaluated) ranging from about 13.2
to >20[micro]g/m\3\ (i.e., U.S. EPA, 2020b, Table 3-3). Given the lack
of available accountability studies
[[Page 16223]]
reporting public health improvements attributable to reductions in
ambient PM2.5 in locations meeting the existing standards,
together with his broader concerns regarding the lack of experimental
studies examining PM2.5 exposures typical of areas meeting
the existing standards, the then-Administrator judged that there was
considerable uncertainty in the potential for increased public health
protection from further reductions in ambient PM2.5
concentrations beyond those achieved under the existing primary
PM2.5 standards (85 FR 82717, December 18, 2020).
When the above considerations were taken together, the then-
Administrator concluded that the scientific evidence assessed in the
2019 ISA, together with the analyses in the 2020 PA based on that
evidence and consideration of CASAC advice and public comments, did not
call into question the adequacy of the public health protection
provided by the existing annual and 24-hour PM2.5 standards.
In particular, the then-Administrator judged that there was
considerable uncertainty in the potential for additional public health
improvements from reducing ambient PM2.5 concentrations
below the concentrations achieved under the existing primary standards
and that, therefore, standards more stringent than the existing
standards (e.g., with lower levels) were not supported. That is, he
judged that more stringent standards would be more than requisite to
protect the public health with an adequate margin of safety. This
judgment reflected the Administrator's consideration of the
uncertainties in the potential implications of the lower end of the air
quality distributions from the epidemiologic studies due in part to the
lack of supporting evidence from experimental studies and retrospective
accountability studies conducted at PM2.5 concentrations
meeting the existing standards (85 FR 82717, December 18, 2020).
In reaching this conclusion in the 2020 review, the then-
Administrator judged that the existing standards provided an adequate
margin of safety. With respect to the annual standard, the level of 12
[micro]g/m\3\ was below the lowest ``starting'' concentration (i.e.,
13.2 [micro]g/m\3\) in the available accountability studies that showed
public health improvements attributable to reductions in ambient
PM2.5. In addition, while the then-Administrator placed less
weight on the epidemiologic evidence for selecting a standard, he noted
that the level of the annual standard was below the reported mean (and
median) concentrations in the majority of the key U.S. epidemiologic
studies using ground-based monitoring data (noting that these means
tend to be 10-20% lower than their corresponding area design values
which is the more relevant metric when considering the level of the
standard) and below the mean of the reported means (or medians) of
these studies (i.e., 13.5 [micro]g/m\3\). In addition, the then-
Administrator recognized that concentrations in areas meeting the
existing 24-hour and annual standards remained well-below the
PM2.5 exposure concentrations consistently shown to elicit
effects in human exposure studies (85 FR 82717-82718, December 18,
2020).
In addition, based on the then-Administrator's review of the
science in the 2020 review, including controlled human exposure studies
examining effects following short-term PM2.5 exposures, the
epidemiologic studies, and accountability studies conducted at levels
just above the existing annual standard, he judged that the degree of
public health protection provided by the existing annual standard is
not greater than warranted. This judgment, together with the fact that
no CASAC member expressed support for a less stringent standard, led
the then- Administrator to conclude that standards less stringent than
the existing standards (e.g., with higher levels) were also not
supported (85 FR 82718, December 18, 2020).
In reaching his final decision in the 2020 review, the then-
Administrator concluded that the scientific evidence and technical
information continued to support the existing annual and 24-hour
PM2.5 standards. This conclusion reflected the then-
Administrator's view that there were important limitations and
uncertainties that remained in the evidence. The then-Administrator
concluded that these limitations contributed to considerable
uncertainty regarding the potential public health implications of
revising the existing primary PM2.5 standards. Given this
uncertainty, and noting the advice from some CASAC members, he
concluded that the primary PM2.5 standards, including the
indicators (PM2.5), averaging times (annual and 24-hour),
forms (arithmetic mean and 98th percentile, averaged over three years)
and levels (12.0 [micro]g/m\3\, 35 [micro]g/m\3\), when taken together,
remained requisite to protect the public health. Therefore, in the 2020
review, the Administrator reached the conclusion that the primary 24-
hour and annual PM2.5 standards, together, were requisite to
protect public health from fine particles with an adequate margin of
safety, including the health of at-risk populations, and retained the
standards, without revision (85 FR 82718, December 18, 2020).
2. Overview of the Health Effects Evidence
The information summarized here and further detailed in section
II.B of the proposal (88 FR 5580, January 27, 2023), is an overview of
the policy-relevant aspects of the health effects evidence available in
this reconsideration; the assessment of this evidence is documented in
the 2019 ISA (U.S. EPA, 2019a) and ISA Supplement (U.S. EPA, 2022a) and
its policy implications are further discussed in the 2022 PA (U.S. EPA,
2022b). While the 2019 ISA provides the broad scientific foundation for
this reconsideration, additional literature has become available since
the cutoff date of the 2019 ISA that expands the body of evidence
related to mortality and cardiovascular effects for both short- and
long-term PM2.5 exposure, which can inform the
Administrator's judgment on the adequacy of the current primary
PM2.5 standards. As such, the ISA Supplement builds on the
information presented within the 2019 ISA with a targeted
identification and evaluation of new scientific information (U.S. EPA,
2022a, section 1.2). The ISA Supplement focuses on PM2.5
health effects evidence where the 2019 ISA concludes a ``causal
relationship,'' because such health effects are given the most weight
in an Administrator's decisions in a NAAQS review. As such, in
selecting the health effects to evaluate within the ISA Supplement
(i.e., newly available evidence related to short- and long-term
PM2.5 exposure and mortality and cardiovascular effects),
the primary rationale is based on the causality determinations for
health effect categories presented in the 2019 PM ISA, and the
subsequent use of the health effects evidence in the 2020 PM PA.
Specifically, U.S. and Canadian epidemiologic studies for mortality and
cardiovascular effects, along with controlled human exposure studies
associated with cardiovascular effects at near ambient concentrations,
were considered to be of greatest utility in informing the
Administrator's conclusions on the adequacy of the current primary
PM2.5 standards. Additionally, studies examining
associations outside the U.S. or Canada reflect air quality and
exposure patterns that may be less typical of the U.S., and thus less
likely to be informative for purposes of reviewing the NAAQS (U.S. EPA,
2022b, p.1-3). While the ISA Supplement does not include information
for health effects other than mortality and cardiovascular effects, the
[[Page 16224]]
scientific evidence for other health effect categories is evaluated in
the 2019 ISA, which in combination with the ISA Supplement represents
the complete scientific record for the reconsideration of the 2020
final decision.
The ISA Supplement also assessed accountability studies because
these types of epidemiologic studies were part of the body of evidence
that was a focus of the 2020 review. Accountability studies inform our
understanding of the potential for public health improvements as
ambient PM2.5 concentrations have declined over time.
Further, the ISA Supplement considered studies that employed
statistical approaches that attempt to more extensively account for
confounders and are more robust to model misspecification (i.e., used
alternative methods for confounder control),\50\ given that such
studies were highlighted by the CASAC and identified in public comments
in the 2020 review. Since the literature cutoff date for the 2019 ISA,
multiple accountability studies and studies that employ alternative
methods for confounder control have become available for consideration
in the ISA Supplement and, subsequently, in this reconsideration.
---------------------------------------------------------------------------
\50\ As noted in the ISA Supplement (U.S. EPA, 2022a, p. 1-3):
``In the peer-reviewed literature, these epidemiologic studies are
often referred to as causal inference studies or studies that used
causal modeling methods. For the purposes of this Supplement, this
terminology is not used to prevent confusion with the main
scientific conclusions (i.e., the causality determinations)
presented within an ISA. In addition, as is consistent with the
weight-of-evidence framework used within ISAs and discussed in the
Preamble to the Integrated Science Assessments, an individual study
on its own cannot inform causality, but instead represents a piece
of the overall body of evidence.''
---------------------------------------------------------------------------
The ISA Supplement also considered recent health effects evidence
that addresses key scientific issues where the literature has expanded
since the completion of the 2019 ISA.\51\ The 2019 ISA evaluated a
couple of controlled human exposure studies that investigated the
effect of exposure to near-ambient concentrations of PM2.5
(U.S. EPA, 2019a, section 6.1.10 and 6.1.13). The ISA Supplement adds
to this limited evidence, including a recent study conducted in young
healthy individuals exposed to near-ambient PM2.5
concentrations (U.S. EPA, 2022a, section 3.3.1). Given the importance
of identifying populations at increased risk of PM2.5-
related effects, the ISA Supplement also included epidemiologic or
exposure studies that examined whether there is evidence of exposure or
risk disparities by race/ethnicity or SES. These types of studies
provide additional information related to factors that may increase
risk of PM2.5-related health effects and provide additional
evidence for consideration by the Administrator in reaching conclusions
regarding the adequacy of the current standards. In addition, the ISA
Supplement evaluated studies that examined the relationship between
short- and long-term PM2.5 exposures and SARS-CoV-2
infection and/or COVID-19 death, as these studies are a new area of
research and were raised by a number of public commenters in the 2020
review.
---------------------------------------------------------------------------
\51\ As with the epidemiologic studies for long- and short-term
PM2.5 exposure and mortality and cardiovascular effects,
epidemiologic studies of exposure or risk disparities and SARS-CoV-2
infection and/or COVID-19 death were limited to those conducted in
the U.S. and Canada.
---------------------------------------------------------------------------
The evidence presented within the 2019 ISA, along with the targeted
identification and evaluation of new scientific information in the ISA
Supplement, provides the scientific basis for the reconsideration of
the 2020 final decision on the primary PM2.5 standards. The
subsections below briefly summarize the nature of PM2.5-
related health effects (II.A.2.a), with a focus on those health effects
for which the 2019 ISA concluded a ``causal'' or ``likely to be
causal'' relationship, the potential public health implications and
populations at risk (II.A.2.b), and PM2.5 concentrations in
key studies reporting health effects (II.A.2.c).
a. Nature of Effects
The evidence base available in the reconsideration includes decades
of research on PM2.5-related health effects (U.S. EPA,
2004b; U.S. EPA, 2009a; U.S. EPA, 2019a), including the full body of
evidence evaluated in the 2019 ISA (U.S. EPA, 2019a), along with the
targeted evaluation of recent evidence in the ISA Supplement (U.S. EPA,
2022a). In considering the available scientific evidence, the sections
below, and in more detail in section II.B.1 of the proposal (88 FR
5580, January 27, 2023), summarize the relationships between long-and
short-term PM2.5 exposures and mortality (II.A.2.a.i),
cardiovascular effects (II.A.2.a.ii), respiratory effects
(II.A.2.a.iii), cancer (II.A.2.a.iv), nervous system effects
(II.A.2.a.v) and other effects (II.A.2.a.vi). For these outcomes, the
2019 ISA concluded that the evidence supports either a ``causal'' or a
``likely to be causal'' relationship.\52\
---------------------------------------------------------------------------
\52\ In this reconsideration of the PM NAAQS, the EPA considers
the full body of health evidence, placing the greatest emphasis on
the health effects for which the evidence has been judged in the
2019 ISA to demonstrate a ``causal'' or ``likely to be causal''
relationship with PM2.5 exposures.
---------------------------------------------------------------------------
i. Mortality
Long-Term PM2.5 Exposures
In the 2012 review, the 2009 ISA reported that the evidence was
``sufficient to conclude that the relationship between long-term
PM2.5 exposures and mortality is causal'' (U.S. EPA, 2009a,
p. 7-96). The strongest evidence supporting this conclusion was
provided by epidemiologic studies, particularly those examining two
seminal cohorts, the American Cancer Society (ACS) cohort and the
Harvard Six Cities cohort. Analyses of the Harvard Six Cities cohort
included evidence indicating that reductions in ambient
PM2.5 concentrations are associated with reduced mortality
risk (Laden et al., 2006) and increases in life expectancy (Pope et
al., 2009). Further support was provided by other cohort studies
conducted in North America and Europe that reported positive
associations between long-term PM2.5 exposure and mortality
(U.S. EPA, 2019a).
Cohort studies, which have become available since the completion of
the 2009 ISA and evaluated in the 2019 ISA, continue to provide
consistent evidence of positive associations between long-term
PM2.5 exposures and mortality. These studies add support for
associations with all-cause and total (non-accidental) mortality,\53\
as well as with specific causes of mortality, including cardiovascular
disease and respiratory disease (U.S. EPA, 2019a, section 11.2.2).
Several of these studies conducted analyses over longer study durations
and periods of follow-up than examined in the original ACS and Harvard
Six Cities cohort studies and continue to report positive associations
between long-term exposure to PM2.5 and mortality (U.S. EPA,
2019a, section 11.2.2.1; Figures 11-18 and 11-19). In addition to
studies focusing on the ACS and Harvard Six Cities cohorts, additional
studies examining other cohorts also provide evidence of consistent,
positive associations between long-term PM2.5 exposure and
mortality across a wide range of demographic groups (e.g., age, sex,
occupation), spatial and temporal extents, exposure assessment metrics,
and statistical techniques (U.S. EPA, 2019a, sections 11.2.2.1, 11.2.5;
U.S. EPA, 2022a, Table 11-8). This includes some of the largest cohort
studies conducted to date, such as analyses of the U.S. Medicare cohort
that includes
[[Page 16225]]
nearly 61 million enrollees and studies that control for a range of
individual and ecological covariates, including race, age, SES, smoking
status, body mass index, and annual weather variables (e.g.,
temperature, humidity) (U.S. EPA, 2019a).
---------------------------------------------------------------------------
\53\ The majority of these studies examined non-accidental
mortality outcomes, though some Medicare studies lack cause-specific
death information and, therefore, examine total mortality.
---------------------------------------------------------------------------
In addition to those cohort studies evaluated in the 2019 ISA,
recent North American cohort studies evaluated in the ISA Supplement
continue to examine the relationship between long-term PM2.5
exposure and mortality and report consistent, positive, and
statistically significant associations. These recent studies also
utilize large and demographically diverse cohorts that are generally
representative of the national populations in both the U.S. and Canada.
These ``studies published since the 2019 ISA support and extend the
evidence base that contributed to the conclusion of a causal
relationship between long-term PM2.5 exposure and
mortality'' (U.S. EPA, 2022a, section 3.2.2.2.1, Figure 3-19, Figure 3-
20).
Furthermore, studies evaluated in the 2019 ISA and the ISA
Supplement that examined cause-specific mortality expand upon previous
research that found consistent, positive associations between
PM2.5 exposure and specific mortality outcomes, which
include cardiovascular and respiratory mortality, as well as other
mortality outcomes. For cardiovascular-related mortality, the evidence
evaluated in the ISA Supplement is consistent with the evidence
evaluated in the 2019 ISA with recent studies reporting positive
associations with long-term PM2.5 exposure. When evaluating
cause-specific cardiovascular mortality, recent studies reported
positive associations for a number of outcomes, such as ischemic heart
disease (IHD) and stroke mortality (U.S. EPA, 2022a, Figure 3-23).
Moreover, recent studies also provide some initial evidence that
individuals with pre-existing health conditions, such as heart failure
and diabetes, are at an increased risk of PM2.5-related
health effects (U.S. EPA, 2022a, section 3.2.2.4) and that these
individuals have a higher risk of mortality overall, which was
previously only examined in studies that used stratified analyses
rather than a cohort of people with an underlying health condition
(U.S. EPA, 2022a, section 3.2.2.4). With regard to respiratory
mortality, epidemiologic studies evaluated in the 2019 ISA and ISA
Supplement continue to provide support for associations between long-
term PM2.5 exposure and respiratory mortality (U.S. EPA,
2019a, section 5.2.10; U.S. EPA, 2022a, Table 3-2).
A series of epidemiologic studies evaluated in the 2019 ISA tested
the hypothesis that past reductions in ambient PM2.5
concentrations are associated with increased life expectancy or a
decreased mortality rate and report that reductions in ambient
PM2.5 are associated with improvements in longevity (U.S.
EPA, 2022a, section 11.2.2.5). Pope et al. (2009) conducted a cross-
sectional analysis using air quality data from 51 metropolitan areas
across the U.S., beginning in the 1970s through the early 2000s, and
found that a 10 [micro]g/m\3\ decrease in long-term PM2.5
concentration was associated with a 0.61-year increase in life
expectancy. In a subsequent analysis, the authors extended the period
of analysis to include 2000 to 2007, a time period with lower ambient
PM2.5 concentrations and found a decrease in long-term
PM2.5 concentration continued to be associated with an
increase in life expectancy, though the magnitude of the increase was
smaller than during the earlier time period (i.e., a 10 [micro]g/m\3\
decrease in long-term PM2.5 concentration was associated
with a 0.35-year increase in life expectancy) (Correia et al., 2013).
Additional studies conducted in the U.S. or Europe similarly report
that reductions in ambient PM2.5 are associated with
improvements in longevity (U.S. EPA, 2022a, section 11.2.2.5).
Since the literature cutoff date for the 2019 ISA, a few
epidemiologic studies were published that examined the relationship
between long-term PM2.5 exposure and life-expectancy (U.S.
EPA, 2022a, section 3.2.1.3) and report results that are consistent
with and expand upon the body of evidence from the 2019 ISA. For
example, Bennett et al. (2019) reported that PM2.5
concentrations above the lowest observed concentration (2.8 [micro]g/
m\3\) were associated with a 0.15 year decrease in national life
expectancy for women and 0.13 year decrease in national life expectancy
for men (U.S. EPA, 2022a, section 3.2.2.2.4, Figure 3-25). Another
study compared participants living in areas with PM2.5
concentrations >12 [micro]g/m\3\ to participants living in areas with
PM2.5 concentrations <12 [micro]g/m\3\ and reported that the
number of years of life lost due to living in areas with higher
PM2.5 concentrations was 0.84 years over a 5-year period
(Ward-Caviness et al., 2020; U.S. EPA, 2022a, section 3.2.2.2.4).
Additionally, a number of accountability studies, which are
epidemiologic studies that evaluate whether an environmental policy or
air quality intervention resulted in reductions in ambient air
pollution concentrations and subsequent reductions in mortality or
morbidity, have emerged and were evaluated in the ISA Supplement (U.S.
EPA, 2022a, section 3.2.2.3). For example, Sanders et al. (2020a)
examined whether policy actions (i.e., the first annual
PM2.5 NAAQS implementation rule in 2005 for the 1997 annual
PM2.5 standard with a 3-year annual average of 15.0 [mu]g/
m\3\) reduced PM2.5 concentrations and mortality rates in
Medicare beneficiaries between 2000-2013, and found that following
implementation of the annual PM2.5 NAAQS, annual
PM2.5 concentrations decreased by 1.59 [mu]g/m\3\ (95% CI:
1.39, 1.80) which corresponded to a 0.93% reduction in mortality rates
among individuals 65 years and older ([95% CI: 0.10%, 1.77%) in non-
attainment counties relative to attainment counties.
The 2019 ISA also evaluated a small number of studies that used
alternative methods for confounder control to further assess
relationship between long-term PM2.5 exposure and mortality
(U.S. EPA, 2019a, section 11.2.2.4). In addition, multiple
epidemiologic studies that implemented alternative methods for
confounder control and were published since the literature cutoff date
of the 2019 ISA were evaluated in the ISA Supplement (U.S. EPA, 2022a,
section 3.2.2.3). These studies used a variety of statistical methods
including generalized propensity score (GPS), inverse probability
weighting (IPW), and difference-in-difference (DID) to reduce
uncertainties related to confounding bias in the association between
long-term PM2.5 exposure and mortality. These studies
reported consistent positive associations between long-term
PM2.5 exposure and total mortality (U.S. EPA, 2022a, section
3.2.2.3), and provided further support for the associations reported in
the cohort studies referenced above.
The 2019 ISA and ISA Supplement also evaluated the degree to which
recent studies examining the relationship between long-term
PM2.5 exposure and mortality addressed key policy-relevant
issues and/or previously identified data gaps in the scientific
evidence, including methods to estimate exposure, methods to control
for confounding (e.g., co-pollutant confounding), the shape of the C-R
relationship, as well as examining whether a threshold exists below
which mortality effects do not occur. With respect to exposure
assessment, based on its evaluation of the evidence, the 2019 ISA
concludes that positive associations between long-term PM2.5
exposures and mortality are robust
[[Page 16226]]
across recent analyses using various approaches to estimate
PM2.5 exposures (e.g., based on monitors, models, satellite-
based methods, or hybrid methods that combine information from multiple
sources) (U.S. EPA, 2019a, section 11.2.5.1). Hart et al. (2015) report
that correction for bias due to exposure measurement error increases
the magnitude of the hazard ratios (confidence intervals widen but the
association remains statistically significant), suggesting that failure
to correct for exposure measurement error could result in attenuation
or underestimation of risk estimates.
The 2019 ISA additionally concludes that positive associations
between long-term PM2.5 exposures and mortality are robust
across statistical models that use different approaches to control for
confounders or different sets of confounders (U.S. EPA, 2019a, sections
11.2.3 and 11.2.5), across diverse geographic regions and populations,
and across a range of temporal periods including periods of declining
PM concentrations (U.S. EPA, 2019a, sections 11.2.2.5 and 11.2.5.3).
Additional evidence further demonstrates that associations with
mortality remain robust in copollutants analyses (U.S. EPA, 2019a,
section 11.2.3), and that associations persist in analyses restricted
to long-term exposures (annual average PM2.5 concentrations)
below 12 [mu]g/m\3\ (Di et al., 2017b) or 10 [mu]g/m\3\ (Shi et al.,
2016), indicating that risks are not disproportionately driven by the
upper portions of the air quality distribution. Recent studies
evaluated in the ISA Supplement further assess potential copollutant
confounding and indicate that while there is some evidence of potential
confounding of the PM2.5-mortality association by
copollutants in some of the studies (i.e., those studies of the
Mortality Air Pollution Associations in Low Exposure Environments
(MAPLE) cohort), this result is inconsistent with other recent studies
evaluated in the 2019 ISA that were conducted in the U.S. and Canada
that found associations in both single and copollutant models (U.S.
EPA, 2019a; U.S. EPA, 2022a, section 3.2.2.4)
Additionally, a few studies use statistical techniques to reduce
uncertainties related to potential confounding to further inform
conclusions on causality for long-term PM2.5 exposure and
mortality, as further detailed in section II.B.1.a.i of the proposal
(88 FR 5582, January 27, 2023), studies by Greven et al. (2011), Pun et
al. (2017), and Eum et al. (2018) completed sensitivity analyses as
part of their Medicare cohort study in which they decompose ambient
PM2.5 into ``spatial'' and ``spatiotemporal'' components in
order to evaluate the potential for bias due to unmeasured spatial
confounding. Pun et al. (2017) observed positive associations for the
``temporal'' variation model and approximately null associations for
the ``spatiotemporal'' variation model for all causes of death except
for COPD mortality. The difference in the results of these two models
for most causes of death suggests the presence of unmeasured
confounding, though the authors do not indicate anything about the
direction or magnitude of this bias. It is important to note that the
``temporal'' and ``spatiotemporal'' coefficients are not directly
comparable to the results of other epidemiologic studies when examined
individually and can only be used in comparison with one another to
evaluate the potential for unmeasured confounding bias. Eum et al.
(2018) and Wu et al. (2020) also attempted to address long-term trends
and meteorological variables as potential confounders and found that
not adjusting for temporal trends could overestimate the association,
while effect estimates in analyses that excluded meteorological
variables remained unchanged compared to the main analyses. While
results of these analyses suggest the presence of some unmeasured
confounding, they do not indicate the direction or magnitude of the
bias.\54\
---------------------------------------------------------------------------
\54\ In public comments on the 2019 draft PA, the authors of the
Pun et al. (2017) study further note that ``the presence of
unmeasured confounding. . .was expected given that we did not
control for several potential confounders that may impact
PM2.5-mortality associations, such as smoking, socio-
economic status (SES), gaseous pollutants, PM2.5
components, and long-term time trends in PM2.5'' and that
``spatial confounding may bias mortality risks both towards and away
from the null'' (Docket ID EPA-HQ-OAR-2015-0072-0065; accessible in
https://www.regulations.gov/).
---------------------------------------------------------------------------
An additional important consideration in characterizing the public
health impacts associated with PM2.5 exposure is whether C-R
relationships are linear across the range of concentrations or if
nonlinear relationships exist along any part of this range. Studies
evaluated in the 2019 ISA and the ISA Supplement examine this issue,
and continue to provide evidence of linear, no-threshold relationships
between long-term PM2.5 exposures and all-cause and cause-
specific mortality (U.S. EPA, 2019a, section 11.2.4; U.S. EPA, 2022a,
section 3.2.2.2.7, Table 3-6). Across the studies evaluated in the 2019
ISA and the ISA Supplement, a variety of statistical methods have been
used to assess whether there is evidence of deviations in linearity
(U.S. EPA, 2019a, Table 11-7; U.S. EPA, 2022a, section 2.2.3.2).
Studies have also conducted cut-point analyses that focus on examining
risk at specific ambient PM2.5 concentrations. Generally,
the evidence remains consistent in supporting a no-threshold
relationship, and in supporting a linear relationship for
PM2.5 concentrations >8 [mu]g/m\3\. However, uncertainties
remain about the shape of the C-R curve at PM2.5
concentrations <8 [mu]g/m\3\, with some recent studies providing
evidence for either a sublinear, linear, or supralinear relationship at
these lower concentrations (U.S. EPA, 2019a, section 11.2.4; U.S. EPA,
2022a, section 2.2.3.2). There was also some limited evidence
indicating that the slope of the C-R function may be steeper
(supralinear) at lower concentrations for cardiovascular mortality
(U.S. EPA, 2022a, section 3.1.1.2.6).
The biological plausibility of PM2.5-attributable
mortality is supported by the coherence of effects across scientific
disciplines (i.e., animal toxicological, controlled human exposure
studies, and epidemiologic) when evaluating respiratory and
cardiovascular morbidity effects, which are some of the largest
contributors to total (nonaccidental) mortality. The 2019 ISA outlines
the available evidence for biologically plausible pathways by which
inhalation exposure to PM2.5 could progress from initial
events (e.g., pulmonary inflammation, autonomic nervous system
activation) to endpoints relevant to population outcomes, particularly
those related to cardiovascular diseases such as ischemic heart
disease, stroke and atherosclerosis (U.S. EPA, 2019a, section 6.2.1),
and to metabolic effects, including diabetes (U.S. EPA, 2019a, section
7.3.1). The 2019 ISA notes ``more limited evidence from respiratory
morbidity'' (U.S. EPA, 2019a, p. 11-101) such as development of chronic
obstructive pulmonary disease (COPD) (U.S. EPA, 2019a, section 5.2.1)
to support the biological plausibility of mortality due to long-term
PM2.5 exposures (U.S. EPA, 2019a, section 11.2.1).
Taken together, epidemiologic studies, including those evaluated in
the 2019 ISA and more recent studies evaluated in the ISA Supplement,
consistently report positive associations between long-term
PM2.5 exposure and mortality across different geographic
locations, populations, and analytic approaches (U.S. EPA, 2019a; U.S.
EPA, 2022a, section 3.2.2.4). As such, these studies reduce key
uncertainties identified in previous reviews, including those related
to potential
[[Page 16227]]
copollutant confounding, and provide additional information on the
shape of the C-R curve. As evaluated in the 2019 ISA, experimental and
epidemiologic evidence for cardiovascular effects, and respiratory
effects to a more limited degree, supports the plausibility of
mortality due to long-term PM2.5 exposures. Overall, studies
evaluated in the 2019 ISA support the conclusion of a causal
relationship between long-term PM2.5 exposure and mortality,
which is supported and extended by evidence from recent epidemiologic
studies evaluated in the ISA Supplement (U.S. EPA, 2022a, section
3.2.2.4).
Short-Term PM2.5 Exposures
The 2009 ISA concluded that ``a causal relationship exists between
short-term exposure to PM2.5 and mortality'' (U.S. EPA,
2009a). This conclusion was based on the evaluation of both multi- and
single-city epidemiologic studies that consistently reported positive
associations between short-term PM2.5 exposure and non-
accidental mortality. These associations were strongest, in terms of
magnitude and precision, primarily at lags of 0 to 1 days. Examination
of the potential confounding effects of gaseous copollutants was
limited, though evidence from single-city studies indicated that
gaseous copollutants have minimal effect on the PM2.5-
mortality relationship (i.e., associations remain robust to inclusion
of other pollutants in copollutant models). The evaluation of cause-
specific mortality found that effect estimates were larger in
magnitude, but also had larger confidence intervals, for respiratory
mortality compared to cardiovascular mortality. Although the largest
mortality risk estimates were for respiratory mortality, the
interpretation of the results was complicated by the limited coherence
from studies of respiratory morbidity. However, the evidence from
studies of cardiovascular morbidity provided both coherence and
biological plausibility for the relationship between short-term
PM2.5 exposure and cardiovascular mortality.
Multicity studies evaluated in the 2019 ISA and the ISA Supplement
provide evidence of primarily positive associations between daily
PM2.5 exposures and mortality, with percent increases in
total mortality ranging from 0.19% (Lippmann et al., 2013) to 2.80%
(Kloog et al., 2013) \55\ at lags of 0 to 1 days in single-pollutant
models. Whereas many studies assign exposures using data from ambient
monitors, other studies employ hybrid modeling approaches, which
estimate PM2.5 concentrations using data from a variety of
sources (i.e., from satellites, land use information, and modeling, in
addition to monitors) and enable the inclusion of less urban and more
rural locations in analyses (e.g., Kloog et al., 2013, Lee et al.,
2015, Shi et al., 2016).
---------------------------------------------------------------------------
\55\ As detailed in the Preface to the ISA, risk estimates are
for a 10 [micro]g/m\3\ increase in 24-hour avg PM2.5
concentrations, unless otherwise noted (U.S. EPA, 2019a).
---------------------------------------------------------------------------
Some studies have expanded the examination of potential confounders
including long-term temporal trends, weather, and co-occurring
pollutants. Mortality associations were found to remain positive,
although in some cases were attenuated, when using different approaches
to account for temporal trends or weather covariates (e.g., U.S. EPA,
2019a, section 11.1.5.1). For example, Sacks et al. (2012) examined the
influence of model specification using the approaches for confounder
adjustment from models employed in several multicity studies within the
context of a common data set (U.S. EPA, 2019a, section 11.1.5.1). These
models use different approaches to control for long-term temporal
trends and the potential confounding effects of weather. The authors
report that associations between daily PM2.5 and
cardiovascular mortality were similar across models, with the percent
increase in mortality ranging from 1.5-2.0% (U.S. EPA, 2019a, Figure
11-4). Thus, alternative approaches to controlling for long-term
temporal trends and for the potential confounding effects of weather
may influence the magnitude of the association between PM2.5
exposures and mortality but have not been found to influence the
direction of the observed association (U.S. EPA, 2019a, section
11.1.5.1). Taken together, the 2019 ISA and the ISA Supplement conclude
that recent multicity studies conducted in the U.S., Canada, Europe,
and Asia continue to provide consistent evidence of positive
associations between short-term PM2.5 exposures and total
mortality across studies that use different approaches to control for
the potential confounding effects of weather (e.g., temperature) (U.S.
EPA, 2019a, section 1.4.1.5.1; U.S. EPA, 2022a, section 3.2.1.2).
With regard to copollutants, studies evaluated in the 2019 ISA
provide additional evidence that associations between short-term
PM2.5 exposures and mortality remain positive and relatively
unchanged in copollutant models with both gaseous pollutants and
PM10-2.5 (U.S. EPA, 2019a, section 11.1.4). Additionally,
the low (r < 0.4) to moderate correlations (r = 0.4-0.7) between
PM2.5 and gaseous pollutants and PM10-2.5
increase the confidence in PM2.5 having an independent
effect on mortality (U.S. EPA, 2019a, section 11.1.4). Consistent with
the studies evaluated in the 2019 ISA, studies evaluated in the ISA
Supplement that used data from more recent years also indicate that
associations between short-term PM2.5 exposure and mortality
remain unchanged in copollutant models. However, the evidence indicates
that the association could be larger in magnitude in the presence of
some copollutants such as oxidant gases (Lavigne et al., 2018; Shin et
al., 2021).
The generally positive associations reported with mortality are
supported by a small group of studies employing alternative methods for
confounder control or quasi-experimental statistical approaches (U.S.
EPA, 2019a, section 11.1.2.1). For example, two studies by Schwartz et
al. report associations between PM2.5 instrumental variables
and mortality (U.S. EPA, 2019a, Table 11-2), including in an analysis
limited to days with 24-hour average PM2.5 concentrations
<30 [mu]g/m\3\ (Schwartz et al., 2015; Schwartz et al., 2017). In
addition to the main analyses, these studies conducted Granger-like
causality tests as sensitivity analyses to examine whether there was
evidence of an association between mortality and PM2.5 after
the day of death, which would support the possibility that unmeasured
confounders were not accounted for in the statistical model. Neither
study reports evidence of an association with PM2.5 after
death (i.e., they do not indicate unmeasured confounding). Yorifuji et
al. (2016) conducted a quasi-experimental study to examine whether a
specific regulatory action in Tokyo, Japan (i.e., a diesel emission
control ordinance) resulted in a subsequent reduction in daily
mortality (Yorifuji et al., 2016). The authors reported a reduction in
mortality in Tokyo due to the ordinance, compared to Osaka, which did
not have a similar diesel emission control ordinance in place. In
another study, Schwartz et al. (2018) utilized three statistical
methods including instrumental variable analysis, a negative exposure
control, and marginal structural models to estimate the association
between PM2.5 and daily mortality (Schwartz et al., 2018).
Results from this study continue to support a relationship between
short-term PM2.5 exposure and mortality. Additional
epidemiologic studies evaluated in the ISA Supplement that employed
alternative methods for confounder control to examine the association
between short-term PM2.5 exposure and
[[Page 16228]]
mortality also report consistent positive associations in studies that
examine effects across multiple cities in the U.S. (U.S. EPA, 2022a).
The positive associations for total mortality reported across the
majority of studies evaluated are further supported by cause-specific
mortality analyses, which generally report consistent, positive
associations with both cardiovascular and respiratory mortality (U.S.
EPA, 2019a, section 11.1.3). Recent multicity studies evaluated in the
ISA Supplement add to the body of evidence indicating a relationship
between short-term PM2.5 exposure and cause-specific
mortality, with more variability in the magnitude and precision of
associations for respiratory mortality (U.S. EPA, 2022a; Figure 3-14).
For both cardiovascular and respiratory mortality, there has been a
limited assessment of potential copollutant confounding, though initial
evidence indicates that associations remain positive and relatively
unchanged in models with gaseous pollutants and PM10-2.5,
which further supports the copollutant analyses conducted for total
mortality. The strong evidence for ischemic events and heart failure,
as detailed in the assessment of cardiovascular morbidity (U.S. EPA,
2019a, Chapter 6), provides biological plausibility for
PM2.5-related cardiovascular mortality, which comprises the
largest percentage of total mortality (i.e., ~33%) (NHLBI, 2017).
Although there is evidence for exacerbations of COPD and asthma, the
collective body of respiratory morbidity evidence provides limited
biological plausibility for PM2.5-related respiratory
mortality (U.S. EPA, 2019a, Chapter 5).
In the 2009 ISA, one of the main uncertainties identified was the
regional and city-to-city heterogeneity in PM2.5-mortality
associations. Studies evaluated in the 2019 ISA examine both city-
specific as well as regional characteristics to identify the underlying
contextual factors that could contribute to this heterogeneity (U.S.
EPA, 2019a, section 11.1.6.3). Analyses focusing on effect modification
of the PM2.5 mortality relationship by PM2.5
components, regional patterns in PM2.5 components and city
specific differences in composition and sources indicate some
differences in the PM2.5 composition and sources across
cities and regions, but these differences do not fully explain the
observed heterogeneity. Additional studies find that factors related to
potential exposure differences, such as housing stock and commuting, as
well as city specific factors (e.g., land use, port volume, and traffic
information), may also explain some of the observed heterogeneity (U.S.
EPA, 2019a, section 11.1.6.3). Collectively, studies evaluated in the
2019 ISA and the ISA Supplement indicate that the heterogeneity in
PM2.5 mortality risk estimates cannot be attributed to one
factor, but instead a combination of factors including, but not limited
to, PM composition and sources as well as community characteristics
that could influence exposures (U.S. EPA, 2019a, section 11.1.12; U.S.
EPA, 2022a, section 3.2.1.2.1).
A number of studies evaluated in the 2019 ISA and ISA Supplement
conducted systematic evaluations of the lag structure of associations
for the PM2.5-mortality relationship by examining either a
series of single day or multiday lags and these studies continue to
support an immediate effect (i.e., lag 0 to 1 days) of short-term
PM2.5 exposures on mortality (U.S. EPA, 2019a, section
11.1.8.1; U.S. EPA, 2022a, section 3.2.1.1). Recent studies also
conducted analyses comparing the traditional 24-hour average exposure
metric with a subdaily metric (i.e., 1-hour max) and provide evidence
of a similar pattern of associations for both the 24-hour average and
1-hour max metric, with the association larger in magnitude for the 24-
hour average metric.
Multicity studies indicate that positive and statistically
significant associations with mortality persist in analyses restricted
to short-term (24-hour average PM2.5 concentrations)
PM2.5 exposures below 35 [mu]g/m\3\ (Lee et al., 2015),\56\
below 30 [mu]g/m\3\ (Shi et al., 2016), and below 25 [mu]g/m\3\ (Di et
al., 2017a), indicating that risks associated with short-term
PM2.5 exposures are not disproportionately driven by the
peaks of the air quality distribution. Additional studies examined the
shape of the C-R relationship for short-term PM2.5 exposure
and mortality and whether a threshold exists below which mortality
effects do not occur (U.S. EPA, 2019a, section 11.1.10). These studies
used various statistical approaches and consistently demonstrate linear
C-R relationships with no evidence of a threshold.
---------------------------------------------------------------------------
\56\ Lee et al. (2015) restrict exposures below 35 [mu]g/m\3\
only in areas with annual average concentrations <12 [mu]g/m\3\.
Additionally, Lee et al. (2015) also report that positive and
statistically significant associations between short-term
PM2.5 exposures and mortality persist in analyses
restricted to areas with long-term concentrations below 12 [mu]g/
m\3\.
---------------------------------------------------------------------------
Moreover, recent studies evaluated in the ISA Supplement provide
additional support for a linear, no-threshold C-R relationship between
short-term PM2.5 exposure and mortality, with confidence in
the shape decreasing at concentrations below 5 [micro]g/m\3\ (Shi et
al., 2016; Lavigne et al., 2018). Recent analyses provide initial
evidence indicating that PM2.5-mortality associations
persist and may be stronger (i.e., a steeper slope) at lower
concentrations (e.g., Di et al., 2017a; Figure 11-12 in U.S. EPA,
2019). However, given the limited data available at the lower end of
the distribution of ambient PM2.5 concentrations, the shape
of the C-R curve remains uncertain at these low concentrations.
Although difficulties remain in assessing the shape of the short-term
PM2.5-mortality C-R relationship, to date, studies have not
conducted systematic evaluations of alternatives to linearity and
recent studies evaluated in the ISA Supplement continue to provide
evidence of a no-threshold linear relationship, with less confidence at
concentrations lower than 5 [micro]g/m\3\.
Overall, epidemiologic studies evaluated in the 2019 ISA and the
ISA Supplement build upon and extend the conclusions of the 2009 ISA
for the relationship between short-term PM2.5 exposures and
total mortality. Supporting evidence for PM2.5-related
cardiovascular morbidity, and more limited evidence from respiratory
morbidity, provide biological plausibility for mortality due to short-
term PM2.5 exposures. The primarily positive associations
observed across studies conducted in diverse geographic locations is
further supported by the results from copollutant analyses indicating
robust associations, along with evidence from analyses examining the C-
R relationship. Overall, studies evaluated in the 2019 ISA support the
conclusion of a causal relationship between short-term PM2.5
exposure and mortality, which is further supported by evidence from
recent epidemiologic studies evaluated in the ISA Supplement (U.S. EPA,
2022a, section 3.2.1.4, p. 3-69).
ii. Cardiovascular Effects
Long-Term PM2.5 Exposures
The scientific evidence reviewed in the 2009 ISA was ``sufficient
to infer a causal relationship between long-term PM2.5
exposure and cardiovascular effects'' (U.S. EPA, 2009a). The strongest
line of evidence comprised findings from several large epidemiologic
studies of U.S. and Canadian cohorts that reported consistent positive
associations between long-term PM2.5 exposure and
cardiovascular mortality (Pope et al., 2004; Krewski et al., 2009;
Miller et al.,
[[Page 16229]]
2007; Laden et al., 2006). Studies of long-term PM2.5
exposure and cardiovascular morbidity were limited in number.
Biological plausibility and coherence with the epidemiologic findings
were provided by studies using genetic mouse models of atherosclerosis
demonstrating enhanced atherosclerotic plaque development and
inflammation, as well as changes in measures of impaired heart
function, following 4- to 6-month exposures to PM2.5
concentrated ambient particles (CAPs), and by a limited number of
studies reporting CAPs-induced effects on coagulation factors, vascular
reactivity, and worsening of experimentally induced hypertension in
mice (U.S. EPA, 2009a).
Consistent with the evidence assessed in the 2009 ISA, the 2019 ISA
concludes that recent studies, together with the evidence available in
previous reviews, support a causal relationship between long-term
exposure to PM2.5 and cardiovascular effects. Additionally,
recent epidemiologic studies published since the completion of the 2019
ISA and evaluated in the ISA Supplement expands the body of evidence
and further supports such a conclusion (U.S. EPA, 2022a). As discussed
above (section II.A.2.a.i), results from U.S. and Canadian cohort
studies evaluated in the 2019 ISA conducted at varying spatial and
temporal scales and employing a variety of exposure assessment and
statistical methods consistently report positive associations between
long-term PM2.5 exposure and cardiovascular mortality (U.S.
EPA, 2019, Figure 6-19, section 6.2.10). Positive associations between
long-term PM2.5 exposures and cardiovascular mortality are
generally robust in copollutant models adjusted for ozone,
NO2, PM10-2.5, or SO2. In addition,
most of the results from analyses examining the shape of the C-R
relationship between long-term PM2.5 exposures and
cardiovascular mortality support a linear relationship and do not
identify a threshold below which mortality effects do not occur (U.S.
EPA, 2019a, section 6.2.16, Table 6-52).
The body of literature examining the relationship between long-term
PM2.5 exposure and cardiovascular morbidity has greatly
expanded since the 2009 ISA, with positive associations reported in
several cohorts evaluated in the 2019 ISA (U.S. EPA, 2019a, section
6.2). Though results for cardiovascular morbidity are less consistent
than those for cardiovascular mortality (U.S. EPA, 2019a, section 6.2),
studies in the 2019 ISA and the ISA Supplement provide some evidence
for associations between long-term PM2.5 exposures and the
progression of cardiovascular disease. Positive associations with
cardiovascular morbidity (e.g., coronary heart disease, stroke,
arrhythmias, myocardial infarction (MI), atherosclerosis progression)
are observed in several epidemiologic studies (U.S. EPA, 2019a,
sections 6.2.2 to 6.2.9; U.S. EPA, 2022a, section 3.1.2.2).
Additionally, studies evaluated in the ISA Supplement report positive
associations among those with pre-existing conditions, among patients
followed after a cardiac event procedure, and among those with a first
hospital admission for heart attacks among older adults enrolled in
Medicare (U.S. EPA, 2022a, sections 3.1.1 and 3.1.2).
Recent studies published since the literature cutoff date of the
2019 ISA and evaluated in the ISA Supplement further assessed the
relationship between long-term PM2.5 exposure and
cardiovascular effects by conducting accountability analyses or by
using alternative methods for confounder control in evaluating the
association between long-term PM2.5 exposure and
cardiovascular hospital admissions (U.S. EPA, 2022a, section 3.1.2.3).
Studies that apply alternative methods for confounder control increase
confidence in the relationship between long-term PM2.5
exposure and cardiovascular effects by using methods that reduce
uncertainties related to potential confounding through statistical and/
or study design approaches. For example, to control for potential
confounding Wei et al. (2021) used a doubly robust additive model
(DRAM) and found an association between long-term exposure to
PM2.5 and cardiovascular effects, including MI, stoke, and
atrial fibrillation, among the Medicare population. For example, an
accountability study by Henneman et al. (2019) utilized a difference-
in-difference (DID) approach to determine the relationship between
coal-fueled power plant emissions and cardiovascular effects and found
that reductions in PM2.5 concentrations resulted in
reductions of cardiovascular-related hospital admissions. Furthermore,
several recent epidemiologic studies evaluated in the ISA Supplement
reported that the association between long-term PM2.5
exposure with stroke persisted after adjustment for NO2 but
was attenuated in the model with O3 and oxidant gases
represented by the redox weighted average of NO2 and
O3 (U.S. EPA, 2022a, section 3.1.2.2.8). Overall, these
studies report consistent findings that long-term PM2.5
exposure is related to increased hospital admissions for a variety of
cardiovascular disease outcomes among large nationally representative
cohorts and provide additional support for a relationship between long-
term PM2.5 exposure and cardiovascular effects.
Positive associations reported in epidemiologic studies are
supported by toxicological evidence evaluated in the 2019 ISA. The
positive associations reported in epidemiologic studies are supported
by toxicological evidence for increased plaque progression in mice
following long-term exposure to PM2.5 collected from
multiple locations across the U.S. (U.S. EPA, 2019a, section 6.2.4.2).
A small number of epidemiologic studies also report positive
associations between long-term PM2.5 exposure and heart
failure, changes in blood pressure, and hypertension (U.S. EPA, 2019a,
sections 6.2.5 and 6.2.7). Associations with heart failure are
supported by animal toxicological studies demonstrating decreased
cardiac contractility and function, and increased coronary artery wall
thickness following long-term PM2.5 exposure (U.S. EPA,
2019a, section 6.2.5.2). Similarly, a limited number of animal
toxicological studies demonstrating a relationship between long-term
PM2.5 exposure and consistent increases in blood pressure in
rats and mice are coherent with epidemiologic studies reporting
positive associations between long-term exposure to PM2.5
and hypertension.
Additionally, a number of studies evaluated in the ISA Supplement
focusing on morbidity outcomes, including those that focused on
incidence of MI, atrial fibrillation (AF), stroke, and congestive heart
failure (CHF), expand the evidence pertaining to the shape of the C-R
relationship between long-term PM2.5 exposure and
cardiovascular effects. These studies use statistical techniques that
allow for departures from linearity (U.S. EPA, 2022a, Table 3-3), and
generally support the evidence characterized in the 2019 ISA showing
linear, no-threshold C-R relationship for most cardiovascular disease
(CVD) outcomes. However, there is evidence for a sublinear or
supralinear C-R relationship for some outcomes (U.S. EPA, 2022a,
section 3.1.2.2.9).\57\
---------------------------------------------------------------------------
\57\ As noted above for mortality, uncertainty in the shape of
the C-R relationship increases near the upper and lower ends of the
distribution due to limited data.
---------------------------------------------------------------------------
Longitudinal epidemiologic analyses also report positive
associations with markers of systemic inflammation (U.S. EPA, 2019a,
section 6.2.11), coagulation (U.S. EPA, 2019a, section 6.2.12), and
[[Page 16230]]
endothelial dysfunction (U.S. EPA, 2019a, section 6.2.13). These
results are coherent with animal toxicological studies generally
reporting increased markers of systemic inflammation, oxidative stress,
and endothelial dysfunction (U.S. EPA, 2019a, section 6.2.12.2 and
6.2.14).
In summary, the 2019 ISA concludes that there is consistent
evidence from multiple epidemiologic studies illustrating that long-
term exposure to PM2.5 is associated with mortality from
cardiovascular causes. Epidemiologic studies evaluated in the ISA
Supplement provide additional evidence of positive associations between
long-term PM2.5 exposure and cardiovascular morbidity (U.S.
EPA, 2022a, section 3.1.2.2). Associations with coronary heart disease
(CHD), stroke and atherosclerosis progression were observed in several
additional epidemiologic studies, providing coherence with the
mortality findings. Results from copollutant models generally support
an independent effect of PM2.5 exposure on mortality.
Additional evidence of the independent effect of PM2.5 on
the cardiovascular system is provided by experimental studies in
animals, which support the biological plausibility of pathways by which
long-term exposure to PM2.5 could potentially result in
outcomes such as CHD, stroke, CHF, and cardiovascular mortality.
Overall, studies evaluated in the 2019 ISA support the conclusion of a
causal relationship between long-term PM2.5 exposure and
cardiovascular effects, which is supported and extended by evidence
from recent epidemiologic studies evaluated in the ISA Supplement (U.S.
EPA, 2022a, section 3.1.2.2).
Short-Term PM2.5 Exposures
The 2009 ISA concluded that ``a causal relationship exists between
short-term exposure to PM2.5 and cardiovascular effects''
(U.S. EPA, 2009a). The strongest evidence in the 2009 ISA was from
epidemiologic studies of emergency department (ED) visits and hospital
admissions for IHD and heart failure (HF), with supporting evidence
from epidemiologic studies of cardiovascular mortality (U.S. EPA,
2009a). Animal toxicological studies provided coherence and biological
plausibility for the positive associations reported with MI, ED visits,
and hospital admissions. These included studies reporting reduced
myocardial blood flow during ischemia and studies indicating altered
vascular reactivity. In addition, effects of PM2.5 exposure
on a potential indicator of ischemia (i.e., ST segment depression on an
electrocardiogram) were reported in both animal toxicological and
epidemiologic panel studies.\58\ Key uncertainties from the last review
resulted from inconsistent results across disciplines with respect to
the relationship between short-term exposure to PM2.5 and
changes in blood pressure, blood coagulation markers, and markers of
systemic inflammation. In addition, while the 2009 ISA identified a
growing body of evidence from controlled human exposure and animal
toxicological studies, uncertainties remained with respect to
biological plausibility.
---------------------------------------------------------------------------
\58\ Some animal studies included in the 2009 ISA examined
exposures to mixtures, such as motor vehicle exhaust or woodsmoke.
In these studies, it was unclear if the resulting cardiovascular
effects could be attributed specifically to the fine particle
component of the mixture.
---------------------------------------------------------------------------
Studies evaluated in the 2019 ISA provide additional support for a
causal relationship between short-term PM2.5 exposure and
cardiovascular effects. This includes generally positive associations
observed in multicity epidemiologic studies of emergency department
visits and hospital admissions for IHD, heart failure (HF), and
combined cardiovascular-related endpoints. In particular, nationwide
studies of older adults (65 years and older) using Medicare records
report positive associations between PM2.5 exposures and
hospital admissions for HF (U.S. EPA, 2019a, section 6.1.3.1).
Moreover, recent multicity studies, published after the literature
cutoff date of the 2019 ISA and evaluated in the ISA Supplement, are
consistent with studies evaluated in the 2019 ISA that report positive
association between short-term PM2.5 exposure and ED visits
and hospital admission for IHD, heart attacks, and HF (U.S. EPA, 2022a,
section 3.1). Epidemiologic studies conducted in single cities
contribute some support to the causality determination, though
associations reported in single-city studies are less consistently
positive than in multicity studies, and include a number of studies
reporting null associations (U.S. EPA, 2019a, sections 6.1.2 and
6.1.3). As a whole, though, the recent body of IHD and HF epidemiologic
evidence supports the evidence from previous ISAs reporting mainly
positive associations between short-term PM2.5
concentrations and emergency department visits and hospital admissions.
Consistent with the evidence assessed in the 2019 ISA, some studies
evaluated in the ISA Supplement report no evidence of an association
with stroke, regardless of stroke subtype. Additionally, as in the 2019
ISA, evidence evaluated in the ISA Supplement continues to indicate an
immediate effect of PM2.5 on cardiovascular-related outcomes
primarily within the first few days after exposure, and that
associations generally persisted in models adjusted for copollutants
(U.S. EPA, 2022a, section 3.1.1.2).
The ISA Supplement includes additional epidemiologic studies,
published since the literature cutoff date for the 2019 ISA, including
accountability analyses and epidemiologic studies that employ
alternative methods for confounder control to evaluate the association
between short-term PM2.5 exposure and cardiovascular-related
effects (U.S. EPA, 2022a, section 3.1.1.3). These studies employ a
number of statistical approaches and report positive associations,
providing additional support for a relationship between short-term
PM2.5 exposure and cardiovascular effects, while also
reducing uncertainties related to potential confounder bias.
A number of controlled human exposure, animal toxicological, and
epidemiologic panel studies provide evidence that PM2.5
exposure could plausibly result in IHD or HF through pathways that
include endothelial dysfunction, arterial thrombosis, and arrhythmia
(U.S. EPA, 2019a, section 6.1.1). The most consistent evidence from
recent controlled human exposure studies is for endothelial
dysfunction, as measured by changes in brachial artery diameter or flow
mediated dilation. Multiple controlled human exposure studies that
examined the potential for endothelial dysfunction report an effect of
PM2.5 exposure on measures of blood flow (U.S. EPA, 2019a,
section 6.1.13.2). However, these studies report variable results
regarding the timing of the effect and the mechanism by which reduced
blood flow occurs (i.e., availability vs sensitivity to nitric oxide).
In addition, some controlled human exposure studies using CAPs report
evidence for small increases in blood pressure (U.S. EPA, 2019a,
section 6.1.6.3). Although not entirely consistent, there is also some
evidence across controlled human exposure studies for conduction
abnormalities/arrhythmia (U.S. EPA, 2019a, section 6.1.4.3), changes in
heart rate variability (HRV) (U.S. EPA, 2019a, section 6.1.10.2),
changes in hemostasis that could promote clot formation (U.S. EPA,
2019a, section 6.1.12.2), and increases in inflammatory cells and
markers (U.S. EPA, 2019a, section 6.1.11.2). A recent study by Wyatt et
al.
[[Page 16231]]
(2020), evaluated in the ISA Supplement, adds to the limited evidence
base of controlled human exposure studies conducted at near ambient
PM2.5 concentrations. The study, completed in healthy young
adults subject to intermittent exercise, found some significant
cardiovascular effects (e.g., systematic inflammation markers,
including C-reactive protein (CRP), and cardiac repolarization). Thus,
when taken as a whole, controlled human exposure studies are coherent
with epidemiologic studies in that they demonstrate that short-term
exposures to PM2.5 may result in the types of cardiovascular
endpoints that could lead to emergency department visits and hospital
admissions for IHD or HF, as well as mortality in some people.
Animal toxicological studies published since the 2009 ISA and
evaluated in the 2019 ISA also support a relationship between short-
term PM2.5 exposure and cardiovascular effects. A study
demonstrating decreased cardiac contractility and left ventricular
pressure in mice is coherent with the results of epidemiologic studies
that report associations between short-term PM2.5 exposure
and heart failure (U.S. EPA, 2019a, section 6.1.3.3). In addition, and
as with controlled human exposure studies, there is generally
consistent evidence in animal toxicological studies for indicators of
endothelial dysfunction (U.S. EPA, 2019a, section 6.1.13.3). Some
studies in animals also provide evidence for changes in a number of
other cardiovascular endpoints following short-term PM2.5
exposure including conduction abnormalities and arrhythmia (U.S. EPA,
2019a, section 6.1.4.4), changes in HRV (U.S. EPA, 2019a, section
6.1.10.3), changes in blood pressure (U.S. EPA, 2019a, section
6.1.6.4), and evidence for systemic inflammation and oxidative stress
(U.S. EPA, 2019a, section 6.1.11.3).
In summary, evidence evaluated in the 2019 ISA extends the
consistency and coherence of the evidence base evaluated in the 2009
ISA and prior assessments. Epidemiologic studies reporting robust
associations in copollutant models are supported by direct evidence
from controlled human exposure and animal toxicologic studies reporting
independent effects of PM2.5 exposures on endothelial
dysfunction as well as endpoints indicating impaired cardiac function,
increased risk of arrhythmia, changes in HRV, increases in BP, and
increases in indicators of systemic inflammation, oxidative stress, and
coagulation (U.S. EPA, 2019, section 6.1.16). For some cardiovascular
effects, there are inconsistencies in results across some animal
toxicological, controlled human exposure, and epidemiologic panel
studies, though this may be due to substantial differences in study
design and/or study populations. Overall, the results from
epidemiologic panel, controlled human exposure, and animal
toxicological studies, in particular those related to endothelial
dysfunction, impaired cardiac function, ST segment depression,
thrombosis, conduction abnormalities, and changes in blood pressure
provide coherence and biological plausibility for the consistent
results from epidemiologic studies observing positive associations
between short-term PM2.5 exposures and IHD and HF, and
ultimately cardiovascular mortality. Overall, studies evaluated in the
2019 ISA support the conclusion of a causal relationship between short-
term PM2.5 exposure and cardiovascular effects, which is
supported and extended by evidence from recent epidemiologic studies
evaluated in the ISA Supplement (U.S. EPA, 2022a, section 3.1.1.4).
iii. Respiratory Effects
Long-Term PM2.5 Exposures
The 2009 ISA concluded that ``a causal relationship is likely to
exist between long-term PM2.5 exposure and respiratory
effects'' (U.S. EPA, 2009a). This conclusion was based mainly on
epidemiologic evidence demonstrating associations between long-term
PM2.5 exposure and changes in lung function or lung function
growth in children. Biological plausibility was provided by a single
animal toxicological study examining pre- and post-natal exposure to
PM2.5 CAPs, which found impaired lung development.
Epidemiologic evidence for associations between long-term
PM2.5 exposure and other respiratory outcomes, such as the
development of asthma, allergic disease, and COPD; respiratory
infection; and the severity of disease was limited, both in the number
of studies available and the consistency of the results. Experimental
evidence for other outcomes was also limited, with one animal
toxicological study reporting that long-term exposure to
PM2.5 CAPs results in morphological changes in nasal airways
of healthy animals. Other animal studies examined exposure to mixtures,
such as motor vehicle exhaust and woodsmoke, and effects were not
attributed specifically to the particulate components of the mixture.
Cohort studies evaluated in the 2019 ISA provided additional
support for the relationship between long-term PM2.5
exposure and decrements in lung function growth (as a measure of lung
development), indicating a robust and consistent association across
study locations, exposure assessment methods, and time periods (U.S.
EPA, 2019a, section 5.2.13). This relationship was further supported by
a retrospective study that reports an association between declining
PM2.5 concentrations and improvements in lung function
growth in children (U.S. EPA, 2019a, section 5.2.11). Epidemiologic
studies also examine asthma development in children (U.S. EPA, 2019a,
section 5.2.3), with prospective cohort studies reporting generally
positive associations, though several are imprecise (i.e., they report
wide confidence intervals). Supporting evidence is provided by studies
reporting associations with asthma prevalence in children, with
childhood wheeze, and with exhaled nitric oxide, a marker of pulmonary
inflammation (U.S. EPA, 2019a, section 5.2.13). Additionally, the 2019
ISA includes an animal toxicological study showing the development of
an allergic phenotype and an increase in a marker of airway
responsiveness supports the biological plausibility of the development
of allergic asthma (U.S. EPA, 2019a, section 5.2.13). Other
epidemiologic studies report a PM2.5-related acceleration of
lung function decline in adults, while improvement in lung function was
observed with declining PM2.5 concentrations (U.S. EPA,
2019a, section 5.2.11). A longitudinal study found declining
PM2.5 concentrations are also associated with an improvement
in chronic bronchitis symptoms in children, strengthening evidence
reported in the 2009 ISA for a relationship between increased chronic
bronchitis symptoms and long-term PM2.5 exposure (U.S. EPA,
2019a, section 5.2.11). A common uncertainty across the epidemiologic
evidence is the lack of examination of copollutants to assess the
potential for confounding. While there is some evidence that
associations remain robust in models with gaseous pollutants, a number
of these studies examining copollutant confounding were conducted in
Asia, and thus have limited generalizability due to high annual
pollutant concentrations.
When taken together, the 2019 ISA concludes that the epidemiologic
evidence strongly supports a relationship with decrements in lung
function growth asthma development in children, as well as increased
bronchitis symptoms in children with asthma. Additionally, the
epidemiologic
[[Page 16232]]
evidence strongly supports a relationship with an acceleration of lung
function decline in adults, and with respiratory mortality and cause-
specific respiratory mortality for COPD and respiratory infection (U.S.
EPA, 2019a, p. 1-34). In support of the biological plausibility of
associations reported in epidemiologic studies associated with
respiratory health effects, animal toxicological studies evaluated in
the 2019 ISA continue to provide direct evidence that long-term
exposure to PM2.5 results in a variety of respiratory
effects, including pulmonary oxidative stress, inflammation, and
morphologic changes in the upper (nasal) and lower airways. Other
results show that changes are consistent with the development of
allergy and asthma, and with impaired lung development. Overall, the
2019 ISA concludes that ``the collective evidence is sufficient to
conclude that a causal relationship is likely to exist between long-
term PM2.5 exposure and respiratory effects'' (U.S. EPA,
2019a, section 5.2.13).
Short-Term PM2.5 Exposures
The 2009 ISA (U.S. EPA, 2009a) concluded that a ``causal
relationship is likely to exist'' between short-term PM2.5
exposure and respiratory effects. This conclusion was based mainly on
the epidemiologic evidence demonstrating positive associations with
various respiratory effects. Specifically, the 2009 ISA described
epidemiologic evidence as consistently showing PM2.5-
associated increases in hospital admissions and ED visits for COPD and
respiratory infection among adults or people of all ages, as well as
increases in respiratory mortality. These results were supported by
studies reporting associations with increased respiratory symptoms and
decreases in lung function in children with asthma, though the
epidemiologic evidence was inconsistent for hospital admissions or
emergency department visits for asthma. Studies examining copollutant
models showed that PM2.5 associations with respiratory
effects were robust to inclusion of CO or SO2 in the model,
but often were attenuated (though still positive) with inclusion of
O3 or NO2. In addition to the copollutant models,
evidence supporting an independent effect of PM2.5 exposure
on the respiratory system was provided by animal toxicological studies
of PM2.5 CAPs demonstrating changes in some pulmonary
function parameters, as well as inflammation, oxidative stress, injury,
enhanced allergic responses, and reduced host defenses. Many of these
effects have been implicated in the pathophysiology for asthma
exacerbation, COPD exacerbation, or respiratory infection. In the few
controlled human exposure studies conducted in individuals with asthma
or COPD, PM2.5 exposure mostly had no effect on respiratory
symptoms, lung function, or pulmonary inflammation. Available studies
in healthy people also did not clearly demonstrate respiratory effects
following short-term PM2.5 exposures.
Epidemiologic studies evaluated in the 2019 ISA continue to provide
strong evidence for a relationship between short-term PM2.5
exposure and several respiratory-related endpoints, including asthma
exacerbation (U.S. EPA, 2019a, section 5.1.2.1), COPD exacerbation
(U.S. EPA, 2019a, section 5.1.4.1), and combined respiratory-related
diseases (U.S. EPA, 2019a, section 5.1.6), particularly from studies
examining ED visits and hospital admissions. The generally positive
associations between short-term PM2.5 exposure and asthma
and COPD as well as ED visits and hospital admissions are supported by
epidemiologic studies demonstrating associations with other
respiratory-related effects such as symptoms and medication use that
are indicative of asthma and COPD exacerbations (U.S. EPA, 2019a,
sections 5.1.2.2 and 5.4.1.2). The collective body of epidemiologic
evidence for asthma exacerbation is more consistent in children than in
adults. Additionally, epidemiologic studies examining the relationship
between short-term PM2.5 exposure and respiratory mortality
provide evidence of consistent positive associations, demonstrating a
continuum of effects (U.S. EPA, 2019a, section 5.1.9).
Epidemiologic studies evaluated in the 2019 ISA expand the
assessment of potential copollutant confounding evaluated in the 2009
ISA. There is some evidence that PM2.5 associations with
asthma exacerbation, combined respiratory-related diseases, and
respiratory mortality remain relatively unchanged in copollutant models
with gaseous pollutants including O3, NO2,
SO2, and with more limited evidence for CO, as well as other
particle sizes (i.e., PM10-2.5) (U.S. EPA, 2019a, section
5.1.10.1).
Insight into whether there is an independent effect of
PM2.5 on respiratory health is also partially addressed by
findings from animal toxicological studies evaluated in the 2019 ISA.
Specifically, short-term exposure to PM2.5 enhanced asthma-
related responses in an animal model of allergic airways disease and
enhanced lung injury and inflammation in an animal model of COPD (U.S.
EPA, 2019a, sections 5.1.2.4.4 and 5.1.4.4.3). The experimental
evidence provides biological plausibility for some respiratory-related
endpoints, including limited evidence of altered host defense and
greater susceptibility to bacterial infection as well as consistent
evidence of respiratory irritant effects. However, animal toxicological
evidence for other respiratory effects is inconsistent. A recent study
evaluated in the ISA supplement by Wyatt et al. (2020) and conducted at
near ambient PM2.5 concentrations, adds to the limited
evidence base of controlled human exposure studies. The study,
completed in healthy young adults subject to intermittent exercise,
found some significant respiratory effects (including decrease in lung
function), however these findings were inconsistent with the controlled
human exposure studies evaluated in the 2019 ISA (U.S. EPA, 2019a,
section 5.1.7.2, 5.1.2.3, and 6.1.11.2.1).
The 2019 ISA concludes that ``[t]he strongest evidence of an effect
of short-term PM2.5 exposure on respiratory effects is
provided by epidemiologic studies of asthma and COPD exacerbation.
While animal toxicological studies provide biological plausibility for
these findings, some uncertainty remains with respect to the
independence of PM2.5 effects'' (U.S. EPA, 2019a, p. 5-155).
When taken together, the 2019 ISA concludes that this evidence ``is
sufficient to conclude that a causal relationship is likely to exist
between short-term PM2.5 exposure and respiratory effects''
(U.S. EPA, 2019a, p. 5-155).
iv. Cancer
The 2009 ISA concluded that the overall body of evidence was
``suggestive of a causal relationship between relevant PM2.5
exposures and cancer'' (U.S. EPA, 2009a). This conclusion was based
primarily on positive associations observed in a limited number of
epidemiologic studies of lung cancer mortality. The few epidemiologic
studies that had evaluated PM2.5 exposure and lung cancer
incidence or cancers of other organs and systems generally did not show
evidence of an association. Toxicological studies did not focus on
exposures to specific PM size fractions, but rather investigated the
effects of exposures to total ambient PM, or other source-based PM such
as wood smoke. Collectively, results of in vitro studies were
consistent with the larger body of evidence demonstrating that ambient
PM and PM from specific combustion sources are mutagenic and genotoxic.
However, animal inhalation studies
[[Page 16233]]
found little evidence of tumor formation in response to chronic
exposures. A small number of studies provided preliminary evidence that
PM exposure can lead to changes in methylation of DNA, which may
contribute to biological events related to cancer.
Since the completion of the 2009 ISA, additional cohort studies
provide evidence that long-term PM2.5 exposure is positively
associated with lung cancer mortality and with lung cancer incidence,
and provide initial evidence for an association with reduced cancer
survival (U.S. EPA, 2019a, section 10.2.5). Re-analyses of the ACS
cohort using different years of PM2.5 data and follow up,
along with various exposure assignment approaches, provide consistent
evidence of positive associations between long-term PM2.5
exposure and lung cancer mortality (U.S. EPA, 2019a, Figure 10-3).
Additional support for positive associations with lung cancer mortality
is provided by recent epidemiologic studies using individual level data
to control for smoking status, as well as by studies of people who have
never smoked (though such studies generally report wide confidence
intervals due to the small number of lung cancer mortality cases within
this population), and in additional analyses of cohorts that relied
upon proxy measures to account for smoking status (U.S. EPA, 2019a,
section 10.2.5.1.1). Although studies that evaluate lung cancer
incidence, including studies of people who have never smoked, are
limited in number, studies in the 2019 ISA generally report positive
associations with long-term PM2.5 exposures (U.S. EPA,
2019a, section 10.2.5.1.2). A subset of the studies focusing on lung
cancer incidence also examined histological subtype, providing some
evidence of positive associations for adenocarcinomas, the predominate
subtype of lung cancer observed in people who have never smoked (U.S.
EPA, 2019a, section 10.2.5.1.2). Associations between long-term
PM2.5 exposure and lung cancer incidence were found to
remain relatively unchanged, though in some cases confidence intervals
widened, in analyses that attempted to reduce exposure measurement
error by accounting for length of time at residential address or by
examining different exposure assignment approaches (U.S. EPA, 2019a,
section 10.2.5.1.2).
To date, relatively few studies have evaluated the potential for
copollutant confounding of the relationship between long-term
PM2.5 exposure and lung cancer mortality or incidence. A
small number of such studies have generally focused on O3
and report that PM2.5 associations remain relatively
unchanged in copollutant models (U.S. EPA, 2019a, section 10.2.5.1.3).
However, available studies have not systematically evaluated the
potential for copollutant confounding by other gaseous pollutants or by
other particle size fractions (U.S. EPA, 2019a, section 10.2.5.1.3).
Compared to total (non-accidental) mortality (U.S. EPA, 2019a,
section 10.2.4.1.4), fewer studies have examined the shape of the C-R
curve for cause-specific mortality outcomes, including lung cancer.
Several studies of lung cancer mortality and incidence have reported no
evidence of deviations from linearity in the shape of the C-R
relationship (Lepeule et al., 2012; Raaschou-Nielsen et al., 2013;
Puett et al., 2014), though authors provided only limited discussions
of results (U.S. EPA, 2019a, section 10.2.5.1.4).
In support of the biological plausibility of an independent effect
of PM2.5 on lung cancer, the 2019 ISA notes evidence from
experimental and epidemiologic studies demonstrating that
PM2.5 exposure can lead to a range of effects indicative of
mutagenicity, genotoxicity, and carcinogenicity, as well as epigenetic
effects (U.S. EPA, 2019a, section 10.2.7). For example, both in vitro
and in vivo toxicological studies have shown that PM2.5
exposure can result in DNA damage (U.S. EPA, 2019a, section 10.2.2).
Although such effects do not necessarily equate to carcinogenicity, the
evidence that PM exposure can damage DNA, and elicit mutations,
provides support for the plausibility of epidemiologic associations
exhibited with lung cancer mortality and incidence. Additional
supporting studies indicate the occurrence of micronuclei formation and
chromosomal abnormalities (U.S. EPA, 2019a, section 10.2.2.3), and
differential expression of genes that may be relevant to cancer
pathogenesis, following PM2.5 exposures. Experimental and
epidemiologic studies that examine epigenetic effects indicate changes
in DNA methylation, providing some support that PM2.5
exposure contributes to genomic instability (U.S. EPA, 2019a, section
10.2.3). Overall, there is limited evidence that long-term
PM2.5 exposure is associated with cancers in other organ
systems, though there is some evidence that PM2.5 exposure
may reduce survival in individuals with cancer (U.S. EPA, 2019a,
section 10.2.7; U.S. EPA, 2022a, section 2.1.1.4.1).
Epidemiologic evidence for associations between PM2.5
and lung cancer mortality and incidence, together with evidence
supporting the biological plausibility of such associations,
contributes to the 2019 ISA's conclusion that the evidence ``is
sufficient to conclude that a causal relationship is likely to exist
between long-term PM2.5 exposure and cancer'' (U.S. EPA,
2019, section 10.2.7).
v. Nervous System Effects
Reflecting the very limited evidence available in the 2012 review,
the 2009 ISA did not make a causality determination for long-term
PM2.5 exposures and nervous system effects (U.S. EPA,
2009c). Since the 2012 review, this body of evidence has grown
substantially (U.S. EPA, 2019, section 8.2). Animal toxicological
studies assessed in in the 2019 ISA report that long-term
PM2.5 exposures can lead to morphologic changes in the
hippocampus and to impaired learning and memory. This evidence is
consistent with epidemiologic studies reporting that long-term
PM2.5 exposure is associated with reduced cognitive function
(U.S. EPA, 2019a, section 8.2.5). Further, while the evidence is
limited, the presence of early markers of Alzheimer's disease pathology
has been demonstrated in rodents following long-term exposure to
PM2.5 CAPs. These findings support reported associations
with neurodegenerative changes in the brain (i.e., decreased brain
volume), all-cause dementia, or hospitalization for Alzheimer's disease
in a small number of epidemiologic studies (U.S. EPA, 2019a, section
8.2.6). Additionally, loss of dopaminergic neurons in the substantia
nigra, a hallmark of Parkinson disease, has been reported in mice (U.S.
EPA, 2019a, section 8.2.4), though epidemiologic studies provide only
limited support for associations with Parkinson's disease (U.S. EPA,
2019a, section 8.2.6). Overall, the lack of consideration of
copollutant confounding introduces some uncertainty in the
interpretation of epidemiologic studies of nervous system effects, but
this uncertainty is partly addressed by the evidence for an independent
effect of PM2.5 exposures provided by experimental animal
studies.
While the findings described above are most relevant to older
adults, several studies of neurodevelopmental effects in children have
also been conducted. Epidemiologic studies provided limited evidence of
an association between PM2.5 exposure during pregnancy and
childhood on cognitive and motor development (U.S. EPA, 2019, section
8.2.5.2). While some studies report positive associations between long-
term
[[Page 16234]]
exposure to PM2.5 during the prenatal period and autism
spectrum disorder (ASD) (U.S. EPA, 2019, section 8.2.7.2), the
interpretation of these epidemiologic studies is limited due to the
small number of studies, their lack of control for potential
confounding by copollutants, and uncertainty related to the critical
exposure windows. Biological plausibility is provided for the ASD
findings by a study in mice that found inflammatory and morphologic
changes in the corpus collosum and hippocampus, as well as
ventriculomegaly (i.e., enlarged lateral ventricles) in young mice
following prenatal exposure to PM2.5 CAPs.
Taken together, the 2019 ISA concludes that studies indicate long-
term PM2.5 exposures can lead to effects on the brain
associated with neurodegeneration (i.e., neuroinflammation and
reductions in brain volume), as well as cognitive effects in older
adults (U.S. EPA, 2019a, Table 1-2). Animal toxicological studies
provide evidence for a range of nervous system effects in adult
animals, including neuroinflammation and oxidative stress,
neurodegeneration, cognitive effects, and effects on neurodevelopment
in young animals. The epidemiologic evidence is more limited, but
studies generally support associations between long-term
PM2.5 exposure and changes in brain morphology, cognitive
decrements and dementia. There is also initial, and limited, evidence
for neurodevelopmental effects, particularly ASD. The consistency and
coherence of the evidence supports the 2019 ISA's conclusion that ``the
collective evidence is sufficient to conclude that a causal
relationship is likely to exist between long-term PM2.5
exposure and nervous system effects'' (U.S. EPA, 2019a, section 8.2.9).
vi. Other Effects
For other health effect categories that were evaluated for their
relationship with PM2.5 exposures (i.e., short-term
PM2.5 exposure and nervous system effects and short- and
long-term PM2.5 exposure and metabolic effects, reproduction
and fertility, and pregnancy and birth outcomes (U.S. EPA, 2022a, Table
ES-1), the currently available evidence is ``suggestive of, but not
sufficient to infer, a causal relationship,'' mainly due to
inconsistent evidence across specific outcomes and uncertainties
regarding exposure measurement error, the potential for confounding,
and potential modes of action (U.S. EPA, 2019a, sections 7.14, 7.2.10,
8.1.6, and 9.1.5). The causality determination for short-term
PM2.5 exposure and nervous system effects in the 2019 ISA
reflects a revision to the causality determination in the 2009 ISA from
``inadequate to infer a causal relationship,'' while this is the first-
time assessments of causality were conducted for long-term
PM2.5 exposure and nervous system effects, as well as short-
and long-term PM2.5 exposure and metabolic effects reflect.
Recent studies evaluated in the 2019 ISA also further explored the
relationship between short-and long-term UFP exposure and health
effects. (i.e., cardiovascular effects and short-term UFP exposures;
respiratory effects and short-term UFP exposures; and nervous system
effects and long- and short-term exposures (U.S. EPA, 2022a, Table ES-
1). The currently available evidence is ``suggestive of, but not
sufficient to infer, a causal relationship'' for short-term UFP
exposure and cardiovascular and respiratory effects and for short- and
long-term UFP exposure and nervous system effects, primarily due to
uncertainties and limitations in the evidence, specifically,
variability across studies in the definition of UFPs and the exposure
metric used (U.S. EPA, 2019a, P.3.1; U.S. EPA, 2022a, section
3.3.1.6.3). The causality determinations for the other health effect
categories evaluated in the 2019 ISA are ``inadequate to infer a causal
relationship.'' Additionally, this is the first time assessments of
causality were conducted for short- and long-term UFP exposure and
metabolic effects and long-term UFP exposure and nervous system effects
(U.S. EPA, 2022a, Table ES-1).
With the advent of the global COVID-19 pandemic, a number of recent
studies evaluated in the ISA Supplement examined the relationship
between ambient air pollution, specifically PM2.5, and SARS-
CoV-2 infections and COVID-19 deaths, including a few studies within
the U.S. and Canada (U.S. EPA, 2022a, section 3.3.2).\59\ Some studies
examined whether daily changes in PM2.5 can influence SARS-
CoV-2 infection and COVID-19 death (U.S. EPA, 2022a, section 3.3.2.1).
Additionally, several studies evaluated whether long-term
PM2.5 exposure increases the risk of SARS-CoV-2 infection
and COVID-19 death in North America (U.S. EPA, 2022a, section 3.3.2.2).
While there is initial evidence of positive associations with SARS-CoV-
2 infection and COVID-19 death, uncertainties remain due to
methodological issues that may influence the results, including: (1)
The use of ecological study design; (2) studies were conducted during
the ongoing pandemic when the etiology of COVID-19 was still not well
understood (e.g., specifically, there are important differences in
COVID-19-related outcomes by a variety of factors such as race and
SES); and (3) studies did not account for crucial factors that could
influence results (e.g., stay-at-home orders, social distancing, use of
masks, and testing capacity) (U.S. EPA, 2022a, chapter 5). Taken
together, while there is initial evidence of positive associations with
SARS-CoV-2 infection and COVID-19 death, uncertainties remain due to
methodological issues.
---------------------------------------------------------------------------
\59\ While there is no exact corollary within the 2019 ISA for
these types of studies, the 2019 ISA presented evidence that
evaluates the potential relationship between short- and long-term
PM2.5 exposure and respiratory infection (U.S. EPA,
2022a, section 5.1.5 and 5.2.6). Studies assessed in the 2019 ISA
report some evidence of positive associations between short-term
PM2.5 and hospital admissions and ED visits for
respiratory infections, however the interpretation of these studies
is complicated by the variability in the type of respiratory
infection outcome examined (U.S. EPA, 2022a, Figure 5-7). In the
2019 ISA, studies of long-term PM2.5 exposure were
limited and while there were some positive associations reported,
there was minimal overlap in respiratory infection outcomes examined
across studies. Exposure to PM2.5 has been shown to
impair host defense, specifically altering macrophage function,
providing a biological pathway by which PM2.5 exposure
could lead to respiratory infection (U.S. EPA, 2022a, sections 5.1.1
and 5.1.5.) There is some additional evidence that PM2.5
exposure can lead to decreases in an individual's immune response,
which can subsequently facilitate replication of respiratory viruses
(Bourdrel et al., 2021).
---------------------------------------------------------------------------
b. Public Health Implications and At-Risk Populations
The public health implications of the evidence regarding
PM2.5-related health effects, as for other effects, are
dependent on the type and severity of the effects, as well as the size
of the population affected. Such factors are discussed below in the
context of our consideration of the health effects evidence related to
PM2.5 in ambient air. This section also summarizes the
current information on population groups at increased risk of the
effects of PM2.5 in ambient air.
The information available in this reconsideration has not altered
our understanding of human populations at risk of health effects from
PM2.5 exposures. As recognized in the 2020 review, the 2019
ISA cites extensive evidence indicating that ``both the general
population as well as specific populations and lifestages are at risk
for PM2.5-related health effects'' (U.S. EPA, 2019a, p. 12-
1). Factors that may contribute to increased risk of PM2.5-
related health effects include lifestage (children and older adults),
pre-existing diseases (cardiovascular disease and
[[Page 16235]]
respiratory disease), race/ethnicity, and SES.\60\
---------------------------------------------------------------------------
\60\ As described in the 2019 ISA, other factors that have the
potential to contribute to increased risk include obesity, diabetes,
genetic factors, smoking status, sex, diet, and residential location
(U.S. EPA, 2019, chapter 12).
---------------------------------------------------------------------------
Children make up a substantial fraction of the U.S. population, and
often have unique factors that contribute to their increased risk of
experiencing a health effect due to exposures to ambient air pollutants
because of their continuous growth and development.\61\ Children may be
particularly at risk for health effects related to ambient
PM2.5 exposures compared with adults because they have (1) a
developing respiratory system, (2) increased ventilation rates relative
to body mass compared with adults, and (3) an increased proportion of
oral breathing, particularly in boys, relative to adults (U.S. EPA,
2019a, section 12.5.1.1). There is strong evidence that demonstrates
PM2.5 associated health effects in children, particularly
from epidemiologic studies of long-term PM2.5 exposure and
impaired lung function growth, decrements in lung function, and asthma
development. However, there is limited evidence from stratified
analyses that children are at increased risk of PM2.5-
related health effects compared to adults. Additionally, there is some
evidence that indicates that children receive higher PM2.5
exposures than adults, and dosimetric differences in children compared
to adults can contribute to higher doses (U.S. EPA, 2019a, section
12.5.1.1).
---------------------------------------------------------------------------
\61\ Children, as used throughout this document, generally
refers to those younger than 18 years old.
---------------------------------------------------------------------------
In the U.S., older adults, often defined as adults 65 years of age
and older, represent an increasing portion of the population and often
have pre-existing diseases or conditions that may compromise biological
function. While there is limited evidence to indicate that older adults
have higher exposures than younger adults, older adults may receive
higher doses of PM2.5 due to dosimetric differences. There
is consistent evidence from studies of older adults demonstrating
generally consistent positive associations in studies examining health
effects from short- and long-term PM2.5 exposure and
cardiovascular or respiratory hospital admissions, emergency department
visits, or mortality (U.S. EPA, 2019a, sections 6.1, 6.2, 11.1, 11.2,
12.5.1.2). Additionally, several animal toxicological, controlled human
exposure, and epidemiologic studies did not stratify results by
lifestage, but instead focused the analyses on older individuals, and
can provide coherence and biological plausibility for the occurrence
among this lifestage (U.S. EPA, 2019a, section 12.5.1.2).
Individuals with pre-existing disease may be considered at greater
risk of an air pollution-related health effect than those without
disease because they are likely in a compromised biological state that
can vary depending on the disease and severity. With regard to
cardiovascular disease, we first note that cardiovascular disease is
the leading cause of death in the U.S., accounting for one in four
deaths, and approximately 12% of the adult population in the U.S. has a
cardiovascular disease (U.S. EPA, 2019a, section 12.3.1). Strong
evidence demonstrates that there is a causal relationship between
cardiovascular effects and long- and short-term exposures to
PM2.5. Some of the evidence supporting this conclusion is
from studies of panels or cohorts with pre-existing cardiovascular
disease, which provide supporting evidence but do not directly
demonstrate an increased risk (U.S. EPA, 2019a, section 12.3.1).
Epidemiologic evidence indicates that individuals with pre-existing
cardiovascular disease may be at increased risk for PM2.5-
associated health effects compared to those without pre-existing
cardiovascular disease. While the evidence does not consistently
support increased risk for all pre-existing cardiovascular diseases,
there is evidence that certain pre-existing cardiovascular diseases
(e.g., hypertension) may be a factor that increases PM2.5-
related risk. Furthermore, there is strong evidence supporting a causal
relationship for long- and short-term PM2.5 exposure and
cardiovascular effects, particularly for IHD (U.S. EPA, 2019a, chapter
6, section 12.3.1).
With regard to respiratory disease, we first note that the most
chronic respiratory diseases in the U.S. are asthma and COPD. Asthma
affects a substantial fraction of the U.S. population and is the
leading chronic disease among children. COPD primarily affects older
adults and contributes to compromised respiratory function and
underlying pulmonary inflammation. The body of evidence indicates that
individuals with pre-existing respiratory diseases, particularly asthma
and COPD, may be at increased risk for PM2.5-related health
effects compared to those without pre-existing respiratory diseases
(U.S. EPA, 2019a, section 12.3.5). There is strong evidence indicating
PM2.5-associated respiratory effects among those with
asthma, which forms the primary evidence base for the likely to be
causal relationship between short-term exposures to PM2.5
and respiratory health effects (U.S. EPA, 2019a, section 12.3.5). For
asthma, epidemiologic evidence demonstrates associations between short-
term PM2.5 exposures and respiratory effects, particularly
evidence for asthma exacerbation, and controlled human exposure and
animal toxicological studies demonstrate support for the biological
plausibility for asthma exacerbation with PM2.5 exposures
(U.S. EPA, 2019a, section 12.3.5.1). For COPD, epidemiologic studies
report positive associations between short-term PM2.5
exposures and hospital admissions and emergency department visits for
COPD, with supporting evidence from panel studies demonstration COPD
exacerbation. Epidemiologic evidence is supported by some experimental
evidence of COPD-related effects, which provides support for the
biological plausibility for COPD in response to PM2.5
exposures (U.S. EPA, 2019a, section 12.3.5.2).
There is strong evidence for racial and ethnic disparities in
PM2.5 exposures and PM2.5-related health risk, as
assessed in the 2019 ISA and with even more evidence available since
the literature cutoff date for the 2019 ISA and evaluated in the ISA
Supplement. There is strong evidence demonstrating that Black and
Hispanic populations, in particular, have higher PM2.5
exposures than non-Hispanic White populations (U.S. EPA, 2019a, Figure
12-2; U.S. EPA, 2022a, Figure 3-38). Black populations or individuals
that live in predominantly Black neighborhoods experience higher
PM2.5 exposures, in comparison to non-Hispanic White
populations. There is also consistent evidence across multiple studies
that demonstrate increased risk of PM2.5-related health
effects, with the strongest evidence for health risk disparities for
mortality (U.S. EPA, 2019a, section 12.5.4). There is also evidence of
health risk disparities for both Hispanic and non-Hispanic Black
populations compared to non-Hispanic White populations for cause-
specific mortality and incident hypertension (U.S. EPA, 2022a, section
3.3.3.2).
Socioeconomic status (SES) is a composite measure that includes
metrics such as income, occupation, or education, and can play a role
in access to healthy environments as well as access to healthcare. SES
may be a factor that contributes to differential risk from
PM2.5-related health effects. Studies assessed in the 2019
ISA and ISA Supplement provide evidence that lower SES communities are
exposed to higher concentrations of PM2.5
[[Page 16236]]
compared to higher SES communities (U.S. EPA, 2019a, section 12.5.3;
U.S. EPA, 2022a, section 3.3.3.1.1). Studies using composite measures
of neighborhood SES consistently demonstrated a disparity in both
PM2.5 exposure and the risk of PM2.5-related
health outcomes. There is some evidence that supports associations
larger in magnitude between mortality and long-term PM2.5
exposures for those with low income or living in lower income areas
compared to those with higher income or living in higher income
neighborhoods (U.S. EPA, 2019a, section 12.5.3; U.S. EPA, 2022a,
section 3.3.3.1.1). Additionally, evidence supports conclusions that
lower SES is associated with cause-specific mortality and certain
health endpoints (i.e., HI and CHF), but less so for all-cause or total
(non-accidental) mortality (U.S. EPA, 2022a, section 3.3.3.1).
The magnitude and characterization of a public health impact is
dependent upon the size and characteristics of the populations
affected, as well as the type or severity of the effects. As summarized
above, lifestage (children and older adults), race/ethnicity and SES
are factors that increase the risk of PM2.5-related health
effects. The American Community Survey (ACS) for 2019 estimates that
approximately 22% and 16% of the U.S. population are children (age<18)
and older adults (age 65+), respectively. For all ages, non-Hispanic
Black and Hispanic populations comprise approximately 12% and 18% of
the overall U.S. population in 2019. Currently available information
that helps to characterize key features of these population is included
in the 2022 PA (U.S. EPA, 2022b, Table 3-2).
As noted above, individuals with pre-existing cardiovascular
disease and pre-existing respiratory disease may also be at increased
risk of PM2.5-related health effects. Currently available
information that helps to characterize key features of populations with
cardiovascular or respiratory diseases or conditions is included in the
2022 PA (U.S. EPA, 2022b, Table 3-3). The National Center for Health
Statistics data for 2018 indicate that, for adult populations, older
adults (e.g., those 65 years and older) have a higher prevalence of
cardiovascular diseases compared to younger adults (e.g., those 64
years and younger). For respiratory diseases, older adults also have a
higher prevalence of emphysema than younger adults, and adults 44 years
or older have a higher prevalence of chronic bronchitis. However, the
prevalence for asthma is generally similar across all adult age groups.
With respect to race, American Indians or Alaskan Native
populations have the highest prevalence of all heart disease and
coronary heart disease, while Black populations have the highest
prevalence of hypertension and stroke. Hypertension has the highest
prevalence across all racial groups compared to other cardiovascular
diseases or conditions, ranging from approximately 22% to 32% of each
racial group. Overall, the prevalence of cardiovascular diseases or
conditions is lowest for Asians compared to Whites, Blacks, and
American Indians or Alaskan Natives. Asthma prevalence is highest among
Black and American Indian or Alaska Native populations, while the
prevalence of chronic bronchitis and emphysema is generally similar
across racial groups. Overall, the prevalence of respiratory diseases
is lowest for Asians compared to Whites, Blacks, and American Indians
or Alaskan Natives. With regard to ethnicity, cardiovascular and
respiratory disease prevalence across all diseases or conditions is
generally similar between Hispanic and non-Hispanic populations,
although non-Hispanics have a slightly higher prevalence compared to
Hispanics.
Taken together, this information indicates that the groups at
increased risk of PM2.5-related health effects represent a
substantial portion of the total U.S. population. In evaluating the
primary PM2.5 standards, an important consideration is the
potential PM2.5-related public health impacts in these
populations.
c. PM2.5 Concentrations in Key Studies Reporting Health
Effects
To inform conclusions on the adequacy of the public health
protection provided by the current primary PM2.5 standards,
the sections below summarize the 2022 PA's evaluation of the
PM2.5 exposures, specifically the concentrations that have
been examined in controlled human exposure studies, animal
toxicological studies, and epidemiologic studies. The 2022 PA places
the greatest emphasis on the health outcomes for which the 2019 ISA
concludes that the evidence supports a ``causal'' or a ``likely to be
causal'' relationship with short- or long-term PM2.5
exposures (U.S. EPA, 2022b, section 3.3.3). As described in greater
detail in section II.A.2 above, this includes short- or long-term
PM2.5 exposures and mortality, cardiovascular effects, and
respiratory effects and long-term PM2.5 exposures and cancer
and nervous system effects. While the causality determinations in the
2019 ISA are informed by studies evaluating a wide range of
PM2.5 concentrations,\62\ the sections below summarize the
considerations in the 2022 PA regarding the degree to which the
evidence assessed in the 2019 ISA and ISA Supplement supports the
occurrence of PM-related health effects at concentrations relevant to
informing conclusions on the primary PM2.5 standards. In so
doing, the 2022 PA focuses on the available studies that are most
directly informative to reaching conclusions regarding the adequacy of
the current primary PM2.5 standards (e.g., epidemiologic
studies with annual mean PM2.5 concentrations near or below
the level of the standard; and controlled human exposure studies at
PM2.5 exposures that elicit consistent effects, as well as
examining PM2.5 exposures at concentrations that are at or
near the level of the standard).
---------------------------------------------------------------------------
\62\ As described in more detail in section 5 of the Preamble to
the ISAs, judgments regarding causality take into consideration a
number of aspects when evaluating the available scientific evidence
(U.S. EPA, 2015, Table I). In reaching conclusions regarding
causality, ``evidence is evaluated for major outcome categories or
groups of related endpoints (e.g., respiratory effects, vegetation
growth), integrating evidence from across disciplines, and
evaluating the coherence of evidence across a spectrum of related
endpoints'' (U.S. EPA, 2015, p. 24). Furthermore, ``[i]n drawing
judgments regarding causality for the criteria air pollutants, the
ISA focuses on evidence of effects in the range of relevant
pollutant exposures or doses and not on determination of causality
at any dose. Emphasis is placed on evidence of effects at doses
(e.g., blood Pb concentration) or exposures (e.g., air
concentrations) that are relevant to, or somewhat above, those
currently experienced by the population. The extent to which studies
of higher concentrations are considered varies by pollutant and
major outcome category, but generally includes those with doses or
exposures in the range of one to two orders of magnitude above
current or ambient conditions to account for intra-species
variability and toxicokinetic or toxicodynamic differences between
experimental animals and humans. Studies that use higher doses or
exposures may also be considered to the extent that they provide
useful information to inform understanding of mode of action, inter-
species differences, or factors that may increase risk of effects
for a population and if biological mechanisms have not been
demonstrated to differ based on exposure concentration. Thus, a
causality determination is based on weight-of-evidence evaluation
for health or welfare effects, focusing on the evidence from
exposures or doses generally ranging from recent ambient
concentrations to one or two orders of magnitude above recent
ambient concentrations'' (U.S. EPA, 2015, p. 24).
---------------------------------------------------------------------------
i. PM2.5 Exposure Concentrations Evaluated in Experimental
Studies
Evidence for a particular PM2.5-related health outcome
is strengthened when results from experimental studies demonstrate
biologically plausible mechanisms through which adverse human health
outcomes could occur (U.S. EPA, 2015, p. 20). Two types of experimental
studies are of particular importance in understanding the effects
[[Page 16237]]
of PM exposures: controlled human exposure and animal toxicological
studies. In such studies, investigators expose human volunteers or
laboratory animals, respectively, to known concentrations of air
pollutants under carefully regulated environmental conditions and
activity levels. Thus, controlled human exposure and animal
toxicological studies can provide information on the health effects of
experimentally administered pollutant exposures under highly controlled
laboratory conditions (U.S. EPA, 2015, p. 11).
Controlled human exposure studies have reported that
PM2.5 exposures lasting from less than one hour up to five
hours can impact cardiovascular function,\63\ and the most consistent
evidence from these studies is for impaired vascular function (U.S.
EPA, 2019a, section 6.1.13.2). In addition, although less consistent,
the 2019 ISA notes that studies examining PM2.5 exposures
also provide evidence for increased blood pressure (U.S. EPA, 2019a,
section 6.1.6.3), conduction abnormalities/arrhythmia (U.S. EPA, 2019a,
section 6.1.4.3), changes in heart rate variability (U.S. EPA, 2019a,
section 6.1.10.2), changes in hemostasis that could promote clot
formation (U.S. EPA, 2019a, section 6.1.12.2), and increases in
inflammatory cells and markers (U.S. EPA, 2019a, section 6.1.11.2). The
2019 ISA concludes that, when taken as a whole, controlled human
exposure studies demonstrate that short-term exposure to
PM2.5 may impact cardiovascular function in ways that could
lead to more serious outcomes (U.S. EPA, 2019a, section 6.1.16). Thus,
such studies can provide insight into the potential for specific
PM2.5 exposures to result in physiological changes that
could increase the risk of more serious effects. Table 3-4 in the 2022
PA summarizes information from the 2019 ISA and 2022 ISA supplement on
available controlled human exposure studies that evaluate effects on
markers of cardiovascular function following exposure to
PM2.5 (U.S. EPA, 2022b). Most of the controlled human
exposure studies in Table 3-4 of the 2022 PA have evaluated average
PM2.5 concentrations at or above about 100 [micro]g/m\3\,
with exposure durations typically up to about two hours. Statistically
significant effects on one or more indicators of cardiovascular
function are often, though not always, reported following 2-hour
exposures to average PM2.5 concentrations at and above about
120 [micro]g/m\3\, with less consistent evidence for effects following
exposures to concentrations lower than 120 [micro]g/m\3\. Impaired
vascular function, the effect identified in the 2019 ISA as the most
consistent across studies (U.S. EPA, 2019a, section 6.1.13.2) is shown
following 2-hour exposures to PM2.5 concentrations at and
above 149 [micro]g/m\3\. Mixed results are reported in the studies that
evaluated longer exposure durations (i.e., longer than 2 hours) and
lower (i.e., near-ambient) PM2.5 concentrations (U.S. EPA,
2022b, section 3.3.3.1). For example, significant effects for some
outcomes were reported following 5-hour exposures to 24 [micro]g/m\3\
in Hemmingsen et al. (2015b), but not for other outcomes following 5-
hour exposures to 24 [micro]g/m\3\ in Hemmingsen et al. (2015a) and not
following 24-hour exposures to 10.5 [micro]g/m\3\ in Br[auml]uner et
al. (2008). Additionally, Wyatt et al. (2020) found significant effects
for some cardiovascular (e.g., systematic inflammation markers, cardiac
repolarization, and decreased pulmonary function) effects following 4-
hour exposures to 37.8 [micro]g/m\3\ in healthy young participants (18-
35 years, n=21) who were subject to intermittent moderate exercise. The
higher ventilation rate and longer exposure duration in this study
compared to most controlled human exposure studies is roughly
equivalent to a 2-hour exposure of 75-100 [micro]g/m\3\ of
PM2.5. Therefore, dosimetric considerations may explain the
observed changes in inflammation in young healthy individuals. Though
this study provides evidence of some effects at lower PM2.5
concentrations, overall, there is inconsistent evidence for
inflammation in other controlled human exposure studies evaluated in
the 2019 ISA (U.S. EPA, 2019a, sections 5.1.7., 5.1.2.3.3, and
6.1.11.2.1; U.S. EPA, 2022a, section 3.3.1).
---------------------------------------------------------------------------
\63\ In contrast, controlled human exposure studies provide
little evidence for respiratory effects following short-term
PM2.5 exposures (U.S. EPA, 2019a, section 5.1, Table 5-
18). Therefore, this section focuses on cardiovascular effects
evaluated in controlled human exposure studies of PM2.5
exposure.
---------------------------------------------------------------------------
While controlled human exposure studies are important in
establishing biological plausibility, it is unclear how the results
from these studies alone and the importance of the effects observed in
these studies, should be interpreted with respect to adversity to
public health. More specifically, impaired vascular function can signal
an intermediate effect along the potential biological pathways for
cardiovascular effects following short-term exposure to
PM2.5 and show a role for exposure to PM2.5
leading to potential worsening of IHD and heart failure followed
potentially by ED visits, hospital admissions, or mortality (U.S. EPA,
2019a, section 6.1 and Figure 6-1). However, just observing the
occurrence of impaired vascular function alone does not clearly suggest
an adverse health outcome. Additionally, associated judgments regarding
adversity or health significance of measurable physiological responses
to air pollutants have been informed by guidance, criteria or
interpretative statements developed within the public health community,
including the American Thoracic Society (ATS) and the European
Respiratory Society (ERS), which cooperatively updated the ATS 2000
statement What Constitutes an Adverse Health Effect of Air Pollution
(ATS, 2000) with new scientific findings, including the evidence
related to air pollution and the cardiovascular system (Thurston et
al., 2017).\64\ With regard to vascular function, the ATS/ERS statement
considers the adversity of both chronic and acute reductions in
endothelial function. While the ATS/ERS statement concluded that
chronic endothelial and vascular dysfunction can be judged to be a
biomarker of an adverse health effect from air pollution, they also
conclude that ``the health relevance of acute reductions in endothelial
function induced by air pollution is less certain'' (Thurston et al.,
2017). This is particularly informative to our consideration of the
controlled human exposure studies which are short-term in nature (i.e.,
generally ranging from 2- to 5-hours), including those studies that are
conducted at near-ambient PM2.5 concentrations.
---------------------------------------------------------------------------
\64\ The ATS/ERS described its 2017 statement as one ``intended
to provide guidance to policymakers, clinicians and public health
professionals, as well as others who interpret the scientific
evidence on the health effects of air pollution for risk management
purposes'' and further notes that ``considerations as to what
constitutes an adverse health effect, in order to provide guidance
to researchers and policymakers when new health effects markers or
health outcome associations might be reported in future.'' The most
recent policy statement by the ATS, which once again broadens its
discussion of effects, responses and biomarkers to reflect the
expansion of scientific research in these areas, reiterates that
concept, conveying that it does not offer ``strict rules or
numerical criteria, but rather proposes considerations to be weighed
in setting boundaries between adverse and nonadverse health
effects,'' providing a general framework for interpreting evidence
that proposes a ``set of considerations that can be applied in
forming judgments'' for this context (Thurston et al., 2017).
---------------------------------------------------------------------------
The 2022 PA also notes that it is important to recognize that
controlled human exposure studies include a small number of individuals
compared to epidemiologic studies. Additionally, these studies tend to
include generally healthy adult individuals, who are at a lower risk of
experiencing health effects.
[[Page 16238]]
These studies, therefore, often do not include children, older adults,
or individuals with pre-existing conditions. As such, these studies are
somewhat limited in their ability to inform at what concentrations
effects may be elicited in at-risk populations.
Nonetheless, to provide some insight into what these controlled
human exposure studies may indicate regarding short-term exposure to
peak PM2.5 concentrations and how concentrations relate to
ambient PM2.5 concentrations, analyses in the 2022 PA (U.S.
EPA, 2022b, Figure 2-19) examine monitored 2-hour PM2.5
concentrations (the exposure window most often utilized in the
controlled human exposure studies) at sites meeting the current primary
PM2.5 standards to evaluate the degree to which 2-hour
ambient PM2.5 concentrations at such locations are likely to
exceed the 2-hour exposure concentrations in the controlled human
exposure studies at which statistically significant effects are
reported in multiple studies for one or more indicators of
cardiovascular function. At sites meeting the current primary
PM2.5 standards, most 2-hour concentrations are below 10
[mu]g/m\3\, and almost never exceed 30 [mu]g/m\3\. The extreme upper
end of the distribution of 2-hour PM2.5 concentrations is
shifted higher during the warmer months (April to September), generally
corresponding to the period of peak wildfire frequency in the U.S. At
sites meeting the current primary PM2.5 standards, the
highest 2-hour concentrations measured tend to occur during the period
of peak wildfire frequency (i.e., 99.9th percentile of 2-hour
concentrations is 62 [mu]g/m\3\ during the warm season considered as a
whole). Most of the sites measuring these very high concentrations are
in the northwestern U.S. and California (U.S. EPA, 2022b, Appendix A,
Figure A-1), where wildfires have been relatively common in recent
years. When the typical fire season is excluded from the analysis, the
extreme upper end of the distribution is reduced (i.e., 99.9th
percentile of 2-hour concentrations is 55 [mu]g/m\3\).\65\ Given these
results, the 2022 PA concludes that PM2.5 exposure
concentrations evaluated in most of these controlled human exposure
studies are well-above the 2-hour ambient PM2.5
concentrations typically measured in locations meeting the current
primary standards.
---------------------------------------------------------------------------
\65\ Similar analyses of 4-hour and 5-hour PM2.5
concentrations are presented in Appendix A, Figure A-2 and Figure A-
3, respectively of the 2022 PA (U.S. EPA, 2022b).
---------------------------------------------------------------------------
With respect to animal toxicological studies, the 2019 ISA relies
on animal toxicological studies to support the plausibility of a wide
range of PM2.5-related health effects. While animal
toxicological studies often examine more severe health outcomes and
longer exposure durations than controlled human exposure studies, there
is uncertainty in extrapolating the effects seen in animals, and the
PM2.5 exposures and doses that cause those effects, to human
populations. The 2022 PA considers these uncertainties when evaluating
what the available animal toxicological studies may indicate with
regard to the current primary PM2.5 standards.
As with controlled human exposure studies, most animal
toxicological studies evaluated in the 2019 ISA have examined effects
following exposure to PM2.5 well above the concentrations
likely to be allowed by the current PM2.5 standards. Such
studies have generally examined short-term exposures to
PM2.5 concentrations ranging from 100 to >1,000 [mu]g/m\3\
and long-term exposures to concentrations from 66 to >400 [mu]g/m\3\
(e.g., see U.S. EPA, 2019a, Table 1-2). Two exceptions are animal
toxicological studies reporting impaired lung development following
long-term exposures (i.e., 24 hours per day for several months
prenatally and postnatally) to an average PM2.5
concentration of 16.8 [mu]g/m\3\ (Mauad et al., 2008) and increased
carcinogenic potential following long-term exposures (i.e., 2 months)
to an average PM2.5 concentration of 17.7 [mu]g/m\3\
(Cangerana Pereira et al., 2011). These two studies report serious
effects following long-term exposures to PM2.5
concentrations similar to the ambient concentrations reported in some
PM2.5 epidemiologic studies (U.S. EPA, 2019a, Table 1-2),
though still above the ambient concentrations likely to occur in areas
meeting the current primary PM2.5 standards. However, noting
uncertainty in extrapolating the effects seen in animals, and the
PM2.5 exposures and doses that cause those effects to human
populations, animal toxicological studies are of limited utility in
informing decisions on the public health protection provided by the
current or alternative primary PM2.5 standards. Therefore,
the animal toxicological studies are most useful in providing further
evidence to support the biological mechanisms and plausibility of
various adverse effects.
ii. Ambient PM2.5 Concentrations in Locations of
Epidemiologic Studies
As summarized in section II.A.2.a above, epidemiologic studies
examining associations between daily or annual average PM2.5
exposures and mortality or morbidity represent a large part of the
evidence base supporting several of the 2019 ISA's ``causal'' and
``likely to be causal'' determinations. The 2022 PA considers the
ambient PM2.5 concentrations present in areas where
epidemiologic studies have evaluated associations with mortality or
morbidity, and what such concentrations may indicate regarding the
adequacy of the primary PM2.5 standards. The use of
information from epidemiologic studies to inform conclusions on the
primary PM2.5 standards is complicated by the fact that such
studies evaluate associations between distributions of ambient
PM2.5 and health outcomes, and do not identify the specific
exposures that can lead to the reported effects. Rather, health effects
can occur over the entire distribution of ambient PM2.5
concentrations evaluated, and epidemiologic studies conducted to date
do not identify a population-level threshold below which it can be
concluded with confidence that PM2.5-associated health
effects do not occur. Therefore, the 2022 PA evaluates the
PM2.5 air quality distributions over which epidemiologic
studies support health effect associations (U.S. EPA, 2022b, section
3.3.3.2). In the absence of discernible thresholds, the 2022 PA
considers the study-reported ambient PM2.5 concentrations
reflecting estimated exposure with a focus around the middle portion of
the PM2.5 air quality distribution, where the bulk of the
observed data reside and which provides the strongest support for
reported health effect associations. The section below, as well as in
more detail in section II.B.3.b.i of the proposal (88 FR 5594, January
27, 2023), describes the consideration of the key epidemiologic studies
and observations from these studies, as evaluated in the 2022 PA (U.S.
EPA, 2022b, section 3.3.3.2).
As an initial matter, in considering the PM2.5 air
quality distributions associated with mortality or morbidity in the key
epidemiologic studies, the 2022 PA recognizes that in previous reviews,
the decision framework used to judge adequacy of the existing
PM2.5 standards, and what levels of any potential
alternative standards should be considered, placed significant weight
on epidemiologic studies that assessed associations between
PM2.5 exposure and health outcomes that were most strongly
supported by the body of scientific evidence. In doing so, the decision
framework recognized that while there is no specific point in the air
quality distribution of any
[[Page 16239]]
epidemiologic study that represents a ``bright line'' at and above
which effects have been observed and below which effects have not been
observed, there is significantly greater confidence in the magnitude
and significance of observed associations for the part of the air
quality distribution corresponding to where the bulk of the health
events in each study have been observed, generally at or around the
mean concentration. This is the case both for studies of daily
PM2.5 exposures and for studies of annual average
PM2.5 exposures (U.S. EPA, 2022b, section 3.3.3.2.1).
As discussed further in the 2022 PA, studies of daily
PM2.5 exposures examine associations between day-to-day
variation in PM2.5 concentrations and health outcomes, often
over several years (U.S. EPA, 2022b, section 3.3.3.2.1). While there
can be considerable variability in daily exposures over a multi-year
study period, most of the estimated exposures reflect days with ambient
PM2.5 concentrations around the middle of the air quality
distributions examined (i.e., ``typical'' days rather than days with
extremely high or extremely low concentrations). Similarly, for studies
of annual PM2.5 exposures, most of the health events occur
at estimated exposures that reflect annual average PM2.5
concentrations around the middle of the air quality distributions
examined. In both cases, epidemiologic studies provide the strongest
support for reported health effect associations for this middle portion
of the PM2.5 air quality distribution, which corresponds to
the bulk of the underlying data, rather than the extreme upper or lower
ends of the distribution. Consistent with this, as noted in the 2022 PA
(U.S. EPA, 2022b, section 3.3.1.1), several epidemiologic studies
report that associations persist in analyses that exclude the upper
portions of the distributions of estimated PM2.5 exposures,
indicating that ``peak'' PM2.5 exposures are not
disproportionately responsible for reported health effect associations.
Thus, in considering PM2.5 air quality data from
epidemiologic studies, consistent with approaches in the 2012 and 2020
reviews (78 FR 3161, January 15, 2013; U.S. EPA, 2011, sections 2.1.3
and 2.3.4.1; 85 FR 82716-82717, December 18, 2020; U.S. EPA, 2020b,
sections 3.1.2 and 3.2.3), the 2022 PA evaluates study-reported means
(or medians) of daily and annual average PM2.5
concentrations as indicators for the middle portions of the air quality
distributions, over which studies generally provide strong support for
reported associations and for which confidence in the magnitude and
significance of associations observed in the epidemiologic studies is
greatest (78 FR 3101, January 15, 2013). In addition to the overall
study means, the 2022 PA also focuses on concentrations somewhat below
the means (e.g., 25th and 10th percentiles), when such information is
available from the epidemiologic studies, which again is consistent
with approaches used in previous reviews. In so doing, the 2022 PA
notes, as in previous reviews, that a relatively small portion of the
health events are observed in the lower part of the air quality
distribution and confidence in the magnitude and significance of the
associations begins to decrease in the lower part of the air quality
distribution. Furthermore, consistent with past reviews, there is no
single percentile value within a given air quality distribution that is
most appropriate or ``correct'' to use to characterize where our
confidence in associations becomes appreciably lower. However, and as
detailed further in the 2022 PA, the range from the 25th to 10th
percentiles is a reasonable range to consider as a region where there
is appreciably less confidence in the associations observed in
epidemiologic studies compared to the means (U.S. EPA, 2022b, p. 3-
69).\66\
---------------------------------------------------------------------------
\66\ As detailed in the 2011 PA, we note the interrelatedness of
the distributional statistics and a range of one standard deviation
around the mean which represents approximately 68% of normally
distributed data, and in that one standard deviation below the mean
falls between the 25th and 10th percentiles (U.S. EPA, 2011, p. 2-
71; U.S. EPA, 2005, p. 5-22).
---------------------------------------------------------------------------
In evaluating the overall study-reported means, and concentrations
somewhat below the means from epidemiologic studies, the 2022 PA
focuses on the form, averaging time and level of the current primary
annual PM2.5 standard. Consistent with the approaches used
in the 2012 and 2020 reviews (78 FR 3161-3162, January 15, 2013; 85 FR
82716-82717, December 18, 2020), the annual standard has been utilized
as the primary means of providing public health protection against the
bulk of the distribution of short- and long-term PM2.5
exposures. Thus, the evaluation of the study-reported mean
concentrations from key epidemiologic studies lends itself best to
evaluating the adequacy of the annual PM2.5 standard (rather
than the 24-hour standard with its 98th percentile form). This is true
for the study-reported means from both long-term and short-term
exposure epidemiologic studies, recognizing that the overall mean
PM2.5 concentrations reported in studies of short-term (24-
hour) exposures reflect averages across the study population and over
the years of the study. Thus, mean concentrations from short-term
exposure studies reflect long-term averages of 24-hour PM2.5
exposure estimates. In this manner, the examination of study-reported
means in key epidemiologic studies in the 2022 PA aims to evaluate the
protection provided by the annual PM2.5 standard against the
exposures where confidence is greatest for associations with mortality
and morbidity. In addition, the protection provided by the annual
standard is evaluated in conjunction with that provided by the 24-hour
standard, with its 98th percentile form, which aims to provide
supplemental protection against the short-term exposures to peak
PM2.5 concentrations that can occur in areas with strong
contributions from local or seasonal sources, even when overall ambient
mean PM2.5 concentrations in an area remain relatively low.
In focusing on the annual standard, and in evaluating the range of
study-reported exposure concentrations for which the strongest support
for adverse health effects exists, the 2022 PA examines exposure
concentrations in key epidemiologic studies to determine whether the
current primary annual PM2.5 standard provides adequate
protection against these exposure concentrations. This means, as in
past reviews, application of a decision framework based on assessing
means reported in key epidemiologic studies must also consider how the
study means were computed and how these values compare to the annual
standard metric (including the level, averaging time and form) and the
use of the monitor with the highest PM2.5 design value in an
area for compliance. In the 2012 review, it was recognized that the key
epidemiologic studies computed the study mean using an average across
monitor-based PM2.5 concentrations. As such, the Agency
noted that this decision framework applied an approach of using maximum
monitor concentrations to determine compliance with the standard, while
selecting the standard level based on consideration of composite
monitor concentrations. Further, the Agency included analyses (Hassett-
Sipple et al., 2010; Frank, 2012) that examined the differences in
these two metrics (i.e., maximum monitor concentrations and composite
monitor concentrations) across the U.S. and in areas included in the
key epidemiologic studies and found that the maximum design value in an
area was generally higher than the monitor average across that area,
with the difference varying
[[Page 16240]]
based on location and concentration. This information was taken into
account in the Administrator's final decision in selecting a level for
the primary annual PM2.5 standard the 2012 review and
discussed more specifically in her considerations on adequate margin of
safety.
Consistent with the approach taken in 2012, in assessing how the
overall mean (or median) PM2.5 concentrations reported in
key epidemiologic studies can inform conclusions on the primary annual
PM2.5 standard, the 2022 PA notes that the relationship
between mean PM2.5 concentrations and the area design value
continues to be an important consideration in evaluating the adequacy
of the current or potential alternative annual PM2.5
standard levels in this reconsideration. In a given area, the area
design value is based on the monitor in an area with the highest
PM2.5 concentrations and is used to determine compliance
with the standard. The highest PM2.5 concentrations
spatially distributed in the area would generally occur at or near the
area design value monitor and the distribution of PM2.5
concentrations would generally be lower in other locations and at
monitors in that area. As such, when an area is meeting a specific
annual standard level, the annual average exposures in that area are
expected to be at concentrations lower than that level and the average
of the annual average exposures across that area are expected (i.e., a
metric similar to the study-reported mean values) to be lower than that
level.\67\
---------------------------------------------------------------------------
\67\ In setting a standard level that would require the design
value monitor to meet a level equal to the study-reported mean
PM2.5 concentrations would generally result in lower
concentrations of PM2.5 across the entire area, such that
even those people living near an area design value monitor (where PM
concentrations are generally highest) will be exposed to
PM2.5 concentrations below the air quality conditions
reported in the epidemiologic studies.
---------------------------------------------------------------------------
Another important consideration is that there are a substantial
number of different types of epidemiologic studies available since the
2012 review, included in both the 2019 ISA and the ISA Supplement, that
make understanding the relationship between the mean PM2.5
concentrations and the area design value even more important (U.S. EPA,
2019a; U.S. EPA, 2022a). While the key epidemiologic studies in the
2012 review were all monitor-based studies, the newer studies include
hybrid modeling approaches, which have emerged in the epidemiologic
literature as an alternative to approaches that only use ground-based
monitors to estimate exposure. As assessed in the 2019 ISA and ISA
Supplement, a substantial number of epidemiologic studies used hybrid
model-based methods in evaluating associations between PM2.5
exposure and health effects (U.S. EPA, 2019a; U.S. EPA, 2022a). Hybrid
model-based studies employ various fusion techniques that combine
ground-based monitored data with air quality modeled estimates and/or
information from satellites to estimate PM2.5 exposures.\68\
Additionally, hybrid modeling approaches tend to broaden the areas
captured in the exposure assessment, and in so doing, tend to report
lower mean PM2.5 concentrations than monitor-based
approaches because they include more suburban and rural areas where
concentrations are lower. While these studies provide a broader
estimation of PM2.5 exposures compared to monitor-based
studies (i.e., PM2.5 concentrations are estimated in areas
without monitors), the hybrid modeling approaches result in study-
reported means that are more difficult to relate to the annual standard
metric and to the use of maximum monitor design values to assess
compliance. In addition, and to further complicate the comparison, when
looking across these studies, variations exist in how exposure is
estimated between such studies, which in turn affects how the study
means are calculated. Two important variations across studies include:
(1) Variability in spatial scale used (i.e., averages computed across
the nation (or large portions of the country) versus a focus on only
CBSAs) and (2) variability in exposure assignment methods (i.e.,
averaging across all grid cells [non-population weighting], averaging
across a scaled-up area like a ZIP code [aspects of population
weighting applied], and/or applying population weighting). To elaborate
further on the variability in exposure assignment methods, studies that
use hybrid modeling approaches can estimate PM2.5
concentrations at different spatial resolutions, including at 1 km x 1
km grid cells, at 12 km x 12 km grid cells, or at the census tract
level. Mean reported PM2.5 concentrations can then be
estimated either by averaging up to a larger spatial resolution that
corresponds to the spatial resolution for which health data exists
(e.g., ZIP code level) and therefore apply aspects of population
weighting. These values are then averaged across all study locations at
the larger spatial resolution (e.g., averaged across all ZIP codes in
the study) over the study period, resulting in the study-reported mean
24-hour average or average annual PM2.5 concentration. Other
studies that use hybrid modeling methods to estimate PM2.5
concentrations may use each grid cell to calculate the study-reported
mean 24-hour average or average annual PM2.5 concentration.
As such, these types of studies do not apply population weighting in
their mean concentrations. In studies that use each grid cell to report
a mean PM2.5 concentration and do not apply aspects of
population weighting, the study mean may not reflect the exposure
concentrations used in the epidemiologic study to assess the reported
association. The impact of the differences in methods is an important
consideration when comparing mean concentrations across studies (U.S.
EPA, 2022b, section 3.3.3.2.1). Thus, the 2022 PA also considers the
methods used to estimate PM2.5 concentrations, which vary
from traditional methods using monitoring data from ground-based
monitors \69\ to those using more complex hybrid modeling approaches
and how these methods calculate the study-reported mean
PM2.5 concentration.\70\
---------------------------------------------------------------------------
\68\ More detailed information about hybrid model methods and
performance is described in section 2.3.3.2 of the 2022 PA (U.S.
EPA, 2022b).
\69\ In those studies that use ground-based monitors alone to
estimate long- or short-term PM2.5 concentrations,
approaches include: (1) PM2.5 concentrations from a
single monitor within a city/county; (2) average of PM2.5
concentrations across all monitors within a city/county or other
defined study area (e.g., CBSA); or (3) population-weighted averages
of exposures. Once the study location average PM2.5
concentration is calculated, the study-reported long-term average is
derived by averaging daily/annual PM2.5 concentrations
across all study locations over the entire study period.
\70\ Detailed information on the methods by which mean
PM2.5 concentrations are calculated in key monitor- and
hybrid model-based U.S. and Canadian epidemiologic studies are
presented in Tables 3-6 through 3-9 in the 2022 PA (U.S. EPA,
2022b).
---------------------------------------------------------------------------
Given the emergence of the hybrid model-based epidemiologic studies
since the 2012 review, the 2022 PA explores the relationship between
the approaches used in these studies to estimate PM2.5
concentrations and the impact that the different methods have on the
study-reported mean PM2.5 concentrations. The 2022 PA
further seeks to understand how the approaches and resulting mean
concentrations compare across studies, as well as what the resulting
mean values represent relative to the annual standard. In so doing, the
2022 PA presents analyses that compare the area annual design values,
composite monitor PM2.5 concentrations, and mean
concentrations from two hybrid modeling approaches, including
evaluation of the means when population weighting is applied and when
population weighting is not
[[Page 16241]]
applied (U.S. EPA, 2022b, section 2.3.3.1).
In the air quality analyses comparing composite monitored
PM2.5 concentrations with annual PM2.5 design
values in U.S. CBSAs, maximum annual PM2.5 design values
were approximately 10% to 20% higher than annual average composite
monitor concentrations (i.e., averaged across multiple monitors in the
same CBSA) (sections I.D.5.a above and U.S. EPA, 2022b, section
2.3.3.1, Figure 2-28 and Table 2-3). The difference between the maximum
annual design value and average concentration in an area can be smaller
or larger than this range (10-20%), depending on a variety of factors
such as the number of monitors, monitor siting characteristics, the
distribution of ambient PM2.5 concentrations, and how the
average concentrations are calculated (i.e., averaged across monitors
versus across modeled grid cells). Results of this analysis suggest
that there will be a distribution of concentrations across an area and
the maximum annual average monitored concentration in an area (at the
design value monitor, used for compliance with the standard), will
generally be 10-20% higher than the average PM2.5
concentration across the other monitors in the area. Thus, in
considering how the annual standard levels would relate to the study-
reported means from key monitor-based epidemiologic studies, the 2022
PA generally concludes that an annual standard level that is no more
than 10-20% higher than monitor-based study-reported mean
PM2.5 concentrations would generally maintain air quality
exposures to be below those associated with the study-reported mean
PM2.5 concentrations, exposures for which the strongest
support for adverse health effects occurring is available.
The 2022 PA also evaluates data from two hybrid modeling approaches
(DI2019 and HA2020) that have been used in several recent epidemiologic
studies (U.S. EPA, 2022b, section 2.3.3.2.4).\71\ The analysis shows
that the means differ when PM2.5 concentrations are
estimated in urban areas only (CBSAs) versus when the averages were
calculated with all or most grid cells nationwide, likely because areas
included outside of CBSAs tend to be more rural and have lower
estimated PM2.5 concentrations. The 2022 PA recognizes the
importance of this variability in the means since the study areas
included in the calculation of the mean, and more specifically whether
a study is focused on nationwide, regional, or urban areas, will affect
the calculation of the study mean based on how many rural areas, with
lower estimated PM2.5 concentrations, are included in the
study area. While the determination of what spatial scale to use to
estimate PM2.5 concentrations does not inherently affect the
quality of the epidemiologic study, the spatial scale can influence the
calculated reported long-term mean concentration across the study area
and period. The results of the analysis show that, regardless of the
hybrid modeling approach assessed, the annual average PM2.5
concentrations in CBSA-only analyses are 4-8% higher than for
nationwide analyses, likely as a result of higher PM2.5
concentrations in more densely populated areas, and exclusion of more
rural areas (U.S. EPA, 2022b, Table 2-4). When evaluating comparisons
between surfaces that estimate exposure using aspects of population
weighting versus surfaces that do not calculate means using population
weighting, surfaces that calculate long-term mean PM2.5
concentrations with population-weighted averages have higher average
annual PM2.5 concentrations, compared to annual
PM2.5 concentrations in analyses that do not apply
population weighting.\72\ Analyses show that average maximum annual
design values are 40 to 50% higher when compared to annual average
PM2.5 concentrations estimated without population weighting
versus 15% to 18% higher when compared to average annual
PM2.5 concentrations estimated with population weighting
applied (similar to the differences observed for the composite monitor
comparison values for the monitor-based epidemiologic studies) (U.S.
EPA, 2022b, section 2.3.3.2.4). Given these results, it is worth noting
that for the studies using the hybrid modeling approaches, the choice
of methodology employed in calculating the study-reported means (i.e.,
using population weighting or not), and not a difference in estimates
of exposure in the study itself, can produce substantially different
study-reported mean values, where approaches that do not apply
population weighting leading to much lower estimated mean
PM2.5 concentrations.
---------------------------------------------------------------------------
\71\ More details on the evaluation of the two hybrid modeling
approaches is provided in section 2.3.3.2.4 of the 2022 PA (U.S.
EPA, 2022b).
\72\ The annual PM2.5 concentrations for the
population-weighted averages ranged from 8.2-10.2 [mu]g/m\3\, while
those that do not apply population weighting ranged from 7.0-8.6
[mu]g/m\3\. Average maximum annual design values ranged from 9.5 to
11.7 [mu]g/m\3\.
---------------------------------------------------------------------------
Based on these results, and similar to conclusions for the monitor-
based studies, the 2022 PA generally concludes that study-reported mean
concentrations in the studies that employ hybrid modeling approaches
and calculate a population-weighted mean are associated with air
quality conditions that would be achieved by meeting annual standard
levels that are 15-18% higher than study-reported means. Therefore, an
annual standard level that is no more than 15-18% higher than the
study-reported means would generally maintain air quality exposures to
be below those associated with the study-reported mean PM2.5
concentrations, exposures for which we have the strongest support for
adverse health effects occurring. For the studies that utilize hybrid
modeling approaches but do not incorporate population weighting in
calculating the mean, the annual design values associated with these
air quality conditions are expected to be much higher (i.e., 40-50%
higher) and this larger difference makes it more difficult to consider
how these studies can be used to determine the adequacy of the
protection afforded by the current or potential alternative annual
standards. Additionally, as noted above in studies that utilize hybrid
modeling approaches and that do not incorporate population weighting in
calculating the mean (e.g., use each grid cell to calculate a mean
PM2.5 concentration), the study mean does not reflect the
exposure concentrations used in the epidemiologic study to assess the
reported association.
The 2022 PA notes that while these analyses can be useful to
informing the understanding of the relationship between study-reported
mean concentrations and the level of the annual standard, some
limitations of this analysis must be recognized (U.S. EPA, 2022a,
section 3.3.3.2.1). First, the comparisons used only two hybrid
modeling approaches. Although these two hybrid modeling surfaces have
been used in a number of recent epidemiologic studies, they represent
just two of the many hybrid modeling approaches that have been used in
epidemiologic studies to estimate PM2.5 concentrations.
These methods continue to evolve, with further development and
improvement to prediction models that estimate PM2.5
concentrations in epidemiologic studies. In addition to differences in
hybrid modeling approaches, epidemiologic studies also use different
methods to assign a population weighted average PM2.5
concentration to their study population, and the assessment presented
in the
[[Page 16242]]
2022 PA does not evaluate all of the potential methods that could be
used.
Additionally, while some of these epidemiologic studies also
provide information on the broader distributions of exposure estimates
and/or health events and the PM2.5 concentrations
corresponding to the lower percentiles of those data (e.g., 25th and/or
10th), the air quality analysis in the 2022 PA focuses on mean
PM2.5 concentrations and a similar comparison for lower
percentiles of data was not assessed. Therefore, any direct comparison
of study-reported PM2.5 concentrations corresponding to
lower percentiles and annual design values is more uncertain than such
comparisons with the mean. Finally, air quality analysis presented in
the 2022 PA and detailed above in section I.D.5 included two hybrid
modeling-based approaches that used U.S.-based air quality information
for estimating PM2.5 concentrations. As such, the analyses
are most relevant to interpreting the study-reported mean
concentrations from U.S. epidemiologic studies and do not provide
additional information about how the mean exposures concentrations
reported in epidemiologic studies in other countries would compare to
annual design values observed in the U.S. In addition, while
information from Canadian studies can be useful in assessing the
adequacy of the annual standard, differences in the exposure
environments and population characteristics between the U.S. and other
countries can affect the study-reported mean value and its relationship
with the annual standard level. Sources and pollutant mixtures, as well
as PM2.5 concentration gradients, may be different between
countries, and the exposure environments in other countries may differ
from those observed in the U.S. Furthermore, differences in population
characteristics and population densities can also make it challenging
to directly compare studies from countries outside of the U.S. to a
design value in the U.S.
As with the experimental studies discussed above, the 2022 PA
focuses on epidemiologic studies assessed in the 2019 ISA and ISA
Supplement that have the potential to be most informative in reaching
decisions on the adequacy of the primary PM2.5 standards.
The 2022 PA focuses on epidemiologic studies that provide strong
support for ``causal'' or ``likely to be causal'' relationships with
PM2.5 exposures in the 2019 ISA. Further, the 2022 PA also
focuses on the health effect associations that are determined in the
2019 ISA and ISA Supplement to be consistent across studies, coherent
with the broader body of evidence (e.g., including animal and
controlled human exposure studies), and robust to potential confounding
by co-occurring pollutants and other factors.\73\ In particular the
2022 PA considers the U.S. and Canadian epidemiologic studies to be
more useful for reaching conclusions on the current standards than
studies conducted in other countries, given that the results of the
U.S. and Canadian studies are more directly applicable for quantitative
considerations, whereas studies conducted in other countries reflect
different populations, exposure characteristics, and air pollution
mixtures. Additionally, epidemiologic studies outside of the U.S. and
Canada generally reflect higher PM2.5 concentrations in
ambient air than are currently found in the U.S., and are less relevant
to informing questions about adequacy of the current standards.\74\
However, and as noted above, the 2022 PA also recognizes that while
information from Canadian studies can be useful in assessing the
adequacy of the annual standard, there are still important differences
between the exposure environments in the U.S. and Canada and
interpreting the data (e.g., mean concentrations) from the Canadian
studies in the context of a U.S.-based standard may present challenges
in directly and quantitatively informing questions regarding the
adequacy of the current or potential alternative the levels of the
annual standard. Lastly, the 2022 PA emphasizes multicity/multistate
studies that examine health effect associations, as such studies are
more encompassing of the diverse atmospheric conditions and population
demographics in the U.S. than studies that focus on a single city or
State. Figures 3-4 through 3-7 in the 2022 PA summarize the study
details for the key U.S. and Canadian epidemiologic studies (U.S. EPA,
2022b, section 3.3.3.2.1).\75\
---------------------------------------------------------------------------
\73\ As described in the Preamble to the ISAs (U.S. EPA, 2015),
``the U.S. EPA emphasizes the importance of examining the pattern of
results across various studies and does not focus solely on
statistical significance or the magnitude of the direction of the
association as criteria of study reliability. Statistical
significance is influenced by a variety of factors including, but
not limited to, the size of the study, exposure and outcome
measurement error, and statistical model specifications. Statistical
significance may be informative; however, it is just one of the
means of evaluating confidence in the observed relationship and
assessing the probability of chance as an explanation. Other
indicators of reliability such as the consistency and coherence of a
body of studies as well as other confirming data may be used to
justify reliance on the results of a body of epidemiologic studies,
even if results in individual studies lack statistical significance.
Traditionally, statistical significance is used to a larger extent
to evaluate the findings of controlled human exposure and animal
toxicological studies. Understanding that statistical inferences may
result in both false positives and false negatives, consideration is
given to both trends in data and reproducibility of results. Thus,
in drawing judgments regarding causality, the U.S. EPA emphasizes
statistically significant findings from experimental studies, but
does not limit its focus or consideration to statistically
significant results in epidemiologic studies.''
\74\ This emphasis on studies conducted in the U.S. or Canada is
consistent with the approach in the 2012 and 2020 reviews of the PM
NAAQS (U.S. EPA, 2011, section 2.1.3; U.S. EPA, 2020b, section
3.2.3.2.1) and with approaches taken in other NAAQS reviews.
However, the importance of studies in the U.S., Canada, and other
countries in informing an ISA's considerations of the weight of the
evidence that informs causality determinations is recognized.
\75\ The cohorts examined in the studies included in Figure 3-4
to Figure 3-7 of the 2022 PA include large numbers of individuals in
the general population, and often also include those populations
identified as at-risk (i.e., children, older adults, minority
populations, and individuals with pre-existing cardiovascular and
respiratory disease).
---------------------------------------------------------------------------
The key epidemiologic studies identified in the 2022 PA indicate
generally positive and statistically significant associations between
estimated PM2.5 exposures (short- or long-term) and
mortality or morbidity across a range of ambient PM2.5
concentrations (U.S. EPA, 2022b, section 3.3.3.2.1), report overall
mean (or median) PM2.5 concentrations, and include those for
which the years of PM2.5 air quality data used to estimate
exposures overlap entirely with the years during which health events
are reported.\76\ Additionally, for studies that estimate
PM2.5 exposure using hybrid modeling approaches, the 2022 PA
also considers the approach used to estimate PM2.5
concentrations and the approach used to validate hybrid model
predictions when evaluating those studies as key epidemiologic studies
\77\ and focuses on those studies that use recent methods based on
surfaces that are with fused with monitored PM2.5
[[Page 16243]]
concentration data (U.S. EPA, 2022b, section 3.3.3.2.1).
---------------------------------------------------------------------------
\76\ For some studies of long-term PM2.5 exposures,
exposure is estimated from air quality data corresponding to only
part of the study period, often including only the later years of
the health data, and are not likely to reflect the full ranges of
ambient PM2.5 concentrations that contributed to reported
associations. While this approach can be reasonable in the context
of an epidemiologic study that is evaluating health effect
associations with long-term PM2.5 exposures, under the
assumption that spatial patterns in PM2.5 concentrations
are not appreciably different during time periods for which air
quality information is not available (e.g., Chen et al., 2016), the
2022 PA focuses on the distribution of ambient PM2.5
concentrations that could have contributed to reported health
outcomes. Therefore, the 2022 PA identifies studies as key
epidemiologic studies when the years of air quality data and health
data overlap in their entirety.
\77\ Such studies are identified as those that use hybrid
modeling approaches for which recent methods and models were used
(e.g., recent versions and configurations of the air quality
models); studies that are fused with PM2.5 data from
national monitoring networks (i.e., FRM/FEM data); and studies that
reported a thorough model performance evaluation for core years of
the study.
---------------------------------------------------------------------------
Figure 1 below (U.S. EPA, 2022b, Figure 3-8) highlights the overall
mean (or median) PM2.5 concentrations reported in key U.S.
studies that use ground-based monitors alone to estimate long- or
short-term PM2.5 exposure.\78\ For the small subset of
studies with available information on the broader distributions of
underlying data, Figure 1 below also identifies the study-period
PM2.5 concentrations corresponding to the 25th and 10th
percentiles of health events \79\ (see Appendix B, Section B.2 of the
2022 PA for more information). Figure 2 (U.S. EPA, 2022a, Figure 3-14)
presents overall means of predicted PM2.5 concentrations for
key U.S. model-based epidemiologic studies that apply aspects of
population-weighting, and the concentrations corresponding to the 25th
and 10th percentiles of estimated exposures or health events \80\ when
available (see Appendix B, section B.3 for additional information).\81\
---------------------------------------------------------------------------
\78\ Canadian studies that use ground-based monitors estimate
long- or short-term PM2.5 exposures are found in Figure
3-9 of the 2022 PA, including concentrations corresponding to the
25th and 10th percentiles of estimated exposures or health events,
when available (U.S. EPA, 2022b).
\79\ That is, 25% of the total health events occurred in study
locations with mean PM2.5 concentrations (i.e., averaged
over the study period) below the 25th percentiles identified in
Figure 3-8 of the 2022 PA and 10% of the total health events
occurred in study locations with mean PM2.5
concentrations below the 10th percentiles identified.
\80\ For most studies in Figure 2 below (Figure 3-14 in the 2022
PA), 25th percentiles of exposure estimates are presented. The
exception is Di et al. (2017b), for which Figure 2 (U.S. EPA, 2022b,
Figure 3-14) presents the short-term PM2.5 exposure
estimates corresponding to the 25th and 10th percentiles of deaths
in the study population (i.e., 25% and 10% of deaths occurred at
concentrations below these concentrations). In addition, the authors
of Di et al. (2017b) provided population-weighted exposure values.
The 10th and 25th percentiles of these population-weighted exposure
estimates are 7.9 and 9.5 [mu]g/m\3\, respectively.
\81\ Overall mean (or median) PM2.5 concentrations
reported in key Canadian studies that use model-based approaches to
estimate long- or short-term PM2.5 concentrations and the
concentrations corresponding to the 25th and 10th percentiles of
estimated exposures or health events, when available are found in
Figure 3-9 of the 2022 PA (U.S. EPA, 2022b).
---------------------------------------------------------------------------
[[Page 16244]]
BILLING CODE 6560-50-P
[GRAPHIC] [TIFF OMITTED] TR06MR24.029
[[Page 16245]]
[GRAPHIC] [TIFF OMITTED] TR06MR24.030
BILLING CODE 6560-50-C
Based on its evaluation of study-reported mean concentrations, the
2022 PA notes that key epidemiologic studies conducted in the U.S. or
Canada report generally positive and statistically significant
associations between estimated PM2.5 exposures (short- or
long-term) and mortality or morbidity across a wide range of ambient
PM2.5 concentrations (U.S. EPA, 2022b, section 3.3.3.2.1).
The 2022 PA makes a number of observations with regard to the study-
reported PM2.5 concentrations in the key U.S. and Canadian
epidemiologic studies.
The 2022 PA first considers the PM2.5 concentrations
from the key U.S. epidemiologic studies. For studies that use monitors
to estimate PM2.5 exposures, overall mean PM2.5
concentrations range between 9.9 [mu]g/m\3\ \82\ to 16.5 [mu]g/m\3\
(Figure 1 above and
[[Page 16246]]
U.S. EPA, 2022b, Figure 3-8). For key U.S. epidemiologic studies that
use hybrid model-predicted exposures and apply aspects of population-
weighting, mean PM2.5 concentrations range from 9.3 [mu]g/
m\3\ to just above 12.2 [micro]g/m\3\ (Figure 2 above and U.S. EPA,
2022b, Figure 3-14). In studies that average up from the grid cell
level to the ZIP code, postal code, or census tract level, mean
PM2.5 concentrations range from 9.8 [micro]g/m\3\ to 12.2
[micro]g/m\3\. The one study that population-weighted the grid cell
prior to averaging up to the ZIP code or census tract level reported
mean PM2.5 concentrations of 9.3 [mu]g/m\3\. Based on air
quality analyses noted above, these hybrid modelled epidemiologic
studies are expected to report means similar to those from monitor-
based studies.
---------------------------------------------------------------------------
\82\ This is generally consistent with, but slightly below, the
lowest study-reported mean PM2.5 concentration from
monitor-based studies available in the 2020 PA, which was 10.7
[mu]g/m\3\ (U.S. EPA, 2020a, Figure 3-7).
---------------------------------------------------------------------------
Other key U.S. epidemiologic studies that use hybrid modeling
approaches estimate mean PM2.5 exposure by averaging each
grid cell across the entire study area, whether that be the nation or a
region of the country. These studies do not weight the estimated
exposure concentrations based on population density or location of
health events. As such, the study mean reported in these studies may
not reflect the exposure concentrations used in the epidemiologic study
to assess the reported association. As a result, these reported mean
concentrations are the most different (and much lower) than the means
reported in monitor-based studies. Due to the methodology employed in
calculating the study-reported means and not necessarily a difference
in estimates of exposure, these epidemiologic studies are expected to
report some of the lowest mean values. For these studies, the reported
mean PM2.5 concentrations range from 8.1 [mu]g/m\3\ to 11.9
[mu]g/m\3\ (U.S. EPA, 2022b, Figure 3-14). As noted above, for studies
that utilize hybrid modeling approaches but do not incorporate
population weighting into the reported mean calculation, the associated
annual design values would be expected to be much higher (i.e., 40-50%
higher) than the study-reported means. This larger difference between
design values and study-reported mean concentrations makes it more
difficult to consider how these studies can be used to determine the
adequacy of the protection afforded by the current or potential
alternative annual standards (U.S. EPA, 2022b, section 3.3.3.2.1).
In addition to the mean PM2.5 concentrations, a subset
of the key U.S. epidemiologic studies report PM2.5
concentrations corresponding to the 25th and 10th percentiles of health
data or exposure estimates to provide insight into the concentrations
that comprise the lower quartile of the air quality distributions. In
studies that use monitors to estimate PM2.5 exposures, 25th
percentiles of health events correspond to PM2.5
concentrations (i.e., averaged over the study period for each study
city) at or above 11.5 [micro]g/m\3\ and 10th percentiles of health
events correspond to PM2.5 concentrations at or above 9.8
[micro]g/m\3\ (i.e., 25% and 10% of health events, respectively, occur
in study locations with PM2.5 concentrations below these
values) (Figure 1 above and U.S. EPA, 2022b, Figure 3-8). Of the key
U.S. epidemiologic studies that use hybrid modeling approaches and
apply population-weighting to estimate long-term PM2.5
exposures, the ambient PM2.5 concentrations corresponding to
25th percentiles of estimated exposures are 9.1 [micro]g/m\3\ (Figure 2
and U.S. EPA, 2022b, Figure 3-14). In key U.S. epidemiologic studies
that use hybrid modeling approaches and apply population-weighting to
estimate short-term PM2.5 exposures, the ambient
concentrations corresponding to 25th percentiles of estimated
exposures, or health events, are 6.7 [micro]g/m\3\ (Figure 2 and U.S.
EPA, 2022b, Figure 3-14). In key U.S. epidemiologic studies that use
hybrid modeling approaches and do not apply population-weighting to
estimate PM2.5 exposures, the ambient concentrations
corresponding to 25th percentiles of estimated exposures, or health
events, range from 4.6 to 9.2 [micro]g/m\3\ (U.S. EPA, 2022b, Figure 3-
14).\83\ In the key epidemiologic studies that apply hybrid modeling
approaches with population-weighting and with information available on
the 10th percentile of health events, the ambient PM2.5
concentration corresponding to that 10th percentile range from 4.7
[micro]g/m\3\ to 7.3 [micro]g/m\3\ (Figure 2 and U.S. EPA, 2022b,
Figure 3-14).
---------------------------------------------------------------------------
\83\ In the one study that reports 25th percentile exposure
estimates of 4.6 [micro]g/m\3\ (Shi et al., 2016), the authors
report that most deaths occurred at or above the 75th percentile of
annual exposure estimates (i.e., 10 [mu]g/m\3\). The short-term
exposure estimates accounting for most deaths are not presented in
the published study.
---------------------------------------------------------------------------
The 2022 PA next considers the PM2.5 concentrations from
the key Canadian epidemiologic studies. Generally, the study-reported
mean concentrations in Canadian studies are lower than those reported
in the U.S. studies for both monitor-based and hybrid model methods.
For the majority of key Canadian epidemiologic studies that use
monitor-based exposure, mean PM2.5 concentrations generally
ranged from 7.0 [micro]g/m\3\ to 9.0 [micro]g/m\3\ (U.S. EPA, 2022b,
Figure 3-9). For these studies, 25th percentiles of health events
correspond to PM2.5 concentrations at or above 6.5 [micro]g/
m\3\ and 10th percentiles of health events correspond to
PM2.5 concentrations at or above 6.4 [micro]g/m\3\ (U.S.
EPA, 2022b, Figure 3-9). For the key Canadian epidemiologic studies
that use hybrid model-predicted exposure, the mean PM2.5
concentrations are generally lower than in U.S. model-based studies
(U.S. EPA, 2022b, Figure 3-10), ranging from approximately 6.0
[micro]g/m\3\ to just below 10.0 [micro]g/m\3\ (U.S. EPA, 2022b, Figure
3-11). The majority of the key Canadian epidemiologic studies that used
hybrid modeling were completed at the nationwide scale, while four
studies were completed at the regional geographic spatial scale. In
addition, all the key Canadian epidemiologic studies apply aspects of
population weighting, where all grid cells within a postal code are
averaged, individuals are assigned exposure at the postal code
resolution, and study mean PM2.5 concentrations are based on
the average of individual exposures. The majority of studies estimating
exposure nationwide range between just below 6.0 [micro]g/m\3\ to 8.0
[micro]g/m\3\ (U.S. EPA, 2022b, Figure 3-11). One study by Erickson et
al. (2020) presents an analysis related immigrant status and length of
residence in Canada versus non-immigrant populations, which accounts
for the four highest mean PM2.5 concentrations which range
between 9.0 [micro]g/m\3\ and 10.0 [micro]g/m\3\ (U.S. EPA, 2022b,
Figure 3-11). The four studies that estimate exposure at the regional
scale report mean PM2.5 concentrations that range from 7.8
[micro]g/m\3\ to 9.8 [micro]g/m\3\ (U.S. EPA, 2022b, Figure 3-11).
Three key Canadian epidemiologic studies report information on the 25th
percentile of health events. In these studies, the ambient
PM2.5 concentration corresponding to the 25th percentile is
approximately 8.0 [micro]g/m\3\ in two studies, and 4.3 [micro]g/m\3\
in a third study (U.S. EPA, 2022b, Figure 3-11).
In addition to the expanded body of evidence from the key U.S.
epidemiologic studies discussed above, there are also a subset of
epidemiologic studies that have emerged that further inform an
understanding of the relationship between PM2.5 exposure and
health effects, including studies with the highest exposures excluded
(restricted analyses), epidemiologic studies that employed statistical
approaches that attempt to more extensively account for confounders and
are more robust to model misspecification (i.e., used alternative
[[Page 16247]]
methods for confounder control),\84\ and accountability studies (U.S.
EPA, 2019a, U.S. EPA, 2021a, U.S. EPA, 2022a).
---------------------------------------------------------------------------
\84\ As noted in the ISA Supplement (U.S. EPA, 2022a, p. 1-3):
``In the peer-reviewed literature, these epidemiologic studies are
often referred to as alternative methods for confounder control. For
the purposes of this Supplement, this terminology is not used to
prevent confusion with the main scientific conclusions (i.e., the
causality determinations) presented within an ISA. In addition, as
is consistent with the weight-of-evidence framework used within ISAs
and discussed in the Preamble to the Integrated Science Assessments,
an individual study on its own cannot inform causality, but instead
represents a piece of the overall body of evidence.''
---------------------------------------------------------------------------
Restricted analyses are studies that examine health effect
associations in analyses with the highest exposures excluded,
restricting analyses to daily exposures less than the 24-hour primary
PM2.5 standard and annual exposures less than the annual
PM2.5 standard. The 2022 PA presents a summary of restricted
analyses evaluated in the 2019 ISA and ISA Supplement (U.S. EPA, 2022b,
Table 3-10). The restricted analyses can be informative in assessing
the nature of the association between long-term exposures (e.g., annual
average concentrations <12.0 [micro]g/m\3\) or short-term exposures
(e.g., daily concentrations <35 [micro]g/m\3\) when looking only at
exposures to lower concentrations, including whether the association
persists in such restricted analyses compared to the same analyses for
all exposures, as well as whether the association is stronger, in terms
of magnitude and precision, than when completing the same analysis for
all exposures. While these studies are useful in supporting the
confidence and strength of associations at lower concentrations, these
studies also have inherent uncertainties and limitations, including
uncertainty in how studies exclude concentrations (e.g., are they
excluded at the modeled grid cell level, the ZIP code level) and in how
concentrations in studies that restrict air quality data relate to
design values for the annual and 24-hour standards. Further, these
studies often do not report descriptive statistics (e.g., mean
PM2.5 concentrations, or concentrations at other
percentiles) that allow for additional consideration of this
information. As such, while these studies can provide additional
supporting evidence for associations at lower concentrations, the 2022
PA notes that there are also limitations in how to interpret these
studies when evaluating the adequacy of the current or potential
alternative standards.
Restricted analyses provide additional information on the nature of
the association between long- or short-term exposures when analyses are
restricted to lower PM2.5 concentrations and indicate that
effect estimates are generally greater in magnitude in the restricted
analyses for long- and short-term PM2.5 exposure compared to
the main analyses. In two U.S. studies that report mean
PM2.5 concentrations in restricted analyses and that
estimate effects associated with long-term exposure to
PM2.5, the effect estimates are greater in the restricted
analyses than in the main analyses. Di et al. (2017a) and Dominici et
al. (2019) report positive and statistically significant associations
in analyses restricted to concentrations less than 12.0 [micro]g/m\3\
for all-cause mortality and effect estimates are greater in the
restricted analyses than effect estimates reported in main analyses. In
addition, both studies report mean PM2.5 concentrations of
9.6 [micro]g/m\3\. While none of the U.S. studies of short-term
exposure present mean PM2.5 concentrations for the
restricted analyses, these studies generally have mean 24-hour average
PM2.5 concentrations in the main analyses below 12.0
[micro]g/m\3\, and report increases in the effect estimates in the
restricted analyses compared to the main analyses. Additionally, in the
one Canadian study of long-term PM2.5 exposure, Zhang et al.
(2021) conducted analyses where annual PM2.5 concentrations
were restricted to concentrations below 10.0 [micro]g/m\3\ and 8.8
[micro]g/m\3\, which presumably have lower mean concentrations than the
mean of 7.8 [micro]g/m\3\ reported in the main analyses, though
restricted analysis mean PM2.5 concentrations are not
reported. Effect estimates for non-accidental mortality are greater in
analyses restricted to PM2.5 concentrations less than 10.0
[micro]g/m\3\, but less in analyses restricted to <8.8 [micro]g/m\3\.
The second type of studies that have recently emerged and further
inform the consideration of the relationship between PM2.5
exposure and health effects in the 2022 PA are those that employ
alternative methods for confounder control. Alternative methods for
confounder control seek to mimic randomized experiments through the use
of study design and statistical methods to more extensively account for
confounders and are more robust to model misspecification. The 2022 PA
presents a summary of the studies that employ alternative methods for
confounder control, and employ a variety of statistical methods, which
are evaluated in the 2019 ISA and ISA Supplement (U.S. EPA, 2022b,
Table 3-11). These studies reported consistent results among large
study populations across the U.S. and can further inform the
relationship between long- and short-term PM2.5 exposure and
total mortality. Studies that employ alternative methods for confounder
control to assess the association between long-term exposure to
PM2.5 and mortality reduce uncertainties related to
confounding and provide additional support for the associations
reported in the broader body of cohort studies that examined long-term
PM2.5 exposure and mortality.
Lastly, there is a subset of epidemiologic studies that assess
whether long-term reductions in ambient PM2.5 concentrations
result in corresponding reductions in health outcomes. These include
studies that evaluate the potential for improvements in public health,
including reductions in mortality rates, increases in life expectancy,
and reductions in respiratory disease as ambient PM2.5
concentrations have declined over time. Some of these studies,
accountability studies, provide insight on whether the implementation
of environmental policies or air quality interventions result in
changes/reductions in air pollution concentrations and the
corresponding effect on health outcomes.\85\ The 2022 PA presents a
summary of these studies, which are assessed in the 2019 ISA and ISA
Supplement (U.S. EPA, 2022b, Table 3-12). These studies lend support
for the conclusion that improvements in air quality are associated with
improvements in public health.
---------------------------------------------------------------------------
\85\ Given the nature of these studies, the majority tend to
focus on time periods in the past during which ambient
PM2.5 concentrations were substantially higher than those
measured more recently (e.g., see U.S. EPA, 2022b, Figure 2-16).
---------------------------------------------------------------------------
More specifically, of the accountability studies that account for
changes in PM2.5 concentrations due to a policy or the
implementation of an intervention and whether there was evidence of
changes in associations with mortality or cardiovascular effects as a
result of changes in annual PM2.5 concentrations, Corrigan
et al. (2018), Henneman et al. (2019) and Sanders et al. (2020a)
present analyses with starting PM2.5 concentrations (or
concentrations prior to the policy or intervention) below 12.0
[micro]g/m\3\. Henneman et al. (2019) explored changes in modeled
PM2.5 concentrations following the retirement of coal fired
power plants in the U.S., and found that reductions from mean annual
PM2.5 concentrations of 10.0 [micro]g/m\3\ in 2005 to mean
annual PM2.5 concentrations of 7.2 [micro]g/m\3\ in 2012
from coal-fueled power plants resulted in corresponding reductions in
the number of cardiovascular-related
[[Page 16248]]
hospital admissions, including for all cardiovascular disease, acute
MI, stroke, heart failure, and ischemic heart disease in those aged 65
and older. Corrigan et al. (2018) examined whether there was a change
in the cardiovascular mortality rate before (2000-2004) and after
(2005-2010) implementation of the first annual PM2.5 NAAQS
implementation based on mortality data from the National Center for
Health Statistics and reported 1.10 (95% confidence interval (CI):
0.37, 1.82) fewer cardiovascular deaths per year per 100,000 people for
each 1 [mu]g/m\3\ reduction in annual PM2.5 concentrations.
When comparing whether counties met the annual PM2.5
standard (attainment counties), there were 1.96 (95% CI: 0.77, 3.15)
fewer cardiovascular deaths for each 1 [mu]g/m\3\ reduction in annual
PM2.5 concentrations between the two periods for attainment
counties, whereas in non-attainment counties (e.g., counties that did
not meet the annual PM2.5 standard), there were 0.59 (95%
CI: - 0.54, 1.71) fewer cardiovascular deaths between the two periods.
And lastly, Sanders et al. (2020a) examined whether policy actions
(i.e., the first annual PM2.5 NAAQS implementation rule in
2005 for the 1997 annual PM2.5 standard with a 3-year annual
average of 15 [mu]g/m\3\) reduced PM2.5 concentrations and
mortality rates in Medicare beneficiaries between 2000-2013. They
report evidence of changes in associations with mortality (a decreased
mortality rate of ~0.5 per 1,000 in attainment and non-attainment
areas) due to changes in annual PM2.5 concentrations in both
attainment and non-attainment areas. Additionally, attainment areas had
starting concentrations below 12.0 [micro]g/m\3\ prior to
implementation of the annual PM2.5 NAAQS in 2005. In
addition, following implementation of the annual PM2.5
NAAQS, annual PM2.5 concentrations decreased by 1.59 [mu]g/
m\3\ (95% CI: 1.39, 1.80) which corresponded to a reduction in
mortality rates among individuals 65 years and older (0.93% [95% CI:
0.10%, 1.77%]) in non-attainment counties relative to attainment
counties. In a life expectancy study, Bennett et al. (2019) reports
increases in life expectancy in all but 14 counties (1325 of 1339
counties) that have exhibited reductions in PM2.5
concentrations from 1999 to 2015. These studies provide support for
improvements in public health following the implementation of policies,
including in areas with PM2.5 concentrations below the level
of the current annual standard, as well as increases in life expectancy
in areas with reductions in PM2.5 concentrations.
d. Uncertainties in the Health Effects Evidence
The 2022 PA recognizes that there are a number of uncertainties and
limitations associated with the available health effects evidence.
Although the epidemiologic studies clearly demonstrate associations
between long- and short-term PM2.5 exposures and health
outcomes, several uncertainties and limitations in the health effects
evidence remain. Epidemiologic studies evaluating short-term
PM2.5 exposure and health effects have reported
heterogeneity in associations between cities and geographic regions
within the U.S. Heterogeneity in the associations observed across
epidemiologic studies may be due in part to exposure error related to
measurement-related issues, the use of central fixed-site monitors to
represent population exposure to PM2.5, and a limited
understanding of factors including exposure error related to
measurement-related issues, variability in PM2.5 composition
regionally, and factors that result in differential exposures (e.g.,
topography, the built environment, housing characteristics, personal
activity patterns). Heterogeneity is expected when the methods or the
underlying distribution of covariates vary across studies (U.S. EPA,
2019a, p. 6-221). Studies assessed in the 2019 ISA and ISA Supplement
have advanced the state of exposure science by presenting innovative
methodologies to estimate PM exposure, detailing new and existing
measurement and modeling methods, and further informing our
understanding of the influence of exposure measurement error due to
exposure estimation methods on the associations between
PM2.5 and health effects reported in epidemiologic studies
(U.S. EPA, 2019a, section 1.2.2; U.S. EPA, 2022a). Data from
PM2.5 monitors continue to be commonly used in health
studies as a surrogate for PM2.5 exposure, and often provide
a reasonable representation of exposures throughout a study area (U.S.
EPA, 2019a, section 3.4.2.2; U.S. EPA, 2022a, section 3.2.2.2.2).
However, an increasing number of studies employ hybrid modeling methods
to estimate PM2.5 exposure using data from several sources,
often including satellites and models, in addition to ground-based
monitors. These hybrid models typically have good cross-validation,
especially for PM2.5, and have the potential to reduce
exposure measurement error and uncertainty in the health effect
estimates from epidemiologic models of long-term exposure (U.S. EPA,
2019a, section 3.5; U.S. EPA, 2022a, section 2.3.3).
While studies using hybrid modeling methods have reduced exposure
measurement error and uncertainty in the health effect estimates, these
studies use a variety of approaches to estimate PM2.5
concentrations and to assign exposure to assess the association between
health outcomes and PM2.5 exposure. This variability in
methodology has inherent limitations and uncertainties, as described in
more detail in section 2.3.3.1.5 of the 2022 PA, and the performance of
the modeling approaches depends on the availability of monitoring data
which varies by location. Factors that likely contribute to poorer
model performance often coincide with relatively low ambient
PM2.5 concentrations, in areas where predicted exposures are
at a greater distance to monitors, and under conditions where the
reliability and availability of key datasets (e.g., air quality
modeling) are limited. Thus, uncertainty in hybrid model predictions
becomes an increasingly important consideration as lower predicted
concentrations are considered.
Regardless of whether a study uses monitoring data or a hybrid
modeling approach when estimating PM2.5 exposures, one key
limitation that persists is associated with the interpretation of the
study-reported mean PM2.5 concentrations and how they
compare to design values, the metric that describes the air quality
status of a given area relative to the NAAQS.\86\ As discussed above in
section II.B.3.b, the overall mean PM2.5 concentrations
reported by key epidemiologic studies reflect averaging of short- or
long-term PM2.5 exposure estimates across location (i.e.,
across multiple monitors or across modeled grid cells) and over time
(i.e., over several years). For monitor-based studies, the comparison
is somewhat more straightforward than for studies that use hybrid
modeling methods, as the monitors used to estimate exposure in the
epidemiologic studies are generally the same monitors that are used to
calculate design values for a given area. It is expected that areas
meeting a PM2.5 standard with a particular level would be
expected to have average PM2.5 concentrations (i.e.,
averaged across space and over time in the area) somewhat below that
standard level., but the difference between the maximum annual design
value and
[[Page 16249]]
average concentration in an area can be smaller or larger than analyses
presented above in section I.D.5.a, likely depending on factors such as
the number of monitors, monitor siting characteristics, and the
distribution of ambient PM2.5 concentrations. For studies
that use hybrid modeling methods to estimate PM2.5
concentrations, the comparison between study-reported mean
PM2.5 concentrations and design values is more complicated
given the variability in the modeling methods, temporal scales (i.e.,
daily versus annual), and spatial scales (i.e., nationwide versus
urban) across studies. Analyses above in section I.D.5.b and detailed
more in the 2022 PA (U.S. EPA, 2022b, section 2.3.3.2.4) present a
comparison between two hybrid modeling surfaces, which explored the
impact of these factors on the resulting mean PM2.5
concentrations and provided additional information about the
relationship between mean concentrations from studies using hybrid
modeling methods and design values. However, the results of those
analyses only reflect two surfaces and two types of approaches, so
uncertainty remains in understanding the relationship between estimated
modeled PM2.5 concentrations and design values more broadly
across hybrid modeling studies. Moreover, this analysis was completed
using two hybrid modeling methods that estimate PM2.5
concentrations in the U.S., thus an additional uncertainty includes
understanding the relationship between modeled PM2.5
concentrations and design values reported in Canada.
---------------------------------------------------------------------------
\86\ For the annual PM2.5 standard, design values are
calculated as the annual arithmetic mean PM2.5
concentration, averaged over 3 years. For the 24-hour standard,
design values are calculated as the 98th percentile of the annual
distribution of 24-hour PM2.5 concentrations, averaged
over three years (Appendix N of 40 CFR part 50).
---------------------------------------------------------------------------
In addition, where PM2.5 and other pollutants (e.g.,
ozone, nitrogen dioxide, and carbon monoxide) are correlated, it can be
difficult to distinguish whether attenuation of effects in some studies
results from copollutant confounding or collinearity with other
pollutants in the ambient mixture (U.S. EPA, 2019a, section 1.5.1; U.S.
EPA, 2022a, section 2.2.1). Studies evaluated in the 2019 ISA and ISA
Supplement further examined the potential confounding effects of both
gaseous and particulate copollutants on the relationship between long-
and short-term PM2.5 exposure and health effects. As noted
in the Appendix to the 2019 ISA (U.S. EPA, 2019a, Table A-1),
copollutant models are not without their limitations, such as instances
for which correlations are high between pollutants resulting in greater
copollutant confounding bias in results. However, the studies continue
to provide evidence indicating that associations with PM2.5
are relatively unchanged in copollutants models (U.S. EPA, 2019a,
section 1.5.1; U.S. EPA, 2022a, section 2.2.1).
Another area of uncertainty is associated with other potential
confounders, beyond copollutants. Some studies have expanded the
examination of potential confounders to not only include copollutants,
but also systematic evaluations of the potential impact of inadequate
control from long-term temporal trends and weather (U.S. EPA, 2019a,
section 11.1.5.1). Analyses examining these covariates further confirm
that the relationship between PM2.5 exposure and mortality
is unlikely to be biased by these factors. Other studies have explored
the use of alternative methods for confounder control to more
extensively account for confounders and are more robust to model
misspecification that can further inform the causality determination
for long-term and short-term PM2.5 and mortality and
cardiovascular effects (U.S. EPA, 2019a, section 11.2.2.4; U.S. EPA,
2022a, sections 3.1.1.3, 3.1.2.3, 3.2.1.2, and 3.2.2.3). These studies
indicate that bias from unmeasured confounders can occur in either
direction, although controlling for these confounders did not result in
the elimination of the association, but instead provided additional
support for associations between long-term PM2.5 exposure
and mortality when accounting for additional confounders (U.S. EPA,
2022a, section 3.2.2.2.6).
Another important limitation associated with the evidence is that,
while epidemiologic studies indicate associations between
PM2.5 and health effects, the currently available evidence
does not identify particular PM2.5 concentrations that do
not elicit health effects. Rather, health effects can occur over the
entire distribution of ambient PM2.5 concentrations
evaluated, and epidemiologic studies conducted to date do not identify
a population-level threshold below which it can be concluded with
confidence that PM2.5-related effects do not occur.
Overall, evidence assessed in the 2019 ISA and ISA Supplement
continues to indicate a linear, no-threshold C-R relationship for
PM2.5 concentrations >8 [mu]g/m\3\. However, uncertainties
remain about the shape of the C-R curve at PM2.5
concentrations <8 [mu]g/m\3\, with some recent studies providing
evidence for either a sublinear, linear, or supralinear relationship at
these lower concentrations (U.S. EPA, 2019a, section 11.2.4; U.S. EPA,
2022a, section 2.2.3.2).
There are also a number of uncertainties and limitations associated
with the experimental evidence (i.e., controlled human exposure studies
and animal toxicological studies). With respect to controlled human
exposure studies, the PA recognizes that these studies include a small
number of individuals compared to epidemiologic studies. Additionally,
these studies tend to include generally healthy adult individuals, who
are at a lower risk of experiencing health effects. These studies,
therefore, often do not include populations that are at increased risk
of PM2.5-related health effects, including children, older
adults, or individuals with pre-existing conditions. As such, these
studies are somewhat limited in their ability to inform at what
concentrations effects may be elicited in at-risk populations. With
respect to animal toxicological studies, while these studies often
examine more severe health outcomes and longer exposure durations and
higher exposure concentrations than controlled human exposure studies,
there is uncertainty in extrapolating the effects seen in animals, and
the PM2.5 exposures and doses that cause those effects, to
human populations.
Consideration of health effects are informed by the epidemiologic,
controlled human exposure, and animal toxicological studies. The
evaluation and integration of the scientific evidence in the ISA
focuses on evaluating the findings from the body of evidence across
disciplines, including evaluating the strengths and weaknesses in the
overall collection of studies across disciplines. Integrating evidence
across disciplines can strengthen causal inference, such that a weak
inference from one line of evidence can be addressed by other lines of
evidence, and coherence of these lines of evidence can add support to a
cause-effect interpretation of the association. Evaluation and
integration of the evidence also includes consideration of
uncertainties that are inherent in the scientific findings (U.S. EPA,
2015, pp. 13-15), some of which are described above.
3. Summary of Exposure and Risk Estimates
Beyond the consideration of the scientific evidence, discussed
above in section II.B, the EPA also considers the extent to which new
or updated quantitative analyses of PM2.5 air quality,
exposure, or health risks could inform conclusions on the adequacy of
the public health protection provided by the current primary
PM2.5 standards. Additionally, the 2022 PA includes an at-
risk analysis that assesses PM2.5-attributable risk
associated with PM2.5 air quality that has been adjusted to
simulate air quality scenarios of policy
[[Page 16250]]
interest (e.g., ``just meeting'' the current or potential alternative
standards). Drawing on the summary in section II.C of the proposal, the
sections below provide a brief overview of key aspects of the
assessment design (II.A.3.a), key limitations and uncertainties
(II.A.3.b), and exposure/risk estimates (II.A.3.c).
a. Key Design Aspects
Risk assessments combine data from multiple sources and involve
various assumptions and uncertainties. Input data for these analyses
includes C-R functions from epidemiologic studies for each health
outcome and ambient annual or 24-hour PM2.5 concentrations
for the study areas utilized in the risk assessment (U.S. EPA, 2022b,
section 3.4.1). Additionally, quantitative and qualitative methods were
used to characterize variability and uncertainty in the risk estimates
(U.S. EPA, 2022b, section 3.4.1.7).
Concentration-response functions used in the risk assessment are
from large, multicity U.S. epidemiologic studies that evaluate the
relationship between PM2.5 exposures and mortality.
Epidemiologic studies and concentration-response studies that were used
in the risk assessment to estimate risk were identified using criteria
that take into account factors such as study design, geographic
coverage, demographic populations, and health endpoints (U.S. EPA,
2022b, section 3.4.1.1).\87\ The risk assessment focuses on all-cause
or nonaccidental mortality associated with long-term and short-term
PM2.5 exposures, for which the 2019 ISA concluded that the
evidence provides support for a ``causal relationship'' (U.S. EPA,
2022b, section 3.4.1.2).\88\
---------------------------------------------------------------------------
\87\ Additional detail regarding the selection of epidemiologic
studies and specification of C-R functions is provided in the 2022
PA (U.S. EPA, 2022b, Appendix C, section C.1.1).
\88\ While the 2019 ISA also found that evidence supports the
determination of a ``causal relationship'' between long- and short-
term PM2.5 exposures and cardiovascular effects,
cardiovascular mortality was not included as a health outcome as it
will be captured in the estimates of all-cause mortality.
---------------------------------------------------------------------------
As described in more detail in the 2022 PA, the risk assessment
first estimated health risks associated with air quality for 2015
adjusted to simulate ``just meeting'' the current primary
PM2.5 standards (i.e., the annual standard with its level of
12.0 [micro]g/m\3\ and the 24-hour standard with its level of 35
[micro]g/m\3\). Air quality modeling was then used to simulate air
quality just meeting an alternative standard with a level of 10.0
[micro]g/m\3\ (annual) and 30 [micro]g/m\3\ (24-hour). In addition to
the model-based approach, for the subset of 30 areas controlled by the
annual standard linear interpolation and extrapolation were employed to
simulate just meeting alternative annual standards with levels of 11.0
(interpolated between 12.0 and 10.0 [mu]g/m\3\), 9.0 [mu]g/m\3\, and
8.0 [mu]g/m\3\ (both extrapolated from 12.0 and 10.0 [mu]g/m\3\) (U.S.
EPA, 2022b, section 3.4.1.3). The 2022 PA notes that there is greater
uncertainty regarding whether a revised 24-hour standard (i.e., with a
lower level) is needed to further limit ``peak'' PM2.5
concentration exposure and whether a lower 24-hour standard level would
most effectively reduce PM2.5-associated health risks
associated with ``typical'' daily exposures. The risk assessment
estimates health risks associated with air quality adjusted to meet a
revised 24-hour standard with a level of 30 [micro]g/m\3\, in
conjunction with estimating the health risks associated with meeting a
revised annual standard with a level of 10.0 [micro]g/m\3\ (U.S. EPA,
2022b, section 3.4.1.3). More details on the air quality adjustment
approaches used in the risk assessment are described in section 3.4.1.4
and Appendix C of the 2022 PA (U.S. EPA, 2022b).
When selecting U.S. study areas for inclusion in the risk
assessment, the available ambient monitors, geographic diversity, and
ambient PM2.5 air quality concentrations were taken into
consideration (U.S. EPA, 2022b, section 3.4.1.4). When these factors
were applied, 47 urban study areas were identified, which include
nearly 60 million people aged 30-99, or approximately 30% of the U.S
population in this age range (U.S. EPA, 2022b, section 3.4.1.5,
Appendix C, section C.1.3). Of the 47 study areas, there were 30 study
areas where just meeting the current standards is controlled by the
annual standard,\89\ 11 study areas where just meeting the current
standards is controlled by the daily standard,\90\ and 6 study areas
where the controlling standard differed depending on the air quality
adjustment approach (U.S. EPA, 2022b, section 3.4.1.5).\91\
---------------------------------------------------------------------------
\89\ For these areas, the annual standard is the ``controlling
standard'' because when air quality is adjusted to simulate just
meeting the current or potential alternative annual standards, that
air quality also would meet the 24-hour standard being evaluated.
\90\ For these areas, the 24-hour standard is the controlling
standard because when air quality is adjusted to simulate just
meeting the current or potential alternative 24-hour standards, that
air quality also would meet the annual standard being evaluated.
Some areas classified as being controlled by the 24-hour standard
also violate the annual standard.
\91\ In these 6 areas, the controlling standard depended on the
air quality adjustment method used and/or the standard scenarios
evaluated.
---------------------------------------------------------------------------
In addition to the overall risk assessment, the 2022 PA also
includes an at-risk analysis and estimates exposures and health risks
of specific populations identified as at-risk that would be allowed
under the current and potential alternative standards to further inform
the Administrator's conclusions regarding the adequacy of the public
health protection provided by the current primary PM2.5
standards. In so doing, the 2022 PA evaluates exposure and
PM2.5 mortality risk for older adults (e.g., 65 years and
older), stratified for White, Black, Asian, Native American, Non-
Hispanic, and Hispanic individuals residing in the same study areas
included in the overall risk assessment. This analysis utilizes a
recent epidemiologic study that provides race- and ethnicity-specific
risk coefficients (Di et al., 2017b).
b. Key Limitations and Uncertainties
Uncertainty in risk estimates (e.g., in the size of risk estimates)
can result from a number of factors, including the assumptions about
the shape of the C-R function with mortality at low ambient PM
concentrations, the potential for confounding and/or exposure
measurement error in the underlying epidemiologic studies, and the
methods used to adjust PM2.5 air quality. More specifically,
the use of air quality modeling to adjust PM2.5
concentrations are limited as they rely on model predictions, are based
on emission changes scaled by fixed percentages, and use only two of
the full set of possible emission scenarios and linear interpolation/
extrapolation to adjust air quality that may not fully capture
potential non-linearities associated with real-world changes in air
quality. Additionally, the selection of case study areas is limited to
urban areas predominantly located CA and in the Eastern U.S. that are
controlled by the annual standard. While the risk assessment does not
report quantitative uncertainty in the risk estimates as exposure
concentrations are reduced, it does provide information on the
distribution of concentrations associated with the risk estimates when
evaluating progressively lower alternative annual standards. Based on
these data, as lower alternative annual standards are evaluated, larger
proportions of the distributions in risk occur at or below 10 [mu]g/
m\3\ (at concentrations below or near most of the study-reported means
from the key U.S. epidemiologic studies) and at or below 8 [mu]g/m\3\
(the concentration at which the ISA reports increasing uncertainty in
the shape of the C-R curve based on the body of epidemiologic
evidence).
[[Page 16251]]
Similarly, the at-risk analysis is also subject to many of these
same uncertainties noted above. Additionally, the at-risk analysis
included C-R functions from only one study (Di et al., 2017b), which
reported associations between long-term PM2.5 exposures and
mortality, stratified by race/ethnicity, in populations age 65 and
older, as opposed to the multiple studies used in the overall risk
assessment to convey risk estimate variability. These and other sources
of uncertainty in the overall risk assessment and the at-risk analyses
are characterized in more depth in the 2022 PA (U.S. EPA, 2022b,
section 3.4.1.7, section 3.4.1.8, Appendix C, section C.3).
c. Summary of Risk Estimates
Although limitations in the underlying data and approaches lead to
some uncertainty regarding estimates of PM2.5-associated
risk, the risk assessment estimates that the current primary
PM2.5 standards could allow a substantial number of
PM2.5-associated deaths in the U.S. For example, when air
quality in the 47 study areas is adjusted to simulate just meeting the
current standards, the risk assessment estimates up to 45,100 deaths in
2015 are attributable to long-term PM2.5 exposures
associated with just meeting the current annual and 24-hour
PM2.5 standards (U.S. EPA, 2022b, section 3.4.2.1).
Additionally, as described in more detail in the 2022 PA, the at-risk
analysis suggests that a lower annual standard level (i.e., below 12
[micro]g/m\3\ and down as low as 8 [micro]g/m\3\) will help to reduce
PM2.5 exposure and may also help to mitigate exposure and
risk disparities in populations identified as particularly at-risk for
adverse effects from PM exposures (i.e., minority populations).
Compared to the current annual standard, meeting a revised annual
standard with a lower level is estimated to reduce PM2.5-
associated health risks in the 30 study areas controlled by the annual
standard by about 7-9% for a level of 11.0 [micro]g/m\3\, 15-19% for a
level of 10.0 [micro]g/m\3\, 22-28% for a level of 9.0 [micro]g/m\3\,
and 30-37% for a level of 8.0 [micro]g/m\3\) (U.S. EPA, 2022b, Table 3-
17). Meeting a revised annual standard with a lower level may also help
to mitigate exposure and risk disparities in populations identified as
particularly at-risk for adverse effects from PM exposures (i.e.,
minority populations) in simulated scenarios just meeting alternative
annual standards. However, though reduced, disparities by race and
ethnicity persist even at an alternative annual standard level of 8
[micro]g/m\3\, the lowest alternative annual standard included in the
risk assessment (U.S. EPA, 2022b, section 3.4.2.4).
Revising the level of the 24-hour standard to 30 [mu]g/m\3\ is
estimated to lower PM2.5-associated risks across a more
limited population and number of areas than revising the annual
standard (U.S. EPA, 2022, section 3.4.2.4). Risk reduction predictions
are largely confined to areas located in the western U.S., several of
which are also likely to experience risk reductions upon meeting a
revised annual standard. In the 11 areas controlled by the 24-hour
standard, when air quality is simulated to just meet the current 24-
hour standard, PM2.5 exposures are estimated to be
associated with as many as 2,570 deaths annual. Compared to just
meeting the current standard, air quality just meeting an alternative
24-hour standard level of 30 [micro]g/m\3\ is associated with
reductions in estimated risk of 9-13% (U.S. EPA, 2022b, section
3.4.2.3).
B. Conclusions on the Primary PM2.5 Standards
In drawing conclusions on the adequacy of the current primary
PM2.5 standards, in view of the advances in scientific
knowledge and additional information now available, the Administrator
has considered the evidence base, information, and policy judgments
that were the foundation of the 2012 and 2020 reviews and reflects upon
the body of evidence and information newly available in this
reconsideration. In so doing, the Administrator has taken into account
both evidence-based and risk-based considerations, as well as advice
from the CASAC and public comments. Evidence-based considerations draw
upon the EPA's integrated assessment of the scientific evidence of
health effects related to PM2.5 exposure presented in the
2019 ISA and ISA Supplement (summarized in the proposal in sections
II.B (88 FR 5580, January 27, 2023) and II.D.2.a (88 FR 5609, January
27, 2023), and also in section II.A.2 above) to address key policy-
relevant questions in the reconsideration. Similarly, the risk-based
considerations draw upon the assessment of population exposure and risk
(summarized in the proposal in sections II.C (88 FR 5605, January 27,
2023) and II.D.2.b (88 FR 5614, January 27, 2023), and also in section
II.A.3 above) in addressing policy-relevant questions focused on the
potential for PM2.5 exposures associated with mortality
under air quality conditions just meeting the current and potential
alternative standards.
The approach to reviewing the primary standards is consistent with
requirements of the provisions of the CAA related to the review of the
NAAQS and with how the EPA and the courts have historically interpreted
the CAA. As discussed in section I.A above, these provisions require
the Administrator to establish primary standards that, in the
Administrator's judgment, are requisite (i.e., neither more nor less
stringent than necessary) to protect public health with an adequate
margin of safety. Consistent with the Agency's approach across all
NAAQS reviews, the EPA's approach to informing these judgments is based
on a recognition that the available health effects evidence generally
reflects a continuum that includes ambient air exposures for which
scientists generally agree that health effects are likely to occur
through lower levels at which the likelihood and magnitude of response
become increasingly uncertain. The CAA does not require the
Administrator to establish a primary standard at a zero-risk level or
at background concentration levels, but rather at a level that reduces
risk sufficiently so as to protect public health, including the health
of sensitive groups, with an adequate margin of safety.
The decisions on the adequacy of the current primary
PM2.5 standards described below is a public health policy
judgment by the Administrator that draws on the scientific evidence for
health effects, quantitative analyses of population exposures and/or
health risks, and judgments about how to consider the uncertainties and
limitations that are inherent in the scientific evidence and
quantitative analyses. The four basic elements of the NAAQS (i.e.,
indicator, averaging time, form, and level) have been considered
collectively in evaluating the public health protection afforded by the
current standards.
Section II.B.2 below briefly summarizes the basis for the
Administrator's proposed decision, drawing from section II.D.3 of the
proposal (88 FR 5617, January 27, 2023). The advice and recommendations
of the CASAC and public comments on the proposed decision are addressed
below in sections II.B.1 and II.B.3, respectively. The Administrator's
final conclusions in this reconsideration regarding the adequacy of the
current primary PM2.5 standards and whether any revisions
are appropriate are described in section II.B.4.
1. CASAC Advice
As part of its review of the 2019 draft PA, the CASAC provided
advice on the adequacy of the public health protection afforded by the
current primary PM2.5 standards. Its advice is documented in
[[Page 16252]]
a letter sent to the EPA Administrator (Cox, 2019b). In this letter,
the committee recommended retaining the current 24-hour
PM2.5 standard but did not reach consensus on whether the
scientific and technical information support retaining or revising the
current annual standard. In particular, though the CASAC agreed that
there is a long-standing body of health evidence supporting
relationships between PM2.5 exposures and various health
outcomes, including mortality and serious morbidity effects, individual
CASAC members ``differ[ed] in their assessments of the causal and
policy significance of these associations'' (Cox, 2019b, p. 8 of
consensus responses). Drawing from this evidence, ``some CASAC
members'' expressed support for retaining the current annual standard
while ``other members'' expressed support for revising that standard in
order to increase public health protection (Cox, 2019b, p.1 of letter).
These views are summarized below.
The CASAC members who supported retaining the current annual
standard expressed the view that substantial uncertainty remains in the
evidence for associations between PM2.5 exposures and
mortality or serious morbidity effects. These committee members
asserted that ``such associations can reasonably be explained in light
of uncontrolled confounding and other potential sources of error and
bias'' (Cox, 2019b, p. 8 of consensus responses). They noted that
associations do not necessarily reflect causal effects, and they
contended that recent epidemiologic studies assessed in the 2019 ISA
that report positive associations at lower estimated exposure
concentrations mainly confirm what was anticipated or already assumed
in setting the 2012 NAAQS. In particular, they concluded that such
studies have some of the same limitations as prior studies and do not
provide new information calling into question the existing standard.
They further asserted that ``accountability studies provide potentially
crucial information about whether and how much decreasing
PM2.5 causes decreases in future health effects'' (Cox,
2019b, p. 10 of consensus responses), and they cited recent reviews
(i.e., Henneman et al., 2017; Burns et al., 2019) to support their
position that in such studies, ``reductions of PM2.5
concentrations have not clearly reduced mortality risks'' (Cox, 2019b,
p. 8 of consensus responses). Thus, the committee members who supported
retaining the current annual standard advise that, ``while the data on
associations should certainly be carefully considered, this data should
not be interpreted more strongly than warranted based on its
methodological limitations'' (Cox, 2019b, p. 8 of consensus responses).
These members of the CASAC further concluded that the quantitative
risk assessment included in the 2019 draft PA does not provide a valid
basis for revising the current standards. This conclusion was based on
concerns that (1) ``the risk assessment treats regression coefficients
as causal coefficients with no justification or validation provided for
this decision;'' (2) the estimated regression concentration-response
functions ``have not been adequately adjusted to correct for
confounding, errors in exposure estimates and other covariates, model
uncertainty, and heterogeneity in individual biological (causal)
[concentration-response] functions;'' (3) the estimated concentration-
response functions ``do not contain quantitative uncertainty bands that
reflect model uncertainty or effects of exposure and covariate
estimation errors;'' and (4) ``no regression diagnostics are provided
justifying the use of proportional hazards . . . and other modeling
assumptions'' (Cox, 2019b, p. 9 of consensus responses). These
committee members also contended that details regarding the derivation
of concentration-response functions, including specification of the
beta values and functional forms, were not well-documented, hampering
the ability of readers to evaluate these design details. Thus, these
members ``think that the risk characterization does not provide useful
information about whether the current standard is protective'' (Cox,
2019b, p. 11 of consensus responses).
Drawing from their evaluation of the evidence and the risk
assessment in the 2019 draft PA, these committee members concluded that
``the Draft PM PA does not establish that new scientific evidence and
data reasonably call into question the public health protection
afforded by the . . . 2012 PM2.5 annual standard'' (Cox,
2019b, p.1 of letter).
In contrast, ``[o]ther members of CASAC conclude[d] that the weight
of the evidence, particularly reflecting recent epidemiology studies
showing positive associations between PM2.5 and health
effects at estimated annual average PM2.5 concentrations
below the current standard, does reasonably call into question the
adequacy of the 2012 annual PM2.5 [standard] to protect
public health with an adequate margin of safety'' (Cox, 2019b, p.1 of
letter). The committee members who supported this conclusion noted that
the body of health evidence for PM2.5 not only includes the
repeated demonstration of associations in epidemiologic studies, but
also includes support for biological plausibility established by
controlled human exposure and animal toxicology studies. They pointed
to recent studies demonstrating that the associations between
PM2.5 and health effects occur in a diversity of locations,
in different time periods, with different populations, and using
different exposure estimation and statistical methods. They concluded
that ``the entire body of evidence for PM health effects justifies the
causality determinations made in the Draft PM ISA'' (Cox, 2019b, p. 8
of consensus responses).
The members of the CASAC who supported revising the current annual
standard particularly emphasized recent findings of associations with
PM2.5 in areas with average long-term PM2.5
concentrations below the level of the annual standard and studies that
show positive associations even when estimated exposures above 12
[mu]g/m\3\ are excluded from analyses. They found it ``highly
unlikely'' that the extensive body of evidence indicating positive
associations at low estimated exposures could be fully explained by
confounding or by other non-causal explanations (Cox, 2019b, p. 8 of
consensus responses). They additionally concluded that ``the risk
characterization does provide a useful attempt to understand the
potential impacts of alternate standards on public health risks'' (Cox,
2019b, p. 11 of consensus responses). These CASAC members concluded
that the available evidence reasonably calls into question the
protection provided by the current primary PM2.5 standards
and supports revising the annual standard to increase that protection
(Cox, 2019b).
As a part of this reconsideration, the CASAC reviewed the 2021
draft PA (developed to support the reconsideration as described in
section I.C.5 above). As a part of their review of the 2021 draft PA,
the CASAC provided advice on the adequacy of the current primary
PM2.5 standards. The range of views summarized here
generally reflects differing judgments as to the relative weight to
place on various types of evidence, the risk-based information, and the
associated uncertainties, as well as differing judgments about the
importance of various PM2.5-related health effects from a
public health perspective.
In its comments on the 2021 draft PA, the CASAC stated that:
``[o]verall the CASAC finds the Draft PA to be well-
[[Page 16253]]
written and appropriate for helping to `bridge the gap' between the
agency's scientific assessments and quantitative technical analyses,
and the judgments required of the Administrator in determining whether
it is appropriate to retain or revise the National Ambient Air Quality
Standards (NAAQS)'' (Sheppard, 2022a, p. 1 of consensus letter). The
CASAC also stated that the ``[d]raft PA adequately captures and
appropriately characterizes the key aspects of the evidence assessed
and integrated in the 2019 ISA and Draft ISA Supplement of
PM2.5-related health effects'' (Sheppard, 2022b, p. 2 of
consensus letter). The CASAC also stated that ``[t]he interpretation of
the risk assessment for the purpose of evaluating the adequacy of the
current primary PM2.5 annual standard is appropriate given
the scientific findings presented'' (Sheppard, 2022a, p. 2 of consensus
letter).
With regard to the adequacy of the current primary annual
PM2.5 standard, ``all CASAC members agree that the current
level of the annual standard is not sufficiently protective of public
health and should be lowered'' (Sheppard, 2022a, p. 2 of consensus
letter). Additionally, ``the CASAC reached consensus that the
indicator, form, and averaging time should be retained, without
revision'' (Sheppard, 2022a, p. 2 of consensus letter). With regard to
the level of the primary annual PM2.5 standard, the CASAC
had differing recommendations for the appropriate range for an
alternative level. The majority of the CASAC ``judge[d] that an annual
average in the range of 8-10 [mu]g/m\3\'' was most appropriate, while
the minority of the CASAC members stated that ``the range of the
alternative standard of 10-11 [mu]g/m\3\ is more appropriate''
(Sheppard, 2022a, p. 16 of consensus responses). The CASAC did
highlight, however, that ``the alternative standard level of 10 [mu]g/
m\3\ is within the range of acceptable alternative standards
recommended by all CASAC members, and that an annual standard below 12
[mu]g/m\3\ is supported by a larger and coherent body of evidence''
(Sheppard, 2022a, p. 16 of consensus responses).
In reaching conclusions on a recommended range of 8-10 [mu]g/m\3\
for the primary annual PM2.5 standard, the majority of the
CASAC placed weight on various aspects of the available scientific
evidence and quantitative risk assessment information discussed in the
2021 draft PA (Sheppard, 2022a, p. 16 of consensus responses). In
particular, these members cited recent U.S.- and Canadian-based
epidemiologic studies that show positive associations between
PM2.5 exposure and mortality with study-reported mean
concentrations below 10 [mu]g/m\3\. Further, these members also noted
that the lower portions of the air quality distribution (i.e.,
concentrations below the mean) provide additional information to
support associations between health effects and PM2.5
concentrations lower than the reported long-term mean concentration. In
addition, the CASAC members recognized that the available evidence has
not identified a threshold concentration, below which an association no
longer remains, pointing to the conclusion in the draft ISA Supplement
that the ``evidence remains clear and consistent in supporting a no-
threshold relationship, and in supporting a linear relationship for
PM2.5 concentrations >8 [mu]g/m\3\'' (Sheppard, 2022a, p. 16
of consensus responses). Finally, these CASAC members placed weight on
the at-risk analysis as providing support for protection of at-risk
demographic groups, including minority populations.
In recommending a range of 10-11 [mu]g/m\3\ for the primary annual
PM2.5 standard, the minority of the CASAC emphasized that
there were few key epidemiologic studies that reported positive and
statistically significant health effects associations for
PM2.5 air quality distributions with overall mean
concentrations below 9.6 [mu]g/m\3\ (Sheppard, 2022a, p. 17 of
consensus responses). In so doing, the minority of the CASAC
specifically noted the variability in the relationship between study-
reported means and area annual design values based on the methods
utilized in the studies, noting that design values are generally higher
than area average exposure levels. Further, the minority of the CASAC
stated that ``uncertainties related to copollutants and confounders
make it difficult to justify a recommendation below 10-11 [mu]g/m\3\''
(Sheppard, 2022a, p. 17 of consensus responses). Finally, the minority
of the CASAC placed less weight on the risk assessment results, noting
large uncertainties, including the approaches used for adjusting air
quality to simulate just meeting the current and alternative standards.
With regard to the current primary 24-hour PM2.5
standard, in their review of the 2021 draft PA, the CASAC did not reach
consensus regarding the adequacy of the public health protection
provided by the current standard. As described further below, the
majority of the CASAC members concluded ``that the available evidence
calls into question the adequacy of the current 24-hour standard''
(Sheppard, 2022a, p. 3 of consensus letter), while the minority of the
CASAC members agreed with ``the EPA's preliminary conclusion [in the
draft PA] to retain the current 24-hour PM2.5 standard
without revision'' (Sheppard, 2022a, p. 4 of consensus letter). The
CASAC recommended that in future reviews, the EPA should also consider
alternative forms for the primary 24-hour PM2.5 standard.
Specifically, the CASAC ``suggests considering a rolling 24-hour
average and examining alternatives to the 98th percentile of the 3-year
average,'' pointing to concerns that computing 24-hour average
PM2.5 concentrations using the current midnight-to-midnight
timeframe could potentially underestimate the effects of high 24-hour
exposures, especially in areas with wood-burning stoves and wintertime
stagnation (Sheppard, 2022a, p. 18 of consensus responses).
As noted above, the majority of the CASAC favored revising the
level of the primary 24-hour PM2.5 standard, suggesting that
a range of 25-30 [mu]g/m\3\ would be adequately protective. In so
doing, the majority of the CASAC placed weight on the available
epidemiologic evidence, including epidemiologic studies that restricted
analyses to 24-hour PM2.5 concentrations below 25 [mu]g/
m\3\. These members also placed weight on results of controlled human
exposure studies with exposures close to the current standard, which
they note provide support for the epidemiologic evidence to lower the
standard. These members noted the limitations in using controlled human
exposure studies alone in considering the adequacy of the 24-hour
standard, recognizing that controlled human exposure studies
preferentially recruit less susceptible individuals and have a typical
exposure duration shorter than 24 hours. These members also placed
``greater weight on the scientific evidence than on the values
estimated by the risk assessment,'' citing their concerns that the risk
assessment ``may not adequately capture areas with wintertime
stagnation and residential wood-burning where the annual standard is
less likely to be protective'' (Sheppard, 2022a, p. 17 of consensus
responses). Furthermore, these CASAC members ``also are less confident
that the annual standard could adequately protect against health
effects of short-term exposures'' (Sheppard, 2022a, p. 17 of consensus
responses).
The minority of the CASAC agreed with the EPA's preliminary
conclusion in the 2021 draft PA to retain the current primary 24-hour
PM2.5 standard. In so doing, the minority of the CASAC
placed greater weight on the risk assessment, noting that the risk
[[Page 16254]]
assessment accounts for both the level and the form of the current
standard and the manner by which attainment with the standard is
determined. Further, the minority of the CASAC stated that the ``risk
assessment indicates that the annual standard is the controlling
standard across most of the urban study areas evaluated and revising
the level of the 24-hour standard is estimated to have minimal impact
on the PM2.5-associated risks'' and therefore, ``the annual
standard can be used to limit both long- and short-term
PM2.5 concentrations'' (Sheppard, 2022a, p. 18 of consensus
responses). Further, the minority of the CASAC placed more weight on
the controlled human exposure studies, which show ``effects at
PM2.5 concentrations well above those typically measured in
areas meeting the current standards'' and which suggest that ``the
current standards are providing adequate protection against these
exposures'' (Sheppard, 2022a, p. 18 of consensus responses).
While the CASAC members expressed differing opinions on the
appropriate revisions to the current standards, they did ``find that
both primary standards, 24-hour and annual, are critical to protect
public health given the evidence on detrimental health outcomes at both
short-term and long-term exposures including peak events'' (Sheppard,
2022a, p. 13 of consensus responses). The comments from the CASAC also
took note of uncertainties that remain in this reconsideration of the
primary PM2.5 standards and they identified a number of
additional areas for future research and data gathering and
dissemination that would inform future reviews of the primary
PM2.5 NAAQS (Sheppard, 2022a, pp. 14-15 of consensus
responses).
2. Basis for the Proposed Decision
In reaching his proposed decisions to revise the level of the
primary annual PM2.5 standard from its current level of 12.0
[micro]g/m\3\ to within the range of 9.0 to 10.0 [micro]g/m\3\, and to
retain the current primary 24-hour PM2.5 standard (88 FR
5558, January 27, 2023), the Administrator carefully considered the
assessment of the current evidence and conclusions reached in the 2019
ISA and ISA Supplement; the currently available exposure and risk
information, including associated limitations and uncertainties,
described in detail in the 2022 PA; the considerations and staff
conclusions and associated rationales presented in the 2022 PA; the
advice and recommendations from the CASAC; and public comments that had
been offered up to that point (88 FR 5558, January 27, 2023).
In reaching his proposed conclusions on whether the currently
available scientific evidence and quantitative risk-based information
support or call into question the adequacy of the public health
protection afforded by the current primary PM2.5 standards,
and as is the case with NAAQS reviews in general, the extent to which
the current primary PM2.5 standards are judged to be
adequate will depend on a variety of factors, including science policy
and public health policy judgments to be made by the Administrator on
the strength and uncertainties of the scientific evidence. The factors
relevant to judging the adequacy of the standards also include the
interpretation of, and decisions as to the weight to place on,
different aspects of the results of the risk assessment for the study
areas included and the associated uncertainties. Thus, in reaching
proposed conclusions of the current standards, the Administrator
recognized that such a determination depends in part on judgments
regarding aspects of the evidence and risk estimates, and judgments
about the degree of protection that is requisite to protect public
health with an adequate margin of safety.
The Administrator's full rationale for his proposed conclusions is
presented in section II.D.3 of proposal (88 FR 5658, January 27, 2023),
but is also briefly summarized here. In reaching the proposed decision
to revise the annual standard level to 9-10 [micro]g/m\3\, the
Administrator placed weight on the full body of scientific information.
He noted that the 2019 ISA finds that exposure to PM2.5
causes mortality and cardiovascular effects and is likely to cause
respiratory effects, cancer, and nervous system effects as detailed
further in section II.B.1 of the proposal. As detailed further in
section II.B.4 of the proposal, he additionally noted that the 2019 ISA
identifies at-risk populations at greater risk of health effects from
exposure to PM2.5, including children, older adults, people
with pre-existing respiratory or cardiovascular disease, minority
populations, and low socioeconomic status (SES) populations.
The Administrator also recognized that epidemiologic studies
provide the strongest scientific evidence when evaluating the adequacy
of the level of the annual standard. He noted that there is no specific
point in the air quality distribution of any epidemiologic study that
represents a `bright line' at and above which effects have been
observed and below which effects have not been observed. In his
proposed decision, he noted previous decision-making frameworks, which
placed weight on values at or near the study-reported mean
PM2.5 concentrations, which is where the most confidence in
the reported association of the epidemiologic study exists. He further
noted that there are a number of epidemiologic studies available in
this reconsideration that use new PM2.5 exposure estimation
techniques (e.g., hybrid modeling) that were not used in epidemiologic
studies that were available in previous reviews. These recent
epidemiologic studies that use new exposure estimation techniques
report long-term mean PM2.5 concentrations that are well
below corresponding design values, which is an important consideration
in reaching decisions on the level of the annual PM2.5
standard.
In reaching his proposed decision, the Administrator noted that a
level of 9-10 [micro]g/m\3\ would near or below the reported 25th
percentiles in key U.S. based epidemiologic studies, while also
recognizing that he has less confidence in the magnitude and
significance of the association at even lower percentiles (e.g., 10th
percentile), where even fewer health events are observed. The
Administrator also noted that a proposed level of 9-10 [micro]g/m\3\
would be near the mean PM2.5 reported in Canadian based
studies, though he also recognized that there are a number of factors
associated with the studies in Canada (e.g., exposure environments)
that make it more difficult to compare mean concnetrations from
Canadian studies to design values, which determine compliance with the
standard in the U.S.
The Administrator took note of additional pieces of scientific
evidence, which were not available in previous reviews, including
restricted analyses, which support that the association seen in
epidemiologic studies does not just occur from the peaks of the
exposure distribution. Additionally, he notes that a level of 9-10
[micro]g/m\3\ would be below the starting concentration in newly
available accountability studies, though he did note that it is more
difficult to interpret these studies in the context of selecting the
level of the annual PM2.5 standard.
Further, the Administrator took into consideration the advice of
the CASAC, noting that all members included 10 [micro]g/m\3\ in their
recommended range, and that the proposed range of 9-10 [micro]g/m\3\
for the level of the primary annual PM2.5 standard was
within the range recommended by the majority of the CASAC.
In reaching the proposed conclusion of a range between 9-10
[micro]g/m\3\, the Administrator noted that a level as high
[[Page 16255]]
as 11 [micro]g/m\3\ might not provide an adequate margin of safety,
given that 11 [micro]g/m\3\ was well above many of the epidemiologic
study-reported mean PM2.5 concentrations. Additionally, the
Administrator noted the uncertainties associated with the scientific
and quantitative information supporting a level as low as 8 [micro]g/
m\3\, which call into question the potential public health improvements
of a standard below 9 [micro]g/m\3\. The Administrator specifically
noted the lack of key U.S. studies with mean concentrations below 9.3
[micro]g/m\3\ and he further noted that the risk assessment suggests
that the risk remaining under a standard of 8 [micro]g/m\3\ would occur
at very low concentrations (e.g., mainly 7 [micro]g/m\3\ and below).
As such, the Administrator's proposed decision noted that the
current PM2.5 annual standard did not adequately provide
requisite protection against exposures to PM2.5 and that a
proposed range of 9-10 [micro]g/m\3\ would provide an adequate margin
of safety.
In his proposed decision to retain the current primary 24-hour
PM2.5 standard with a level of 35 [micro]g/m\3\, the
Administrator first considered the scientific information related to
short-term exposures to PM2.5 and health effects. He noted
that the controlled human exposure studies are the strongest line of
evidence for informing his conclusions regarding the adequacy of the
current 24-hour standard. In so doing, the Administrator recognized
that controlled human exposure studies are conducted with healthy adult
volunteers and that these studies do not include individuals who may be
at increased risk of PM2.5-related health effects (i.e.,
children, older adults, people with pre-existing diseases). He also
noted that the effects observed in the controlled human exposure
studies (e.g., changes in vascular function) are not effects that are
judged to be clearly adverse. He recognized the most consistent
evidence of effects in these studies occurs at higher concentrations
(e.g., >120 [micro]g/m\3\) following 1-5 hour exposures, and that one
study observed effects at concentrations as low as 38 [micro]g/m\3\
following 4-hour exposures. However, the Administrator reiterated that
these studies do not tell us at exactly what concentrations an adverse
effect might occur, especially for at-risk populations. As noted above
in section II.A.2.c, controlled human exposure studies tend to include
generally healthy adult individuals who are at a lower risk of
experiencing health effects, and often do not include at-risk
populations (e.g., children, older adults, or individuals with pre-
existing conditions). As such, the Administrator recognized that these
studies are somewhat limited in their ability to inform at what
concentrations effects may be elicited in in at-risk populations. The
Administrator also considered air quality analyses in the 2022 PA that
demonstrate that there will be very few, if any, days with
PM2.5 concentrations at levels evaluated in controlled human
exposure studies that are associated with effects in areas that meet
the current primary 24-hour PM2.5 standard.
The Administrator also noted that as, in previous PM NAAQS reviews,
the protection provided by the suite of standards (e.g., annual and 24-
hour standards) is evaluated together. He noted that the annual
standard is the controlling standard in most areas of the country. He
also considered air quality analyses in the 2022 PA that suggest that
revision of the annual standard to a level between 9-10 [micro]g/m\3\
would also control 24-hour PM2.5 concentrations in most
areas to, or below, 30 [micro]g/m\3\. Finally, the Administrator noted
the agreement with the advice from the minority of CASAC and
additionally noted the limited rationale and evidence provided by the
majority CASAC's recommendation to support revision of the 24-hour
standard. As such, the Administrator proposed to retain the current 24-
hour standard with its level of 35 [micro]g/m\3\.
Additionally, the Administrator proposed to conclude that it is
appropriate to retain all other elements (i.e., indicator, averaging
time, and form) of the annual and 24-hour standards.
3. Comments on the Proposed Decision
With respect to the adequacy of the primary annual PM2.5
standard, a number of commenters, primarily those from industry and
industry groups, non-governmental organizations, and some State and
local governments, disagree with the EPA's proposed decision to revise
the level of the primary annual PM2.5 standard. These
commenters generally expressed the view that the current standards
provide the requisite degree of public health protection and should be
retained, consistent with the 2020 final decision. In supporting their
view, these commenters assert that the scientific evidence available in
this reconsideration is essentially unchanged since the 2020 final
decision and that the additional scientific evidence and quantitative
risk information available for the reconsideration does not support
strengthening the primary annual PM2.5 standard. These
commenters also assert that uncertainties associated with the available
scientific evidence have not changed since the 2020 final decision, and
they note that these uncertainties were essential factors in the then-
Administrator's decision to retain the primary annual PM2.5
standard. These commenters argue that, while the current Administrator
acknowledges these uncertainties, he does not place enough weight on
them in reaching his conclusions regarding the current standard. The
commenters specifically highlight uncertainties related to exposure
misclassification, confounding, and other sources of potential bias,
which they claim supports retaining the current level of the annual
standard. These commenters also note that these uncertainties were
emphasized by the minority of the CASAC in their review of the 2021
draft PA, and the commenters further suggest that the lack of consensus
from the CASAC on the appropriate level for the primary annual
PM2.5 standard show that the research is unclear. The
commenters contend that there is not support in this reconsideration
for deviating from the then-Administrator's decision in 2020.
In contrast, other commenters, primarily from public health and
environmental organizations, some State and local elected
representatives, and some State and local government agencies agree
with the EPA's proposed decision that the primary annual
PM2.5 standard is not adequate. These commenters support
revising the level of the primary annual PM2.5 standard and
emphasize that the available scientific evidence, in particular
epidemiologic studies, along with the CASAC's advice in their review
for the 2021 draft PA, provide strong support for the proposed
decision. In particular, these commenters agree with the EPA's
conclusions about the strength of the scientific evidence, including
uncertainties, and they emphasize that the CASAC reached consensus in
their review of the 2021 draft PA that the current primary annual
PM2.5 standard is not adequate. Some of these commenters
also note that a revised primary annual PM2.5 standard would
result in significant public health benefits by reducing morbidity and
mortality associated with PM2.5 exposure, especially for at-
risk populations.
The EPA agrees with commenters that the primary annual
PM2.5 standard is not adequate. The EPA recognizes the
longstanding body of health evidence supporting relationships between
PM2.5 exposures (short- and long-term) and both mortality
and serious morbidity effects. The evidence available in this
reconsideration (i.e., the studies
[[Page 16256]]
assessed in the 2019 ISA and ISA Supplement summarized above in section
II.A.2.a) reaffirms, and in some cases strengthens, the conclusions
from the 2009 ISA regarding the health effects of PM2.5
exposures. As noted above, epidemiologic studies demonstrate generally
positive and often statistically significant associations between
PM2.5 exposures and health effects. Such studies report
associations between estimated PM2.5 exposures and non-
accidental, cardiovascular, or respiratory mortality; cardiovascular or
respiratory hospitalizations or emergency room visits; and other
mortality/morbidity outcomes (e.g., lung cancer mortality or incidence,
asthma development). Recent experimental evidence, as well as evidence
from epidemiologic panel studies, strengthens support for potential
biological pathways through which PM2.5 exposures could lead
to the serious effects reported in many population-level epidemiologic
studies, including support for pathways that could lead to
cardiovascular, respiratory, nervous system, and cancer-related
effects. Moreover, these recent epidemiologic studies strengthen
support for health effect associations at PM2.5
concentrations lower than in those evaluated in epidemiologic studies
available at the time of previous reviews.
Additionally, as discussed in more detail in section I.C.5.b above,
the ISA Supplement focused on studies that were most likely to inform
decisions on the appropriate standard, but not to reassess areas that,
based on the assessment of available science published since the cutoff
date of the 2019 ISA and through 2021, were judged unlikely to have new
information that would be useful for the Administrator's decision
making. The ISA Supplement included U.S. and Canadian epidemiologic
studies for health effect categories where the 2019 ISA concluded a
causal relationship (i.e., short- and long-term PM2.5
exposure and cardiovascular effects and mortality), as well as U.S. and
Canadian epidemiologic studies that employed alternative methods for
confounder control or conducted accountability analyses (i.e., studies
that examined the effect of a policy on reducing PM2.5
concentrations). These studies, summarized in section II.A.2.a above,
examine both short- and long-term PM2.5 exposure and
cardiovascular effects and mortality. Additionally, studies that employ
alternative methods for confounder control, as described in II.A.2.a
above and in Table 3-11 and of the 2022 PA (U.S. EPA, 2022b), use a
variety of statistical methods to control for confounding bias. These
studies consistently report positive associations, which further
supports the broader body of epidemiologic evidence for both
cardiovascular effects and mortality.
In addition, there are epidemiologic studies that provide
supplemental information for consideration in reaching conclusions that
the current suite of PM2.5 standards is not adequate. These
studies include analyses that restrict annual average PM2.5
concentrations to concentrations below 12 [micro]g/m\3\ and provide
support for positive and statistically significant associations with
mortality and cardiovascular morbidity at mean PM2.5
concentrations below the current level of the primary annual
PM2.5 standard (described above in section II.A.2.c.ii and
in Table 3-10 of the 2022 PA (U.S. EPA, 2022b)). Recent accountability
studies that have starting annual PM2.5 concentrations at or
below 12 [micro]g/m\3\ suggest public health improvements may occur at
concentrations below 12 [micro]g/m\3\. These studies indicate positive
and statistically significant associations with mortality and morbidity
(e.g., cardiovascular hospital admissions) and reductions in
PM2.5 concentrations in ambient air (described above in
section II.A.2.c.ii and in Table 3-12 of the 2022 PA (U.S. EPA,
2022b)).
Thus, in considering the available scientific evidence to inform
conclusions on the adequacy of the primary PM2.5 standards,
the Administrator recognizes that the 2019 ISA and the ISA Supplement
together provides a strong scientific foundation for concluding that
the current primary PM2.5 standards are not adequate.
In addition to the scientific evidence above, the risk assessment
estimates that the current primary annual PM2.5 standard
could allow a substantial number of deaths in the U.S. Although the
Administrator recognizes that while the risk estimates can help to
place the evidence for specific health effects into a broader public
health context, they should be considered along with the inherent
uncertainties and limitations of such analyses when informing judgments
about the potential for additional public health protection associated
with PM2.5 exposures and related health effects. The
Administrator takes into consideration these uncertainties, which are
described in more detail in section II.A.3.b above, but notes that the
general magnitude of risk estimates supports the potential for
significant public health impacts, particularly for lower alternative
annual standard levels.
In the CASAC's review of the 2019 draft PA, the CASAC did not reach
consensus on whether the current annual standard is adequate, with the
majority of the CASAC recommending that the annual standard be retained
and the minority of the CASAC recommending that the standard be
revised. In their review of the 2021 draft PA, the CASAC unanimously
recommended that the current annual standard is not sufficiently
protective of public health (Sheppard, 2022a, p. 2 of consensus
letter).
The EPA disagrees with the commenters who state that the available
scientific and quantitative information available in this
reconsideration does not provide support for the current Administrator
to reach a different decision than the then-Administrator reached in
the 2020 final action. The EPA agrees with these commenters that there
are uncertainties associated with the currently available scientific
evidence. The EPA has considered these uncertainties extensively both
in reaching conclusions in the 2022 PA (U.S. EPA, 2022b, sections
3.4.3, 3.6.1, and 4.6.3) and in the proposal (88 FR 5604, 5609, January
27, 2023), and the EPA addresses more detailed public comments about
these uncertainties, including those related to copollutant
confounding, unmeasured confounding, and temporal and spatiotemporal
confounding, in the Response to Comments document. However, we disagree
with the commenters that the evidence does not provide support for the
Administrator's conclusion that the current primary annual
PM2.5 standard is not adequate to protect public health with
an adequate margin of safety, and should be revised. As described
above, epidemiologic studies in the 2019 ISA and the ISA Supplement
support and extend the evidence evaluated in the 2009 ISA, through
studies conducted in diverse populations and geographic locations,
using various statistical models and approaches to control for
potential confounders, and using a variety of exposure assessment
methodologies. Therefore, the consistent, positive associations
reported across studies (U.S. EPA, 2019a, Figures 11-1 and 11-18; U.S.
EPA, 2022a) are unlikely to be to be the result of unmeasured
confounding and other biases are unlikely to account for the consistent
positive associations observed across epidemiologic studies.
Additionally, this reconsideration includes epidemiologic studies
that were not before the then-Administrator for consideration in
reaching his final decisions at the time of the 2020 decision and that
specifically evaluate
[[Page 16257]]
confounding using alternative methods for confounder control). These
recent epidemiologic studies provide support for the current
Administrator's conclusion that the suite of primary PM2.5
standards are not adequate. While confounding was an uncertainty noted
by the then-Administrator in the 2020 decision, he recognized ``that
methodological study designs to address confounding, such as causal
inference methods, are an emerging field of study'' (85 FR 82710,
December 18, 2020). The ISA Supplement considered studies that employed
statistical approaches that attempt to more extensively account for
confounders and are more robust to model misspecification (i.e., used
alternative methods for confounder control),\92\ given that such
studies were highlighted by the CASAC in their review of the 2019 draft
PA and identified in public comments on the 2020 proposal. Since the
literature cutoff date for the 2019 ISA, multiple studies that employ
alternative methods for confounder control have become available for
consideration in the ISA Supplement and, subsequently, in this
reconsideration. For example, one study before the Administrator in
this reconsideration that was not available in the 2019 ISA is Schwartz
et al. (2021), which used a causal modeling approach focused on
exposure changes and controls for measured confounders by design in
order to evaluate the association between long-term PM2.5
exposure and mortality in the Medicare population. The study authors
found significant associations of PM2.5 with increased
mortality rates using a causal modeling approach robust to omitted
confounding. The results of this study and other studies in the ISA
Supplement that employ alternative methods to control for confounders
lend support to the robustness of positive associations between
PM2.5 exposure and multiple morbidity and mortality
endpoints exhibited across epidemiologic studies, and also indicate
that unmeasured confounding and other biases are unlikely to account
for the consistent positive associations observed across epidemiologic
studies (U.S. EPA, 2022b, sections 3.1.1.3, 3.1.2.3, 3.2.1.3, and
3.2.2.3).
---------------------------------------------------------------------------
\92\ As noted in the ISA Supplement: ``In the peer-reviewed
literature, these epidemiologic studies are often referred to as
causal inference studies or studies that used causal modeling
methods. For the purposes of this Supplement, this terminology is
not used to prevent confusion with the main scientific conclusions
(i.e., the causality determinations) presented within an ISA. In
addition, as is consistent with the weight-of-evidence framework
used within ISAs and discussed in the Preamble to the Integrated
Science Assessments, an individual study on its own cannot inform
causality, but instead represents a piece of the overall body of
evidence'' (U.S. EPA, 2022a, p. 1-3).
---------------------------------------------------------------------------
Further, the EPA disagrees with the commenters who argue that the
Administrator did not appropriately consider the strengths and
limitations of the health evidence in reaching his decision to revise
the current primary annual PM2.5 standard in this
reconsideration. In reaching his proposed decision, the Administrator
considered the entire body of evidence and how to appropriately weigh
the uncertainties associated with the health evidence (88 FR 5617,
January 27, 2023). Such an approach is consistent with setting
standards that are neither more nor less stringent than necessary,
recognizing that ``Congress provided that the Administrator is to use
his judgment in setting air quality standards precisely to permit him
to act in the face of uncertainty,'' the Administrator must set
standards on ``the frontiers of scientific and medical knowledge'' and
``Congress directed the Administrator to err on the side of caution in
making the necessary decisions.'' Lead Indus. Ass'n, Inc. v. EPA, 647
F.2d 1130, 1155 & n.50 (D.C. Cir. 1980) (quoting H.R. Rep. No. 95-294,
at 50). As such, a determination of identifying a specific level at
which the standard should be set necessarily requires the
Administrator's judgement (e.g., weighing the uncertainties and margin
of safety).
Additionally, the EPA disagrees with the commenters that contend
that there is no basis in this reconsideration for deviating from the
previous Administrator's decision in 2020. It is well-established that
in CAA section 109 Congress specifically left the determination of the
requisite NAAQS to the judgment of the Administrator and, moreover,
that ``decisions about the appropriate NAAQS level must `necessarily .
. . rest largely on policy judgments.' '' Mississippi v. EPA, 744 F.3d
1344, 1357 (D.C. Cir. 2013) (quoting Lead Industries Ass'n v. EPA, 647
F.2d 1130, 1147 (D.C. Cir. 1980)). As the Court of Appeals for the D.C.
Circuit has noted, ``Every time EPA reviews a NAAQS, it (presumably)
does so against contemporary policy judgments and the existing corpus
of scientific knowledge.'' Id., at 1343.
In this reconsideration, both the existing corpus of scientific
knowledge as well as the Administrator's policy judgments about how to
interpret and weigh that evidence to protect public health with an
adequate margin of safety have changed. The expansion of the air
quality criteria to encompass additional studies, information and
analyses in the ISA Supplement and 2022 PA, as well as the additional
consideration of the scientific record by the CASAC and the public
provided the Administrator with significant additional information on
which to base his decision.\93\ In addition, in this reconsideration,
the Administrator is reaching different judgments about how to weigh
the epidemiologic evidence, including the uncertainties in the
scientific evidence, and how to ensure an adequate margin of safety to
protect against uncertain harms, compared to the approach in the 2020
final decision. For example, as discussed in greater detail above in
section II.A.1 and in the 2020 notice of final rulemaking (85 FR 82717,
December 18, 2020), in considering the epidemiologic evidence as part
of his decision to retain the current primary annual PM2.5
standard in the 2020 decision, the then-Administrator placed weight on
the mean of the study-reported means (or medians) (i.e., 13.5 [micro]g/
m\3\) from key U.S. epidemiologic studies that are monitor-based being
above the level of the current primary annual PM2.5 standard
of 12.0 [micro]g/m\3\. By contrast, in this reconsideration, the
current Administrator has taken an approach more similar to how the EPA
has considered study-reported mean PM2.5 concentrations
relative to the level of the primary annual PM2.5 standard
in other recent PM NAAQS reviews. In so doing, in reaching his decision
to revise the level of the primary annual PM2.5 standard to
9.0 [micro]g/m\3\, he is using an approach that places weight on
selecting a level for the standard that is below the study-reported
mean PM2.5 concentrations reported in key U.S. epidemiologic
studies, including recent epidemiologic studies that use hybrid model-
based methods, as well as being near or below the 25th percentile
PM2.5 concentrations in those key U.S. epidemiologic studies
that report these concentrations.
---------------------------------------------------------------------------
\93\ The EPA notes that, in considering the additional
scientific evidence available in this reconsideration, one member of
the CASAC who reviewed both the 2019 draft PA and the 2021 draft PA
found that the available scientific and quantitative information
available in this reconsideration supported revising the level of
the primary annual PM2.5 standard, whereas he recommended
retaining the standard during the review of the 2019 draft PA.
---------------------------------------------------------------------------
As such and further detailed in section II.B.4 below, in
considering the adequacy of the current primary PM standards in this
reconsideration, the Administrator has carefully considered the: (1)
Policy-relevant evidence and conclusions contained in the 2019 ISA and
2022 ISA Supplement; (2) the quantitative information presented and
[[Page 16258]]
assessed in the 2022 PA; (3) the evaluation of this evidence, the
quantitative information, and the rationale and conclusions presented
in the 2022 PA; (4) the advice and recommendations from the CASAC; and
(5) public comments. The Administrator concludes that the current suite
of primary PM2.5 standards are not adequate to protect
public health with an adequate margin of safety.
The four basic elements of the NAAQS (indicator, averaging time,
form, and level) are considered collectively in evaluating the health
protection afforded by a standard. The EPA received relatively few
comments on the averaging time and form for the primary
PM2.5 standards, but those who did provide comments on these
elements were primarily from public health and environmental
organizations, State and local elected representatives, and State and
local government agencies. Some commenters assert that the current 24-
hour averaging time for the primary 24-hour PM2.5 standard
does not adequately protect against short-term peaks. These commenters
further state that the 24-hour averaging time protects against chronic
exposures but does not adequately protect against serious acute risks
from certain sources such as prescribed burning. Also, a few commenters
explicitly recommend that a subdaily averaging time would be more
appropriate, although none of the commenters recommended a specific
averaging time for consideration. Additionally, some commenters cite to
the CASAC's advice in their review of the 2021 draft PA that future
reviews of the PM NAAQS should include evaluation of alternative forms
and averaging times of the current primary 24-hour PM2.5
standard.
The EPA disagrees with commenters that the current primary 24-hour
PM2.5 standard, with its 24-hour averaging time, does not
adequately protect against short-term peaks and disagrees that that
there is sufficient information to conclude that a subdaily averaging
time would be more appropriate than a 24-hour averaging time. The EPA
has reviewed the currently available scientific evidence and finds that
it does not indicate that alternative averaging times would be more
appropriate for the primary PM2.5 standards. Accordingly,
the EPA concludes that it is appropriate to retain both the annual and
24-hour averaging times for standards meant to protect against long-
and short-term PM2.5.
As noted in the proposal, the 2019 ISA and ISA Supplement found
that the scientific evidence continues to provide strong support for
health effect associations with both long-term (e.g., annual or multi-
year) and short-term (e.g., mostly 24-hour) exposures to
PM2.5. Epidemiologic studies continue to provide strong
support for health effects associated with short-term PM2.5
exposures based on 24-hour PM2.5 averaging periods, and we
note that subdaily effect estimates are less consistent and, in some
cases, smaller in magnitude (88 FR 5618, January 27, 2023). Controlled
human exposure and panel-based studies of subdaily exposures typically
examine subclinical effects rather than the more serious population-
level effects that have been reported to be associated with 24-hour
exposures (e.g., mortality, hospitalizations). Collectively, the 2019
ISA concludes that epidemiologic studies do not indicate that subdaily
averaging periods are more closely associated with health effects than
the 24-hour average exposure metric (U.S. EPA, 2019a, section 1.5.2.1).
Additionally, the EPA notes that while recent controlled human exposure
studies provide consistent evidence for cardiovascular effects
following PM2.5 exposures for less than 24 hours (i.e., <30
minutes to 5 hours), exposure concentrations in these studies are well-
above the ambient concentrations typically measured in locations
meeting the current standards (U.S. EPA, 2022a, section 3.3.3.1).
Therefore, this information does not indicate that a revision to the
averaging time is needed to provide additional protection against
subdaily PM2.5 exposures, beyond that provided by the
current primary standards. This conclusion is also supported by the
advice given to EPA by the CASAC in their review of the 2021 draft PA,
which reached consensus that averaging times for the standards should
be retained, without revision (Sheppard, 2022a, p. 2 of consensus
letter).\94\ For all of these reasons, the Administrator concludes that
the currently available evidence does not support considering
alternatives to the annual and 24-hour averaging times for standards
meant to protect against long- and short-term PM2.5
exposures.
---------------------------------------------------------------------------
\94\ In providing advice on the 2019 draft PA, the CASAC did not
weigh in specifically on the averaging time of the primary 24-hour
PM2.5 standard but did recommend that the standard be
retained because the available evidence does not call into question
its adequacy (Cox, 2019b, p. 3 of consensus letter).
---------------------------------------------------------------------------
Multiple commenters, primarily from public health and environmental
organizations, recommend revising the form of the primary 24-hour
PM2.5 standard to a 99th percentile to provide increased
public health protection against peak PM2.5 exposures,
particularly for at-risk populations. These commenters express concern
that the current 98th percentile form allows 7 exceedances per year and
contend that a 99th percentile form that would allow half that number
is more appropriate. Commenters also cite to the CASAC's advice in
their review of the 2021 draft PA, which recommended that the EPA
consider alternative percentiles for the form of the primary 24-hour
PM2.5 standard in the future.
The EPA disagrees that the current 98th percentile form does not
provide the requisite public health protection against peak
PM2.5 exposures and concludes that the 98th percentile,
averaged over three years, remains appropriate for the primary 24-hour
PM2.5 standard. As noted in previous reviews and in the
proposal, the EPA has set both an annual standard and a 24-hour
standard to provide protection from health effects associated with both
long- and short-term exposures to PM2.5 (62 FR 38667, July
18, 1997; 88 FR 5620, January 27, 2023). With respect to the form of
the 24-hour standard, as described just above, the epidemiologic
studies continue to provide strong support for health effect
associations with short-term (e.g., mostly 24-hour) PM2.5
exposures and controlled human exposure studies provide evidence for
health effects following single short-term ``peak'' PM2.5
exposures (88 FR 5619, January 27, 2023). Both the 98th and the 99th
percentile form provide a very high degree of control of peak
concentrations. As the commenters point out, a 99th percentile would
reduce the number of allowable exceedances to four days per year. The
EPA anticipates, however, that such a revision to the form would make
the attainment status of an area more subject to change from
unpredictable nonanthropogenic factors, such as meteorological events.
The EPA has often noted that frequent shifts in attainment status that
are unrelated to long-term air quality trends is inconsistent with
providing a stable target for air quality planning and risk management
programs, which in turn provides for the most effective public health
protection in the long run (78 FR 3127, January 15, 2013; 80 FR 65351,
October 26, 2015). Thus, the EPA's interest in an appropriate degree of
stability is to ensure that the State air quality programs are
effective in controlling pollution and that the public health
protections of the standard are achieved. As discussed above, while
recent controlled human exposure studies provide consistent evidence
for cardiovascular effects following PM2.5
[[Page 16259]]
exposures for less than 24 hours (i.e., < 30 minutes to 5 hours),
exposure concentrations in these studies are well-above the ambient
concentrations typically measured in locations meeting the current
standards (U.S. EPA, 2022a, section 3.3.3.1), and the 98th percentile
form is very effective at limiting occurrences of exposures of concern.
Taking into consideration the available scientific information and
quantitative information, the EPA therefore concludes that the 98th
percentile form provides an appropriate balance between limiting the
occurrence of peak 24-hour PM2.5 concentrations and
identifying a stable target for risk management programs. This
conclusion is also supported by the advice given to the EPA by the
CASAC in their review of the 2021 draft PA, where they reached
consensus that the form for the standards should be retained, without
revision (Sheppard, 2022a, p. 2 of consensus letter).\95\
---------------------------------------------------------------------------
\95\ The CASAC did not provide advice or recommendations
regarding the forms of the primary PM2.5 standards in
their review of the 2019 draft PA (Cox, 2019b).
---------------------------------------------------------------------------
Additionally, the EPA recognizes the CASAC's advice in their review
of the 2021 draft PA, where they recommended ``that in future reviews,
the EPA provide a more comprehensive assessment of the 24-hour standard
that includes the form as well as the level'' (Sheppard, 2022a, p. 4 of
consensus letter). This advice is reflected in the proposal by the EPA,
which noted ``that it would be appropriate to gather additional air
quality and scientific information and further consider these issues in
future reviews'' (88 FR 5619, January 27, 2023). The EPA will consider
the information provided by the commenters regarding the form of the
24-hour PM2.5 standard in the next review of the PM NAAQS.
A number of commenters who support revising the level of the
primary annual PM2.5 standard, particularly those who
support a revised level of 8 [micro]g/m\3\, disagree with how the EPA
has emphasized the mean PM2.5 concentrations reported in key
epidemiologic studies to inform conclusions on the level of the primary
PM2.5 standard. These commenters argue that, in this
reconsideration, the EPA is arbitrarily emphasizing uncertainties in
key epidemiologic studies in the focus on mean concentrations. Many of
these commenters recommend that the EPA consider the full distribution
of PM2.5 concentrations from the key epidemiologic studies
in reaching conclusions on the appropriate level for the primary annual
PM2.5 standards, in particular concentrations below the
mean, such as the 25th percentile. In supporting this view, commenters
point to the CASAC's advice in their review of the 2021 draft PA, where
the majority of the CASAC stated that the ``use of the mean to define
where the data provide the most evidence is conservative since robust
data clearly indicate effects below the mean in concentration-response
functions'' (Sheppard, 2022a, p. 16 of consensus responses), and that
``[e]pidemiologic studies require consideration of distribution around
the mean of exposure to identify effects and thus lower levels than the
mean must be considered as part of the range where the data provide
higher confidence'' (Sheppard, 2022a, p. 13 of consensus responses).
As an initial matter, consistent with some previous approaches and
as detailed by the Administrator in reaching conclusions on the level
of the primary annual PM2.5 standard in section II.B.4
below, the EPA considers the long-term study-reported mean
PM2.5 concentrations from key epidemiologic studies and sets
the level of the standard to somewhat below the lowest long-term mean
PM2.5 concentration. Additionally, as discussed further
below, the EPA also considers the available information from a subset
of epidemiologic studies that report exposure estimates or health
events at the 25th and 10th percentiles of PM2.5
concentrations. The Administrator gives some weight to the 25th
percentile data, although he recognizes that his confidence in the
magnitude and significance in the reported concentrations, and their
ability to inform decisions on the appropriate level of the annual
standard, decreases with reduced data (below the mean) and diminishes
further at percentiles that are even further below the mean and the
25th percentile. Therefore, the Administrator places weight on the
reported 25th percentiles concentrations, rather than the reported 10th
percentile concentrations, for the subset of studies that report lower
percentile PM2.5 concentrations in reaching his conclusions
regarding the appropriate level for the primary annual PM2.5
standard.
In considering the available scientific evidence to reach decisions
on the adequacy of the suite of primary PM2.5 standards, the
EPA notes that in previous PM NAAQS reviews (including the 1997, 2006
and 2012 reviews), evidence-based approaches were used that focused on
identifying standard levels near or somewhat below long-term mean
concentrations reported in key epidemiologic studies. These approaches
were supported by the CASAC in previous reviews and were supported in
this reconsideration by the CASAC in their review of the 2021 draft
PA.\96\
---------------------------------------------------------------------------
\96\ The Administrator notes that, in their review of the 2021
draft PA, a majority of members of the CASAC noted that there are
some limitations for this approach ``for the purpose of informing
the adequacy of the standards'' (Sheppard, 2022a, p. 8 of consensus
responses) and advised that future reviews should include evaluation
of other metrics, including the distribution of concentrations
reported in epidemiologic studies and in analyses restricting
concentrations to below the current standard level. The
Administrator also notes that, in their review of the 2019 draft PA,
the CASAC lacked consensus on the inferences to be drawn from the
epidemiologic evidence, with a majority of CASAC having concerns
about confounding, error and bias and concluding that newer studies
did not provide a basis for revising the current standards, while a
minority concluded that the evidence, including more recent studies
showing associations in areas with average long-term
PM2.5 concentrations below the current annual standard,
supported their conclusion that the current standards are inadequate
(Cox, 2019b, pp. 8-9 of consensus responses).
---------------------------------------------------------------------------
In considering the available scientific evidence, the EPA notes the
strength of the epidemiologic evidence which includes multiple studies
that consistently report positive associations for short- and long-term
PM2.5 exposures and mortality and cardiovascular effects.
Some available studies also use a variety of statistical methods to
control for confounding bias and report similar associations, which
further supports the broader body of epidemiologic evidence for both
mortality and cardiovascular effects. Additionally, the EPA notes that
recent epidemiologic studies strengthen support for health effect
associations at PM2.5 concentrations lower than in those
evaluated in epidemiologic studies available at the time of previous
reviews.
While these epidemiologic studies evaluate associations between
distributions of ambient PM2.5 concentrations and health
outcomes, they do not identify the specific exposures that led to the
reported effects. As such, there is no specific point in the air
quality distribution of any epidemiologic study that represents a
``bright line'' at and above which effects have been observed and below
which effects have not been observed.
Studies of daily PM2.5 exposures examine associations
between day-to-day variation in PM2.5 concentrations and
health outcomes, often over several years. While there can be
considerable variability in daily exposures over a multi-year study
period, most of the estimated exposures reflect days with
[[Page 16260]]
ambient PM2.5 concentrations around the middle of the air
quality distributions examined (i.e., ``typical'' days rather than days
with extremely high or extremely low concentrations). Similarly, for
studies of annual PM2.5 exposures, most of the health events
occur at estimated exposures that reflect annual average
PM2.5 concentrations around the middle of the air quality
distributions examined. In both cases, epidemiologic studies provide
the strongest support for reported health effect associations for this
middle portion of the PM2.5 air quality distribution, which
corresponds to the bulk of the underlying data, rather than the extreme
upper or lower ends of the distribution. Therefore, in the absence of
discernible thresholds, long-term study-reported means--that is, the
study-reported ambient PM2.5 concentrations in the
epidemiologic studies that reflect estimated exposures with a focus
around the middle portion of the PM2.5 air quality
distribution where the bulk of the observed data reside--provide the
strongest support for reported health effect associations in
epidemiologic studies.
Based on the air quality criteria for this reconsideration, as
described in the 2019 ISA, ISA Supplement, 2022 PA and the proposal,
the EPA believes it is appropriate to continue to use the mean
PM2.5 concentrations from the key epidemiologic studies to
inform conclusions regarding the appropriate level for the primary
annual PM2.5 standard.
There are a large number of key epidemiologic studies available in
this reconsideration to inform conclusions regarding the level of the
primary annual PM2.5 standard. For the key U.S.
epidemiologic studies, the study-reported mean PM2.5
concentrations range from 9.9-16.5 [mu]g/m\3\ for monitor-based studies
(Figure 1 above) and range from 9.3-12.2 [mu]g/m\3\ for hybrid
modeling-based studies (Figure 2 above).
In addition to the study-reported mean PM2.5
concentrations, the EPA agrees with the CASAC's advice in their review
of the 2021 draft PA and public comments that information on other
percentiles below the mean can also be informative, and the EPA notes
that the CASAC advised that for the purpose of informing the adequacy
of the standards, future reviews should include an evaluation of other
metrics, including the distribution of concentrations reported in
epidemiologic studies (Sheppard, 2022a, p. 9 of consensus responses).
As such, in reaching conclusions in this reconsideration, the EPA takes
note of the additional study-reported PM2.5 concentrations
below the means (e.g., 25th and 10th percentiles) that are available
from a limited subset of key U.S. epidemiologic studies. As shown in
Figures 1 and 2 above, six key U.S. epidemiologic studies report
information on other percentiles (e.g., 10th and 25th percentiles of
PM2.5 concentrations or 10th and 25th percentiles of
PM2.5 concentrations associated with health events) that are
below the mean.\97\ Three of the studies are monitor-based and three
are hybrid model-based.
---------------------------------------------------------------------------
\97\ The Wang et al. (2017) study only reports the 25th
percentile of the estimated PM2.5 concentrations, not the
10th percentile.
---------------------------------------------------------------------------
The key U.S. epidemiologic studies that report percentiles below
the mean that are monitor based are older studies. These studies
included smaller numbers of people than the newer hybrid model-based
studies. For the three older, monitor-based studies, because the
cohorts were smaller in size, a relatively smaller portion of the
health events were observed in the lower part of the air quality
distribution. As such, our confidence in the magnitude and significance
of the associations begins to decrease in the lower part of the air
quality distribution of those older, monitor-based studies.
The three newer, hybrid model-based studies have larger cohort
sizes than the older, monitor-based studies and, as noted by
commenters, have more health events in the lower part of the air
quality distribution. For these reasons, the EPA notes that we have
more confidence in the reported association at concentrations lower
than the reported mean in these more recent hybrid model-based studies,
particularly at the 25th percentile compared to the 10th percentile.
While the cohort sizes in the more recent, hybrid model-based studies
are larger than the older, monitor-based studies, the EPA notes that
the 10th percentiles are well below the middle portion of the air
quality distribution for which we have the greatest confidence, and as
noted above, our confidence in the magnitude and significance of
associations in the lower parts of the air quality distribution begins
to decrease. While we have more confidence in the lower percentiles
because of the larger cohort sizes in the more recent hybrid model-
based studies, we also have more confidence in the 25th percentiles
than in the 10th percentiles, which are further from the means and
closer to the lower end of the air quality distribution.
In considering how the six studies that report percentiles lower
than the mean can be used to inform conclusions regarding the level of
the primary annual PM2.5 standard, the EPA first notes that
the three monitor-based epidemiologic studies (Bell et al., 2008;
Franklin et al., 2007; Zanobetti and Schwartz, 2009) report 25th
percentile concentrations that are at or above 11.5 [micro]g/m\3\. For
two of the more recent hybrid model-based studies (Di et al., 2017b;
Wang et al., 2017), the 25th percentile of estimated PM2.5
concentrations are just above 9 [micro]g/m\3\, while one study (Di et
al., 2017a) reports a PM2.5 concentrations corresponding to
25th percentiles of health events of just below 7 [micro]g/m\3\. For
the Di et al. (2017a) study, the 25th percentile PM2.5
concentration (6.7 [micro]g/m\3\) is based on the PM2.5
concentration at which the 25th percentile of deaths occur in the
study, while the reported mean (11.6 [micro]g/m\3\) is based on
estimated PM2.5 exposure concentrations. Additionally, the
25th percentiles of the other two recently available hybrid model-based
studies (Di et al., 2017b; Wang et al., 2017) are based on estimated
PM2.5 concentrations. As such, the PM2.5
concentration at which the 25th percentile of health events occur may
be different from the estimated 25th percentile PM2.5
concentration in this study (Di et al., 2017a), creating an uncertain
basis for comparison with the studies by Di et al. (2017b) and Wang et
al. (2017). The 25th percentiles from these studies, in particular
those that are more recently available, help to inform the
Administrator's judgments regarding the appropriate level for the
primary annual PM2.5 standard.
Some commenters disagree with the EPA's consideration of the
relationship between mean PM2.5 concentrations reported in
the key epidemiologic studies and design values to inform conclusions
on the appropriate level for the primary annual PM2.5
standards. Commenters contend that setting the level of the primary
annual standard below the design values in the epidemiologic studies,
rather than below the study-reported mean concentrations, might keep
overall mean PM2.5 concentrations throughout an area below
the study-reported means but allow PM2.5 concentrations in
some parts of the area, including near the ``design value monitor'' to
remain above the study-reported mean PM2.5 concentrations,
which are the concentrations where the evidence of health effects is
strongest. Commenters contend that such a decision framework would not
result in a standard that would provide requisite protection with an
adequate margin of safety,
[[Page 16261]]
particularly for at-risk populations. These commenters further support
this view by citing the CASAC's advice in their review of the 2021
draft PA, where the majority of CASAC stated that ``even if a design
value is somewhat higher than the area average, it reflects actual
exposure levels and thus any portion of the population living near the
design value monitor does experience exposures at that level and
consequent health effects of exposure to that higher concentration''
(Sheppard, 2022a, p. 14 of consensus responses). Additionally, these
commenters suggest that the EPA should not deviate from the approach
taken in the 2012 review, which was to set the standard at a level
``somewhat below'' the lowest mean PM2.5 concentration in
the key epidemiologic studies.
To the extent that commenters are suggesting that the EPA is
setting the level of the primary annual PM2.5 standard below
the design values in the epidemiologic studies, rather than below the
study-reported mean PM2.5 concentrations, we disagree with
the commenters. In reaching conclusions on the level of the primary
annual PM2.5 standard, the EPA considers the long-term
study-reported mean PM2.5 concentrations from key
epidemiologic studies and sets the level of the standard to somewhat
below the lowest long-term mean PM2.5 concentration, not
below the design values in the epidemiologic studies. Additionally, the
EPA also considers the available information from a subset of
epidemiologic studies that report exposure estimates or health events
at the 25th and 10th percentiles of PM2.5 concentrations.
The EPA particularly considers the 25th percentile data, while
recognizing that our confidence in the magnitude and significance in
the reported concentrations, and the ability of the lower percentile
PM2.5 concentrations to inform decisions on the appropriate
level of the annual standard, decreases with reduced data (below the
mean) and diminishes further at percentiles that are even further below
the mean and the 25th percentile.
However, the EPA notes that it is important to understand, and to
not ignore, the relationship between the study-reported mean
PM2.5 concentrations reported in key epidemiologic studies
and the area design value. As an initial matter, the NAAQS consists of
all four elements of the standard (indicator, averaging time, form, and
level) and setting a standard that is requisite to protect public
health includes consideration of all four elements together. Following
implementation of the NAAQS, the design value is the metric used to
determine compliance with the standard and is the statistic that
describes the air quality status of a given location relative to the
level of the primary annual PM2.5 NAAQS. The design value is
different from the study-reported mean PM2.5 concentrations.
This is because the study-reported mean PM2.5 concentrations
are an annual average PM2.5 concentration, similar to the
level of the standard, but the epidemiologic studies do not report
statistics that take into account the other elements of the standard
(i.e., averaging time and form). Therefore, when considering the
appropriate revisions to the annual PM2.5 standard, the EPA
must consider the protection provided by a revised standard taking into
account all of the elements of the standard, not just the annual
average PM2.5 concentration alone.
In considering the annual standard, and in assessing the range of
study-reported exposure concentrations for which we have the strongest
support for adverse health effects observed in epidemiologic studies,
the EPA focuses on whether the current primary annual PM2.5
standard provides adequate protection against these exposure
concentrations or if the level of the standard should be revised to
provide the appropriate public health protection. This means that, as
in some previous reviews, it is important to consider how the study
means were computed and how these concentrations compare to the annual
standard metric (including the level, averaging time and form) which
must be met at the monitor with the highest PM2.5 design
value in an area for compliance with the NAAQS. This approach is based
on the application of a decision framework based on assessing means (as
well as the lower distribution of reported PM2.5
concentration, as noted above) reported in key epidemiologic studies.
In the 2012 review, the available key epidemiologic studies computed
the mean PM2.5 concentrations using an average across
monitor-based PM2.5 concentrations. As such, at that time,
the decision framework used an approach based on maximum monitor
concentrations to determine compliance with the standard, while
selecting the standard level based on consideration of composite
monitor concentrations (i.e., selecting the standard level of 12.0
[micro]g/m\3\ was just below the long-term study-reported mean
PM2.5 concentrations in key epidemiologic studies). Further,
the EPA conducted analyses that examined the differences in these two
metrics (i.e., maximum monitor concentrations, which is how compliance
with the standard is assessed and composite monitor concentrations,
which is how key epidemiologic studies report their mean
concentrations) across the U.S. and in areas included in the key
epidemiologic studies and found that the maximum design value in an
area was generally higher than the monitor average across that area,
with the amount of difference between the two metrics varying based on
location and concentration (Hassett-Sipple et al., 2010; Frank, 2012).
This information was taken into account by the then-Administrator's
final decision in selecting a level of 12.0 [micro]g/m\3\ for the
primary annual PM2.5 standard in the 2012 review and
discussed more specifically in her considerations on adequate margin of
safety.
The relationship between the mean PM2.5 concentrations
and the area design value continues to be an important consideration in
evaluating the adequacy of the current or potential alternative annual
standard levels in this reconsideration. Again, in a given area, the
area design value is based on the monitor in an area with the highest
PM2.5 concentrations and is used to determine compliance
with the standard, including the averaging time and form of the
standard (i.e., an annual average over 3-years must not exceed the
level of the of the annual PM2.5 standard). The highest
PM2.5 concentrations spatially distributed in the area would
generally occur at or near the area design value monitor and the
distribution of PM2.5 concentrations would generally be
lower in other locations and at monitors in that area. As such, when an
area is meeting a specific annual standard level (e.g., 9.0 [micro]g/
m\3\), we would expect the annual average exposures (i.e., a metric
similar to the study-reported mean values) in that area to be at
concentrations lower than that level (e.g., lower than 9.0 [micro]g/
m\3\).
However, as described in section II.A.2.c.ii, we note that there
are a substantial number of different types of epidemiologic studies
available since the 2012 review, as assessed in both the 2019 ISA and
the ISA Supplement, that make understanding the relationship between
the mean PM2.5 concentrations and the area design value an
even more important consideration in this reconsideration (U.S. EPA,
2019a; U.S. EPA, 2022a). While the key epidemiologic studies in the
2012 review were all monitor-based studies, the recent epidemiologic
studies in this reconsideration include hybrid modeling approaches that
have emerged in the epidemiologic literature as an
[[Page 16262]]
alternative to approaches that only use ground-based monitors to
estimate PM2.5 exposure. As assessed in the 2019 ISA and ISA
Supplement, a substantial number of epidemiologic studies used hybrid
model-based methods in evaluating associations between PM2.5
exposure and health effects. Hybrid model-based studies employ various
fusion techniques that combine ground-based monitored data with air
quality modeled estimates and/or information from satellites to
estimate PM2.5 exposures. While these studies provide a
broader estimation of PM2.5 exposures compared to monitor-
based studies (i.e., PM2.5 concentrations are estimated in
areas without monitors), the hybrid modeling approaches result in
study-reported means that are more difficult to relate to the annual
standard metric and to the maximum monitor design values used to assess
compliance. In addition, to further complicate the comparison, when
looking across these studies, we find variations in how exposure is
estimated between such studies, and thus, how the study means are
calculated. Two important variations across studies include: (1)
Variability in spatial scale used (i.e., averages computed across the
national (or large portions of the country) versus a focus on only
CBSAs); and (2) variability in exposure assignment methods (i.e.,
averaging across all grid cells, averaging across a scaled-up area like
a ZIP code, and population weighting). The differences in these
approaches can result in studies reporting different study means, even
though the association between PM2.5 exposure and health
effects outcomes are similar.
To emphasize the importance of the differences between the studies,
we revisit the simplified example in the State of Georgia from the 2022
PA that evaluates monitors and hybrid modeling approaches, noting that
this example is useful to exhibit how the differences in the methods
used to estimate exposure can lead to differences in the reported mean
concentrations (U.S. EPA, 2022b, p. 3-71). In this example, for all
monitors within the Atlanta-Sandy Springs-Roswell CBSA, the average
PM2.5 concentration is 9.3 [micro]g/m\3\, while the area
design value (based on the highest monitored PM2.5
concentration in the area) is 10.4 [micro]g/m\3\. This comparison helps
to illustrate the fact that composite monitor values tend to be
somewhat lower than the highest area monitor values, consistent with
the key points made in the 2012 review. This example also illustrates
how monitors are sited to represent the higher concentrations within
the area and that the area's annual design value, which is used for
compliance with the standard, is calculated based on the highest
monitor in the area. Next, in this example, mean PM2.5
concentrations were calculated using similar approaches to those used
in hybrid modeling-based epidemiologic studies to compute study-
reported means, including (1) the average concentration across the
entire State of Georgia; (2) the population-weighted average across the
entire State; (3) the average concentration across the Atlanta-Sandy
Springs-Roswell CBSA; and (4) the population-weighted average across
the Atlanta-Sandy Springs-Roswell CBSA. At the urban level (e.g.,
Atlanta-Sandy Springs-Roswell CBSA), the average PM2.5
concentration when taking the mean of all grid cells is 9.2 [micro]g/
m\3\, whereas the population-weighted mean is 9.6 [micro]g/m\3\. Across
Georgia, the average PM2.5 concentration using the hybrid
approach and averaged across each grid cell is 8.3 [micro]g/m\3\, which
is lower than the population-weighted statewide average of 9.1
[micro]g/m\3\. While this is a simple example completed in one State
and one CBSA, it suggests that the lowest mean values tend to result
from the approaches that use concentrations from all or most grid cells
(e.g., did not apply population weighting), both urban and rural,
across the study area to compute the mean. Higher mean values are
observed when the approach focuses on the urban areas alone or when the
approach incorporates population weighting. Overall, this example
suggests that the means from studies using hybrid modeling approaches
are generally lower than the means from monitor-based approaches, and
means from both approaches are lower than the annual design values for
the same area. Population weighting tends to increase the calculated
mean concentration, likely because more densely populated areas also
tend to have higher PM2.5 concentrations. In other words,
this simplified example exhibits how not all reported mean
PM2.5 concentrations from key epidemiologic studies are the
same; some reported means are from monitored studies and some reported
means are from hybrid modeling studies, while some reported means
include only urban areas, and other reported means include both urban
and rural areas, and some reported means include aspects of population
weighting while others do not.
As detailed above in section I.D.5, in the air quality analyses
comparing composite monitored PM2.5 concentrations with
annual PM2.5 design values in U.S. CBSAs, maximum annual
PM2.5 design values were approximately 10% to 20% higher
than annual average composite monitor concentrations (i.e., averaged
across multiple monitors in the same CBSA). Based on these results,
this analysis suggests that there will be a distribution of
concentrations and the maximum annual average monitored concentration
in an area (at the design value monitor, used for compliance with the
standard), will generally be 10-20% higher than the average across the
other monitors in the area. Thus, in considering how the annual
standard levels would relate to the study-reported means from monitor-
based studies, we can generally conclude that an annual standard level
that is no more than 10-20% higher than monitor-based study-reported
mean PM2.5 concentrations would generally maintain air
quality exposures to be below those associated with the study-reported
mean PM2.5 concentrations, exposures for which we have the
strongest support for adverse health effects occurring.
Air quality analyses described in section I.D.5 above also consider
information from the epidemiologic studies that utilized the hybrid
modeling approaches. Analyses show that average maximum annual design
values are 40-50% higher when compared to annual average
PM2.5 concentrations estimated without population weighting
and are 15-18% higher when compared to average annual PM2.5
concentrations with population weighting applied. Given these results,
it is worth noting that for the studies using the hybrid modeling
approaches, the choice of methodology employed in calculating the
study-reported means (i.e., using population weighting versus not
applying aspects of population weighting), and not a difference in
estimates of exposure in the study itself, can produce substantially
different study-reported mean values, with the approach that does not
employ population weighting producing a much lower reported mean
PM2.5 concentration. Therefore, the impact of the
differences in methods is an important consideration when comparing
mean concentrations across studies.
Because of the differences in the methods employed by the key
epidemiologic studies, and as demonstrated by the example and air
quality analyses above, the application of any decision framework that
considers the study-reported mean PM2.5 concentrations, and
evaluates whether the current annual standard provides adequate
protection against these reported exposure concentrations, is more
complicated than the
[[Page 16263]]
approaches used in past reviews. As such, the EPA disagrees with
commenters who argue that the EPA's consideration of the relationship
between mean PM2.5 concentrations reported in key
epidemiologic studies and design values is not appropriate and should
be ignored.
In considering the information from the epidemiologic studies,
while the EPA does not dispute the reported associations of
epidemiologic studies in hybrid modeling studies that report long-term
mean concentrations and do not apply aspects of population weighting,
using the reported long-term mean concentration from these studies in
informing an appropriate level of the annual PM2.5 standard
is more uncertain. Given this, hybrid modeling studies that do not
apply aspects of population weighting provide less information on
conclusions regarding the appropriate level of the primary annual
PM2.5 standard. In support of this, some commenters also
noted this consideration and suggested that the Administrator place
lower weight on U.S. studies that did not use population weighting.
In considering the relationship between study-reported mean
PM2.5 concentrations and the design values, the EPA agrees
with commenters that setting the level of the primary annual standard
below the design values, rather than below the study-reported mean
concentrations, might allow PM2.5 concentrations in some
part of the area near the design value monitor to remain above the
study-reported mean PM2.5concentration, where evidence of
health effects is strongest. As discussed in the proposal and in
section II.B.4 below, the Administrator specifically notes that that
the highest PM2.5 concentrations spatially distributed in
the area would generally occur at or near the area design value monitor
and that PM2.5 concentrations will be equal to or lower at
other monitors in the area. Furthermore, since monitoring strategies
aim to site monitors in areas with higher PM2.5
concentrations, monitored areas will generally have higher
concentrations compared to areas without monitors. Therefore, by
setting the level of the standard to 9.0 [micro]g/m\3\ and just below
the lowest study-reported mean PM2.5 concentration (e.g.,
9.3 [micro]g/m\3\), the highest possible design value in a given area
would be just below the study-reported mean PM2.5
concentration, the concentration where we have the most confidence in
the reported health effect association, and we anticipate that, based
on our assessment of air quality data, the distribution of
PM2.5 concentrations would decrease even further with
distance from the highest monitor (i.e., the ``design value monitor'')
(see, for example, U.S. EPA, 2022a, section 2.3.3.2.4 and pp. 3-71 to
3-77). The Administrator further notes that when an epidemiologic study
reports a mean PM2.5 concentration that reflects the average
of annual average monitor-based concentrations across an area, the area
design value will generally be higher than the study-reported mean.
Similarly, he observes that when a study reports a mean that reflects
the average of annual average concentrations estimated at across an
area using a hybrid modeling approach, the area design value will
generally be higher. As such, by evaluating the difference between the
study-reported mean PM2.5 concentrations and design values,
the Administrator seeks to set the level of the standard below the
lowest study-reported mean, while ensuring that the primary annual
PM2.5 standard, including its averaging time and form,
provides protection against the exposures associated with health
effects observed in the key epidemiologic studies.
Additionally, the EPA disagrees with commenters who contend that
the approach taken may allow PM2.5 near the design value
monitor to remain above the study-reported mean PM2.5
concentrations. In following this approach of setting the annual
standard level somewhat below the lowest reported mean PM2.5
concentration, setting a standard level that requires the design value
monitor (which is the highest monitor in an area) to be just below the
lowest study-reported mean across key studies will generally result in
distributions of even lower concentrations of PM2.5 across
the entire area, such that even those people living near an area design
value monitor (where PM2.5 concentrations are generally
highest) will be exposed to PM2.5 concentrations below the
PM2.5 concentrations reported in the epidemiologic studies
where there is the highest confidence of an association. In their
review of the 2021 draft PA, the majority of the CASAC had some
concerns about the approach for comparing study means and design
values, questioning whether such an approach would provide adequate
protection for people who live in areas with higher concentrations,
such as those living in areas with higher concentrations (e.g., near
the design value monitor) (Sheppard, 2022a, p. 8 of consensus
responses). The minority of the CASAC, in considering the relationship
between the study-reported mean PM2.5 concentration and
design values, stated that ``the form of the standard and the way
attainment with the standard is determined (i.e., highest design value
in the CBSA) are important factors when determining the appropriate
level for the standard'' and noted that that design values are
generally higher than area average exposure levels (Sheppard, 2022a, p.
17 of consensus responses). For all of the reasons discussed above, and
consistent with the minority of the CASAC's advice in their review of
the 2021 draft PA, we disagree with the commenters that areas near the
design value monitors would be expected to experience PM2.5
concentrations above the study-reported mean concentrations.
Several commenters assert that epidemiologic studies that restrict
PM2.5 concentration to below 12 [micro]g/m\3\ provide
additional support for revising the level of the primary annual
PM2.5 standard to 8 [micro]g/m\3\. Some commenters disagree
with the EPA's assertion that the studies that employ restricted
analyses do not provide enough information to understand how the
studies were restricted to certain PM2.5 concentrations,
with commenters providing additional information on the methods for
restricted analyses. The commenters state that for the long-term
studies at issue here, the study authors simply examined their database
that linked subjects to long-term PM2.5 concentrations above
12 [mu]g/m\3\, removed those data from the analysis, and reran the
analysis. Additionally, one commenter provided an explanation of how
the restricted analyses were conducted in studies for which he was an
author. The commenter notes that for each year a subject was in the
study, annual PM2.5 concentrations were assigned at the ZIP
code level. If they moved, they were assigned the ZIP code level
PM2.5 concentration for the new ZIP code. The commenter
notes that these restricted analyses only included subjects whose
annual PM2.5 exposure never exceeded that restricted
concentration for any year of follow-up in the study. The commenter
suggested that the EPA may be concerned as to how PM2.5
concentrations in restricted analyses related to a design value since
these are exposures for individuals who may have relocated during the
study but argue that that is not the point. The commenters assert that
while the analyses were restricted to people never exposed above
certain concentrations over longer periods of time, the actual
PM2.5 exposure was one year of exposure in most of these
studies. Commenters also suggest that, since the
[[Page 16264]]
EPA has deviated from its approach from the 2012 review for considering
study-reported mean PM2.5 concentrations, the EPA should
dismiss its concerns regarding being able to relate the mean
PM2.5 concentrations from these studies to design values.
First, the EPA agrees with commenters that studies that employ
restricted analyses can be used for informing conclusions regarding the
appropriate level of the primary annual PM2.5 standard.
However, the EPA disagrees that the information provided by the
commenters provides a sufficient basis for an annual standard level of
8 [mu]g/m\3\. Restricted analyses provide additional support for
effects at lower concentrations, exhibiting associations for mean
concentrations presumably below the mean concentrations for the main
analyses. However, even though commenters note that any individual with
exposures over the restricted analyses is excluded from restricted
analyses, uncertainties remain with regard to how the mean
PM2.5 concentrations in restricted analyses compare to
design values, particularly in light of the removal of entire ZIP codes
from analyses. Design values are calculated based on all measured
PM2.5 concentrations. When an analysis is restricted below a
certain level, some parts of the air quality distribution are removed,
but comparing the restricted mean to a design value is not possible
because these are two different metrics. For example, in a study that
restricts concentrations below 12 [micro]g/m\3\, that represents only
part of the air quality distribution, whereas a design value for that
study area would include all PM2.5 concentrations, not just
the ones below 12 [micro]g/m\3\. Therefore, in contrast to means from
the main (unrestricted) analysis, it is not possible to compare mean
concentrations from restricted analyses to design values. Further, it
is unclear how one could evaluate such a relationship between design
values and mean PM2.5 concentrations from studies that use
restricted analyses because the standard is set based on all of its
elements (indicator, averaging time, form, and level) and removing
PM2.5 concentrations from the calculation of the design
value for such a comparison would result in a metric that is no longer
a design value that would provide the intended protection of the
standard. This leads to greater uncertainty in how to use the mean
PM2.5 concentrations from these studies that use restricted
analyses in a similar decision framework as the epidemiologic studies
that report long-term mean PM2.5 concentrations for health
effect associations for the full distribution of PM2.5
concentrations.
As described in reaching his conclusions in the section below, the
Administrator judges that, despite these uncertainties and limitations,
studies that use restricted analyses can provide supplemental
information for consideration in reaching conclusions regarding both
the adequacy and level of the standard. He notes two studies (Di et
al., 2017b and Dominici et al., 2019) are available in this
reconsideration that report means in their restricted analyses
(restricting annual average PM2.5 exposure below 12 [mu]g/
m\3\) and used population-weighted approaches to estimate
PM2.5 exposures and these studies report mean
PM2.5 concentrations of 9.6 [micro]g/m\3\. He recognizes
that these studies are just one line of evidence for consideration and
that along with the broader evidence base, including the key
epidemiologic studies, these studies provide support that the level of
the primary annual PM2.5 standard should be set below 10
[micro]g/m\3\.
We disagree with the commenters that concerns about relating the
mean PM2.5 concentrations from restricted analyses to design
values are not valid. As an initial matter, restricted analyses were
not available and did not inform the 2012 decision to revise the annual
PM2.5 standard level to 12.0 [micro]g/m\3\. The approach in
2012 in revising the annual standard was to set the level to somewhat
below the mean of key epidemiologic studies. As noted above, while the
EPA believes that restricted analyses can help inform conclusions
regarding the adequacy and the level of the primary annual
PM2.5 standard, in the context of placing the studies in a
decision framework to inform the appropriate level of the annual
PM2.5 standard, the EPA has not deviated from its approach
from the 2012 review. Given that restricted analyses are new since the
2012 review, the EPA disagrees with commenters that uncertainties
associated with these studies should not be considered, and that these
studies should be used in a similar manner to their main analyses in
taking an approach to set a level of the standard somewhat below the
lowest long-term reported mean PM2.5 concentration.
Specifically, as detailed above there are uncertainties and limitations
associated with relating the mean PM2.5 concentrations from
these studies to design values for studies that use restricted
analyses, and many of these studies did not expressly report a mean
PM2.5 concentration for the restricted analysis which makes
it impossible to make such a comparison.
Several commenters contend that in considering the accountability
studies, the EPA inappropriately reached conclusions regarding the
level of the primary annual PM2.5 standard based on the
starting PM2.5 concentrations of these studies, rather than
the ending concentrations (i.e., concentrations after a policy was
implemented). The commenters assert that these studies provide support
for revising the level of the primary annual PM2.5 standard
to below the proposed range of 9-10 [micro]g/m\3\ to protect public
health with an adequate margin of safety.
Accountability studies examine the effect of a policy on reducing
PM2.5 concentrations in ambient air and evaluate whether
such reductions were observed to also lead to reductions in
PM2.5- associated health outcomes (e.g., mortality).
Additionally, accountability studies can reduce uncertainties related
to residual confounding of temporal and spatial factors (U.S. EPA,
2022a, p. 3-25). Prior to implementation of the policies, three
accountability studies newly available in this reconsideration and
assessed in the ISA Supplement, report mean PM2.5
concentrations below the level of the current annual standard level
(12.0 [micro]g/m\3\) and ranged from 10.0 [micro]g/m\3\ to 11.1
[micro]g/m\3\ (Sanders et al., 2020b; Corrigan et al., 2018; and
Henneman et al., 2019). These studies suggest that public health
improvements may occur following the implementation of a policy that
reduces annual average PM2.5 concentrations below the level
of the current standard of 12.0 [micro]g/m\3\, and potentially below
the lowest ``starting'' concentrations in these studies of 10.0
[micro]g/m\3\. However, while the small number of studies may provide
limited information related to informing the adequacy and level of the
annual PM2.5 standard, we note that accountability studies
are only one line of evidence, and that these studies provide
supplemental information for consideration in addition to the full body
of evidence. Further, the EPA does not believe it would be appropriate
to determine the level of the standard by reference to ending
concentrations in accountability studies. Accountability studies are
most informative in demonstrating that public health improvements may
occur following the implementation of a policy that reduces annual
average PM2.5 concentrations below the level of the current
standard of 12.0 [micro]g/m\3\, and potentially below the lowest
``starting'' concentrations in these studies of 10.0 [micro]g/m\3\.
However, the EPA finds the available information from accountability
studies is too limited to support a conclusion that the
[[Page 16265]]
appropriate level at which to set the primary annual PM2.5
standard would be equal to the ending concentrations of those studies,
as the commenters suggest. These studies demonstrate that there are
reductions in health outcomes when PM2.5 concentrations are
reduced in these studies from the starting concentration to the ending
concentration, but do not provide support for health effect
associations at or below the ending concentrations that would warrant a
more stringent standard.
Commenters disagree with the Administrator placing less weight on
the epidemiologic studies conducted in Canada when reaching conclusions
regarding the level of the primary annual PM2.5 standard.
These commenters argue that the Canadian epidemiologic studies provide
support for setting the level at the lowest end of the proposed range
(i.e., 8 [micro]g/m\3\) because they report mean PM2.5
concentrations, in some cases, below 8 [micro]g/m\3\. Commenters
disagree with the EPA's reasoning for placing less weight on the
Canadian epidemiologic studies, suggesting it conflicts with the
approaches in previous PM NAAQS reviews and arguing that the findings
of the Canadian epidemiologic studies can be directly translated into a
primary annual PM2.5 standard. Additionally, while the
commenters disagree with the EPA's approach for considering the study-
reported mean PM2.5 concentrations and design values in
general, they note that the CASAC, in their review of the 2021 PA,
noted that ``while there may be no design value in Canada, there are
data that indicate what a U.S. design value would be if an area average
like that found in the Canadian studies were to occur in the U.S.''
(Sheppard, 2022a, p. 13 of consensus responses). The commenters contend
that the EPA failed to acknowledge this advice from the CASAC,
specifically noting that the majority of the CASAC highlighted Canadian
epidemiologic studies as a part of their rationale for revising the
level of the primary annual PM2.5 standard to within the
range of 8-10 [micro]g/m\3\.
In considering the information from the epidemiologic studies in
reaching his conclusions, the Administrator considered the full body of
evidence, including studies conducted in the U.S. and Canada. However,
as described in the proposal and in section II.B.4 below, the
Administrator also recognizes that the exposure environments in the
U.S. are different from those in Canada. In particular, the U.S.
population density is approximately 43 people per square kilometer in
the contiguous U.S.\98\ compared to Canada, which has one of the lowest
population densities on the Earth with 4.2 people per square kilometer
(Statistics Canada, 2023). This difference in population density
between the U.S. and Canada was not as apparent, and did not need to be
highlighted, in the 2012 review given that the available Canadian
epidemiologic studies used population-weighting and focused on urban
areas where monitors were available and population densities were more
comparable with those in the U.S. Given this, the study-reported mean
concentrations from U.S. and Canadian studies in the 2012 review were
very similar. The recent epidemiologic evidence available in this
reconsideration, however, includes studies that utilize approaches that
highlight the importance of considering the differences between the two
exposure environments in the U.S. versus Canada. When focusing on the
recently available Canadian monitor-based epidemiologic studies in this
reconsideration, the information indicates that these studies, unlike
the studies available in the 2012 review, do not apply population
weighting (e.g., Lavigne et al., 2018; Liu et al., 2019). As noted in
responding to other public comments above, the absence of population
weighting is an important consideration that limits the utility of
these studies in informing the appropriate level of the primary annual
PM2.5 standard. In addition, there are recently available
studies in the 2019 ISA and ISA Supplement that expand the geographical
extent of the epidemiologic study areas by estimating exposure
concentrations in areas where there are no monitors. To do this, these
studies use either a statistical extrapolation of monitored values or
use air quality modeling and other forms of data (e.g., hybrid model-
based approaches). For these Canadian studies, the EPA notes two
important considerations in using the information to directly translate
to policy decisions regarding the level of the annual standard in the
U.S. The first is that in incorporating a larger portion of Canada into
these recent studies, more rural areas are included, and as such, the
population densities and exposure environment differences become more
important. The second is that in analyses that evaluate and validate
hybrid models, there is less certainty in PM2.5 exposure
estimates in more rural areas, which are further from air quality
monitors and where PM2.5 concentrations in the ambient air
tend to be lower (U.S. EPA, 2022b, pp. 2-51 and 2-63). Additionally, it
is unclear what portion of the PM2.5 concentrations from
rural areas are contributing to the study reported mean. Given this,
studies that incorporate more rural areas into the epidemiologic
studies highlight the importance of considering the differences between
the population exposures in the studies themselves and in the U.S.
versus Canadian study areas, as well as the influence these differences
have on the interpretation of the epidemiologic study results. For
these reasons, while the Canadian epidemiologic studies provide
additional support for associations between PM2.5
concentrations and health effects, the long-term means from Canadian
epidemiologic studies are a less certain basis for informing the EPA's
selection of the annual standard level, given that it is a U.S.-based
standard.
---------------------------------------------------------------------------
\98\ All of the key U.S. epidemiologic studies considered in
this reconsideration focus on all or subsections of the continental
U.S.
---------------------------------------------------------------------------
With respect to the CASAC's advice in their review of the 2021
draft PA, the EPA recognizes that the majority of the CASAC pointed to
the Canadian studies as supporting their recommendation to revise the
annual standard level to within the range of 8-10 [micro]g/m\3\.
However, the EPA also notes that the CASAC did not advise the EPA to
revise the annual standard to a level that was below the study-reported
means in the key Canadian epidemiologic studies. Indeed, the CASAC
noted that some of the Canadian studies showed associations below 8
[micro]g/m\3\, but did not recommend that the Administrator consider
levels below 8 [micro]g/m\3\ for the annual standard. Further, based on
the CASAC's advice, the Administrator is not excluding Canadian studies
from his consideration in this reconsideration, but he is considering
them in light of the limitations and challenges presented and in the
context of the full body of available scientific evidence.
Lastly, the EPA disagrees with commenters that the findings of the
Canadian epidemiologic studies can be directly translated into a
primary annual PM2.5 standard based on the evaluation of the
relationship between U.S. study-reported mean PM2.5
concentrations and U.S. design values. It is unclear whether the
relationship between U.S. study-reported mean PM2.5
concentrations and U.S. design values (which, in the case of U.S.
hybrid model-based studies, indicates that design values are 15-18%
greater than area mean PM2.5 concentrations) would apply to
the Canadian epidemiologic studies and their reported mean
PM2.5
[[Page 16266]]
concentrations, given that these studies generally report lower
PM2.5 concentrations than the U.S.-based studies. As such,
interpreting the study-reported mean concentrations from the Canadian
studies in the context of a U.S.-based standard may present challenges
in directly and quantitatively informing decisions regarding potential
alternative levels of the annual standard, particularly noting the
different in exposure relationships in the U.S. versus Canada given the
large difference in population densities between the two countries.
Further, as mentioned above, while the CASAC advised the EPA to
consider the Canadian studies as relevant evidence and found that
placing weight on the Canadian studies supported their recommendation
to revise the annual standard level to within the range of 8-10
[micro]g/m\3\, the lower end of their recommended range for the level
of the annual standard did not extend below the lower study-reported
means from those studies.
Commenters who supported retaining and revising the primary annual
PM2.5 standard both raised concerns regarding how the EPA
used the scientific evidence and quantitative risk assessment related
to disparities in PM2.5 exposure and risk in informing
conclusions on the standard. Commenters who supported retaining the
standard assert that the available scientific evidence that
demonstrates disparities for minority populations do not support
revising the standard, noting that these studies are in areas that tend
to have large minority populations and more sources of PM. These
commenters contend that because the studies conclude that minority
populations experience more effects than others living in the same area
that something other than PM2.5 concentrations in ambient
air is causing the disproportionate impact on minority populations,
providing proximity to a source as an example. The commenters note that
it is unclear how a national standard will reduce exposure disparities
for population groups living in the same area, and further assert that
studies of exposure disparities among minority populations were
considered in reaching the 2020 final decision to retain the standards.
Conversely, commenters who support revising the standard assert
that the at-risk analyses conducted in the 2022 PA provide support for
revising the primary annual PM2.5 standard to a level of 8
[mu]g/m\3\. In particular, these commenters state that the at-risk
analysis demonstrated that while disparities in mortality risk remain
at a standard level of 9.0 [mu]g/m\3\, disparities in exposure are
significantly reduced for an alternative standard level of 8.0 [mu]g/
m\3\ (U.S. EPA, 2022b, p. 3-162).
As discussed in section I above, the primary (health-based) NAAQS
are established at a level that is requisite to protect public health,
including the health of sensitive or at-risk groups, with an adequate
margin of safety.\99\ In so doing, decisions on the NAAQS are based on
an explicit and comprehensive assessment of the current scientific
evidence and associated risk analyses. More specifically, the EPA
expressly considers the available information regarding health effects
among at-risk populations in decisions on the primary NAAQS. Where
populations with disparities in exposure and risk are among the at-risk
populations, the decision on the standards is based on providing
requisite protection for these and other at-risk populations and
lifestages.
---------------------------------------------------------------------------
\99\ The legislative history of section 109 indicates that a
primary standard is to be set at ``the maximum permissible ambient
air level . . . which will protect the health of any [sensitive]
group of the population,'' and that for this purpose ``reference
should be made to a representative sample of persons comprising the
sensitive group rather than to a single person in such a group.'' S.
Rep. No. 91-1196, 91st Cong., 2d Sess. 10 (1970); see also, e.g.,
Am. Lung Ass'n v. EPA, 134 F.3d 388, 389 (D.C. Cir. 1998) (``If a
pollutant adversely affects the health of these sensitive
individuals, EPA must strengthen the entire national standard'').
---------------------------------------------------------------------------
The Administrator expressly considered the available information
regarding health effects among at-risk populations in reaching the
proposed decisions that the current primary annual PM2.5
standard is not requisite to protect public health with an adequate
margin of safety, and should be revised. The 2019 ISA and ISA
Supplement identified children, older adults, people with pre-existing
diseases (cardiovascular disease and respiratory disease), minority
populations, and low SES populations as at-risk populations. The
Administrator is thus, in his final decision, establishing primary
PM2.5 standards which, in his judgment, will provide
protection for these at-risk populations, including minority
populations, with an adequate margin of safety.
With respect to the risk assessment, while the EPA notes that the
analyses support the conclusion that the primary PM2.5
standards are not adequate, as detailed further in the proposal and
above in section II.A.3, the EPA also cautions against an over-
interpretation of the absolute results. The quantitative risk
assessment provides estimates of PM2.5-attributable
mortality based on input data that include C-R functions from
epidemiologic studies that do not quantitatively account for
uncertainties in associations between PM2.5 exposure and
health effects at lower concentrations and are based on an air quality
adjustment approach that incorporates proportional decreases in
PM2.5 concentrations to meet lower alternative standard
levels. As a result, simulated air quality improvements used in the
risk assessment will always lead to proportional decreases in risk
(i.e., each additional [micro]g/m\3\ reduction produces additional
benefits with no clear stopping point), without considering the
substantially greater uncertainties associated with the relationship
between PM2.5 exposures and health effects at lower
concentrations.
The same is true for the new at-risk analysis in the risk
assessment presented in the 2022 PA that is based on a recent
epidemiologic study that is available in this reconsideration that
provides mortality risk coefficients for older adults (i.e., 65 years
and older) based on PM2.5 exposure and stratified by racial
and ethnic demographics. Generally, the results of at-risk analyses can
vary greatly depending on the inputs to the analyses, including the
representativeness of the populations and demographics captured by the
study areas that are a part of the analyses, as well as the available
C-R functions from epidemiologic studies that stratify by race and
ethnicity and the air quality adjustment approaches that are used to
simulate air quality at different standard levels. In fact, for this
at-risk analysis, the results are even more uncertain than similar
estimates from the overall risk assessment due to additional sources of
uncertainty specific to the at-risk analysis, such as using C-R
functions derived from smaller epidemiologic sample sizes along with
the sources of uncertainty that apply to the overall risk assessment
(U.S. EPA, 2022b, section 3.4.1.8). Additionally, in characterizing at-
risk populations, the at-risk analysis only used one of the air quality
adjustment approaches used in the overall risk assessment, which
decreases the potential representativeness of the PM2.5
concentrations across the study areas (U.S. EPA, 2022b, section
3.4.1.8). Lastly, this at-risk analysis relies on the stratified risk
coefficients from only one epidemiologic study.\100\ For these reasons,
the Administrator places little
[[Page 16267]]
weight on the absolute results of the risk assessment, including the
at-risk analysis, for purposes of selecting the level of the annual
standard that is requisite.
---------------------------------------------------------------------------
\100\ Additional information on all available at-risk
epidemiologic studies in this reconsideration are available in
section 3.4 and Appendix C of the 2022 PA (U.S. EPA, 2022b, section
3.4, Figure 3-17, and Appendix C, section C.3.2).
---------------------------------------------------------------------------
While there are substantial uncertainties in the absolute results
of the quantitative risk assessment, the EPA also notes that recent
scientific evidence evaluated in the ISA Supplement, which built upon
the 2019 PM ISA conclusions, found that the evidence ``[c]ontinue[s] to
support disparities in PM2.5 exposure and health risks by
race and ethnicity'' while studies of SES ``provide additional support
indicating there may be disparities in PM2.5 exposure and
health risk by SES'' (U.S. EPA, 2022a, p. 5-4). Thus, in light of the
statutory requirement to provide protection for at-risk populations, it
is not surprising that the stratified population results of the risk
assessment suggest that meeting a revised standard would result in
higher risk reductions for minority and low SES populations.
In conclusion, the EPA recognizes that the at-risk analysis was
based on one epidemiologic study that stratified by race/ethnicity for
older adults (e.g., 65+ years old) and that there is increasing
uncertainty in quantitative estimates of stratified risk estimates at
the lower end of the range of standard levels assessed. Moreover, the
EPA finds that the goal of the NAAQS is to provide the requisite
protection to at-risk groups, and where minority populations are
included among the at-risk groups, providing requisite protection to
minority populations will also result in protecting the public health
of other populations. Thus, in setting the NAAQS to protect the health
of at-risk groups with an adequate margin of safety, the Administrator
is selecting the standard that will provide requisite protection,
including for minority populations and other at-risk populations, which
also generally results in protecting the public health of other
populations and reducing risk disparities.
A number of commenters, primarily from industries and industry
groups and some States, support the EPA's proposed decision to retain
the primary 24-hour PM2.5 standard. Many of these commenters
contend that the available scientific evidence and quantitative
information has not significantly changed since the 2020 final decision
and note that important uncertainties remain. The commenters agree with
the EPA's conclusions regarding the controlled human exposure studies
and their relationship to short-term peak PM2.5
concentrations in ambient air. These commenters also noted the primary
annual and 24-hour PM2.5 standards work together to provide
public health protection, with the 98th percentile form of the 24-hour
standard effectively limiting peak daily concentrations. The commenters
agree with the EPA that the current suite of standards maintain
subdaily concentrations below the higher concentrations in controlled
human exposure studies where more consistent health effects are
observed. Commenters also agree with the EPA's conclusions that the
epidemiologic studies are not useful for informing decisions on the
level of the primary 24-hour PM2.5 standard because the
standard focuses on reducing peak exposures with its 98th percentile
form, while the epidemiologic studies often focus on the mean or median
as the percentile for which associations with short-term exposures are
observed. These commenters also agree with the EPA's focus on U.S.-
based studies because of differences compared to Canadian studies. The
commenters also generally agree with the Administrator's judgment that
it was appropriate to place less weight on the risk assessment, noting
that the annual standard is controlling in most areas of the country
and revising the annual standard would have the most potential to
reduce risk related to PM2.5 exposures and would reduce both
average (annual) and peak (daily) PM2.5 concentrations.
Finally, these commenters note that the CASAC did not reach consensus
on whether the current primary 24-hour PM2.5 standard should
be revised, and they agree with the minority of the CASAC's
recommendation in their review of the 2021 draft PA that the primary
24-hour primary PM2.5 standard should be retained. These
commenters also note the CASAC's support in their review of the 2019
draft PA for retaining the primary 24-hour PM2.5 standard.
A number of commenters, primarily from public health and
environmental organizations and some States, oppose the EPA's proposed
decision to retain the primary 24-hour PM2.5 standard. These
commenters support revising the level of the primary 24-hour
PM2.5 standard, contending that a more stringent standard is
necessary to provide requisite public health protection with an
adequate margin of safety, particularly for at-risk groups. In so
doing, these commenters place weight on the same aspects of the
available scientific evidence as the majority of the CASAC in their
review of the 2021 draft PA, and generally advocate for revising the
level of the standard to within the range of 25-30 [micro]g/m\3\ as
recommended by the majority of the CASAC. Some of these commenters
support a level no higher than 25 [micro]g/m\3\ and others support a
level of 20 [micro]g/m\3\. These commenters generally cite to the
available scientific evidence, including evidence of disproportionate
exposures and risks for certain at-risk groups, and the CASAC's advice
in support for their recommendation. Some of these commenters also
suggest that decisions regarding the primary 24-hour PM2.5
standard should not be related to decisions on the primary annual
PM2.5 standard.
As an initial matter, the EPA disagrees with commenters who suggest
that decisions regarding the primary 24-hour PM2.5 standard
should not be related to decisions on the primary annual
PM2.5 standard. In reviewing the adequacy of the public
health protection afforded by the primary PM2.5 standards,
the Administrator's consistent past practice has been to evaluate the
combination of the annual and 24-hour standards together. In 2012, the
then-Administrator concluded that the most effective and efficient way
to reduce total population risk associated with both long- and short-
term PM2.5 exposures was to set a generally controlling
annual standard, and to provide supplemental protection by means of a
24-hour standard set at the appropriate level. In so doing, the then-
Administrator explicitly recognized that potential air quality changes
associated with meeting a revised annual standard (with a level of 12
[micro]g/m\3\) would result in lowering risks associated with both
long- and short-term PM2.5 exposures by lowering the overall
distribution of air quality concentrations, and that retaining a 24-
hour standard at the appropriate level would ensure an adequate margin
of safety against short-term effects in areas with high peak-to-mean
ratios (78 FR 3163, January 15, 2013). In this reconsideration, also,
the Administrator considers it appropriate to rely on the annual
standard (arithmetic mean, averaged over three years) for targeting
protection against both long- and short-term PM2.5
exposures, noting that the annual standard is typically controlling,
while the 24-hour standard (98th percentile, averaged over three years)
can provide supplemental protection against the occurrence of peak 24-
hour PM2.5 concentrations (U.S. EPA, 2022b, section 3.6.3).
Further, the Administrator notes that, as in the 2012 review, changes
in PM2.5 air quality to meet a revised annual standard would
[[Page 16268]]
affect the entire distribution of long- and short-term concentrations,
thus likely resulting not only in lower short- and long-term
PM2.5 concentrations near the middle of the air quality
distribution, but also in fewer and lower short-term peak
PM2.5 concentrations.\101\ Thus, the Administrator continues
to conclude it is appropriate to consider whether the annual and 24-
hour standards together provide requisite protection of public health,
rather than considering each standard in isolation.
---------------------------------------------------------------------------
\101\ Similarly, the Administrator recognizes that changes in
air quality to meet a 24-hour standard, would result not only in
fewer and lower peak 24-hour PM2.5 concentrations, but
also in lower annual average PM2.5 concentrations.
However, as noted in 2012, an approach that relied on setting the
level of the 24-hour standard such that the 24-hour standard was
generally controlling would be less effective and result in less
uniform protection across the U.S. than an approach that focuses on
setting a generally controlling annual standard (78 FR 3163, January
15, 2013).
---------------------------------------------------------------------------
Regarding the appropriate basis for determining the level of the
24-hour standard, a number of commenters who support revising the
primary 24-hour PM2.5 standard to a lower level contend that
the EPA should not rely on the controlled human exposure studies in
evaluating the adequacy of the public health protection afforded by the
primary 24-hour PM2.5 standard. These commenters support
this view by citing the CASAC comments in their review of the 2019
draft PA which advised that controlled human exposure studies have
limitations that may impact their ability to inform conclusions on the
adequacy of the public health protection afforded by the primary 24-
hour PM2.5 standard. Commenters noted that these studies do
not include the most vulnerable populations and often involve exposure
to only one pollutant to elicit a response, and therefore are not
representative of real-world exposures.
Other commenters support the EPA's use of the controlled human
exposure studies to inform the adequacy of the public health protection
and note that the 24-hour standard must at least provide protection
against the health effects observed in controlled human exposure
studies. Some of the commenters cite the Wyatt et al. (2020) study that
demonstrated cardiovascular effects following 2-hour exposures to 120
[micro]g/m\3\ and 4-hour exposures to 37.8 [micro]g/m\3\. Some of these
commenters contend that the current primary 24-hour PM2.5
standard allows PM2.5 exposures comparable to those observed
to elicit effects in the controlled human exposure studies, and
therefore, the EPA must revise the level of the current standard to
protect public health. To support this view, some commenters submitted
an analysis of monitoring data from 2017-2020, which compares the
number of days per year where maximum daily PM2.5
concentrations exceed 120 [micro]g/m\3\ and 37.8 [micro]g/m\3\.
Additionally, other commenters assert that the EPA should focus
less on peak PM2.5 concentrations ``typically measured'' in
areas meeting the current primary PM2.5 standards even if
they do not exceed the concentrations in the controlled human exposure
studies because, in their view, the standard needs to protect against
atypical exposures to atypical peak PM2.5 concentrations.
These commenters conclude that, when considered together, the
controlled human exposure studies and the epidemiologic studies warrant
strengthening the level of the primary 24-hour PM2.5
standard.
The EPA generally disagrees with commenters who contend that it is
inappropriate to rely on the controlled human exposures studies in
evaluating the adequacy of the public health protection afforded by the
primary 24-hour PM2.5 standard. The Agency considers these
studies informative both for establishing biological plausibility and
for determining an appropriate level for the 24-hour standard. When
looking to the experimental studies, the EPA finds that the 2019 ISA
and ISA Supplement included controlled human exposure studies that
report statistically significant effects on one or more indicators of
cardiovascular function following 2-hour exposures to PM2.5
concentrations at and above 120 [mu]g/m\3\ (and at and above 149 [mu]g/
m\3\ for vascular impairment, the effect shown to be most consistent
across studies). As noted in the 2019 ISA, these studies are important
in establishing biological plausibility for PM2.5 exposures
causing more serious health effects, such as those seen in short-term
exposure epidemiologic studies, and they provide support that more
adverse effects may be experienced following longer exposure durations
and/or exposure to higher concentrations. Additionally, one controlled
human exposure study assessed in the ISA Supplement reports evidence of
some effects for cardiovascular markers at lower PM2.5
concentrations, 4-hour exposures to 37.8 [micro]g/m\3\ (Wyatt et al.,
2020). However, there is inconsistent evidence for inflammation in
other controlled human exposure studies evaluated in the 2019 ISA. The
EPA notes that although the controlled human exposure studies do not
provide a threshold below which no effects occur, the observed effects
in these controlled human exposures studies are ones that signal an
intermediate effect in the body, likely due to short-term exposure to
PM2.5, and typically would not, by themselves, be judged as
adverse (88 FR 5620, January 27, 2023) 102 103
---------------------------------------------------------------------------
\102\ Judgments regarding adversity or health significance of
measurable physiological responses to air pollutants have been
informed by guidance, criteria or interpretative statements
developed within the public health community, including the American
Thoracic Society (ATS) and the European Respiratory Society (ERS),
which cooperatively updated the ATS 2000 statement What Constitutes
an Adverse Health Effect of Air Pollution (ATS, 2000) with new
scientific findings, including the evidence related to air pollution
and the cardiovascular system (Thurston et al., 2017).
\103\ The ATS/ERS described its 2017 statement as one ``intended
to provide guidance to policymakers, clinicians and public health
professionals, as well as others who interpret the scientific
evidence on the health effects of air pollution for risk management
purposes'' and further notes that ``considerations as to what
constitutes an adverse health effect, in order to provide guidance
to researchers and policymakers when new health effects markers or
health outcome associations might be reported in future.'' The most
recent policy statement by the ATS, which once again broadens its
discussion of effects, responses and biomarkers to reflect the
expansion of scientific research in these areas, reiterates that
concept, conveying that it does not offer ``strict rules or
numerical criteria, but rather proposes considerations to be weighed
in setting boundaries between adverse and nonadverse health
effects,'' providing a general framework for interpreting evidence
that proposes a ``set of considerations that can be applied in
forming judgments'' for this context (Thurston et al., 2017).
---------------------------------------------------------------------------
The EPA notes that the majority of the CASAC, in their review of
the 2021 draft PA, commented that these controlled human exposure
studies generally do not include populations with substantially
increased risk from exposure to PM2.5, such as children,
older adults, or those with more severe underlying illness, and often
involve exposure to only one pollutant to elicit a response. However,
both the majority and the minority of the CASAC explained that, even
taking into consideration their limitations, the controlled human
exposure studies provide some support for assessing the adequacy of the
24-hour standard.\104\
---------------------------------------------------------------------------
\104\ In their review of the 2021 draft PA, the majority of the
CASAC advised that ``evidence of effects from controlled human
exposure studies with exposures close to the current standard
support epidemiologic evidence for lowering the standard''
(Sheppard, 2022a, p. 4 of consensus letter). The minority of the
CASAC also advised that it was appropriate to place ``more emphasis
on the controlled human exposure studies, showing effects at
PM2.5 concentrations well above those typically measured
in areas meeting the current standards'' (Sheppard, 2022a, p. 4 of
consensus letter), in evaluating adequacy of the 24-hour standard.
---------------------------------------------------------------------------
The EPA agrees with the CASAC that the controlled human exposure
studies generally do not include populations with substantially
increased risk from exposure to PM2.5, like children, older
adults, or those with pre-existing severe illness, like cardiovascular
effects. As such, and as an initial note, these
[[Page 16269]]
studies are therefore somewhat limited in their ability to inform at
what concentrations effects may be elicited in at-risk populations. In
spite of this limitation, the EPA also agrees with the CASAC, that even
taking into consideration the limitations of the controlled human
exposure studies, these studies can provide some support for evaluating
the adequacy of the 24-hour standard. However, the EPA further notes
that while the controlled human exposure studies are important in
establishing biological plausibility, the health outcomes observed in
these controlled human exposure studies are often ``intermediate''
outcomes (i.e., not always clearly adverse) and therefore it is unclear
how the importance of the effects observed in the studies should be
interpreted with respect to adversity to public health. The EPA finds
that it is appropriate to consider these study limitations in assessing
the information provided by controlled human exposure studies in
evaluating the adequacy of the primary 24-hour PM2.5
standard.
The EPA agrees with commenters that the primary 24-hour
PM2.5 standard must at least provide protection against the
health effects consistently observed in controlled human exposure
studies. As discussed in the proposal, the EPA looks at whether the
exposures that elicit a response following exposure to PM2.5
in the controlled human exposure studies occur under recent air quality
conditions in areas meeting the current standards. Based on these air
quality analyses, the EPA concludes that these types of exposures very
rarely occur when the current standards are being met.
The EPA did receive multiple comments questioning these results and
the approach in the EPA's analyses. For example, some commenters
submitted an analysis of monitoring data from 2017-2020, which compares
the number of days per year where maximum daily PM2.5
concentrations exceed 120 [micro]g/m\3\ and 37.8 [micro]g/m\3\ and
evaluate the number of days subset by groups of monitors with 4-year
average PM2.5 concentrations close to the levels of
combinations of current and proposed annual (+/- 0.2 [micro]g/m\3\) and
24-hour (+/-2 [micro]g/m\3\) PM2.5 standards. To support
their view that the primary PM2.5 standards should be
revised, the commenters describe decreases in days per monitor per year
with 2-hour maximum concentrations greater than 120 [micro]g/m\3\ and
4-hour maximum concentrations greater than 37.8 [micro]g/m\3\ when
comparing monitors that achieve close to 10 and 30 [micro]g/m\3\ versus
monitors that meet close to 8 [micro]g/m\3\ and 25 [micro]g/m\3\. The
commenters noted decreases in the number of days per monitor per year
with 2-hour maximum concentrations over 120 [micro]g/m\3\ and 4-hour
max concentration over 37.8 [micro]g/m\3\ were also seen when comparing
monitors close to achieving 24-hour standards with levels of 35
[micro]g/m\3\ versus 25 [micro]g/m\3\.
First, the EPA notes that this analysis submitted by commenters was
limited to a very small number of monitors and did not include a
national perspective. Second, the EPA notes that this analysis focused
on number of days (rather than the number of times) where there was a
2-hour maximum concentration over 120 [micro]g/m\3\ or a 4-hour max
concentration over 37.8 [micro]g/m\3\. In order to evaluate the
protection provided by the current 24-hour standard against peak
exposures, including exposures with 2-hour concentrations greater than
120 [micro]g/m\3\ and 4-hour concentrations greater than 37.8 [micro]g/
m\3\, the EPA considers it more informative and appropriate from a
public health perspective to assess the number of times a subdaily
exposure of concern occurs in a year, rather than the number of days on
which they occur because the former identifies more potential exposures
of concern and provides more information about the scale and scope of
the occurrences of those exposures. Lastly, the analyses allowed
monitors somewhat above the standards to be included. Therefore, it is
unclear whether the exceedances of the 2-hour or 4-hour benchmarks
would still have occurred if the area had actually been meeting the
current primary PM2.5 standards. However, in considering the
analyses submitted by the commenters, the EPA conducted new analyses
\105\ that looked at all individual monitors across the U.S. and
evaluated the percentage of times the monitors experienced a 2-hour
maximum concentration over 120 [micro]g/m\3\ or a 4-hour max
concentration over 37.8 [micro]g/m\3\ when that monitor was meeting the
current standards. Further, given that the Administrator concludes that
the level of the current primary annual PM2.5 is not
adequate and that it should be revised to 9.0 [micro]g/m\3\, the new
analysis evaluates the percentage of times during a recent 3-year
period (i.e. 2019-2021) that individual monitors experienced a 2-hour
maximum concentration over 120 [micro]g/m\3\ or a 4-hour max
concentration over 37.8 [micro]g/m\3\ when that monitor was meeting the
current primary 24-hour PM2.5 standard with its level of 35
[micro]g/m\3\ and a revised primary annual PM2.5 standard of
9.0 [micro]g/m\3\.
---------------------------------------------------------------------------
\105\ Jones et al. (2023). Comparison of Occurrence of
Scientifically Relevant Air Quality Observations Between Design
Value Groups. Memorandum to the Rulemaking Docket for the Review of
the National Ambient Air Quality Standards for Particulate Matter
(EPA-HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
In evaluating the results from the new analyses, it is important to
keep in mind that the 2019 ISA and ISA Supplement concluded that the
most consistent evidence from the controlled human exposures studies is
for impaired vascular function following 2-hour exposures to average
PM2.5 concentrations at and above about 120 [micro]g/m\3\,
with less consistent evidence for effects following exposures to
concentrations lower than 120 [micro]g/m\3\. The new analyses show that
across all monitors, on average, only 0.029 percent of 2-hour
observations reach PM2.5 concentrations higher than 120
[micro]g/m\3\ in areas meeting the current 24-hour standard and a
revised annual standard of 9.0 [micro]g/m\3\. Further, recognizing that
one purpose of the 24-hour standard is to protect against exposure in
areas with high peak-to-mean ratios, when assessing the monitors
individually across the U.S. under these same conditions, the monitors
reporting the highest PM2.5 concentrations have only 0.47
percent of 2-hour observations reach PM2.5 concentrations
higher than 120 [micro]g/m\3\.
Additionally, the analyses also evaluated the frequency of
reporting a 4-hour maximum concentration over 37.8 [micro]g/m\3\ when
monitors were meeting the current 24-hour standard and a revised annual
standard of 9.0 [micro]g/m\3\. For this part of the analysis, the EPA
finds that across all monitors, on average, only 0.41 percent of 4-hour
observations reach PM2.5 concentrations higher than 37.8
[micro]g/m\3\ in areas meeting the current 24-hour standard and a
revised annual standard of 9.0 [micro]g/m\3\. Further, when assessing
the monitors individually across the U.S. under these same conditions,
the monitors reporting the highest PM2.5 concentrations have
only 2.6 percent of 4-hour observations reach PM2.5
concentrations higher than 37.8 [micro]g/m\3\. Thus, the EPA disagrees
with commenters that the current primary 24-hour PM2.5
standard typically allows PM2.5 exposures at or above those
observed to cause health effects in controlled human exposure studies.
Furthermore, the EPA notes that in light of the small number of
occurrences and the intermediate nature of the effects observed in
Wyatt et al. (2020) at concentrations of 37.8 [micro]g/m\3\ (i.e.,
effects that typically would not, by themselves, be judged as adverse),
there is substantial basis to doubt whether further improvements in
public health
[[Page 16270]]
would be achieved by further reducing these exposures. In drawing this
conclusion, the EPA notes the lack of evidence of effects from
controlled human exposure studies at levels below the current 24-hour
standard and the fact that the results of Wyatt et al. (2020) are
inconsistent with other currently available studies, and this study
only observes intermediate effects.
In response to commenters that cited the majority of the CASAC's
view that, in general, ``[t]here is . . . less confidence that the
annual standard could adequately protect against health effects of
short-term exposures'' (Sheppard, 2022a, p. 4 of consensus letter), the
EPA disagrees with the majority of CASAC, noting that the results of
the EPA's analysis suggest that high peak concentrations are extremely
infrequent in areas meeting an annual standard of 9.0 [micro]g/m\3\,
occurring less than 0.029-0.41 percent of the time (for 2-hour
concentrations >120 [micro]g/m\3\ and 4-hour concentrations >37.8
[micro]g/m\3\, respectively). This suggests that in most locations,
even the upper tail of the distribution would be controlled quite well
under a revised annual standard. With regard to the likelihood that the
current standards would allow peak concentrations that are clearly of
concern from a health perspective, therefore, the EPA concludes that
such occurrences are extremely infrequent--and will be even less
frequent under the improved air quality conditions associated with
meeting a revised annual PM2.5 standard of 9.0 [micro]g/
m\3\.
A number of commenters who support revising the primary 24-hour
PM2.5 standard to a lower level suggest that the available
epidemiologic evidence provides support for such a revision. To support
their view, the commenters note that the currently available evidence,
including a number of epidemiologic studies that demonstrate
associations between short-term PM2.5 exposures and health
effects, provides support for causal relationships for short-term
PM2.5 exposures and health effects as described in the 2019
ISA and ISA Supplement. The commenters further note that the available
epidemiologic studies include diverse populations that are broadly
representative of the U.S. population, including at-risk populations,
which they assert is an advantage over the controlled human exposure
studies and the risk assessment, which are not as broadly
representative.
These commenters highlight a number of specific epidemiologic
studies that they suggest provide support for revising the level of the
24-hour standard. Additionally, commenters contend that there are
epidemiologic studies using restricted analyses that show that positive
and statistically significant associations between short-term
PM2.5 exposure and mortality persist at daily mean
concentrations below 25 [micro]g/m\3\. The commenters also cite several
studies that provide no evidence of a threshold. These commenters also
point to the CASAC advice in their review of the 2021 draft PA, where
the majority of the CASAC cited epidemiologic studies using restricted
analyses as offering support for revision. The commenters argue that
the EPA cannot base discretion on uncertainties related to the methods
used in restricted analyses in the epidemiologic studies. In so doing,
these commenters disagree with the EPA that it is important to take
into consideration that these studies do not consider the form or
averaging time of the 24-hour standard. Finally, the commenters claim
that while the EPA stated that the study-reported means from
epidemiologic studies that use restricted analyses are more useful for
identifying impacts from typical 24-hour exposures than for peak 24-
hour exposures, the commenters assert that the studies also indicate
that there are health risks at relatively high concentrations below the
current level of the primary 24-hour PM2.5 standard that
must be addressed.
As noted by the commenters, epidemiologic studies that show
positive and statistically significant associations between short-term
PM2.5 exposure and mortality provide support for the causal
determination in the 2019 ISA. The EPA also agrees that the available
epidemiologic studies include diverse populations that are broadly
representative of the U.S. population, including at-risk populations.
Further, the EPA agrees that studies evaluated in the 2019 ISA and the
ISA Supplement continue to provide evidence of linear, no-threshold
concentration-response relationships, but with less certainty in the
shape of the curve at lower concentrations (i.e., below about 8
[micro]g/m\3\), with some recent studies providing evidence for either
a sublinear, linear, or supralinear relationship at these lower
concentrations (U.S. EPA, 2019a, section 11.2.4; U.S. EPA, 2022a,
section 2.2.3.2).
However, findings of positive, significant associations in short-
term epidemiologic studies do not directly indicate that short-term
effects would occur in areas meeting the 24-hour standard and
therefore, do not directly address the question of whether the current
24-hour standard is adequate. While short-term epidemiologic studies
evaluate associations between distributions of ambient PM2.5
and health outcomes, they do not identify the specific exposures (i.e.,
a specific 24-hour concentration) that can lead to the reported
effects. Short-term epidemiologic studies evaluate the association
between day-to-day variation in daily (24-hour) PM2.5
exposure and health endpoints (e.g., mortality) to understand how these
changes in air pollution concentrations are associated with changes in
health outcomes. But these studies do not report daily concentrations;
rather, they report the long-term mean concentration of the 24-hour
PM2.5 concentrations over the entire multi-year period of
the study, and typically report their results as a relative risk (e.g.,
for each 10 [micro]g/m\3\ increase in PM2.5, the risk of
mortality or cardiovascular hospital admissions increases by a certain
percentage, across the full range of the 24-hour PM2.5
concentrations in the study). This means that there is no specific
point in the air quality distribution of any epidemiologic study that
represents a ``bright line'' at and above which effects have been
observed and below which effects have not been observed. Nor, as noted
above, do these studies allow for any direct inferences about health
impacts associated with the short-term ``peak'' exposures that the
primary 24-hour standard is designed to protect against. While there
can be considerable variability in daily exposures over a multi-year
study period, most of the estimated exposures in these epidemiologic
studies reflect days with ambient PM2.5 concentrations
around the mean or middle of the air quality distributions examined
(i.e., ``typical'' days rather than days with extremely high or
extremely low concentrations). This is true of long-term epidemiologic
studies as well. The difference between epidemiologic studies examining
associations with long-term exposures and short-term exposures is
comparing different levels of exposure over different exposure
durations (i.e., long-term studies exposures are defined as those that
are annual or multi-year, while short-term exposures are defined as
those that are mostly 24-hour) (U.S. EPA, 2019a, section P.3.1). Thus,
in both cases, and in the absence of a discernible threshold,
epidemiologic studies of short-term and long-term exposures provide the
strongest support and confidence for reported health effect
associations around the middle portion of the PM2.5 air
quality distribution (e.g., the study-reported mean PM2.5
[[Page 16271]]
concentration), which corresponds to the bulk of the underlying data,
rather than at the extreme upper or lower ends of the distribution.
However, the difference between the annual standard and the 24-hour
standard, aside from averaging times, is that the form of the annual
standard is a mean PM2.5 concentration, which is based on
the bulk of the air quality data, while the form of the 24-hour
standard is a 98th percentile form, which is based on peak
concentrations. Both long-term and short-term epidemiologic studies are
informative for determining the appropriate level of the annual
PM2.5 standard, which is designed to control ``typical''
daily exposures and risks, because these studies most often report
long-term mean (or median) PM2.5 concentrations that are
representative of ``typical'' exposures that are associated with health
effects. In contrast, while the short-term epidemiologic studies
examine health effects associated with shorter exposure durations
(e.g., mostly 24-hour exposures), these studies are less informative
for determining the appropriate level of the 24-hour PM2.5
standard because these studies do not report the 98th percentile
PM2.5 concentrations,\106\ which is more directly comparable
to the form of the 24-hour standard. Additionally, if the 98th
percentile of data were reported, the EPA would consider the peak
concentrations observed in these studies (which by definition rarely
occur) in conjunction with other supporting evidence. However, as
already noted, there is an absence of new information in this
reconsideration (either from controlled human exposure studies or
epidemiologic studies) suggesting that peak concentrations just below
the level of the current 24-hour standard (with its level of 35
[micro]g/m\3\) are associated with adverse effects. Instead, the
evidence links risk to more typical daily exposures near the middle of
the air quality distribution--exposures most effectively controlled
through a strengthening of the annual standard. As noted in the 2012
final rule, ``reducing the annual standard is the most efficient way to
reduce the risks from short-term exposures . . . as the bulk of the
risk comes from the large number of days across the bulk of the air
quality distribution, not the relatively small number of days with peak
concentrations'' (78 FR 3156, January 15, 2013).
---------------------------------------------------------------------------
\106\ In the 2022 PA, the EPA has identified a number of key
areas for additional research and data collection for
PM2.5, based on the uncertainties and limitations that
remain in the scientific evidence and technical information. In
addition to research and data collection, the EPA specifically
highlights additional information that could be reported in the
epidemiologic studies that may help inform future reviews of the
primary PM2.5 standards, including additional descriptive
statistics in the upper percentiles of the air quality distribution
(i.e., from the 95th to the 99th percentile), as well as the number
of days of concentrations and/or health events within each of these
percentiles (U.S. EPA, 2022a, section 3.7).
---------------------------------------------------------------------------
As noted above, in evaluating the adequacy of the current
standards, the EPA has consistently considered the annual standard
(based on arithmetic mean concentrations) and 24-hour standard (based
on 98th percentile concentrations) together in evaluating the public
health protection provided by the standards against the full
distribution of short- and long-term PM2.5 exposures.
Moreover, the EPA has previously noted that the annual standard is
generally controlling in most parts of the country, providing an
effective and efficient way to reduce total population risk to both
long- and short-term PM2.5 exposures, while the 24-hour
standard, with its 98th percentile form, provides supplemental
protection, particularly for areas with high peak-to-mean ratios of 24-
hour PM2.5 concentrations (78 FR 3158, January 15, 2013). In
such areas, annual average PM2.5 concentrations could be
quite low, and the 24-hour standard provides a means of ensuring
control of episodic peaks possibly associated with strong local or
seasonal sources, or PM2.5-related effects that may be
associated with shorter-than daily exposure periods. The approach taken
in evaluating the adequacy and alternative levels of the annual
standard has been to evaluate the long-term mean PM2.5
concentrations of both long-term and short-term key epidemiologic
studies, where we have the most confidence in the reported health
effects association, while also giving some consideration to lower
percentiles of the air quality distribution (e.g., 25th percentiles).
However, using a similar approach to evaluate the adequacy of the
current and any potential alternative levels of the 24-hour standard
with short-term epidemiologic studies, as the majority of CASAC and
some commenters are suggesting, presents challenges.
Short-term epidemiologic studies, including those that use
restricted analyses, often report metrics that include mean
PM2.5 concentrations, with some studies also reporting lower
percentiles, such as the 25th percentile. As previously noted above,
for studies of daily PM2.5 exposure, which examine
associations between day-to-day variation in PM2.5
concentrations and health outcomes, often over several years, most of
the estimated exposures reflect days with ambient PM2.5
concentrations around the middle of the air quality distributions
examined (i.e., the mean or median). However, there is not a metric or
statistic reported in short-term epidemiologic studies that allows for
a direct comparison to the current 24-hour PM2.5 standard
and its 98th percentile form. While a 98th percentile of
PM2.5 concentrations is a metric that might be more closely
compared to the 24-hour standard level, 98th percentile
PM2.5 concentrations were not reported in key epidemiologic
studies. Consistent with the Administrator's final decision in 2012,
the EPA notes that even if 98th percentile values were reported, it
would be inappropriate to focus on these concentrations without also
considering the impact of a revised annual standard on short-term
concentrations, since many areas would be expected to experience
decreasing short- and long-term PM2.5 concentrations in
response to a revised annual standard (78 FR 3156, January 15, 2013).
Furthermore, in light of the scarcity of days at the very upper end of
the distribution, and to avoid placing undue reliance on the peak
concentrations observed in these studies (which by definition rarely
occur), the EPA finds that such values would need to be considered in
conjunction with other supporting evidence. In addition, as described
above, the other lines of evidence available for consideration by the
EPA do not indicate that the current primary 24-hour standard requires
revision to protect public health with an adequate margin of safety.
The EPA notes again the lack of corroborating evidence from controlled
human exposure studies. While the EPA agrees with the CASAC that the
controlled human exposure studies are limited in their ability to speak
to the concentrations at which effects may be elicited in at-risk
populations, as discussed above the lowest concentration associated
with effects is 37.8 [micro]g/m\3\ and the effects observed were
``intermediate'' outcomes that are not by themselves considered
adverse. We also note that, as detailed in section II.A.2.a above, the
study that observed intermediate effects at concentrations of 37.8
[micro]g/m\3\ was evaluated in the ISA Supplement and the results of
this study were inconsistent with the controlled human exposure studies
evaluated in the 2019 ISA. Additionally, as noted above, the EPA finds
that across all monitors, on average, only 0.41 percent of 4-hour
observations reach PM2.5 concentrations higher than 38
[micro]g/m\3\ in areas meeting the current 24-hour
[[Page 16272]]
standard and a revised annual standard of 9.0 [micro]g/m\3\. Given the
rarity of these occurrences and the fact that the effects associated
with exposures to this PM2.5 concentration have not been
found to be adverse in and of themselves, the EPA finds it reasonable
to conclude that this pattern of air quality will protect at-risk
populations, even though such populations were not in the study groups.
The EPA concludes that further evidence would be needed at specific
short-term (i.e., hourly or daily) concentrations below the level of
the current 24-hour standard to support any revision to the current 24-
hour standard.
With regard to the data that are available from the short-term
epidemiologic studies (which, as noted, do not include 98th percentile
values), the EPA considers it inappropriate to utilize the study-
reported means from the short-term epidemiologic evidence to assess the
adequacy of the 24-hour standard, with its 98th percentile form,
considering that the study-reported mean concentrations do not provide
meaningful insight regarding the frequency or health significance of
peak concentrations occurring during the study period. As indicated in
the 2022 PA, the study-reported means of short-term epidemiologic
studies do not serve a purpose in determining a level at which we can
confidently attribute effects to the impact of ``peak'' exposures. The
24-hour standard is intended to provide supplemental protection against
short-term peak exposures and while there is a general relationship
between mean concentrations and 98th percentile concentrations in
individual locations, such relationships vary by location and there is
not an established relationship that can be relied upon to predict 98th
percentile concentrations based on mean PM2.5 concentrations
reported in multi-city epidemiologic studies. Instead, mean
concentrations from short-term epidemiologic studies are more useful in
addressing questions regarding the effects of ``typical'' or average
24-hour exposures, which are addressed through the annual standard. For
this reason, the EPA does consider the mean concentrations of short-
term studies (as well as the means from the long-term studies) in
evaluating the level of the annual standard, which the EPA recognizes
as the generally controlling standard for both long- and short-term
exposures. However, the EPA does not agree with commenters that it is
appropriate to use means from short-term epidemiologic studies as the
basis for a decision-making framework to determine the adequacy of the
current 24-hour standard, with its 98th percentile form.
As described in the proposal (88 FR 5613, January 27, 2023), the
2022 PA also noted the epidemiologic studies that restrict 24-hour
average PM2.5 concentrations to values of less than 35
[micro]g/m\3\, and in some cases less than 25 [micro]g/m\3\, and annual
average PM2.5 concentrations less than 12 [micro]g/m\3\.
Restricted analyses use a subset of data from their main analyses and
conduct an epidemiologic study with health events that occur at
concentrations below a certain concentration (e.g., 25 [micro]g/m\3\).
While some of these studies do not report the mean PM2.5
concentration for the restricted analysis, the mean of the restricted
analysis is presumably less than the mean PM2.5
concentration in the main analysis. Restricted analyses from long-term
and short-term exposure epidemiologic studies are informative in
providing support that the health effects associations are not driven
by just the upper peaks of the PM2.5 air quality
distributions and provide support for revision to the level of the
annual PM2.5 standard. Short-term restricted analyses also
report positive associations between short-term PM2.5
exposure and morbidity and mortality. As an example, in a restricted
analysis evaluating the association between short-term exposures and
PM2.5 concentrations less than 25 [micro]g/m\3\, Di et al.
(2017a) removed 6.3 percent of the data from their main analyses,
(i.e., all PM2.5 concentrations greater than 25 [micro]g/
m\3\), and still found a positive and significant association between
short-term PM2.5 exposure and mortality. This study provides
additional support that the association between short-term exposure to
PM2.5 and mortality in the main epidemiologic analysis is
not driven by the upper peaks of the PM2.5 air quality
distribution, which in turn supports the conclusion that lowering the
entire distribution of air quality concentrations through a revised
annual standard is an appropriate means of protecting against adverse
effects from short-term exposure, as discussed further below.
In their review of the 2021 draft PA, the majority of the CASAC
highlighted three U.S.-based epidemiologic studies that restricted 24-
hour average PM2.5 concentrations below 25 [mu]g/m\3\ as a
part of their rationale for recommending that the EPA revise the level
of the primary 24-hour PM2.5 standard. Similarly, in
evaluating positive associations in restricted analyses, some
commenters also suggest that because an association exists at 24-hour
concentrations below 25 [micro]g/m\3\, the 24-hour standard level
should be set at the concentration at which the analysis was restricted
(e.g., 25 [micro]g/m\3\). However, the EPA notes that neither the CASAC
nor public commenters provided any detail regarding, how, in their
view, these studies demonstrate that the level of the current 24-hour
standard is not adequate, and/or how these studies demonstrate what
revised level of the 24-hour standard would provide requisite public
health protection with an adequate margin of safety. The EPA considers
that such an approach would have several important limitations. First,
the approach assumes that a specific point on the air quality
distribution (e.g., the point at which the analysis was restricted) is
where health effects are exhibited and where we have the most
confidence in the reported association. However, in addition to the
limitations associated with the short-term epidemiologic studies
outlined above, the EPA does not agree that it would be appropriate to
identify the requisite level of the primary 24-hour PM2.5
standard based on the specific concentration at which the analyses
restrict their studies. The choice to restrict the data at a particular
concentration is in effect arbitrary, and does not establish that any
particular effects are attributable to that concentration as opposed to
other concentrations within the restricted analysis.
Further, these restricted analyses do not report the
PM2.5 concentration at the 98th percentile of data or other
metrics relating to the upper end of the distribution that could
provide information about health risks associated with peak exposures.
For example, the CASAC does not provide a discussion of what the
comparable 98th percentile concentration is in the distribution of
remaining 24-hour PM2.5 concentrations of restricted
analyses (because such data is not reported by the study authors) and
what degree of confidence the Administrator should place on those upper
percentile values (e.g., 98th percentile values). In order to identify
a level of the 24-hour standard based on associations between the
``upper end'' of exposures, either in the unrestricted or the
restricted analyses, and adverse health effects, it would be necessary
to have both greater detail on the distribution of air quality in the
study and greater confidence in the reported association at the peak
concentrations such as the 98th percentile--in other words, a better
understanding of how specific 24-hour concentrations correspond to the
frequency and total number of observed health events in the study.
Further, the EPA notes that when resulting analyses based on the
[[Page 16273]]
restricted dataset continue to find positive associations between the
remaining air quality distribution and health effects, it suggests that
the relationship was in fact not driven primarily by the upper tail
(now removed from the dataset) but rather by lower portions of the
distribution of air quality. In other words, we have no confidence that
the remaining upper end of the air quality distribution is driving the
remaining associations reported in the restricted analyses, as opposed
to the vast array of health events at and around the mean
PM2.5 concentration. In fact, it is reasonable to conclude
that to effectively address the health effects observed in the study,
it is necessary to control not just the peak concentrations but to
reduce the bulk of the exposures (occurring near the mean), a task more
effectively achieved, as noted above through a tightening of the annual
standard, which has the effect of shifting the entire distribution of
PM2.5 concentrations downward (both peaks and means).
Therefore, while the EPA agrees that both short- and long-term
epidemiologic studies that completed restricted analyses and reported
the resulting study means could be used to inform conclusions regarding
the adequacy of the annual standard, given that the resulting study
means (when reported) could be evaluated in the context of the decision
framework described above for informing decisions on the level of the
annual standard, the EPA considers that current short-term
epidemiologic studies that restrict analyses are subject to the same
limitations outlined above for current short-term epidemiologic studies
in how they can be used in a decision-making framework to inform the
adequacy and alternative level of the primary 24-hour PM2.5
standard. As such, while the available short-term epidemiologic studies
that restrict their analyses are useful for informing conclusions
regarding the strength of the associations for health outcomes, they
are not, as currently designed, as useful for informing conclusions
regarding the adequacy of the current primary 24-hour PM2.5
standard. In reaching this conclusion, the EPA notes that the majority
of the CASAC did not address the limitations of these studies outlined
in the 2021 draft PA, particularly in the context of the 24-hour
standard with its 98th percentile form. Among the future research needs
identified by the EPA in the 2022 final PA, the Agency noted a number
of gaps in the currently available information reported in the
epidemiologic studies of short-term exposure, including ``descriptive
statistics of PM2.5 concentrations at individual percentiles
from the 95th percentile to the 99th percentile, as well as the number
of days of concentrations and/or health events within each of these
percentiles'' and other descriptive statistics and details regarding
analytical design in studies employing restricted analyses (U.S. EPA,
2022b, pp. 3-225 to 3-226). Such information could significantly
improve the EPA's ability to draw conclusions from these studies with
regard to the adequacy of the current primary 24-hour PM2.5
standard.
Due to the limitations and uncertainties outlined above, in
reaching his decision on the primary 24-hour PM2.5 standard,
the Administrator judges that the information from currently available
short-term epidemiologic studies, including those that use restricted
analyses, is inadequate to inform decisions regarding the adequacy of
the current 24-hour standard. Additionally, consistent with the final
decision in 2012, the EPA continues to view an approach that focuses on
setting a generally controlling annual standard as the most effective
and efficient way to reduce total population risk associated with both
long- and short-term PM2.5 exposures. Potential air quality
changes associated with meeting an annual standard level of 9.0
[micro]g/m\3\ will result in lowering risk associated with both long-
and short-term PM2.5 exposure by lowering the overall air
quality distribution. As discussed above, reducing the annual standard
is the most efficient way to reduce the risks from short-term exposures
identified in the epidemiologic studies, as the available evidence
suggests the bulk of the risk comes from the large number of days
across the bulk of the air quality distribution, not the relatively
small number of days with peak concentrations. However, as in the 2012
review, the Administrator recognizes that an annual standard alone
would not be expected to offer sufficient protection with an adequate
margin of safety against the effects of short-term PM2.5
exposures in all parts of the country, particularly in areas with high
peak-to-mean ratios, and concludes that it is appropriate to continue
to provide supplemental protection by means of a 24-hour standard. In
so doing, the Administrator concludes that retaining the level of the
primary 24-hour PM2.5 standard of 35 [micro]g/m\3\ will
provide requisite protection against short-term peak PM2.5
concentrations, in conjunction with a revised annual standard level of
9.0 [micro]g/m\3\.
4. Administrator's Conclusions
This section summarizes the Administrator's considerations and
conclusions related to the adequacy of the current primary
PM2.5 standards and presents his decision to revise the
primary annual PM2.5 standard to a level of 9.0 [micro]g/
m\3\ and retain the primary 24-hour PM2.5 standard. In
establishing primary standards under the Act that are ``requisite'' to
protect public health with an adequate margin of safety, the
Administrator is seeking to establish standards that are neither more
nor less stringent than necessary for this purpose. He recognizes that
the requirement to provide an adequate margin of safety was intended to
address uncertainties associated with inconclusive scientific and
technical information and to provide a reasonable degree of protection
against hazards that research has not yet identified. However, the Act
does not require that primary standards be set at a zero-risk level;
rather, the NAAQS must be sufficiently protective, but not more
stringent than necessary.
Given these requirements, the Administrator's final decision in
this reconsideration is a public health policy judgment drawing upon
scientific and technical information examining the health effects of
PM2.5 exposures, including how to consider the range and
magnitude of uncertainties inherent in that information. This public
health policy judgment is based on an interpretation of the scientific
and technical information that neither overstates nor understates its
strengths and limitations, nor the appropriate inferences to be drawn,
and is informed by the Administrator's consideration of advice from the
CASAC and public comments received on the proposal.
The initial issue to be addressed in the reconsideration of the
primary PM2.5 standards is whether, in view of the advances
in scientific knowledge and other information reflected in the 2019
ISA, ISA Supplement, and 2022 PA, the current standards are requisite
to protect public health with an adequate margin of safety. In
considering the adequacy of the current suite of primary
PM2.5 standards, the Administrator has considered the large
body of evidence presented and assessed in the 2019 ISA and ISA
Supplement, the conclusions presented in the 2022 PA, the views
expressed by the CASAC, and public comments. The Administrator has
taken into account both evidence- and risk-based considerations in
developing final conclusions on the adequacy of the current primary
PM2.5 standards. The Administrator has additionally
[[Page 16274]]
considered the associated public health policy judgments and judgments
about the uncertainties inherent in the scientific evidence and
quantitative analyses that are integral to the conclusions on the
adequacy of the current primary PM2.5 standards.
In evaluating the adequacy of the current standards, the
Administrator first recognizes the longstanding body of health evidence
supporting relationships between PM2.5 exposures (short- and
long-term) and mortality and serious morbidity effects. The evidence
available in this reconsideration (i.e., that assessed in the 2019 ISA
and ISA Supplement) and summarized above in section II.A.2.a reaffirms,
and in some cases strengthens, the conclusions from the 2009 ISA
regarding the health effects of PM2.5 exposures. Recent
epidemiologic studies demonstrate generally positive and often
statistically significant associations between PM2.5
exposures and a number of health effects, including non-accidental,
cardiovascular, or respiratory mortality; cardiovascular or respiratory
hospitalizations or emergency room visits; and other mortality/
morbidity outcomes (e.g., lung cancer mortality or incidence, asthma
development). Recent controlled human exposure and animal toxicological
studies, as well as evidence from epidemiologic panel studies,
strengthens support for potential biological pathways through which
PM2.5 exposures could lead to the serious effects reported
in many population-level epidemiologic studies, including support for
pathways that could lead to cardiovascular, respiratory, nervous
system, and cancer-related effects. In considering the available
scientific evidence, and consistent with approaches employed in past
NAAQS reviews, the Administrator places the most weight on evidence
supporting ``causal'' or ``likely to be causal'' relationship with long
or short-term PM2.5 exposures. In addition, the
Administrator also takes note of those populations identified to be at
greater risk of PM2.5-related health effects, as
characterized in the 2019 ISA and ISA Supplement, and the potential
public health implications.
In evaluating what existing or revised standards may be requisite
to protect public health, as described above in section II.A.2, the
Administrator's approach recognizes that the current annual standard
(based on arithmetic mean concentrations) and 24-hour standard (based
on 98th percentile concentrations), together, are intended to provide
public health protection against the full distribution of short- and
long-term PM2.5 exposures. This approach recognizes that
changes in PM2.5 air quality designed to meet either the
annual or the 24-hour standard would likely result in changes to both
long-term average and short-term peak PM2.5 concentrations.
Further, consistent with the approach adopted in 2012, the
Administrator concludes that the most effective and efficient way to
reduce total population risk associated with both long- and short-term
PM2.5 exposures is to set a generally controlling annual
standard, and to provide supplemental protection against the occurrence
of peak 24-hour PM2.5 concentrations by means of a 24-hour
standard set at the appropriate level. In reaching this conclusion, the
Administrator explicitly recognizes that air quality changes associated
with meeting a revised annual standard would result in lowering risks
associated with both long- and short-term PM2.5 exposures by
lowering the overall distribution of air quality concentrations,
leading to not only in lower short- and long-term PM2.5
concentrations near the middle of the air quality distribution, but
also in fewer and lower short-term peak PM2.5
concentrations. Similarly, the Administrator recognizes that changes in
air quality to meet a 24-hour standard, would result not only in fewer
and lower peak 24-hour PM2.5 concentrations, but also in
lower annual average PM2.5 concentrations. However, as noted
in 2012, he also recognizes that an approach that relies on setting the
level of the 24-hour standard such that the 24-hour standard is
generally controlling would be less effective and result in less
uniform protection across the U.S. than an approach that focuses on
setting a generally controlling annual standard. Thus, he concludes
that relying on a revised annual standard as the controlling standard
will reduce aggregate risks associated with both long- and short-term
exposures more consistently than a generally controlling 24-hour
standard. He further concludes that retaining a 24-hour standard at the
appropriate level will ensure an adequate margin of safety against
short-term effects in areas with high peak-to-mean ratios.
In light of his focus on the annual standard as the generally
controlling standard, in considering whether the primary
PM2.5 standards are adequate, the Administrator first
considers information available to inform his final conclusions
regarding the primary annual PM2.5 standard. In so doing, he
notes that in this reconsideration, a large number of key U.S.
epidemiologic studies report positive and statistically significant
associations for air quality distributions with overall mean
PM2.5 concentrations that are well below the current level
of the annual standard of 12.0 [mu]g/m\3\. He further recognizes that
there is additional scientific evidence assessed in the 2019 ISA and
newly assessed in this reconsideration in the ISA Supplement that can
provide supplemental information to inform his decisions. In addition
to the key U.S. epidemiologic studies, the Administrator also
recognizes that key Canadian epidemiologic studies also demonstrate
positive and statistically significant associations at concentrations
below 12 [mu]g/m\3\. He also recognizes that epidemiologic studies that
restrict annual average PM2.5 concentrations to below 12
[mu]g/m\3\ also provide support for positive and statistically
significant associations at lower mean PM2.5 concentrations,
as do accountability studies that also suggest public health
improvements may occur at concentrations below 12 [mu]g/m\3\.
With regard to the available scientific evidence to inform his
final decisions on the adequacy of the current 24-hour standard, the
Administrator finds that there is less information available to support
decisions on the 24-hour standard than that summarized above for the
annual standard. The Administrator first notes that controlled human
exposure studies, including those newly available in this
reconsideration, demonstrate effects following short-term
PM2.5 exposures at concentrations higher than the current
24-hour standard. The Administrator also considers air quality analyses
conducted in the 2022 PA and in responding to public comments, as
described above in section II.B.3, that evaluate PM2.5
concentrations in ambient air for similar durations to the controlled
human exposure studies. As noted above, these air quality analyses
indicate that the current 24-hour standard, particularly in conjunction
with the revised level of the annual standard, provides a high degree
of protection against subdaily PM2.5 concentrations that
have been shown to elicit effects in controlled human exposure studies.
The Administrator considers a limited number of available epidemiologic
studies that report associations with health effects when the analyses
are restricted to daily PM2.5 concentrations below 35 [mu]g/
m\3\. As described above, although these studies are useful in
demonstrating that health effects are associated with exposure to daily
PM2.5 concentrations in the lower part of the air quality
distribution, they do not provide information about health effects
associated with the short-term
[[Page 16275]]
``peak'' exposures that the 24-hour standard is designed to protect
against. Accordingly, these studies have limited relevance in informing
a decision about the appropriate level of the 24-hour standard.
In addition to the scientific evidence, the Administrator also
considers the information from the risk assessment. In so doing, he
notes that the risk assessment estimates that the current primary
annual PM2.5 standard could allow a substantial number of
deaths in the U.S. With respect to the 24-hour standard, the
Administrator recognizes that there are only a small number of study
areas where the 24-hour standard is controlling and changes in the 24-
hour standard level are estimated to have a much smaller impact on
public health. The Administrator recognizes that while the risk
estimates can help to place the evidence for specific health effects
into a broader public health context, they should be considered along
with the inherent uncertainties and limitations of such analyses when
informing judgments about the potential for additional public health
protection associated with PM2.5 exposure and related health
effects. While the Administrator recognizes that these uncertainties
are important, he also notes that the general magnitude of the risk
estimates provide support for significant public health impacts,
particularly for lower alternative annual standard levels.
In reaching his final conclusions regarding the adequacy of the
primary PM2.5 standards, the Administrator also considers
the CASAC's advice and recommendations, as well as public comments.
With respect to the CASAC's advice, the Administrator recognizes that,
in their review of the 2021 draft PA, the CASAC reached consensus that
the current primary annual PM2.5 standard is not adequate
and that it is not sufficiently protective of public health. The
Administrator also takes note of the CASAC's advice in their review of
the 2019 draft PA, where the CASAC did not reach consensus on the
adequacy of the primary annual PM2.5 standard, with the
minority recommending revision and the majority recommending the
standard be retained. Furthermore, he recognizes that in reviewing the
2019 draft PA, the CASAC reached consensus regarding the adequacy of
the primary 24-hour PM2.5 standard, concluding that the
standard should be retained. Conversely, in their review of the 2021
draft PA, the majority of the CASAC advised that the current primary
24-hour PM2.5 standard is not adequate and recommended
revising the level of the standard, while the minority of the CASAC
concluded that the standard was adequate and should be retained.
However, in considering the advice of the CASAC collectively in the
context of this reconsideration, the Administrator recognizes that the
2021 draft PA included scientific evidence and quantitative risk
information that was not available in the 2019 draft PA, and therefore,
the advice and recommendations of the CASAC in their review of the 2021
draft PA are based on consideration of the full body of scientific
evidence available in this reconsideration, including the evidence
evaluated in the 2019 ISA and the ISA Supplement.
The Administrator recognizes that much of the scientific evidence
available in this reconsideration was also available in the 2019 ISA
and was considered by the then-Administrator when he decided that the
current primary PM2.5 standards are requisite to protect
public health with an adequate margin of safety. However, as described
in section I.C.5.b above, in reaching his decision to reconsider the
2020 final decision, the Administrator also recognized that there were
a number of studies published since the literature cutoff date of the
2019 ISA that were raised by some members of the CASAC in their review
of the 2019 draft PA, in public comments on the 2020 proposal, and in
the petitions for reconsideration. As such, the expansion of the air
quality criteria in this reconsideration to encompass both the 2019 ISA
and the additional scientific evidence evaluated in the ISA Supplement,
along with evidence and updated quantitative analyses in the 2022 PA
also provided an expanded record for the CASAC's review and public
comments as a part of this reconsideration. Taken together, the 2019
ISA, ISA Supplement, and 2022 PA, along with the CASAC's advice and
recommendations and public comments, provide the Administrator with
additional information for consideration in reaching his final
conclusions in this reconsideration. As a result, the record before him
notably expands upon and strengthens the basis for the conclusions of
the 2019 ISA while reducing some uncertainties that were identified in
the 2020 final action.
In considering the available information in this reconsideration,
the current Administrator reached different conclusions regarding the
appropriate weight to place on certain aspects of the evidence than the
then-Administrator in the 2020 final decision. For example, in reaching
his conclusions on the primary annual PM2.5 standard in
2020, the then-Administrator concluded that it was appropriate to place
more weight on epidemiologic studies that used ground-based monitors
and to place less weight on the studies that used hybrid model-based
approaches, citing to increased uncertainties associated with this new
and emerging approach to estimating exposure. In placing more weight on
the key U.S. monitor-based studies, the then-Administrator noted that
the majority of these studies had mean concentrations at or above the
level of the annual standard (12.0 [micro]g/m\3\). However, unlike the
approach for considering such studies in the 2012 review, the then-
Administrator concluded that it was appropriate to consider the study-
reported means collectively, and in so doing, he placed weight on the
average of the study-reported means (or medians) across the U.S.
monitor-based studies of 13.5 [micro]g/m\3\, and noted that this
concentration was above the level of the standard (85 FR 82717,
December 18, 2020). In contrast, in this reconsideration, the current
Administrator judges that it is appropriate to consider the individual
study-reported mean PM2.5 concentrations from not only the
U.S. monitor-based epidemiologic studies, but also the U.S. hybrid
model-based epidemiologic studies, which are an advancement in the
available science since the completion of the 2009 ISA. The current
Administrator also adopts an approach similar to some previous
approaches for the PM NAAQS in which he judges it most appropriate to
set the level of the standard to somewhat below the lowest long-term
study-reported mean PM2.5 concentration reported in key U.S.
epidemiologic studies, which is 9.3 [micro]g/m\3\. The study that
reports the long-term mean PM2.5 concentration of 9.3
[micro]g/m\3\ is newly available in this reconsideration and is
evaluated in the ISA Supplement. In the 2019 ISA, the lowest long-term
study-reported mean PM2.5 concentrations for U.S.-based
studies that use ground-based monitors and hybrid model-based
approaches are 9.9 [micro]g/m\3\ and 10.7 [micro]g/m\3\, respectively.
In judging that it is appropriate to consider both monitor- and hybrid
model-based epidemiologic studies and that it is appropriate to adopt
an approach to set the level of the standard to somewhat below the
lowest long-term mean PM2.5 concentration, the current
Administrator judges that the available scientific evidence--evaluated
in both the 2019 ISA and in the ISA Supplement--provide support for his
conclusion that that current primary
[[Page 16276]]
PM2.5 standard is not adequate and should be revised.
In addition to adopting a different approach than the previous
Administrator for considering the long-term mean PM2.5
concentrations from key U.S. epidemiologic studies (one more consistent
with the approach of the EPA in other prior reviews), the current
Administrator both has information newly available in this
reconsideration before him and is reaching different conclusions about
how to weigh the evidence before him in reaching his final conclusions.
For example, in reaching his final decision in 2020, the then-
Administrator was concerned about placing too much weight on
epidemiologic studies to inform his conclusions on the adequacy of the
primary PM2.5 standards, noting that the epidemiologic
studies do not identify particular PM2.5 concentrations that
cause effects and cannot alone identify a specific level at which to
set the standard. In so doing, the then-Administrator placed greater
weight on the uncertainties and limitations associated with the
epidemiologic studies, including exposure measurement error, potential
confounding by copollutants, increased uncertainty of associations at
lower PM2.5 concentrations, and heterogeneity of effects
across different cities or regions (85 FR 82716, December 18, 2020).
The Administrator recognizes that in reaching these judgments, the
then-Administrator took into consideration the views of some members of
the CASAC, who, in their advice on the 2019 draft PA, expressed the
view that the current PM NAAQS should be retained because reported
associations between short- and long-term PM2.5 exposures
and adverse health outcomes ``can reasonably be explained in light of
uncontrolled confounding and other potential sources of error and
bias'' (Cox, 2019b, p. 8 of consensus responses).
In this reconsideration, the current Administrator notes that the
ISA Supplement evaluates additional studies that employed statistical
approaches that attempted to more extensively account for confounders
and are more robust to model misspecification (i.e., used alternative
methods for confounder control, which are sometimes referred to as
causal modeling or causal inference methods) that build upon those
studies available and evaluated in the 2019 ISA (U.S. EPA, 2019,
sections 11.1.2.1 and 11.2.2.4). These studies report consistent
positive associations between long-term and short-term PM2.5
exposure and total mortality and cardiovascular effects (U.S. EPA,
2022a, section 3.2.2.3). In considering the epidemiologic evidence
evaluated in the 2019 ISA, along with the newly available studies
evaluated in the ISA Supplement, the current Administrator also
recognizes that there are uncertainties and limitations associated with
the epidemiologic studies, but judges that it is appropriate to place
less weight on these uncertainties than the then-Administrator placed
on them in reaching his final decision in 2020, given the strength of
the longstanding large body of epidemiologic evidence, employing a
variety of study designs, that demonstrates associations between long-
and short-term PM2.5 exposures and health effects across
multiple U.S. cities and in diverse populations, including in studies
examining populations and lifestages that may be at comparatively
higher risk of experiencing a PM2.5-related health effect
(e.g., older adults, children).
In reaching this final decision, the Administrator recognizes he is
differing not only with the prior Administrator but also with the
advice some members of the CASAC provided during their review of the
2019 draft PA. Specifically, taking into consideration the strength of
the evidence providing support for causality determinations, the advice
of other members of the CASAC and the need to protect public health
with an adequate margin of safety, the current Administrator disagrees
with these members of CASAC regarding the weight to be given to
epidemiologic evidence ``based on its methodological limitations''
(Cox, 2019b, p. 8 of consensus responses), such as the possibility
``that such associations could reasonably be explained by uncontrolled
confounding and other potential sources of error and bias'' (Cox,
2019b, p. 8 of consensus responses).
As another example of information that was not available to the
CASAC in providing advice to the Administrator in reaching his final
decision in 2020, the then-Administrator noted in his final decision
that, while some members of the CASAC and public commenters highlighted
a number of accountability studies that examined past reductions in
ambient PM2.5 concentrations and the degree to which those
reductions have resulted in public health improvements, the small
number of available accountability studies did not examine air quality
with starting concentrations meeting the primary annual
PM2.5 standard of 12.0 [micro]g/m\3\. The then-
Administrator took into consideration the absence of such
accountability studies, as part of his consideration of the full body
of scientific evidence, in reaching his judgment that there was
considerable uncertainty in the potential for increased public health
protection from further reductions in ambient PM2.5
concentrations beyond those achieved under the existing primary
PM2.5 NAAQS (85 FR 82717, December 18, 2020). However, there
are several accountability studies available since the literature
cutoff date of the 2019 ISA and evaluated in the ISA Supplement in this
reconsideration that have starting concentrations (or concentrations
prior to the policy or intervention) below 12.0 [micro]g/m\3\ (Corrigan
et al, 2018; Henneman et al., 2019; Sanders et al., 2020a). The current
Administrator concludes that, while the number of available
accountability studies is limited, he recognizes that these studies
provide supplemental information for consideration for informing
decisions on the appropriate level of the primary annual
PM2.5 standard along with the full body of evidence.
As EPA has frequently noted throughout this document, the extent to
which the current primary PM2.5 standards are judged to be
adequate depends in part on science policy and public health policy
judgments to be made by the Administrator on the strength and
uncertainties of the scientific evidence, such as how to consider
epidemiologic evidence and the need for an adequate margin of safety in
setting the standards. Thus, it would be pure speculation to guess
whether the then-Administrator would have reached the same or different
conclusions in the 2020 final decision had the record before him
included the newly available information in this reconsideration.\107\
However, the current Administrator concludes that, for the reasons
explained herein that, in his judgment, based on the record before him
in this reconsideration, it is necessary and appropriate to revise the
primary annual PM2.5 NAAQS to provide requisite protection
of public health with an adequate margin of safety.
---------------------------------------------------------------------------
\107\ The EPA notes that, in considering the additional
scientific evidence available in this reconsideration, one member of
the CASAC who reviewed both the 2019 draft PA and the 2021 draft PA
found that the available scientific and quantitative information
available in this reconsideration supported revising the level of
the primary annual PM2.5 standard to within the range of
10-11 [micro]g/m\3\, whereas he recommended retaining the standard
during the review of the 2019 draft PA.
---------------------------------------------------------------------------
Based on the available scientific evidence and quantitative
information, as well as consideration of the CASAC's advice and public
comments, the Administrator concludes that the
[[Page 16277]]
current primary annual PM2.5 standard is not adequate to
protect public health with an adequate margin of safety. In addition,
he finds the available information insufficient to call into question
the adequacy of the public health protection afforded by the current
primary 24-hour PM2.5 standard.
In considering how to revise the current suite of primary
PM2.5 standards in order to achieve the requisite protection
for public health, with an adequate margin of safety, against long- and
short-term PM2.5 exposures the Administrator considers the
four basic elements of the NAAQS (indicator, averaging time, form, and
level) collectively. With respect to indicator, the Administrator
recognizes that the scientific evidence in this reconsideration, as in
previous reviews, continues to provide strong support for health
effects associated with PM2.5 mass. He notes the 2022 PA
conclusion that the available information continues to support the
PM2.5 mass-based indicator and remains too limited to
support a distinct standard for any specific PM2.5 component
or group of components, and too limited to support a distinct standard
for the ultrafine fraction of PM (U.S. EPA, 2022b, section 3.6.3.2.1).
In its advice on the adequacy of the current primary PM2.5
standards in their review of the 2021 draft PA, the CASAC reached
consensus that the PM2.5 mass-based indicator should be
retained, without revision (Sheppard, 2022a, p. 2 of consensus
letter).\108\ Additionally, there was no information in the public
comments that provided a rationale for an alternative indicator. For
all of these reasons, the Administrator concludes that it is
appropriate to retain PM2.5 mass as the indicator for the
primary standards for fine particles.
---------------------------------------------------------------------------
\108\ The CASAC did not provide advice or recommendations
regarding the indicator of the primary PM2.5 standards in
their review of the 2019 draft PA (Cox, 2019b).
---------------------------------------------------------------------------
Consistent with his proposed conclusions regarding averaging time,
the Administrator notes that the scientific evidence continues to
provide strong support for health effect associations with both long-
and short-term PM2.5 exposures (88 FR 5618, January 27,
2023). Epidemiologic studies continue to provide strong support for
health effects associated with short-term PM2.5 exposures
based on 24-hour averaging periods, and associations in epidemiologic
studies with subdaily estimates are less consistent and, in some cases,
smaller in magnitude (88 FR 5618, January 27, 2023). Taken together,
the 2019 ISA concludes that epidemiologic studies do not indicate that
subdaily averaging periods are more closely associated with health
effects than the 24-hour average exposure metric (U.S. EPA, 2019a,
section 1.5.2.1). In addition, controlled human exposure and panel-
based studies of subdaily exposures typically examine subclinical
effects rather than the more serious population-level effects that have
been reported to be associated with 24-hour exposures (e.g., mortality,
hospitalizations). While recent controlled human exposure studies
provide consistent evidence for cardiovascular effects following
PM2.5 exposures for less than 24 hours (i.e., <30 minutes to
5 hours), air quality analyses have shown that the current averaging
times can effectively protect against the exposure concentrations in
these studies. This information does not indicate that a revision to
the averaging time is necessary to provide additional protection
against subdaily PM2.5 exposures, beyond that provided by
the current primary annual and 24-hour PM2.5 standards. The
Administrator also notes that this conclusion is also support by the
CASAC's advice in their review of the 2021 draft PA where they reached
consensus that averaging times for the primary PM2.5
standards should be retained, without revision (Sheppard, 2022a, p. 2
of consensus letter).\109\ The Administrator also considers the
relatively few public comments received that support a subdaily
averaging time, but concludes that the currently available information
does not provide support for an alternate averaging time. Consistent
with his proposed decision, the Administrator concludes that it is
appropriate to retain the annual and 24-hour averaging times for the
primary PM2.5 standards to protect against long- and short-
term PM2.5 exposures.
---------------------------------------------------------------------------
\109\ The CASAC did not provide advice or recommendations
regarding the averaging times of the primary PM2.5
standards in their review of the 2019 draft PA (Cox, 2019b).
---------------------------------------------------------------------------
With regard to form, the Administrator first notes that the EPA has
set both an annual standard and a 24-hour standard to provide
protection from health effects associated with both long- and short-
term exposures to PM2.5 (62 FR 38667, July 18, 1997; 88 FR
5620, January 27, 2023). With regard to the form of the annual
standard, the Administrator recognizes that a large majority of the
recently available epidemiologic studies continue to report
associations between health effects and annual average PM2.5
concentrations. These studies of annual average PM2.5
concentrations provide support for retaining the current form of the
annual standard to provide protection against long- and short-term
PM2.5 exposures. In its review of the 2021 draft PA, the
CASAC reached consensus that the form of the annual standard (i.e.,
annual mean, averaged over 3 years) should be retained, without
revision (Sheppard, 2022a, p. 2 of consensus letter).\110\ The
Administrator also notes that there were no public comments that
recommended an alternative form for the primary annual PM2.5
standard.
---------------------------------------------------------------------------
\110\ The CASAC did not provide advice or recommendations
regarding the forms of the primary PM2.5 standards in
their review of the 2019 draft PA (Cox, 2019b).
---------------------------------------------------------------------------
With regard to the form of the 24-hour standard (98th percentile,
averaged over three years), epidemiologic studies continue to provide
strong support for health effect associations with short-term (e.g.,
mostly 24-hour) PM2.5 exposures, and controlled human
exposure studies provide evidence for health effects following single
short-term ``peak'' PM2.5 exposures (88 FR 5618, January 27,
2023). Therefore, the Administrator concludes that the evidence
supports retaining a standard focused on providing supplemental
protection against short-term peak exposures and supports a 98th
percentile form for a 24-hour standard, in combination with a primary
annual PM2.5 standard with its annual mean averaged over
three years form. As described in the proposal and in responding to
comments in section II.B.3 above, the Administrator further notes that
the 98th percentile, averaged over three years, form also provides an
appropriate balance between limiting the occurrence of peak 24-hour
PM2.5 concentrations and identifying a stable target for
risk management programs (U.S. EPA, 2022b, section 3.6.3.2.3).
Furthermore, the Administrator notes that the multi-year percentile
form (i.e., averaged over three years) offers greater stability to the
air quality management process by reducing the possibility that
statistically unusual indicator values will lead to transient
violations of the standard. This conclusion is also supported by the
CASAC's advice in their review of the 2021 draft PA, where they reached
consensus that the form for the primary PM2.5 standards
should be retained, without revision (Sheppard, 2022a, p. 2 of
consensus letter).\111\
---------------------------------------------------------------------------
\111\ The CASAC did not provide advice or recommendations
regarding the forms of the primary PM2.5 standards in
their review of the 2019 draft PA (Cox, 2019b).
---------------------------------------------------------------------------
The Administrator also recognizes that the CASAC recommended that
in future reviews, the EPA also consider alternative forms for the
primary 24-hour PM2.5 standard (Sheppard, 2022a,
[[Page 16278]]
p. 18 of consensus responses). Based on the CASAC's advice, the
proposal solicited comment on alternatives to the current form for
consideration in future reviews (88 FR 5619, January 27, 2023). The
Administrator recognizes that there were a limited number of public
comments related to the form of the primary PM2.5 standards
as discussed in section II.D.3 above and in the Response to Comments
document, and notes that, the EPA will consider the information
provided by the commenters regarding the form of the 24-hour
PM2.5 standard in the next review of the PM NAAQS.
Consistent with his proposed decision, in considering the information
summarized above, the Administrator concludes that it is appropriate to
retain the forms of the current annual and 24-hour PM2.5
standards.
In considering how to revise the current suite of PM2.5
standards to provide the requisite public health protection with an
adequate margin of safety, the Administrator next evaluates the
appropriate levels of the primary PM2.5 standards, beginning
with the annual PM2.5 standard. In having carefully
considered public comments related to the primary annual
PM2.5 standard, the Administrator believes that the
fundamental conclusions regarding the scientific evidence and
quantitative information that supported his proposed conclusions (as
described in the 2019 ISA, ISA Supplement, 2022 PA, and the proposal)
remain valid. In considering the level at which the primary annual
PM2.5 standard should be set, the Administrator considers
the entire body of evidence and information, giving appropriate weight
to each part of that body of evidence and information. He continues to
place the greatest weight in this reconsideration on the available
scientific evidence that provides support for associations between
health effects and long- and short-term PM2.5 exposures. In
conjunction with his decisions to retain the current indicator,
averaging time, and form as described above, the Administrator is
revising the level of the primary annual PM2.5 standard to
9.0 [micro]g/m\3\. In so doing, he is selecting a primary annual
PM2.5 standard that, together with the primary 24-hour
PM2.5 standard, provides requisite public health protection
with an adequate margin of safety, based on his judgments about and
interpretation of the scientific evidence and quantitative risk
information.
The Administrator's decision to revise the level of the primary
annual PM2.5 standard to 9.0 [micro]g/m\3\ builds upon his
conclusion that the overall body of scientific evidence and
quantitative risk information calls into question the adequacy of
public health protection afforded by the current standard, particularly
for at-risk populations. Consistent with his consideration of the
available information in reaching his proposed decisions, the
Administrator's final decision on the level of the primary annual
PM2.5 standard places the greatest emphasis on key U.S.
epidemiologic studies that report associations between long- and short-
term PM2.5 exposures and mortality and morbidity. As in the
proposal, and as discussed further below, he views additional
epidemiologic studies (i.e., studies that employ alternative methods
for confounding control, studies that employ restricted analyses, and
accountability studies), the controlled human exposure studies, and the
risk assessment as providing supplemental information in support of his
decision to revise the current annual standard, but recognizes that
some of these lines of evidence and information provide a more limited
basis for selecting a particular standard level among a range of
options. See Mississippi, 744 F. 3d at 1351-52 (studies can
legitimately support a decision to revise the standard, but not provide
sufficient information to justify their use in setting the level of a
revised standard).
Given his consideration of the scientific evidence, quantitative
risk information, advice from the CASAC, and public comments, the
Administrator judges that a primary annual PM2.5 standard
with a level of 9.0 [micro]g/m\3\ is requisite to protect public health
with an adequate margin of safety. He notes that the determination of
what constitutes an adequate margin of safety is expressly left to the
judgment of the EPA Administrator. See Lead Industries Association v.
EPA, 647 F.2d at 1161-62; Mississippi, 744 F.3d at 1353. He further
notes that in evaluating how particular standards address the
requirement to provide an adequate margin of safety, it is appropriate
to consider such factors as the nature and severity of the health
effects, the size of the at-risk populations, and the kind and degree
of the uncertainties present. In considering the need for an adequate
margin of safety, the Administrator notes that a primary annual
PM2.5 standard with a level of 9.0 [micro]g/m\3\ would be
expected to provide substantial improvements in public health compared
to the current annual standard, including for at-risk groups such as
children, older adults, people with preexisting conditions, minority
populations, and low SES populations.
Consistent with his conclusions on the need for revision of the
current annual standard, in reaching a decision on level, the
Administrator places the most weight on information from epidemiologic
studies. In so doing, the Administrator notes that these studies
provide consistent evidence of positive and statistically significant
associations between long- and short-term exposure to PM2.5
and mortality and morbidity (88 FR 5624, January 27, 2023). The
Administrator recognizes that placing weight on the information from
the epidemiologic studies allows for examination of the entire
population, including those that may be at comparatively higher risk of
experiencing a PM2.5-related health effects (e.g., children,
older adults, minority populations) (88 FR 5624, January 27, 2023). The
Administrator also recognizes that recent epidemiologic studies
continue to support a no-threshold relationship, meaning that there is
no ``bright line'' below which no effects have been found. These
studies also support a linear relationship between health effects and
PM2.5 exposures at PM2.5 concentrations greater
than 8 [mu]g/m\3\, though uncertainties remain about the shape of the
C-R curve at PM2.5 concentrations less than 8 [mu]g/m\3\,
with some recent studies providing evidence for either a sublinear,
linear, or supralinear relationship at these lower concentrations (U.S.
EPA, 2019a, section 11.2.4; U.S. EPA, 2022a, section 2.2.3.2; 88 FR
5625, January 27, 2023).
As at the time of proposal, the Administrator notes that some
recent epidemiologic studies have adopted a broad range of approaches
to examine confounding and the results of those examinations support
the robustness of reported associations seen in epidemiologic studies.
These include studies that employ alternative methods for confounder
control and studies that evaluate the uncertainty related to exposure
measurement error, both of which continue to support associations
between PM2.5 exposures and health effects while taking
approaches to address uncertainties.
In considering the epidemiologic evidence, the Administrator judges
that, in reaching his decision on an appropriate level for the annual
standard that will protect public health with an adequate margin of
safety, in the absence of any discernible population-level thresholds,
and in recognizing the need to weigh uncertainties associated with the
epidemiologic evidence, it is most appropriate to examine where the
evidence of associations observed in the
[[Page 16279]]
epidemiologic studies is strongest and, conversely, to place less
weight where he has less confidence in the associations observed in the
epidemiologic studies. As at the time of proposal, the Administrator
notes that in previous reviews, evidence-based approaches noted that
the evidence of an association in any epidemiologic study is
``strongest at and around the long-term average where the data in the
study are most concentrated'' (78 FR 3140, January 15, 2013). Given
this, these approaches focused on identifying standard levels near or
somewhat below long-term mean concentrations reported in key
epidemiologic studies. These approaches were supported by previous
CASAC advice as well as the CASAC's advice in their review of the 2021
draft PA as a part of this reconsideration.
Additionally, the Administrator acknowledges that in the 2020 final
action, the then-Administrator decided to retain the standard based in
part on concerns about placing reliance on the epidemiologic studies
and his judgment that even if he did rely on them, the majority of the
studies had means or medians, as well as the mean of all of the key
study-reported means or medians, above the level of the current annual
standard. However, after considering the evidence, the advice of CASAC,
and public comments the Administrator judges that this approach is
insufficient to protect public health with an adequate margin of
safety. The Administrator's decision to reach a different judgment
about the appropriate level of the annual standard reflects the updated
and expanded scientific record available to the Administrator in this
reconsideration, as well as the additional advice from the CASAC and
the public comments based on this newly available information. In
addition, the Administrator observes the decision in this action to
place weight on the epidemiologic studies, and to revise the annual
primary standard to a level below the lowest long-term mean in the
U.S.-based epidemiologic studies, is consistent with the EPA's past
practice in PM NAAQS reviews.
In this reconsideration, the Administrator is considering the
scientific record which has been expanded and updated since the 2020
final action, as well as the additional advice from the CASAC and the
public comments that are based on the newly available information that
expands upon the information previously available. In addition, the
Administrator is exercising his judgment about how to interpret and
weigh the expanded evidence in a way that is more consistent with the
approaches used in prior PM NAAQS reviews. As a result, the
Administrator has concluded on reconsideration that the level of the
primary annual standard is not adequate and should be revised to
protect public health with an adequate margin of safety.
Consistent with his proposed decisions, in reaching conclusions on
the level of the primary annual PM2.5 standard, the
Administrator considers the long-term \112\ study-reported mean
PM2.5 concentrations from key long- and short-term
epidemiologic studies and sets the level of the standard to somewhat
below the lowest long-term mean PM2.5 concentration.\113\ He
notes that in previous PM NAAQS reviews (including the 1997, 2006 and
2012 reviews), evidence-based approaches focused on identifying
standard levels near or somewhat below long-term mean concentrations
reported in key long- and short-term epidemiologic studies. These
approaches were supported by the CASAC in previous reviews and were
supported in this reconsideration by the CASAC in their review of the
2021 draft PA. In considering the available scientific evidence to
inform such an approach, the Administrator notes the strength of the
epidemiologic evidence which includes multiple studies that
consistently report positive associations for short- and long-term
PM2.5 exposure and mortality and cardiovascular effects.
Some available studies also use a variety of statistical methods to
control for confounding bias and report similar associations, which
further supports the broader body of epidemiologic evidence for both
mortality and cardiovascular effects. Additionally, he notes that
recent epidemiologic studies available for consideration in reaching
his final decision strengthen support for health effect associations at
PM2.5 concentrations lower than in those evaluated in
epidemiologic studies available at the time of previous reviews. The
Administrator does recognize, however, that while these epidemiologic
studies evaluate associations between distributions of ambient
PM2.5 concentrations and health outcomes, they do not
identify the specific exposures that led to the reported effects. As
such, he notes that there is no specific point in the air quality
distribution of any epidemiologic study that represents a ``bright
line'' at and above which effects have been observed and below which
effects have not been observed. The Administrator further notes that
the epidemiologic studies provide the strongest support for reported
health effect associations for this middle portion of the
PM2.5 air quality distribution, which corresponds to the
bulk of the underlying data, rather than the extreme upper or lower
ends of the distribution, and concludes that the long-term study-
reported means from both long- and short-term studies provide the
strongest support for reported health effect associations in
epidemiologic studies. For these reasons, as described in the proposal
and in responding to public comments in section II.B.3 above, the
Administrator concludes that it is appropriate to continue to employ an
approach that focuses on the mean PM2.5 concentrations from
the key epidemiologic studies to inform his conclusions regarding the
appropriate level for the primary annual PM2.5 standard.
---------------------------------------------------------------------------
\112\ ``Long-term'' represents PM2.5 exposures and
concentrations that are annual or multi-year.
\113\ As described in section II.A.2.c above, key epidemiologic
studies are those that report overall mean (or median)
PM2.5 concentrations and for which the years of
PM2.5 air quality data used to estimate exposures overlap
entirely with the years during which health events are reported.
---------------------------------------------------------------------------
In adopting such an approach, the Administrator considers the long-
term mean concentrations reported in two types of key epidemiologic
studies: (1) Monitor-based studies \114\ (epidemiologic studies that
used ground-based monitors to estimate exposure, similar to approaches
used in past reviews), and (2) hybrid modeling-based studies \115\
(epidemiologic studies that used hybrid modeling approaches and apply
aspects of population weighting to estimate exposures). In reaching
conclusions regarding the level of a standard that would provide
requisite protection with an adequate margin of safety, the
Administrator recognizes that he must use his judgment regarding the
appropriate weight to place on the available evidence and technical
information, including uncertainties. As shown in Figures 1 and 2
above, for the key U.S. monitor-based epidemiologic studies,
[[Page 16280]]
the study-reported mean concentrations range from 9.9-16.5 [mu]g/m\3\,
and for the key U.S. hybrid modeling-based epidemiologic studies, the
mean concentrations range from 9.3-12.2 [mu]g/m\3\. The Administrator
also recognizes that, in their review of the 2021 draft PA, both the
majority and minority of the CASAC emphasized the epidemiologic studies
in support of their recommendations for the level of the annual
standard, but they weighed the studies in different ways (Sheppard,
2022a, p. 16-17 of consensus responses).
---------------------------------------------------------------------------
\114\ Reported mean PM2.5 concentrations in monitor-
based studies are averaged across monitors in each study area with
multiple monitors, referred to as a composite monitor concentration,
in contrast to the highest concentration monitored in the study
area, referred to as a maximum monitor concentration (i.e., the
``design value'' concentration), which is used to determine whether
an area meets a given standard.
\115\ Studies that use hybrid modeling approaches employ methods
to estimate ambient PM2.5 concentrations across large
geographical areas, including areas without monitors, and thus, when
compared to monitor-based studies, require additional information to
inform the relationship between the estimated PM2.5
concentrations across an area and the maximum monitor design values
used to assess compliance.
---------------------------------------------------------------------------
Based on this information, and in considering the CASAC's advice in
their review of the 2021 draft PA, the Administrator judges that it is
appropriate to set the level of the primary PM2.5 standard
at least as low as the lowest mean PM2.5 concentration from
these key U.S.-based epidemiologic studies, which is 9.3 [micro]g/m\3\.
The Administrator additionally notes that setting the annual standard
level at 9.0 [micro]g/m\3\, which is below the lowest study-reported
mean PM2.5 concentration of 9.3 [micro]g/m\3\, would be
expected to shift the distribution of PM2.5 concentrations
in an area such that the area's highest monitor would generally be at
or below 9.0 [micro]g/m\3\ annually, when meeting the annual standard.
In this situation, the resulting average or mean PM2.5
concentration for the entire area (measured across a number of
monitors) would be even further below the study-reported means,\116\
and will provide adequate protection not only in areas where the
highest allowable concentrations would be expected (i.e., near design
value monitors) but also in other parts of the area where
PM2.5 concentrations would be expected to be maintained even
lower.
---------------------------------------------------------------------------
\116\ Analyses in the 2022 PA suggest that the highest monitored
value would be expected to be greater than the study-reported mean
values by 10-20% for monitor-based studies and 15-18% for hybrid
modeling studies that apply aspects of population weighting.
---------------------------------------------------------------------------
As noted above, however, the Administrator must exercise his
judgment regarding the appropriate weight to place on the available
scientific evidence and quantitative information, including
uncertainties, in determining what level of the annual standard is
sufficient to protect public health with an adequate margin of safety.
In so doing, he considers other information available in this
reconsideration to inform his judgments, including study-reported
PM2.5 concentrations at lower percentiles in key
epidemiologic studies, supplemental information from other types of
epidemiologic studies, study-reported PM2.5 concentrations
from key Canadian epidemiologic studies, and the results from the
quantitative risk assessment.
In weighing the evidence in considering the requisite level of the
annual standard, the Administrator also takes into account additional
information from the key long- and short-term U.S. epidemiologic
studies available that provide study-reported PM2.5
concentrations below the mean and, in particular, the subset of
epidemiologic studies that report 25th and 10th percentile
concentrations. Consistent with his proposed conclusions, as well as
the CASAC's advice in their review of the 2021 draft PA and public
comments, the Administrator judges that it is appropriate to place some
weight on these lower percentiles in reaching his conclusions on the
level of the primary annual standard. There are six key U.S.
epidemiologic studies that report information on other percentiles
(e.g., 10th and 25th percentiles of PM2.5 concentrations or
10th and 25th percentiles of PM2.5 concentrations associated
with health events) that are below the mean.\117\ In considering the
information from these studies, the Administrator first notes that the
three older, monitor-based studies that report lower percentiles of
PM2.5 concentrations have smaller cohort sizes than the
three hybrid model-based studies. Thus, the Administrator recognizes
that the older, monitor-based studies had a relatively smaller portion
of the health events that were observed in the lower part of the air
quality distribution because of the generally smaller size of the
cohorts. He further notes that the recent hybrid model-based studies
have larger cohort sizes than the older, monitor-based studies, and
therefore, have more health events in the lower part of the air quality
distribution. Because of the larger cohort sizes and having a larger
portion of health events that are observed across the air quality
distribution, the Administrator has more confidence in the magnitude
and significance of the associations in the lower parts of the air
quality distribution for the recent, hybrid model-based studies
compared to the older, monitor-based studies. Given this, the
Administrator judges that it is appropriate to place weight on the 25th
percentile concentrations reported in the recently available hybrid
model-based studies in reaching his conclusions regarding the
appropriate level for the primary annual PM2.5 standard.
However, the Administrator also recognizes that his confidence in the
magnitude and significance in the reported concentrations, and their
ability to inform decisions on the appropriate level of the annual
standard, starts to diminish at percentiles that are even further below
the mean and the 25th percentile. For these reasons, the Administrator
places weight on the reported 25th percentiles concentrations in the
recent hybrid model-based studies, rather than the reported 10th
percentile concentrations, in reaching his conclusions regarding the
appropriate level for the primary annual PM2.5 standard.
---------------------------------------------------------------------------
\117\ The Wang et al. (2017) study only reports the 25th
percentile of the estimated PM2.5 concentrations, not the
10th percentile.
---------------------------------------------------------------------------
In considering the information from these studies, as described in
section II.A.2.c and in responding to public comments in section II.B.3
above, the Administrator notes that there are two hybrid model-based
studies with large cohort sizes that apply population weighting and
report lower percentile values. These studies are Di et al. (2017b) and
Wang et al. (2017) and the reported 25th percentile concentration is
9.1 [micro]g/m\3\ for both studies.\118\ In considering these studies,
the Administrator concludes that it is appropriate to place weight on
the 25th percentile concentrations of these newer hybrid model-based
studies (of 9.1 [micro]g/m\3\) such that setting the level of the
standard near these 25th percentile concentrations would provide
requisite protection. The Administrator observes that an annual
standard level of 9.0 [micro]g/m\3\ would be near the reported 25th
percentile concentrations in these studies.
---------------------------------------------------------------------------
\118\ There is a third hybrid model-based study, as described in
the 2022 PA and in section II.B.3 above in responding to public
comments, but it is not referenced here because it reports a 25th
percentile PM2.5 concentration based on the 25th
percentile of health events that occur in the study (Di et al.,
2017a) rather than report the 25th percentile based on air quality
concentrations.
---------------------------------------------------------------------------
As at the time of proposal, the Administrator also takes note of
the study-reported long-term mean PM2.5 concentrations in
long- and short-term Canadian epidemiologic studies, which ranged from
6.9 to 13.3 [micro]g/m\3\ for monitor-based studies and 5.9 to 9.8
[micro]g/m\3\ for hybrid model-based studies. While the Administrator
notes that these studies provide additional support for associations
between PM2.5 concentrations and health effects, he is also
mindful that there are important differences between the exposure
environments in the U.S. and Canada and that interpreting the data
(e.g., study-reported mean concentrations)
[[Page 16281]]
from the Canadian studies in the context of a U.S.-based standard may
present challenges in directly and quantitatively informing decisions
regarding potential alternative levels of the annual standard. For
example, in terms of people per square kilometer, the U.S. population
density is nearly 10 times in the contiguous U.S. compared to Canada.
As described in more detail in responding to public comments in section
II.B.3 above, in this reconsideration, the Administrator recognizes
that this difference in population density between the U.S. and Canada
is more apparent than in previous reviews because the studies available
in this reconsideration use different approaches than those previously
available. In the 2012 review, the available Canadian epidemiologic
studies used population-weighting and focused on urban areas where
monitors were available and population densities were more comparable
with those in the U.S., and at that time, the U.S. and Canadian studies
reported similar mean PM2.5 concentrations. However, in this
reconsideration, the Administrator takes note that for the new Canadian
epidemiologic studies: (1) The Canadian monitor-based studies available
in this reconsideration do not apply population weighting as the
previously available studies did; and (2) some of the studies now use
hybrid modeling approaches for estimating exposure. The Administrator
recognizes that these differences are important to consider in reaching
conclusions on how these Canadian epidemiologic studies should be
interpreted regarding decisions on the requisite level of the primary
annual PM2.5 standard. Specifically, the Administrator notes
that the more recent Canadian studies that use hybrid modeling
incorporate larger portions of the country, and therefore include more
rural areas. The more rural areas that are included in the study using
the hybrid modeling approaches, the more important it is to consider
how the population densities and exposure environments differ between
the U.S. and Canada. Additionally, the Administrator notes that for
hybrid modeling-based studies there is less certainty in
PM2.5 exposure estimates in more rural areas, which are
further from air quality monitors and where PM2.5
concentrations in the ambient air tend to be lower. For these hybrid
model-based studies, the portion of the rural areas that are
contributing to the study-reported mean PM2.5 concentrations
in these studies is unclear. For these reasons, the Administrator
concludes that it is important to consider the differences between the
population exposures in the U.S. and Canadian study areas and how these
differences influence the interpretation of the epidemiologic study
results.
Thus, the Administrator considers the Canadian studies to inform
his judgments on what level for the annual standard is requisite in
light of the limitations and challenges presented. The Administrator
also recognizes that the majority of the CASAC in their review of the
2021 draft PA, as well as a number of public commenters, place weight
on the Canadian epidemiologic studies in recommending that the level of
the primary annual PM2.5 standard be revised to 8-10
[micro]g/m\3\. The Administrator further notes while the majority of
the CASAC advised the EPA to consider the Canadian studies in revising
the annual standard level to within the range of 8.0-10.0 [micro]g/
m\3\, they did not advise the EPA to set the annual standard level
below the study-reported means from those studies. Given these
considerations, the Administrator judges that it is appropriate to set
the level of annual standard within the range of 8-10 [micro]g/m\3\ to
be consistent with the majority of the CASAC's advice in their
consideration of these studies.
The Administrator also recognizes that information from
epidemiologic studies that included analyses that restrict annual
average PM2.5 concentrations to concentrations below the
level of the current annual standard can be useful for informing
conclusions regarding the appropriate level of the primary annual
PM2.5 standard. In so doing, he particularly notes the two
key U.S. epidemiologic studies (Di et al., 2017b and Dominici et al.,
2019) that restrict annual average PM2.5 concentrations to
less than 12 [micro]g/m\3\ and report positive and statistically
significant associations with all-cause mortality and mean
PM2.5 concentrations of 9.6 [micro]g/m\3\. He also considers
these results along with the uncertainties and limitations associated
with studies that restricted analyses below certain PM2.5
concentrations. As described in responding to comments in section
II.B.3 above, uncertainties associated with how the studies exclude
PM2.5 concentrations from the analyses (e.g., at what
spatial resolution are concentrations being excluded), make it
difficult to understand how to interpret the results of the restricted
analyses in the context of the approach employed in this
reconsideration, which takes into consideration the relationship
between mean PM2.5 concentrations and design values.
The Administrator also recognizes that, in their review of the 2021
draft PA, the CASAC noted that epidemiologic studies that restrict
analyses below certain PM2.5 concentrations represent one
area for which the evidence has expanded in this reconsideration,
stating that these studies provide support for mortality effects at
concentrations below the current PM NAAQS (Sheppard, 2022a, p. 5 of
consensus responses). In their recommendations on alternative levels
for the primary annual PM2.5 standard, the majority of the
CASAC cited to studies that restrict PM2.5 concentrations to
below 12 [micro]g/m\3\ as a part of their rationale for supporting a
level within the range of 8-10 [micro]g/m\3\ (Sheppard, 2022a p. 16 of
consensus responses). Additionally, the Administrator notes that some
members of the CASAC, in their review of the 2019 draft PA, concluded
that the epidemiologic studies that restrict analyses below 12
[micro]g/m\3\ and show positive associations with health effects, along
with other aspects of the scientific evidence, provide support for
their conclusion that the primary annual PM2.5 standard is
not adequate (Cox, 2019b, p. 9 of consensus responses). Furthermore,
the Administrator takes note of public commenters who also noted that
the epidemiologic studies that restrict PM2.5 concentrations
to below the current standard provide support, along with the other
available information, for lowering the level of the primary annual
PM2.5 standard. In considering the studies that include
restricted analyses, along with the CASAC's advice and public comments
on these types of studies, the Administrator concludes that, although
there are inherent uncertainties associated with this limited body of
evidence, these studies that apply restricted analyses provide support
for serious effects (e.g., mortality) at concentrations below 10.0
[micro]g/m\3\. Given this, the Administrator concludes that it is
appropriate to place some weight on these studies, and in doing so,
notes that a standard level of 9.0 [micro]g/m\3\ would be below the
reported mean PM2.5 concentrations of 9.6 [micro]g/m\3\ in
these studies and would, thus, be expected to provide protection
against exposures related to these reported mean concentrations.
The Administrator also takes into consideration recent U.S.
accountability studies, which assess the health effects associated with
actions that improve air quality (e.g., air quality policies or
implementation of an intervention). These types of studies can also
reduce uncertainties related to residual confounding of temporal and
spatial factors (U.S. EPA, 2022a, p. 3-25). The
[[Page 16282]]
Administrator notes that in the 2020 review, the available
accountability studies had ``starting'' annual average PM2.5
concentrations (i.e., mean concentration prior to reductions being
evaluated) from 13.2-31.5 [micro]g/m\3\, and the then-Administrator
cited the lack of accountability studies in areas where the
``starting'' concentration met the current primary PM2.5
standards as part of his rationale for retaining the standards. As at
the time of proposal, the current Administrator notes that in three
studies newly available in this reconsideration and assessed in the ISA
Supplement, prior to implementation of the policies, mean
PM2.5 concentrations in these studies were below the level
of the current annual standard level (12.0 [micro]g/m\3\) and ranged
from 10.0 [micro]g/m\3\ to 11.1 [micro]g/m\3\. These studies report
positive and significant associations between mortality and
cardiovascular morbidity and reductions in ambient PM2.5
following the implementation of a policy (Henneman et al., 2019;
Corrigan et al., 2018; Sanders et al., 2020a; 88 FR 5627, January 27,
2023). These studies suggest that public health improvements may occur
following the implementation of a policy that reduces annual average
PM2.5 concentrations below the level of the current standard
of 12.0 [micro]g/m\3\. The Administrator recognizes that in their
review of the 2021 draft PA, the CASAC noted that the availability of
recent accountability studies was one area where the evidence had been
strengthened and that the studies assessed in the ISA Supplement
provide evidence of mortality effects at annual average
PM2.5 concentrations below the current NAAQS (Sheppard,
2022a, p. 5 of consensus responses). The Administrator recognizes that
the CASAC also concluded that, along with other lines of evidence, the
accountability studies with starting concentrations below the levels of
the current standards are appropriate to consider for informing
conclusions on alternative standard levels (Sheppard, 2022a, p. 13 of
consensus responses). The Administrator also notes the advice of the
CASAC in their review of the 2019 draft ISA, where they suggested that
accountability studies be taken into account and such studies provide
potentially crucial information about whether and how much decreasing
PM2.5 causes decreases in future health effects, which
reflects the primary purpose of the NAAQS (Cox, 2019b, p. 8 and 10 of
consensus responses). The Administrator also notes that in their review
of the 2019 draft ISA, some members of the CASAC cautioned against
placing more weight on the data from accountability studies based on
the methodological limitations of the studies (Cox, 2019b, p. 8 of
consensus responses). The Administrator notes that the CASAC did not
explicitly cite to accountability studies in their reviews of the 2019
draft PA or 2021 draft PA as support for their recommendations on the
adequacy of the primary annual PM2.5 standard or potential
alternative standard levels. A number of public commenters who support
revising the level of the standard to 8 [micro]g/m\3\ cite these
accountability studies, along with the broader evidence base, as
support for a more protective standard. The Administrator, in
considering the evidence, the advice from the CASAC, and public
comment, first recognizes that accountability studies are just one line
of evidence to be considered in the broader evaluations of the
information available to inform conclusions on the level of the
standard. In so doing, he notes that public health improvements may
occur following the implementation of a policy that reduces annual
average PM2.5 concentrations below the level of the current
standard of 12.0 [micro]g/m\3\, and potentially below the lowest
``starting'' concentrations in these studies of 10.0 [micro]g/m\3\.
However, the Administrator concludes that the limited number of
accountability studies provide limited information for informing
decisions on the appropriate level of the primary annual
PM2.5 standard but recognizes that these studies provide
supplemental information for consideration along with the full body of
evidence. Taken together, the Administrator notes a revised annual
standard level of 9.0 [micro]g/m\3\ is at or below the lowest starting
concentration of these accountability studies (i.e., 10.0 [micro]g/
m\3\), and judges that it is appropriate to place some weight on these
studies, particularly for informing his public policy judgments
regarding an adequate margin of safety.
In addition to his consideration of and conclusions regarding the
available scientific evidence, the Administrator also considers the
results of the quantitative risk assessment to inform his conclusions
regarding the appropriate level for the primary annual PM2.5
standard. The Administrator recognizes that the risk estimates can help
to place the evidence for specific health effects into a broader public
health context, but should be considered along with the inherent
uncertainties and limitations of such analyses when informing judgments
about the potential for additional public health protection associated
with PM2.5 exposure and related health effects. The
Administrator recognizes that the overall risk assessment estimates
suggest that the current primary annual PM2.5 standard could
allow a substantial number of PM2.5-associated deaths in the
U.S. The Administrator also recognizes that the CASAC concurred with
the 2021 draft PA's assessment that meaningful risk reductions will
result from lowering the annual PM2.5 standard (Sheppard,
2022a, p. 16 of consensus responses).
Additionally, with respect to the results of the quantitative risk
assessment, the Administrator recognizes that the 2022 PA also provides
information on the distribution of concentrations associated with the
estimated mortality risk at each alternative standard level assessed
(U.S. EPA, 2022b, sections 3.4.2.2 and 3.6.2.2, Figure 3-18 and 3-19).
When meeting an annual standard of 9.0 [micro]g/m\3\ at the design
value monitor, the exposure concentrations within an area are estimated
to be below 9 [micro]g/m\3\, with the majority of those exposures being
at concentrations of below 8 [micro]g/m\3\. The Administrator notes
that this range of concentrations is below the lowest means in the key
long- and short-term epidemiologic studies (concentrations at which the
evidence is the strongest in supporting an association between exposure
to PM2.5 and adverse health effects observed in the key
epidemiologic studies available in this reconsideration). Thus, the
Administrator concludes that the results of the quantitative risk
assessment suggest that a revised annual standard level of 9.0
[micro]g/m\3\ is estimated to reduce PM2.5 exposures to fall
within the range of concentrations in which there is the most
confidence in the associations and thus, confidence that estimated risk
reductions will actually occur.
The Administrator also notes the information provided by the
quantitative risk assessment on the distribution of concentrations
associated with the estimated mortality risk for a higher annual
standard level of 10.0 [micro]g/m\3\ and a lower standard level of 8.0
[micro]g/m\3\ (U.S. EPA, 2022b, sections 3.4.2.2 and 3.6.2.2, Figure 3-
18 and 3-19). The Administrator finds that, for an annual standard
level of 10.0 [micro]g/m\3\, the quantitative risk assessment estimates
that the standard would allow multiple exposures at concentrations
above the lowest means in the key epidemiologic studies, and therefore,
calls into question whether a standard level of 10.0 [micro]g/m\3\
would provide enough public health protection. Additionally, the
Administrator also finds that, for a lower annual standard level of 8.0
[micro]g/
[[Page 16283]]
m\3\, the quantitative risk assessment estimates the exposure
concentrations to be below 8 [micro]g/m\3\, with the majority of those
exposures being at concentrations of below 7 [micro]g/m\3\. The
Administrator observes that the majority of exposure concentrations
under this air quality scenario are estimated to fall outside of the
range of concentrations in which he has the most confidence in the
associations and that the additional risk reductions will actually
occur.
Recognizing and building upon the above considerations and
judgments, and with consideration of advice from the CASAC and public
comment, the Administrator concludes that the current body of
scientific evidence and quantitative risk assessment support his
judgment that the level of the primary annual PM2.5 standard
should be revised to a level of 9.0 [micro]g/m\3\. Revising the level
of the primary annual PM2.5 standard will, in the
Administrator's judgment, provide requisite public health protection
with an adequate margin of safety.
The Administrator recognizes that placing weight on the information
from the epidemiologic studies allows for examination of the entire
population, including those that may be at comparatively higher risk of
experiencing a PM2.5-related health effects (e.g., children,
older adults, minority populations) (88 FR 5624, January 27, 2023). In
considering the epidemiologic evidence, the Administrator judges that,
in reaching his decision on an appropriate level for the annual
standard that will protect public health with an adequate margin of
safety, in the absence of any discernible population-level thresholds,
and in recognizing the need to weigh uncertainties associated with the
epidemiologic evidence, it is most appropriate to examine where the
evidence of associations observed in the epidemiologic studies is
strongest and, conversely, to place less weight where he has less
confidence in the associations observed in the epidemiologic studies.
The Administrator notes that in previous reviews, evidence-based
approaches noted that the evidence of an association in any
epidemiologic study is ``strongest at and around the long-term average
where the data in the study are most concentrated'' (78 FR 3140,
January 15, 2013). These approaches were supported by previous CASAC
advice as well as the CASAC's advice in their review of the 2021 draft
PA as a part of this reconsideration. Given this, the Administrator
notes that in revising the annual PM2.5 standard to a level
of 9.0 [micro]g/m\3\, he is setting the standard at a level below the
long-term mean PM2.5 concentrations in the key long- and
short-term epidemiologic studies, including the lowest study reported
mean of 9.3 [micro]g/m\3\, following an approach that is consistent
with previous PM NAAQS reviews. The Administrator additionally notes
that air quality analyses in the 2022 PA demonstrate that areas meeting
a revised annual standard of 9.0 [micro]g/m\3\ would be expected to
shift the distribution of PM2.5 exposure concentrations in
an area such that the area's highest monitor would generally be at or
below 9.0 [micro]g/m\3\ annually, and most of the resulting
PM2.5 concentrations across the area would be even further
below the study-reported means.119 120 Thus, a standard
level of 9.0 [micro]g/m\3\ is expected to provide sufficient protection
not only in areas where the highest allowable concentration would be
located (i.e., near design value monitors) but also in other parts of
the area where PM2.5 concentrations would be expected to be
maintained even lower.
---------------------------------------------------------------------------
\119\ Analyses in the 2022 PA suggest that the highest monitored
value would be expected to be greater than the study-reported mean
values by 10-20% for monitor-based studies and 15-18% for hybrid
modeling studies that apply aspects of population weighting (U.S.
EPA, 2022b, section 2.3.3.2.4).
\120\ The risk assessment in the 2022 PA used air quality
adjustments to simulate just meeting the current primary
PM2.5 standards, as well as alternative standard levels
(U.S. EPA, 2022b, section 3.4.1.4 and Appendix C, section C.1.4).
---------------------------------------------------------------------------
Furthermore, the Administrator recognizes the CASAC's advice in
their review of the 2021 draft PA, as well as public comments, that
weight should be placed on study-reported PM2.5
concentrations that are somewhat below the mean, particularly for some
of the newer epidemiologic studies with larger cohort sizes. In
weighing uncertainties associated with using these data to inform a
revised annual standard level, as well as noting the limited studies
for which this information is available, the Administrator judges that
some weight should be placed on these data, but they should not receive
the same weight as the study-reported mean concentrations. Thus, the
Administrator concludes that it would be appropriate to set the annual
standard level near the 25th percentile PM2.5 concentrations
in the two newer key epidemiologic studies for which these values were
reported. In doing so, the Administrator notes that a decision to
revise the annual standard to 9.0 [micro]g/m\3\ would set a level of
the standard near and somewhat below the reported 25th percentile
PM2.5 concentrations of 9.1 [micro]g/m\3\ in these two more
recent hybrid model-based studies.
The Administrator also takes note of the study-reported long-term
mean PM2.5 concentrations in the key Canadian epidemiologic
studies. While the Administrator notes that these studies provide
additional support for associations between PM2.5
concentrations and health effects, he is also mindful that there are
important differences between the exposure environments in the U.S. and
Canada that affect interpretation of the data in the context of
informing decisions regarding potential alternative levels of the
annual standard. The Administrator also recognizes that the majority of
the CASAC in their review of the 2021 draft PA, as well as a number of
public commenters, placed weight on the Canadian epidemiologic studies
in recommending that the level of the primary annual PM2.5
standard be revised to 8-10 [micro]g/m\3\. The Administrator notes that
a decision to revise the annual standard to 9.0 [micro]g/m\3\ would set
the level of the standard within the range of levels recommended by the
majority of CASAC in their consideration of these studies.
Additionally, the Administrator also considers the information
provided by epidemiologic studies that use restricted analyses, as well
as accountability studies. With respect to the restricted analyses, the
Administrator, in considering the CASAC's advice in their review of the
2021 draft PA and many public comments on these types of studies,
concludes that, although there are inherent uncertainties associated
with this limited body of evidence, the studies that apply restricted
analyses provide support for serious effects (e.g., mortality) at
concentrations below 10.0 [micro]g/m\3\. Additionally, in considering
accountability studies, the Administrator concludes that while the
small number of these studies provide limited information for informing
decisions on the appropriate level of the primary annual
PM2.5 standard, these studies provide supplemental
information for consideration along with the full body of evidence. The
Administrator further notes that these studies suggest that public
health improvements may occur following the implementation of a policy
that reduces annual average PM2.5 concentrations below the
level of the current standard of 12.0 [micro]g/m\3\, and potentially
below the lowest ``starting'' concentrations in these studies of 10.0
[micro]g/m\3\. Taken together, the Administrator judges that it is
appropriate to place some weight on these types of studies,
particularly for informing his public policy judgments regarding an
adequate margin
[[Page 16284]]
of safety, and notes that a revised annual standard level of 9.0
[micro]g/m\3\ is below the lowest starting concentration of the
accountability studies (i.e., 10.0 [micro]g/m\3\), and below the
concentration at which studies that apply restricted analyses provide
support for serious effects (i.e., 9.6 [micro]g/m\3\).
The Administrator also judges that the results of the quantitative
risk assessment provide support for a primary annual PM2.5
standard with a level of 9.0 [micro]g/m\3\. The results of the risk
assessment suggest that when meeting an annual standard of 9.0
[micro]g/m\3\, PM2.5 exposures are maintained below 9
[micro]g/m\3\ at the design value monitor, with the majority of those
exposures being at concentrations below 8 [micro]g/m\3\. Thus, the
Administrator notes that an annual standard level of 9.0 [micro]g/m\3\
would be expected to provide protection from exposures where he has the
greatest confidence in the associations between health effects and
PM2.5 exposures (i.e. the long-term mean PM2.5
concentrations in the key U.S. epidemiologic studies, of which the
lowest is 9.3 [micro]g/m\3\) and would provide an adequate margin of
safety by maintaining most PM2.5 exposures even further
below 9.0 [micro]g/m\3\.
When considering adequate margin of safety, the Administrator notes
that in his decision to revise the annual standard level to 9.0
[micro]g/m\3\, he is placing weight on the information from the
epidemiologic studies which allows for examination of the entire
population, including those that may be at comparatively higher risk of
experiencing a PM2.5-related health effects (e.g., children,
older adults, minority populations). Additionally, as discussed above,
the Administrator also recognizes that setting the annual standard
level at 9.0 [micro]g/m\3\, which is below concentrations at which the
evidence is the strongest in supporting an association between exposure
to PM2.5 and adverse health effects observed in the key
epidemiologic studies available in this reconsideration, would be
expected to shift the distribution of PM2.5 exposure
concentrations in an area such that the area's highest monitor would
generally be at or below 9.0 [micro]g/m\3\ annually, and most of the
resulting PM2.5 concentrations across the area would be even
lower. In considering these air quality relationships, the
Administrator judges that a revised annual standard level of 9.0
[micro]g/m\3\ would provide requisite protection with adequate margin
of safety, for all populations, including those most at-risk.
In reaching this conclusion, the Administrator recognizes that in
establishing primary standards under the Act that are requisite to
protect public health with an adequate margin of safety, he is seeking
to establish standards that are neither more nor less stringent than
necessary for this purpose. The Act does not require that primary
standards be set at a zero-risk level or to protect the most sensitive
individual, but rather at a level that avoids unacceptable risks to
public health. In this context, the Administrator's conclusion is that
revised primary annual standard, in conjunction with the 24-hour
standard, provides the appropriate degree of protection, and that more
or less stringent standards would not be requisite.
In considering the requirement for an adequate margin of safety,
the Administrator notes that the determination of what constitutes an
adequate margin of safety is expressly left to the judgment of the EPA
Administrator. See Lead Industries Association v. EPA, 647 F.2d at
1161-62; Mississippi, 744 F.3d at 1353. He further notes that in
evaluating how particular standards address the requirement to provide
an adequate margin of safety, it is appropriate to consider such
factors as the nature and severity of the health effects, the size of
sensitive population(s) at risk, and the kind and degree of the
uncertainties present. Consistent with past practice and long-standing
judicial precedent, and as described in this section, the Administrator
takes the need for an adequate margin of safety into account as an
integral part of his decision making on a standard. See, e.g., NRDC v.
EPA, 902 F. 2d 962, 973-74 (D.C. Cir. 1990).
Given all of the evidence and information discussed above, the
Administrator judges that a standard with a level of 9.0 [micro]g/m\3\
is requisite to protect public health with an adequate margin of
safety. In so doing, he first recognizes that a less stringent standard
would allow the occurrence of higher long- and short-term
PM2.5 concentrations at a level at or above the mean
PM2.5 concentrations in key U.S. epidemiologic studies. That
is, a less stringent standard would be expected to allow more
PM2.5 exposures at concentrations at or above which the key
U.S. epidemiologic studies have reported associations between mean
PM2.5 concentrations and serious health effects and would
deviate from some past approaches for selecting the appropriate level
of the annual standard. A less stringent standard would also not
provide requisite protection with an adequate margin of safety against
PM2.5 exposures in the lower percentiles of the air quality
distribution (i.e., 25th percentile) for which associations with health
effects have been observed in a limited number of epidemiologic
studies. Furthermore, the Administrator notes that the primary annual
and 24-hour PM2.5 standards, together, are intended to
provide public health protection against the full distribution of long-
and short-term PM2.5 exposures. As noted above, the
Administrator recognizes that the changes in PM2.5 air
quality designed to meet a less stringent annual standard would likely
result in higher exposures across the distribution of air quality,
including both higher average (or typical) concentrations as well as
higher short-term peak PM2.5 concentrations. Taking into
consideration both the full evidence base for associations of
PM2.5 with mortality and other adverse health effects,
including the reported mean PM2.5 concentrations from key
long- and short-term U.S. epidemiologic studies, information from
epidemiologic studies that report 25th percentile PM2.5
concentrations, supplemental information from other epidemiologic
studies (i.e., epidemiologic studies that use restricted analyses,
accountability studies, and Canadian epidemiologic studies), and the
results of the risk assessment, as well as the advice from the CASAC
and public comments, the Administrator concludes that a less stringent
standard would allow risks of mortality and other adverse health
effects that are too great, and thus would not provide sufficient
protection for public health as required by the CAA.
Additionally, in considering a less stringent standard, the
Administrator recognizes that through its control of long- and short-
term PM2.5 concentrations, the annual standard provides a
margin of safety for less well-studied exposure levels and population
groups for which the evidence is limited or lacking. In so doing, he
recognizes that our understanding of the relationships between the
presence of a pollutant in ambient air and associated health effects is
based on a broad body of information encompassing not only more
established aspects of the evidence, such as the conclusion that long-
and short-term exposures to PM2.5 are causally related to
mortality and cardiovascular effects and likely to be causally related
to respiratory effects, but also aspects with which there may be
substantial uncertainty. In particular, the Administrator notes that
there are other categories of effects with causality determinations
that are suggestive of, but not sufficient to infer, a causal
[[Page 16285]]
relationship between PM2.5 exposure and health outcomes.
These include, but are not limited t,o short-term exposure and nervous
system effects, as well as long- and short-term exposure and pregnancy
and birth outcomes, where the evidence is less certain but which
represent potentially substantial additional risk to public health from
exposure to PM2.5. He recognizes the CAA requirement that
requires primary standards to provide an adequate margin of safety was
intended to address uncertainties associated with inconclusive
scientific and technical information as well as to provide a reasonable
degree of protection against hazards that research has not yet
identified and in his judgment, the primary NAAQS must be set at a
level that is adequately protective against these and other effects
which research has not yet identified. Thus, even if the Administrator
had somewhat greater concerns about the possibility of confounding,
error and bias in the epidemiologic studies, which reduced his
confidence in finding that PM2.5 is causally related to
mortality and cardiovascular effects, he would still find it
appropriate to set the primary NAAQS below the means of key U.S.
epidemiologic studies given the strength of the evidence providing
support for the association, as well as additional evidence linking
PM2.5 to other endpoints of substantial public health
concern, and the need to protect public health with an adequate margin
of safety. In considering the uncertainties in both the epidemiologic
evidence and the controlled human exposures studies, the Administrator
recognizes that collectively, the health effects evidence generally
reflects a continuum, consisting of levels at which scientists
generally agree that health effects are likely to occur, through lower
levels at which the likelihood and magnitude of the response become
increasingly uncertain. In light of these uncertainties, the
Administrator recognizes that the CAA requirement that primary
standards provide an adequate margin of safety, as summarized in
section I.A above, is intended to address uncertainties associated with
inconclusive scientific and technical information, as well as to
provide a reasonable degree of protection against hazards that research
has not yet identified. The Administrator has taken the need to provide
for an adequate margin of safety into account as an integral part of
his decision-making on the appropriate standards in setting the
standard at a level below the level where available epidemiologic
studies, which include diverse populations that are broadly
representative of the U.S. population including at-risk populations,
have provided the strongest evidence supporting effects, and in other
ways as well. For example, consideration of a margin of safety is
reflected in the approach of setting the level of the annual standard
near and somewhat below the 25th percentile PM2.5
concentrations from key U.S. epidemiologic studies (i.e., 9.1 [micro]g/
m\3\), as well as recognition that attaining a design value will
generally result in significantly broader and greater improvements of
air quality across an area (including but certainly not limited to
areas near the design value monitor) (U.S. EPA, 2022a, sections
2.3.3.2.4 and 3.3.3.2.1, Table 3-5). Based on all of the considerations
noted here, and considering the current body of evidence, including the
associated limitations and uncertainties, in combination with the
exposure/risk information, the Administrator concludes that a less
stringent standard than the current standard would not provide the
requisite protection of public health, including an adequate margin of
safety.
Having concluded that a less stringent standard would not provide
the requisite protection of public health, the Administrator next
considers whether a more stringent standard would be appropriate. In so
doing, he notes that a decision to set the level of the annual standard
to below 9.0 [micro]g/m\3\ would place a large amount of the emphasis
on potential public health importance of further reducing the
occurrence of PM2.5 concentrations of concern, though the
exposures about which he is most concerned are well controlled with an
annual standard level of 9.0 [micro]g/m\3\, as demonstrated by the
quantitative risk assessment. Such a decision would also place greater
weight on (1) further reducing ambient PM2.5 concentrations
relative to those observed in long-and short-term epidemiologic
studies, including those that he had judged to have significant
uncertainties, including Canadian studies, studies using restricted
analyses, and accountability studies; (2) shifting the air quality
distribution in areas such that the highest exposure concentrations are
reduced to below PM2.5 concentrations observed in
epidemiologic studies to be in the 25th or lower percentile, for which
the evidence is limited; and (3) further shifting exposure
concentrations to those shown at the lower end of the distribution in
the quantitative risk assessment, despite the important uncertainties
in the overall risk assessment. As discussed in this section and in
responses to significant comments above and in the Response to Comments
document, the Administrator has concluded that placing a large emphasis
on these factors and revising the standard to a level below 9.0
[micro]g/m\3\ would result in a standard that is more stringent than
the evidence indicates to be sufficient to protect public health with
an adequate margin of safety. Compared to a primary annual
PM2.5 standard set at a level of 9.0 [micro]g/m\3\, the
Administrator concludes that the extent to which lower standard levels
could result in further public health improvements becomes notably less
certain.
Thus, having carefully considered the scientific evidence,
quantitative information, CASAC advice, and public comments relevant to
his decision on the level of the primary annual PM2.5
standard, as discussed above and in the Response to Comments document,
the Administrator is revising the level of the primary annual
PM2.5 standard to 9.0 [micro]g/m\3\. In the Administrator's
judgment, based on the currently available evidence and information, an
annual standard set at this level and using the specified indicator,
averaging time, and form, in conjunction with the other primary PM
standards, would be requisite to protect public health with an adequate
margin of safety. The Administrator judges that such a standard would
protect, with an adequate margin of safety, the health of at-risk
populations, including children, older adults, those with pre-existing
cardiovascular and respiratory diseases, minority populations, and low
SES populations. The Administrator believes that a standard set at 9.0
[micro]g/m\3\ would be sufficient to protect public health with a
margin of safety, and believes that a lower standard would be more than
what is necessary to provide this degree of protection. This judgment
by the Administrator appropriately considers the degree of protection
that is neither more nor less stringent than necessary for this purpose
and recognizes that the CAA does not require that primary standards be
set at a zero-risk level, but rather at a level that reduces risk
sufficiently so as to protect public health with an adequate margin of
safety.
In reaching his conclusions on adequacy of the current suite of
primary PM2.5 standards, based on consideration of the
available scientific evidence and quantitative information, the CASAC's
advice and public comments, the Administrator finds that the available
information is insufficient to call into question the adequacy of the
public
[[Page 16286]]
health protection afforded by the current primary 24-hour
PM2.5 standard. As described earlier in this section, the
Administrator concludes that it is appropriate to retain the current
indicator (PM2.5), averaging time (24-hour), and form (98th
percentile, averaged over three years) for the primary 24-hour
PM2.5 standard and below explains the basis for his final
decision that is also appropriate to retain the current level of the
primary 24-hour PM2.5 standard.
In reaching his conclusion to retain the current primary 24-hour
PM2.5 standard the Administrator does so in light of the
conclusion that the epidemiologic evidence supports associations
between short- and long-term PM2.5 exposures and adverse
health effects, but that the epidemiologic evidence does not identify
specific concentrations at which those effects occur and the
Administrator has greatest confidence in effects where the bulk of the
data is reported (i.e., the mean PM2.5 concentration, with
some consideration for the 25th percentile of the air quality
distribution). Thus, in considering the epidemiologic evidence, the
Administrator concludes it is appropriate to focus on setting a
generally controlling annual standard as the most effective and
efficient way to reduce total population risk associated with both
long- and short-term PM2.5 exposures, and that it is
appropriate to revise the level of the annual standard level to 9.0
[micro]g/m\3\. In addition to the epidemiologic evidence, the
Administrator also considers the available controlled human exposure
studies, which provide evidence for health effects following single,
short-term PM2.5 exposures to concentrations that typically
correspond to upper end of the PM2.5 air quality
distribution in the U.S. (i.e., ``peak'' concentrations). In so doing,
the Administrator notes that these studies report statistically
significant effects on one or more indicators of cardiovascular
function following 2-hour exposures to PM2.5 concentrations
at and above 120 [mu]g/m\3\ and at and above 149 [mu]g/m\3\ for
vascular impairment, the effect shown to be most consistent across
studies. In particular, the Administrator notes that a single study is
assessed in the ISA Supplement that reports effects following 4-hour
exposures at 37.8 [micro]g/m\3\, although the results of this study are
inconsistent with the results of the controlled human exposure studies
assessed in the 2019 ISA. Along with the inconsistent results from the
controlled human exposure studies, the Administrator also recognizes
that effects observed in these studies are intermediate effects which
are not typically considered adverse and that the study participants
were healthy individuals. Taking into consideration the available
scientific evidence, including the uncertainties and limitations, along
with the CASAC's advice, the Administrator concludes that it is
appropriate to maintain a primary 24-hour PM2.5 standard to
protect against peak exposures.
Thus, the Administrator considers what primary 24-hour
PM2.5 standard is requisite to provide supplemental
protection against peak exposures. While having confidence that the
revised annual standard will result in lowering risk associated with
both long- and short-term PM2.5 exposure by lowering the
overall air quality distribution, as in the 2012 review, the
Administrator recognizes that an annual standard alone would not be
expected to offer sufficient protection with an adequate margin of
safety against the effects of short-term PM2.5 exposures in
all parts of the country. Therefore, he continues to conclude that it
is appropriate to continue to provide supplemental protection by means
of a 24-hour standard, in conjunction with a revised annual standard
level of 9.0 [micro]g/m\3\.
In considering the available scientific evidence assessed in the
2019 ISA and ISA Supplement, the Administrator first considers the
controlled human exposure studies for informing his decisions on the
primary 24-hour PM2.5 standard. In so doing, he notes that
in their review of the 2021 draft PA, the majority of CASAC members
expressed the view that controlled human exposure studies are not the
best evidence to use for justifying retaining the 24-hour standard
without revision, in part because these studies preferentially recruit
less susceptible individuals and have a typical exposure duration much
shorter than 24 hours. Thus, in the view of the majority, ``the
evidence of effects from controlled human exposure studies with
exposures close to the current 24-hour standard supports
epidemiological evidence for lowering the standard'' (Sheppard, 2022a,
p. 3-4 of consensus letter). In reviewing the controlled human exposure
studies, the Administrator agrees with the majority of CASAC that these
controlled human exposure studies generally do not include populations
with substantially increased risk from exposure to PM2.5,
such as children, older adults, or those with more severe underlying
illness. However, he disagrees with any conclusion that they should not
be used to inform a decision about the adequacy of the current
standard. The Administrator finds the information available from these
studies to be useful, noting that the recently available controlled
human exposure studies provide evidence for health effects following
single, short-term exposures to PM2.5 concentrations that
are greater than those allowed under the current standard. The results
of the controlled human exposure studies are inconsistent, particularly
at lower PM2.5 concentrations, but some studies do report
statistically significant effects on one or more indicators of
cardiovascular function following 2-hour exposures to PM2.5
concentrations at and above 120 [mu]g/m\3\ (and at and above 149 [mu]g/
m\3\ for vascular impairment, the effect shown to be most consistent
across studies). Additionally, one controlled human exposure study
assessed in the ISA Supplement reports evidence of some effects for
cardiovascular markers following 4-hour exposures to 37.8 [micro]g/m\3\
(Wyatt et al., 2020). However, there is inconsistent evidence for
inflammation in other controlled human exposure studies evaluated in
the 2019 ISA. The Administrator finds these studies are important in
establishing biological plausibility for PM2.5 exposures
causing more serious health effects, such as those seen in short-term
exposure epidemiologic studies, and they provide support that more
adverse effects may be experienced following longer exposure durations
and/or exposure to higher concentrations. As described in more detail
in responding to public comments in section II.B.3 above, he notes that
although the controlled human exposure studies do not provide a
threshold below which no effects occur, the observed effects in these
controlled human exposures studies are ones that signal an intermediate
effect in the body, likely due to short-term exposure to
PM2.5, and typically would not, by themselves, be judged as
adverse. As noted in sections II.A.2 and II.B.3 above, associated
judgments regarding adversity or health significance of measurable
physiological responses to air pollutants in previous NAAQS reviews
have been informed by guidance, criteria or interpretative statements
developed within the public health community. This type of information
on adversity of effects is particularly informative to the
Administrator's judgments regarding the adversity of the effects
observed in the controlled human exposure studies which are short-term
in nature (i.e., generally ranging from 2- to 5-hours), including those
studies that are
[[Page 16287]]
conducted at near-ambient PM2.5 concentrations. Based on the
observation that the effects observed in Wyatt et al. (2020) are not by
themselves adverse, and the fact that the findings of this study are
inconsistent with other currently available evidence regarding the
level at which effects are observed, the Administrator disagrees with
the view expressed by the majority of CASAC that this study supports
epidemiologic evidence for lowering the 24-hour standard.
Consistent with his approach in reaching his proposed decision and
taking into consideration these points as well as balancing these
limitations (i.e., that the health outcomes observed in these
controlled human exposure studies are not clearly adverse and that the
studies generally do not include those at increased risk from
PM2.5 exposure), the Administrator still considers it
appropriate to ensure that the 24-hour PM2.5 standard
provides protection against health effects consistently observed in the
controlled human exposure studies. He next examines the air quality
analyses, described in more detail in section II.A.c.i above, to assess
whether during recent air quality conditions, areas meeting the current
standards would experience PM2.5 concentrations reported in
these controlled human exposure studies. He observes that air quality
analyses demonstrate that the PM2.5 exposures shown to cause
consistent effects in the controlled human exposure studies are well
above the ambient concentrations typically measured in locations
meeting the current primary standards, and therefore suggest that the
current primary PM2.5 standards provide protection against
these ``peak'' concentrations. In fact, at air quality monitoring sites
meeting the current primary PM2.5 standards (i.e., the 24-
hour standard of 35 [mu]g/m\3\ and the annual standard of 12 [mu]g/
m\3\), the 2-hour concentrations generally remain below 10 [mu]g/m\3\,
and rarely exceed 30 [mu]g/m\3\. Though two-hour concentrations are
higher at monitoring sites violating the current standards, they
generally remain below 16 [mu]g/m\3\ and rarely exceed 80 [mu]g/m\3\,
still below concentrations in CHE studies where consistent effects are
observed (e.g., greater than 120 [mu]g/m\3\) (U.S. EPA, 2022b, section
2.3.2.2.3, Figure 2-19, and section 3.3.3.1). Additionally, and in
response to public comments, the Administrator notes additional air
quality analyses conducted by the EPA,\121\ that provide a more refined
analysis of whether areas that meet the current standards experience
peak concentrations reported in controlled human exposure studies. He
notes that 2-hour observations greater than 120 [mu]g/m\3\ and 4-hour
observations greater than 38 [mu]g/m\3\ rarely occur (e.g., 0.025% of
rolling 2-hour observations are greater than 120 [mu]g/m\3\ and 0.78%
of rolling 4-hour observations greater than 38 [mu]g/m\3\). Based on
this information, the Administrator finds that the current suite of
standards maintains subdaily concentrations of PM2.5 in
ambient air far below the exposure concentrations in controlled human
exposure studies where consistent effects have been observed, and notes
that while these studies generally do not include the most at-risk
individuals, the exposure concentrations in these studies also do not
elicit adverse effects.
---------------------------------------------------------------------------
\121\ Jones et al. (2023). Comparison of Occurrence of
Scientifically Relevant Air Quality Observations Between Design
Value Groups. Memorandum to the Rulemaking Docket for the Review of
the National Ambient Air Quality Standards for Particulate Matter
(EPA-HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
Further, in light of the Administrator's emphasis on the annual
standard as the controlling standard, with the 24-hour standard
providing supplemental protection against peak concentrations, he next
considers the potential impact of a revised annual standard of 9.0
[micro]g/m\3\ on the occurrence of peak sub-daily PM2.5
concentrations. Specifically, the Administrator takes note of the new
air quality analyses \122\ where he observes that lower percentages of
concentrations greater than 120 [micro]g/m\3\ and 38 [micro]g/m\3\
occur in areas meeting an annual standard of 9.0 [micro]g/m\3\ and a
24-hour standard of 35 [micro]g/m\3\, versus an annual standard of 12.0
[micro]g/m\3\ and a 24-hour standard of 35 [micro]g/m\3\. Thus, he
concludes that an annual standard that is controlling across most areas
of the country will continue to effectively limit peak daily
concentrations in conjunction with the existing 24-hour standard, with
its level of 35 [micro]g/m\3\ and 98th percentile form, which continues
to provide supplemental protection against peak concentrations.
---------------------------------------------------------------------------
\122\ Jones et al. (2023). Comparison of Occurrence of
Scientifically Relevant Air Quality Observations Between Design
Value Groups. Memorandum to the Rulemaking Docket for the Review of
the National Ambient Air Quality Standards for Particulate Matter
(EPA-HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
In addition, the Administrator also notes that the majority of the
CASAC in their review of the 2021 draft PA, as well as a number of
public commenters, support their recommendation to revise the current
24-hour standard by pointing to ``substantial epidemiologic evidence
from both morbidity and mortality studies'' which ``includes three U.S.
air pollution studies with analyses restricted to 24-hour
concentrations below 25 [mu]g/m\3\'' (Sheppard, 2022a, p. 17 consensus
responses). The Administrator notes that the epidemiologic evidence
available in this reconsideration, including the studies that restrict
short-term PM2.5 exposures (i.e., 24-hour PM2.5
concentrations) to levels below 25 [mu]g/m\3\, provides support for
positive and statistically significant associations between short-term
exposure to PM2.5 and all-cause mortality (Di et al., 2017a)
and CVD hospital admissions (deSouza et al., 2021; Di et al., 2017a).
He agrees that these studies help to provide additional support for
reaching conclusions on causality in the 2019 ISA. He further agrees
that the available epidemiologic studies provide important information
that it is appropriate to consider in this reconsideration, including
information on associations between health effects and PM2.5
exposures in diverse populations that are broadly representative of the
U.S. population, and include populations identified as at-risk (e.g.,
older adults, minority populations), as well as evidence of linear, no-
threshold concentration-response relationships in those associations,
although with less certainty in the shape of the curve at long-term
average concentrations below about 8 [mu]g/m\3\.
However, the Administrator also notes significant limitations in
the currently available epidemiologic information that limit his
ability to draw conclusions from the key short-term studies, including
those that employ restricted analyses, to inform his decision regarding
the level of the 24-hour PM2.5 standard. As a result of
these limitations, the Administrator does not find that the short-term
epidemiologic studies, or the other evidence such as the controlled
human exposure studies or the risk assessment, provide a sufficient
justification for revising the 24-hour standard.
First, he notes that short-term epidemiologic studies examine
associations between day-to-day variations in PM2.5
concentrations and health outcomes, often over multi-year study
periods. As such, these studies report long-term mean 24-hour
PM2.5 concentrations (e.g., mean 24-hour PM2.5
concentrations over multi-year study periods), rather than at specific
points in the distribution (i.e., 90th or 98th percentile 24-hour
concentrations) at which effects occur. Further, he notes
[[Page 16288]]
that while there can be considerable variability in daily exposures
over a multi-year study period, the bulk of the observations reflect
days with ambient PM2.5 concentrations in the middle of the
air quality distribution (i.e., ``typical'' days rather than days with
extremely low or extremely high concentrations). As a result, the
results of these studies are more directly applicable to decisions
regarding the annual standard (which is based on the long-term mean of
both short- and long-term epidemiologic studies), and the fact that
they do not report other air quality statistics, such as the 98th
percentile concentrations which might be more directly compared to the
level of the 24-hour standard, makes them less useful for informing
decisions on the 24-hour standard. As discussed in responding to
comments above, the form of the annual standard is based on the annual
mean PM2.5 concentration averaged over three years,\123\
which makes it better suited as a basis for controlling air quality to
avoid effects observed in both long-term and short-term epidemiologic
studies. By contrast, the form of the 24-hour standard is the 98th
percentile averaged over three years, which makes it appropriate for
controlling short-term peak concentrations. However, based on the
available air quality information, including distribution statistics of
PM2.5 concentrations and health events reported in the
short-term epidemiologic studies, these studies are too limited in
their ability to identify health effects attributable to specific
short-term peak concentrations that are necessary to evaluate whether
the 24-hour standard with its 98th percentile form should be revised
(e.g., restricted epidemiologic studies do not report the number or the
percentile of health events or the percentile of PM2.5
concentrations across the highest part of the restricted air quality
distribution, including the 98th percentile). Thus, the Administrator
does not consider it appropriate to use the reported means from short-
term studies to determine the appropriate level for a 24-hour standard
with a 98th percentile form.
---------------------------------------------------------------------------
\123\ The annual mean is calculated by averaging daily values in
a calendar quarter and then averaging calendar quarters. See 40 CFR
part 50 Appendix N, section 4.4.
---------------------------------------------------------------------------
Similarly, the Administrator does not consider the results of the
restricted analyses to be well suited to informing the choice of level
for a 24-hour standard. Restricted analyses use a subset of data from
their main analyses to evaluate health events that occur at
concentrations below a certain concentration (e.g., 25 [mu]g/m\3\). The
Administrator notes that the associations between the health effects
(e.g., mortality and cardiovascular morbidity) and PM2.5
concentrations remain even after excluding higher concentrations in the
restricted analyses, and he also recognizes that the magnitude of the
effect is generally greater in the restricted analyses compared to the
associations reported in the main analysis. He considers such analyses
to be informative in indicating that the health effects association
reported in the main (unrestricted) analysis are not driven only by the
upper peaks of the PM2.5 air quality distribution, but
rather persist at lower portions of the distribution (consistent with
his emphasis on the annual standard, which is focused on exposures near
the mean concentration, where the bulk of the exposure distribution is
concentrated). Indeed, he notes that if peak concentrations were the
principal driver of health effects associated with PM2.5
exposure, one might expect the associations to become weaker as the
upper portion of the data is excluded in the restricted analyses, which
is not what is reported by the analyses (e.g., the restricted analyses
generally report associations that are greater in magnitude compared to
the main analyses). However, he disagrees with the assertion by the
CASAC in their review of the 2021 draft PA and some public commenters
that it would be appropriate to focus on the specific PM2.5
concentration (e.g., 25 or 30 [mu]g/m\3\) at which the analysis was
restricted as the basis for choosing a 24-hour standard level. The
Administrator recognizes that in restricted analyses, while an
association continues to persist across the full range of the air
quality distribution, and that the cutpoint concentration at which the
analysis was restricted (e.g., 25 or 30 [micro]g/m\3\) becomes the
maximum PM2.5 concentration in the distribution, he also
notes that these studies do not provide information related to the
distribution of health events and PM2.5 concentrations, and
as such, he is more uncertain where the bulk of the data are and where
he has confidence in the reported association.\124\ He notes that no
evidence exists to support a conclusion that the PM2.5
concentration chosen as the cutpoint in a restricted analysis has any
bearing on the concentration at which effects are likely to occur (or
not occur). He notes that, as with long-term studies, the evidence does
not suggest there is a specific point in the air quality distribution
of these short-term studies that represents a ``bright line'' at and
above which effects have been observed and below which effects have not
been observed. In order to identify a level of the 24-hour standard
based on associations between the ``upper end'' of exposures, either in
the unrestricted or the restricted analyses, and adverse health
effects, it would be necessary to have a better understanding of how
specific 24-hour concentrations correspond to the frequency and total
number of observed health events in the study. Currently, such
information, including 98th percentile statistics, are not reported in
the key short-term epidemiologic studies (and if they were reported,
the Administrator would have to carefully consider how to weigh the
data). As such, in reaching his decision on the primary 24-hour
PM2.5 standard, the Administrator judges that the currently
available information from short-term epidemiologic studies, including
those that employ restricted analyses, does not provide a sufficient
basis to revise the current 24-hour standard, given that the 24-hour
standard focuses on reducing ``peak'' exposures (with its 98th
percentile form), but rather that such information supports his
judgment that it is appropriate to focus on revising the annual
standard for purposes of reducing all exposures, across the entire
distribution of air quality, to increase public health protection.
---------------------------------------------------------------------------
\124\ These studies do not report information about the
distribution of the health events and PM2.5
concentrations (e.g., means, medians, other percentiles) in the
restricted analyses.
---------------------------------------------------------------------------
In reaching final decisions regarding the adequacy of the primary
24-hour PM2.5 standard, the Administrator continues to view
an approach that focuses on setting a generally controlling annual
standard as the most effective and efficient way to reduce total
population risk associated with both long- and short-term
PM2.5 exposures. Additionally, he emphasizes that
improvements in air quality associated with meeting an annual standard
level of 9.0 [micro]g/m\3\ will result in lowering risk associated with
both long- and short-term PM2.5 exposure by lowering the
overall air quality distribution. The Administrator concludes that
reducing the annual standard is the most efficient way to reduce the
risks from short-term exposures identified in the epidemiologic
studies, as the available evidence suggests the bulk of the risk comes
from the large number of days across the bulk of the air quality
distribution, not the relatively small number of days with peak
concentrations. However, as in the 2012
[[Page 16289]]
review, the Administrator recognizes that an annual standard alone
would not be expected to offer sufficient protection with an adequate
margin of safety against the effects of short-term PM2.5
exposures in all parts of the country and concludes that, in
conjunction with a revised annual standard level of 9.0 [micro]g/m\3\,
it is appropriate to continue to provide supplemental protection by
means of a 24-hour standard, particularly for areas with high peak-to-
mean ratios possibly associated with strong local or seasonal sources.
In selecting the level of a 24-hour standard designed to provide
supplemental protection against peak exposures (in conjunction with a
revised annual standard of 9.0 [micro]g/m\3\), the Administrator
considers the information from the controlled human exposure studies
and the EPA's analysis of peak concentrations observed in areas meeting
the current standard of 35 [micro]g/m\3\ in conjunction with a revised
standard of 9.0 [micro]g/m\3\ to be of particular relevance. He notes
the controlled human exposure evidence includes studies reporting
effects on one or more indicators of cardiovascular function following
2-hour exposures at and above 120 [micro]g/m\3\, including effects
reported at and above 149 [micro]g/m\3\ for vascular impairment, the
effect shown to be most consistent across studies, and less consistent
effects at lower concentrations, including a single study at near
ambient concentrations (Wyatt et al., 2020) reporting effects following
4-hour exposures at 37.8 [micro]g/m\3\. He recognizes that the effects
observed (in those studies that observed effects) are ones that signal
an intermediate effect in the body, likely due to short-term exposure
to PM2.5, and typically would not, by themselves, be judged
as adverse, and the study participants were healthy individuals.
He notes in particular that, in the EPA's analysis, in areas
meeting the current 24-hour standard and the revised annual standard
0.029 percent of 2-hour observations and 0.41 percent of 4-hour
observations reach PM2.5 concentrations higher than 120
[micro]g/m\3\ and 37.8 [micro]g/m\3\, respectively. He also notes the
lack of evidence of effects from controlled human exposure studies at
levels below the current 24-hour standard and the fact that the results
of Wyatt et al. (2020) are inconsistent with other available studies,
as well as the intermediate nature of effects observed in this study.
In his judgment, the small number of occurrences of peak exposures
indicate that, in conjunction with a revised annual standard of 9.0
[micro]g/m\3\, the current 24-hour standard of 35 [micro]g/m\3\ remains
requisite to protect public health with an adequate margin of safety,
and that there is substantial basis to doubt whether further
improvements in public health would be achieved by further reducing
these exposures. Furthermore, the Administrator concludes that due to
the limitations and uncertainties outlined above, the information from
recent short-term epidemiologic studies, including those that use
restricted analyses, is inadequate to inform decisions regarding the
adequacy of the current 24-hour standard. Thus, in reaching his
decision on the primary 24-hour PM2.5 standard, the
Administrator concludes that currently available evidence does not call
into question the adequacy of the current standard.
In addition to the scientific evidence, the Administrator also
considers the risk assessment in evaluating the appropriate level of
the 24-hour PM2.5 standard. The risk assessment indicates
that the annual standard is the controlling standard across most of the
urban study areas evaluated (i.e., when air quality related to the
annual average PM2.5 concentrations decrease, daily average
PM2.5 concentrations are also expected to decrease). When
air quality is adjusted to just meet an alternative 24-hour standard
level of 30 [mu]g/m\3\ in the areas where the 24-hour standard is
controlling, the risk assessment estimates reductions in
PM2.5-associated risks across a more limited population and
number of areas compared to when air quality is adjusted to simulate
alternative levels for the annual standard (i.e., where the annual
standard is controlling), and these predictions are largely confined to
areas located in the western U.S., several of which are also likely to
experience risk reductions upon meeting a revised annual standard. With
respect to the CASAC's advice in their review of the 2021 draft PA, the
Administrator notes that the minority of CASAC advised that these
results suggest that the annual standard can be used to limit both
long- and short-term PM2.5 concentrations and views these
risk assessment results as supporting the conclusion that the current
24-hour standard is adequate (Sheppard, 2022a, p. 4 of consensus
letter). In contrast, the majority of CASAC members in their review of
the 2021 draft PA, as well as a number of public commenters that
support revision of the 24-hour standard, placed greater weight on the
evidence-based considerations (e.g. scientific evidence, like the
restricted analyses) than on the values estimated by the risk
assessment, noting the potential for uncertainties in how the risk
assessment was able to ``capture areas with wintertime stagnation and
residential wood-burning where the annual standard is less likely to be
protective'' (Sheppard, 2022a, p. 4 of consensus letter).
In considering the application of the risk assessment to judgments
about the adequacy of the current primary 24-hour PM2.5
standard, the Administrator again notes that the risk assessment
analyses of PM2.5-attributable mortality use input data that
include C-R functions from epidemiologic studies that have no threshold
and a linear C-R relationship down to zero, as well an air quality
adjustment approach that incorporates proportional decreases in
PM2.5 concentrations to meet lower standard levels. As such,
the Administrator notes that this quantitative approach does not
incorporate any elements of uncertainty in associations of health
effects at lower concentrations and that simulated air quality
improvements will always lead to proportional decreases in risk (i.e.,
each additional [micro]g/m\3\ reduction produces additional benefits
with no clear stopping point at any PM2.5 concentration).
Therefore, the Administrator recognizes that while the risk estimates
can help to place the evidence for specific health effects into a
broader public health context, the results should be considered along
with the inherent uncertainties and limitations of such analyses when
informing judgments about the potential for additional public health
protection associated with PM2.5 exposure and related health
effects. Further, the Administrator notes additionally that air quality
analyses have also been considered in looking at the adequacy of the
24-hour standard in controlling peak PM2.5 concentrations of
potential concern,\125\ and that those analyses included monitoring
information from across the entire U.S., specifically highlighting
areas with higher peak concentrations and including areas impacted by
wintertime stagnation and residential wood-burning. Thus, while the
risk assessment may have focused on a subset of areas across the U.S.
based on the study area selection criteria, the Administrator is
considering a broader set of information in reaching his conclusions
regarding the appropriateness of the current 24-
[[Page 16290]]
hour standard to control peak concentrations.
---------------------------------------------------------------------------
\125\ Jones et al. (2023). Comparison of Occurrence of
Scientifically Relevant Air Quality Observations Between Design
Value Groups. Memorandum to the Rulemaking Docket for the Review of
the National Ambient Air Quality Standards for Particulate Matter
(EPA-HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
The Administrator also considers the advice from the CASAC in their
reviews of the 2019 draft PA and 2021 draft PA. In their review of the
2019 draft PA, the CASAC ``agrees with the EPA and finds that the
available evidence does not call into question the adequacy of public
health protection afforded by the current 24-hour PM2.5
standard and concurs that it be retained'' (Cox, 2019b, p. 3 of
letter). He also notes that in their review of the 2021 draft PA, the
CASAC did not reach consensus on whether the current 24-hour standard
is adequate, with the majority of the CASAC recommending that the 24-
hour standard be revised and the minority of the CASAC recommending
that the standard be retained. The majority of the CASAC members
further stated that ``[t]here is also less confidence that the annual
standard could adequately protect against health effects of short-term
exposures. A range of 25-30 [mu]g/m\3\ for the 24-hour PM2.5
standard would be adequately protective'' (Sheppard, 2022a, p. 4 of
consensus letter). The Administrator also acknowledges that some public
commenters agreed with the majority of the CASAC in supporting a
revision to the level of the 24-hour standard to a range between 25-30
[micro]g/m\3\. These commenters cite a number of reasons, including:
(1) Results from controlled human exposure studies at near ambient
concentrations; (2) aspects of the scientific evidence, including
restricted analyses that report positive and significant associations
below 35 [micro]g/m\3\; and (3) quantitative risk analyses that show
decreasing risk with decreasing PM2.5 concentrations. In
responding to these comments, the Administrator recognizes that some
commenters have different interpretations of the evidence, air quality
information, and quantitative results from the risk assessment in this
review and would make different judgments about the weight to place on
the relative strength and limitations of the currently available
scientific evidence and information and how such information could be
used in making public health policy decisions on the 24-hour standard.
However, as outlined above, the Administrator has carefully considered
the information available from controlled human exposure studies and
short-term epidemiologic studies, and weighed the strengths and
limitations of this evidence in formulating his decisions. Furthermore,
as discussed above the Administrator has noted significant
uncertainties and limitations inherent in the risk estimates, as well
as noting that very few areas were included. In addition, he has given
careful consideration to the majority of the CASAC's advice in their
review of the 2021 draft PA, but has drawn different conclusions with
respect to how currently available evidence and air quality information
inform the selection of level for the 24-hour primary PM2.5
standard.
In considering the advice of the majority of CASAC, the
Administrator notes that a decision to set the level of the 24-hour
standard to below 35 [micro]g/m\3\ would place a large amount of
emphasis on the potential public health importance of further reducing
the occurrence of peak PM2.5 concentrations. However, the
Administrator concludes that there is insufficient basis to conclude
that a more stringent standard to further reduce peak concentrations is
needed or would benefit public health. As discussed above, he judges
that the PM2.5 exposures in controlled human exposure
studies that correspond to peak concentrations will already be well
controlled via the combination of the revised annual standard, with a
level of 9.0 [micro]g/m\3\, and the 24-hour standard with its level 35
[micro]g/m\3\ and its 98th percentile form. Taking into consideration
the inconsistent results reported in controlled human exposure studies,
the intermediate nature of the health effects observed in the
controlled human exposure studies that are not typically considered
adverse, the health status of the study participants, and how
infrequently peak concentrations of potential concern are anticipated
to occur in areas meeting the revised primary annual PM2.5
standard, he judges that the current 24-hour standard is requisite to
protect against the effects reported in these studies with an adequate
margin of safety. Likewise, he judges that neither the epidemiologic
studies (including the studies that use restricted analyses) nor the
risk assessment provide a sufficient basis for revising the 24-hour
standard. As discussed above, the epidemiologic studies, including
short-term studies and those with restricted analyses, are not well-
suited for identifying a level for a 24-hour standard to address health
effects associated with peak concentrations. The restricted analyses
support the conclusion that the health effects associated with
PM2.5 is not associated primarily with exposure to higher
concentrations of the main analyses, but like other epidemiologic
studies they typically report only long-term mean 24-hour
concentrations (e.g., restricted epidemiologic studies do not report
the number or the percentile of health events or the percentile of
PM2.5 concentrations across the highest part of the
restricted air quality distribution, including the 98th percentile) and
do not identify any particular concentration within the air quality
distribution above which effects have been observed and below which
effects have not been observed. Similarly, the risk assessment
highlights that the annual standard is controlling across much of the
U.S. and is generally more effective at reducing risk than the 24-hour
standard and, taking into account the limitations and assumptions of
the risk assessment discussed above, does not provide a basis for
revising the 24-hour standard. For the reasons discussed herein, the
Administrator judges that the uncertainties as to whether there would
be public health benefits from a more stringent 24-hour standard are
too great to justify revising the standard.
Thus, having carefully considered the scientific evidence,
quantitative information, CASAC advice, and public comments, the
Administrator is retaining the current primary 24-hour PM2.5
standard, with its level of to 35 [micro]g/m\3\ and its 98th percentile
form. In the Administrator's judgment, based on the currently available
evidence and information, a 24-hour standard set at this level and
using the specified indicator, averaging time, and form would be
requisite to protect public health with an adequate margin of safety,
in conjunction with the annual standard. As noted, in evaluating the
adequacy of the current standards, the Administrator focuses on
evaluating the public health protection afforded by the annual and 24-
hour standards, taken together, against adverse health effects
associated with long- or short-term PM2.5 exposures. A 24-
hour standard set at a level of 35 [micro]g/m\3\, in conjunction with a
revised annual standard level of 9.0 [micro]g/m\3\, in the judgment of
the Administrator, provides an appropriate level of public health
protection, for both long- and short-term PM2.5 exposures.
The Administrator believes that a 24-hour standard set at 35 [micro]g/
m\3\ would continue to be sufficient to protect public health with a
margin of safety, and believes that a lower standard would be more than
what is necessary to provide this degree of protection when considered
in conjunction with a revised annual standard. The Administrator
concludes the current 24-hour standard at a level of 35 [micro]g/m\3\,
in conjunction with a revised annual standard level of 9.0 [micro]g/
m\3\, will provide appropriate protection
[[Page 16291]]
in areas in which the long-term mean concentrations are already
relatively low (i.e., below 9 [micro]g/m\3\) but where there may be
elevated short-term peak PM2.5 concentrations, often
associated with strong local or seasonal sources. This judgment by the
Administrator appropriately considers the degree of protection that is
neither more nor less stringent than necessary for this purpose and
recognizes that the CAA does not require that primary standards be set
at a zero-risk level, but rather at a level that reduces risk
sufficiently so as to protect public health with an adequate margin of
safety.
In making this decision to retain the current level of the primary
PM2.5 24-hour standard at 35 [micro]g/m\3\ in conjunction
with revising the annual standard level from 12.0 [micro]g/m\3\ to 9.0
[micro]g/m\3\, given all of the evidence and information discussed
above, the Administrator judges that the revised suite of primary
PM2.5 standards and the rationale supporting these levels
appropriately reflects consideration of the strength of the available
evidence and other information and its associated uncertainties as well
as the advice of CASAC and consideration of public comments. He
additionally judges that this suite of primary PM2.5
standards is requisite to protect public health, including at-risk
populations, with an adequate margin of safety from effects associated
with long and short-term exposures to fine particles. This judgment by
the Administrator appropriately considers the requirement for standards
that are requisite to protect public health but are neither more nor
less stringent than necessary.
C. Decisions on the Primary PM2.5 Standards
For the reasons discussed above, and taking into account the
information and assessments presented in the 2019 ISA and ISA
Supplement, the scientific and quantitative risk information in the
2022 PA, the advice and recommendations of the CASAC, and public
comments, the Administrator revises the current suite of primary
PM2.5 standards. Specifically, the Administrator revises the
level of the primary annual PM2.5 standard to 9.0 [micro]g/
m\3\ while retaining its form, indicator and averaging time. In
conjunction with revising the primary annual PM2.5 standard
level to provide protection from effects associated with long- and
short-term PM2.5 exposures, the Administrator retains the
level of 35 [micro]g/m\3\ and the 98th percentile form, indicator and
averaging time of the primary 24-hour PM2.5 standard to
continue to provide supplemental protection for areas with high peak
PM2.5 concentrations. The Administrator concludes that this
suite of standards is requisite to protect public health with an
adequate margin of safety against health effects potentially associated
with long- and short-term PM2.5 exposures.
III. Rationale for Decisions on the Primary PM10 Standard
This section presents the rationale for the Administrator's
decision to retain the existing primary PM10 standard. This
decision is based on a thorough review of the latest scientific
information, published through January 2018 \126\ and evaluated in the
2019 ISA, on human health effects associated with PM10-2.5
in ambient air. As described in section I above and in section 1.2 of
the ISA Supplement, the scope of the updated scientific evaluation of
the health effects evidence is based on those PM size fractions,
exposure durations, and health effects category combinations where the
2019 ISA concluded a causal relationship exists (U.S. EPA, 2019a, U.S.
EPA, 2022b).). Therefore, because the 2019 ISA did not conclude a
causal relationship for PM10-2.5 for any exposure durations
or health effect categories, the ISA Supplement does not include an
evaluation of additional studies for PM10-2.5. As a result,
the 2019 ISA continues to serve as the scientific foundation for
assessing the adequacy of the primary PM10 standard in this
reconsideration of the 2020 final decision (U.S. EPA, 2019a, section
1.7; U.S. EPA, 2022a). The Administrator's decision also takes into
account the 2022 PA evaluation of the policy-relevant information in
the 2019 ISA, CASAC advice and recommendations, and public comments.
---------------------------------------------------------------------------
\126\ In addition to the review's opening ``call for
information'' (79 FR 71764, December 3, 2014), the 2019 ISA
identified and evaluated studies and reports that have undergone
scientific peer review and were published or accepted for
publication between January 1, 2009, through approximately January
2018 (U.S. EPA, 2019a, p. ES-2). References cited in the 2019 ISA,
the references considered for inclusion but not cited, and
electronic links to bibliographic information and abstracts can be
found at: https://hero.epa.gov/hero/particulate-matter.
---------------------------------------------------------------------------
In presenting the rationale for the Administrator's final decision
and its foundations, Section III.A provides background on the 2020
final decision to retain the primary PM10 and a brief
summary of key aspects of the currently available health effects
information. Section III.B summarizes the CASAC advice and the
Administrator's proposed conclusions to retain the existing primary
PM10 standard, addresses public comments received on the
proposal, and presents the Administrator's conclusions on the adequacy
of the current standard, drawing on consideration of information in the
2019 ISA and the 2022 PA, advice from the CASAC, and comments from the
public. Section III.C summarizes the Administrator's decision on the
primary PM10 standard.
A. Introduction
The general approach for this reconsideration of the 2020 final
decision on the primary PM10 standard relies on the
scientific information available for this review, as well as the
Administrator's judgments regarding the available public health effects
evidence, and the appropriate degree of public health protection for
the existing standards. With the 2020 decision, the then-Administrator
retained the existing primary 24-hour PM10 standard, with
its level of 150 [micro]g/m\3\ and its one-expected-exceedance form on
average over three years, to continue to provide public health
protection against short-term exposures to PM10-2.5 (85 FR
82725, December 18, 2020).
1. Background on the Current Standard
Consistent with the 2009 ISA, the 2019 ISA concluded that the
available epidemiologic, controlled human exposure, and animal
toxicological studies, including uncertainties, provided support for
the causality determinations of ``suggestive of, but not sufficient to
infer, a causal relationship'' between short-term exposures to
PM10-2.5 and cardiovascular effects, respiratory effects,
and mortality (U.S. EPA, 2019a, section 1.4.2). The 2019 ISA also
reached the conclusion that the evidence supports a ``suggestive of,
but not sufficient to infer, a causal relationship'' between short-term
PM10-2.5 exposures and metabolic effects, an endpoint that
was not evaluated in the 2009 ISA (U.S. EPA, 2019a, section 1.4.2).
Compared to the 2009 ISA, the 2019 ISA includes expanded evidence
for the relationships between long-term exposures and cardiovascular
effects, metabolic effects, nervous system effects, cancer, and
mortality. The 2019 ISA concluded that the small number of
epidemiologic and experimental studies, including uncertainties,
contribute to the determination that, ``the evidence is suggestive of,
but not sufficient to infer, a causal relationship between long-term
PM10-2.5 exposure and cardiovascular effects, metabolic
effects, nervous system effects, cancer, and mortality and cancer (U.S.
EPA, 2019a, p. 10-87). For long-term exposures and cardiovascular
effects,
[[Page 16292]]
cardiovascular effects, and cancer, this is an upgrade from the
``inadequate to infer the presence or absence of a causal
relationship'' conclusions in the 2009 ISA (U.S. EPA, 2019a, section
1.4.2). This determination is also the first for long-term exposures
and metabolic effects, as the 2009 ISA did not include metabolic
effects as an endpoint (U.S. EPA, 2019a section 1.4.2).
In considering the available body of evidence, it was noted in the
2020 review there were considerable uncertainties and limitations
associated with the experimental evidence for PM2.5
exposures and health effects, and as such more weight was placed on the
available epidemiologic evidence. Therefore, the primary focus in the
2020 review was on multi-city and single-city epidemiologic studies
that evaluated associations between short-term PM10-2.5 and
mortality, cardiovascular effects (hospital admissions and emergency
department visits, as well as blood pressure and hypertension), and
respiratory effects. Despite differences in the approaches \127\ used
to estimate ambient PM10-2.5 concentrations, the majority of
the studies reported positive, though often not statistically
significant, associations with short-term PM10-2.5
exposures. Most PM10-2.5 effect estimates remained positive
in copollutant models that included either gaseous pollutants or other
particulate matter size fractions (e.g., PM2.5). In U.S.
study locations likely to have met the PM10 standard during
the study period, a few studies reported positive associations between
PM10-2.5 and mortality that were statistically significant
and remained so in copollutant models (U.S. EPA, 2019a). In addition to
the epidemiologic studies, there were a small number of controlled
human exposure studies evaluated in the 2019 ISA that reported
alterations in heart rate variability or increased pulmonary
inflammation following short-term exposure to PM10-2.5,
providing some support for the associations in the epidemiologic
studies. Animal toxicological studies examined the effect of short-term
PM10-2.5 exposures using non-inhalation (e.g., intratracheal
instillation) route.\128\ Therefore, these studies provided limited
evidence for the biological plausibility of PM10-2.5-induced
effects (U.S. EPA, 2019a). Although the scientific evidence available
in the 2019 ISA expanded the understanding of health effects associated
with PM10-2.5 exposures, a number of important uncertainties
remained. These uncertainties, and their implications for interpreting
the scientific evidence, include the following:
---------------------------------------------------------------------------
\127\ As discussed further below, methods employed by the
epidemiologic studies to estimate ambient PM10-2.5
concentrations include: (1) Calculating the difference between
PM10 and PM2.5 at co-located monitors, (2)
calculating the difference between county-wide averages of monitored
PM10 and PM2.5 based on monitors that are not
necessarily co-located, and (3) direct measurement of
PM10-2.5 using a dichotomous sampler (U.S. EPA, 2019a,
section 1.4.2).
\128\ Non-inhalation exposure experiments (i.e., intratracheal
[IT] instillation) are informative for size fractions (e.g.,
PM10-2.5) that cannot penetrate the airway of a study
animal and may provide information relevant to biological
plausibility and dosimetry (U.S. EPA, 2019a, section A-12).
---------------------------------------------------------------------------
The potential for confounding by copollutants, notably
PM2.5, was addressed with copollutant models in a relatively
small number of PM10-2.5 epidemiologic studies (U.S. EPA,
2019a). This was particularly important given the relatively small body
of experimental evidence (i.e., controlled human exposure and animal
toxicological studies) available to support the independent effect of
PM10-2.5 on human health. This increases the uncertainty
regarding the extent to which PM10-2.5 itself, rather than
one or more copollutants, is responsible for the mortality and
morbidity effects reported in epidemiologic studies.
There was greater spatial variability in
PM10-2.5 concentrations than PM2.5
concentrations, resulting in the potential for increased exposure error
for PM10-2.5 (U.S. EPA, 2019a). Available measurements did
not provide sufficient information to adequately characterize the
spatial distribution of PM10-2.5 concentrations (U.S. EPA,
2019a). The limitations in estimates of ambient PM10-2.5
concentrations ``would tend to increase uncertainty and make it more
difficult to detect effects of PM10-2.5 in epidemiologic
studies'' (U.S. EPA, 2019a).
Estimation of PM10-2.5 concentrations over
which reported health outcomes occur remain highly uncertain. When
compared with PM2.5, there is uncertainty spanning all
epidemiologic studies examining associations with PM10-2.5
including deficiencies in the existing monitoring networks, the lack of
a systematic evaluation of the various methods used to estimate
PM10-2.5 concentrations and the resulting uncertainty in the
spatial as well as the temporal variability in PM10-2.5
concentration (U.S. EPA, 2019a).). Given these limitations in routine
monitoring, epidemiologic studies employed a number of different
approaches for estimating PM10-2.5 concentrations, including
(1) calculating the difference between PM10 and
PM2.5 at co-located monitors, (2) calculating the difference
between county-wide averages of monitored PM10 and
PM2.5 based on monitors that are not necessarily co-located,
and (3) direct measurement of PM10-2.5 using a dichotomous
sampler (U.S. EPA, 2019a, section 1.4.2). Given the relatively small
number of PM10-2.5 monitoring sites, the relatively large
spatial variability in ambient PM10-2.5 concentrations, the
use of different approaches to estimating ambient PM10-2.5
concentrations across epidemiologic studies, and the limitations
inherent in such estimates, the distributions of PM10-2.5
concentrations over which reported health outcomes occur remain highly
uncertain (U.S. EPA, 2019a).
There was relatively little information available to characterize
potential exposure differences that may inform the apparent variability
in associations between short-term PM10-2.5 exposures and
health effects across study locations (U.S. EPA, 2019a). Specifically,
the potential spatial and temporal variability in PM10-2.5
exposures complicates the interpretation of results between study
locations as well as the relative lack of information on the chemical
and biological composition of PM10-2.5 (U.S. EPA, 2009a U.S.
EPA, 2019a).
In reaching his decision in 2020 to retain the existing 24-hour
primary PM10 standard, the then-Administrator specifically
noted that, while the health effects evidence was somewhat expanded
since the prior reviews, the overall conclusions in the 2019 ISA,
including uncertainties and limitations, were generally consistent with
what was considered in the 2012 review (85 FR 82725, December 18,
2020). In addition, the then-Administrator recognized that there were
still a number of uncertainties and limitations associated with the
available evidence.
With regard to the evidence on PM10-2.5-related health
effects, the then-Administrator noted that epidemiologic studies
continued to report positive associations with mortality and morbidity
in cities across North America, Europe, and Asia, where
PM10-2.5 sources and composition were expected to vary
widely. While significant uncertainties remained in the 2020 review,
the then-Administrator recognized that this expanded body of evidence
had broadened the range of effects that have been linked with
PM10-2.5 exposures. The studies evaluated in the 2019 ISA
expanded the scientific foundation presented in the 2009 ISA and led to
revised causality determinations (and new determinations) for long-term
PM10-2.5
[[Page 16293]]
exposures and mortality, cardiovascular effects, metabolic effects,
nervous system effects, and cancer (85 FR 82726, December 18, 2020).
Drawing from his consideration of this evidence, the then-Administrator
concluded that the scientific information available since the time of
the last review supported a decision to maintain a primary
PM10 standard to provide public health protection against
PM10-2.5 exposures, regardless of location, source of
origin, or particle composition (85 FR 82726, December 18, 2020). With
regard to uncertainties in the available evidence, the then-
Administrator first noted that a number of limitations were identified
in the 2012 review related to: (1) Estimates of ambient
PM10-2.5 concentrations used in epidemiologic studies; (2)
limited evaluation of copollutant models to address the potential for
confounding; and (3) limited experimental studies supporting biological
plausibility for PM10-2.5-related effects. Despite the
expanded body of evidence for PM10-2.5 exposures and health
effects, the then-Administrator recognized that uncertainties in the
2020 review continued to include those associated with the exposure
estimates used in epidemiologic studies, the independence of the
PM10-2.5 health effect associations, and the biologically
plausible pathways for PM10-2.5 health effects (85 FR 82726,
December 18, 2020). These uncertainties contributed to the 2019 ISA
determinations that the evidence is at most ``suggestive of, but not
sufficient to infer'' causal relationships (85 FR 82726, December 18,
2020). In considering the available evidence in his basis for the
decision, the then-Administrator emphasized evidence supporting
``causal'' and ``likely to be causal'' relationships, and therefore,
judged that the PM10-2.5-related health effects evidence
provided an uncertain scientific foundation for making standard-setting
decisions. He further judged limitations in the evidence raised
questions as to whether additional public health improvements would be
achieved by revising the existing PM10 standard (85 FR
24126, April 30, 2020). In the 2020 decision, for all of the reasons
discussed above and recognizing the CASAC conclusion that the evidence
provided support for retaining the current standard, the then-
Administrator concluded that it was appropriate to retain the existing
primary PM10 standard, without revision. His decision was
consistent with the CASAC advice related to the primary PM10
standard. Specifically, the CASAC agreed with the 2020 PA conclusions
that, while these effects are important, the ``evidence does not call
into question the adequacy of the public health protection afforded by
the current primary PM10 standard'' and ``supports
consideration of retaining the current standard in this review'' (Cox,
2019b, p. 3 of consensus letter). Thus, the then-Administrator
concluded that the primary PM10 standard (in all of its
elements (i.e., indicator, averaging time, form, and level)) was
requisite to protect public health with an adequate margin of safety
against effects that have been associated with PM10-2.5. In
light of this conclusion, the EPA retained the existing PM10
standard.
2. Overview of the Health Effects Evidence
The information summarized here is based on the scientific
assessment of the health effects evidence available in this
reconsideration; this evaluation is documented in the 2019 ISA and its
policy implications are discussed further in the 2022 PA. As noted
above, the ISA Supplement does not include an evaluation of studies for
PM10-2.5, and the 2019 ISA continues to serve as the
scientific foundation for this reconsideration.
a. Nature of Effects
For the health effect categories and exposure duration combinations
evaluated, the 2019 ISA concludes that the evidence supports causality
determinations for PM10-2.5 that are at most ``suggestive
of, but not sufficient to infer, a causal relationship''. While the
evidence supporting the causal nature of relationships between exposure
to PM10-2.5 has been strengthened for some health effect
categories since the completion of the 2009 ISA, the 2019 ISA concludes
that overall ``the uncertainties in the evidence identified in the 2009
ISA have, to date, still not been addressed'' (U.S. EPA, 2019a, section
1.4.2, p. 1-41; U.S. EPA, 2022b, section 4.3.1). Specifically,
epidemiologic studies available in the 2012 review relied on various
methods to estimate PM10-2.5 concentrations, and these
methods had not been systematically compared to evaluate spatial and
temporal correlations in PM10-2.5 concentrations. Methods
included: (1) Calculating the difference between PM10 and
PM2.5 concentrations at co-located monitors, (2) calculating
the difference between county-wide averages of monitored
PM10- and PM2.5-based on monitors that are not
necessarily co-located, and (3) direct measurement of
PM10-2.5 using a dichotomous sampler (U.S. EPA, 2019a,
section 1.4.2). As described in the 2019 ISA, there continues to be
variability across epidemiologic studies in the approaches used to
estimate PM10-2.5 concentrations. Additionally, some studies
estimate long-term PM10-2.5 exposures as the difference
between PM10 and PM2.5 concentrations based on
information from spatiotemporal or land use regression (LUR) models, in
addition to monitors. The various methods used to estimate
PM10-2.5 concentrations have not been systematically
evaluated (U.S. EPA, 2019a, section 3.3.1.1), contributing to
uncertainty regarding the spatial and temporal correlations in
PM10-2.5 concentrations across methods and in the
PM10-2.5 exposure estimates used in epidemiologic studies
(U.S. EPA, 2019a, section 2.5.1.2.3). Given the greater spatial and
temporal variability of PM10-2.5 and the lower number of
PM10-2.5 monitoring sites, compared to PM2.5,
this uncertainty is particularly important for the coarse size
fraction. Beyond the uncertainty associated with PM10-2.5
exposure estimates in epidemiologic studies, the limited information on
the potential for confounding by copollutants and the limited support
available for the biological plausibility of health effects following
PM10-2.5 exposures also continue to contribute to
uncertainty in the PM10-2.5 health evidence. Uncertainty
related to potential confounding stems from the relatively small number
of epidemiologic studies that have evaluated PM10-2.5 health
effect associations in copollutants models with both gaseous pollutants
and other PM size fractions. On the other hand, uncertainty related to
the biological plausibility of effects attributed to
PM10-2.5 exposures results from the small number of
controlled human exposure and animal toxicological studies that have
evaluated the health effects of experimental PM10-2.5
inhalation exposures. The evidence supporting the 2019 ISA's
``suggestive of, but not sufficient to infer, a causal relationship''
causality determinations for PM10-2.5, including
uncertainties in this evidence, is summarized below in sections
III.B.1.a through III.B.1.f.
i. Mortality
Due to the dearth of studies examining the association between
long-term PM10-2.5 exposure and mortality, the 2009 ISA
concluded that the evidence was ``inadequate to determine if a causal
relationship exists'' (U.S. EPA, 2009a). As reported in the 2019 ISA,
some cohort studies conducted in the U.S. and Europe report positive
associations between long-term PM10-2.5 exposure and total
(nonaccidental)
[[Page 16294]]
mortality, though results are inconsistent across studies (U.S. EPA,
2019a, Table 11-11). The examination of copollutant models in these
studies remains limited and, when included, PM10-2.5 effect
estimates are often attenuated after adjusting for PM2.5
(U.S. EPA, 2019a, Table 11-11). Across studies, PM10-2.5
exposure concentrations are estimated using a variety of approaches,
including direct measurements from dichotomous samplers, calculating
the difference between PM10 and PM2.5
concentrations measured at collocated monitors, and calculating
difference of area-wide concentrations of PM10 and
PM2.5. As discussed above, temporal and spatial correlations
between these approaches have not been evaluated, contributing to
uncertainty regarding the potential for exposure measurement error
(U.S. EPA, 2019a, section 3.3.1.1 and Table 11-11). The 2019 ISA
concludes that this uncertainty ``reduces the confidence in the
associations observed across studies'' (U.S. EPA, 2019a, p. 11-125).
The 2019 ISA additionally concludes that the evidence for long-term
PM10-2.5 exposures and cardiovascular effects, respiratory
morbidity, and metabolic disease provide limited biological
plausibility for PM10-2.5-related mortality (U.S. EPA,
2019a, sections 11.4.1 and 11.4). Taken together, the 2019 ISA
concludes that, ``this body of evidence is suggestive, but not
sufficient to infer, that a causal relationship exists between long-
term PM10-2.5 exposure and total mortality'' (U.S. EPA,
2019a, p. 11-125).
With regard to short-term PM10-2.5 exposures and
mortality, the 2009 ISA concluded that the evidence is ``suggestive of
a causal relationship between short-term exposure to
PM10-2.5 and mortality'' (U.S. EPA, 2009a). The 2019 ISA
included multicity epidemiologic studies conducted primarily in Europe
and Asia that continue to provide consistent evidence of positive
associations between short-term PM10-2.5 exposure and total
(nonaccidental) mortality (U.S. EPA, 2019a, Table 11-9). Although these
studies contribute to increasing confidence in the PM10-2.5-
mortality relationship, the use of various approaches to estimate
PM10-2.5 exposures continues to contribute uncertainty to
the associations observed. Recent studies expand the assessment of
potential copollutant confounding of the PM10-2.5-mortality
relationship and provide evidence that PM10-2.5 associations
generally remain positive in copollutant models, though associations
are attenuated in some instances (U.S. EPA, 2019a, section 11.3.4.1,
Figure 11-28, Table 11-10). The 2019 ISA concludes that, overall, the
assessment of potential copollutant confounding is limited due to the
lack of information on the correlation between PM10-2.5 and
gaseous pollutants and the small number of locations in which
copollutant analyses have been conducted. Associations with cause-
specific mortality (i.e., cardiovascular and respiratory mortality)
provide some support for associations with total (nonaccidental)
mortality, though associations with respiratory mortality are more
uncertain (i.e., wider confidence intervals) and less consistent (U.S.
EPA, 2019a, section 11.3.7). The 2019 ISA concludes that the evidence
for PM10-2.5-related cardiovascular effects provides only
limited support for the biological plausibility of a relationship
between short-term PM10-2.5 exposure and cardiovascular
mortality (U.S. EPA, 2019a, section 11.3.7). Based on the overall
evidence, the 2019 ISA concludes that, ``this body of evidence is
suggestive, but not sufficient to infer, that a causal relationship
exists between short-term PM10-2.5 exposure and total
mortality'' (U.S. EPA, 2019a, p. 11-120).
ii. Cardiovascular Effects
In the 2009 ISA, the evidence describing the relationship between
long-term exposure to PM10-2.5 and cardiovascular effects
was characterized as ``inadequate to infer the presence or absence of a
causal relationship.'' The limited number of epidemiologic studies
reported contradictory results and experimental evidence demonstrating
an effect of PM10-2.5 on the cardiovascular system was
lacking (U.S. EPA, 2019a, section 6.4).
The evidence relating long-term PM10-2.5 exposures to
cardiovascular mortality remains limited, with no consistent pattern of
associations across studies and, as discussed above, uncertainty
stemming from the use of various approaches to estimate
PM10-2.5 concentrations (U.S. EPA, 2019a, Table 6-70). The
evidence for associations with cardiovascular morbidity has grown and,
while results across studies are not entirely consistent, some
epidemiologic studies report positive associations with ischemic heart
disease (IHD) and MI (U.S. EPA, 2019a, Figure 6-34); stroke (U.S. EPA,
2019a, Figure 6-35); atherosclerosis (U.S. EPA, 2019a, section 6.4.5);
venous thromboembolism (VTE) (U.S. EPA, 2019a, section 6.4.7); and
blood pressure and hypertension (U.S. EPA, 2019a, Section 6.4.6).
PM10-2.5 cardiovascular mortality effect estimates are often
attenuated, but remain positive, in copollutants models that adjust for
PM2.5. For morbidity outcomes, associations are inconsistent
in copollutant models that adjust for PM2.5, NO2,
and chronic noise pollution (U.S. EPA, 2019a, p. 6-276). The lack of
toxicological evidence for long-term PM10-2.5 exposures
represents a data gap (U.S. EPA, 2019a, section 6.4.10), resulting in
the 2019 ISA conclusion that ``evidence from experimental animal
studies is of insufficient quantity to establish biological
plausibility'' (U.S. EPA, 2019a, p. 6-277). Based largely on the
observation of positive associations in some epidemiologic studies, the
2019 ISA concludes that ``evidence is suggestive of, but not sufficient
to infer, a causal relationship between long-term PM10-2.5
exposure and cardiovascular effects'' (U.S. EPA, 2019a, p. 6-277).
With regard to short-term PM10-2.5 exposures and
cardiovascular effects, the 2009 ISA found that the available evidence
for short-term PM10-2.5 exposure and cardiovascular effects
was ``suggestive of a causal relationship.'' This conclusion was based
on several epidemiologic studies reporting associations between short-
term PM10-2.5 exposure and cardiovascular effects, including
IHD hospitalizations, supraventricular ectopy, and changes in heart
rate variability (HRV). In addition, dust storm events resulting in
high concentrations of crustal material were linked to increases in
total cardiovascular disease emergency department visits and hospital
admissions. However, the 2009 ISA noted the potential for exposure
measurement error primarily due to the different methods used across
studies to estimate PM10-2.5 concentrations and copollutant
confounding in these epidemiologic studies. In addition, there was only
limited evidence of cardiovascular effects from a small number of
experimental studies (e.g. animal toxicological studies and controlled
human exposure studies) that examined short-term PM10-2.5
exposures (U.S. EPA, 2009a, section 6.2.12.2). In the 2019 ISA, key
uncertainties included the potential for exposure measurement error,
copollutant confounding, and limited evidence of biological
plausibility for cardiovascular effects following inhalation exposure
(U.S. EPA, 2019a, section 6.3.13).
The evidence for short-term PM10-2.5 exposure and
cardiovascular outcomes has expanded since the 2009 ISA, though
important uncertainties remain. The 2019 ISA notes that there are a
small number of epidemiologic studies reporting positive associations
between short-term exposure to PM10-2.5 and cardiovascular-
related morbidity
[[Page 16295]]
outcomes. However, the 2019 ISA notes that there is limited evidence to
support that these associations are biologically plausible, or
independent of copollutant confounding. The 2019 ISA also concludes
that it remains unclear how the approaches used to estimate
PM10-2.5 concentrations in epidemiologic studies compare
amongst one another and subsequently how exposure measurement error
varies between each method. Specifically, it is unclear how well-
correlated PM10-2.5 concentrations are both temporally and
spatially across these methods and therefore whether exposure
measurement error varies across these methods. Taken together, the 2019
ISA concludes that ``the evidence is suggestive of, but not sufficient
to infer, a causal relationship between short-term PM10-2.5
exposures and cardiovascular effects'' (U.S. EPA, 2019a, p. 6-254).
iii. Respiratory Effects
With regard to short-term PM10-2.5 exposures and
respiratory effects, the 2009 ISA (U.S. EPA, 2009a) concluded that the
relationship between short-term exposure to PM10-2.5 and
respiratory effects is ``suggestive of a causal relationship'' based on
a small number of epidemiologic studies observing associations with
some respiratory effects and limited evidence from experimental studies
to support biological plausibility. Epidemiologic findings were
consistent for respiratory infection and combined respiratory-related
diseases, but not for COPD. Studies were characterized by overall
uncertainty in the exposure assignment approach and limited information
regarding potential copollutant confounding. Controlled human exposure
studies of short-term PM10-2.5 exposures found no lung
function decrements and inconsistent evidence for pulmonary
inflammation. Animal toxicological studies were limited to those using
non-inhalation (e.g., intra-tracheal instillation) routes of
PM10-2.5 exposure.
Recent epidemiologic findings consistently link PM10-2.5
exposure to asthma exacerbation and respiratory mortality, with some
evidence that associations remain positive (though attenuated in some
studies of mortality) in copollutant models that include
PM2.5 or gaseous pollutants. Epidemiologic studies provide
limited evidence for positive associations with other respiratory
outcomes, including COPD exacerbation, respiratory infection, and
combined respiratory-related diseases (U.S. EPA, 2019a, Table 5-36). As
noted above for other endpoints, an uncertainty in these epidemiologic
studies is the lack of a systematic evaluation of the various methods
used to estimate PM10-2.5 concentrations and the resulting
uncertainty in the spatial and temporal variability in
PM10-2.5 concentrations compared to PM2.5 (U.S.
EPA, 2019a, sections 2.5.1.2.3 and 3.3.1.1). Specifically, the existing
monitoring networks do not provide a good characterization of how well
correlated concentrations are both spatially and temporally across the
PM10-2.5 estimation methods and overall spatial and temporal
patterns in PM10-2.5 concentrations. Taken together, the
2019 ISA concludes that ``the collective evidence is suggestive of, but
not sufficient to infer, a causal relationship between short-term
PM10-2.5 exposure and respiratory effects'' (U.S. EPA,
2019a, p. 5-270).
iv. Cancer
In the 2012 review, little information was available from studies
of cancer following inhalation exposures to PM10-2.5. Thus,
the 2009 ISA determined the evidence was ``inadequate to evaluate the
relationship between long-term PM10-2.5 exposures and
cancer'' (U.S. EPA, 2009a). The scientific information evaluated in the
2019 ISA of long-term PM10-2.5 exposure and cancer remains
limited, with a few recent epidemiologic studies reporting positive,
but imprecise, associations with lung cancer incidence (U.S. EPA,
2019a). Moreover, uncertainty remains in these studies with respect to
exposure measurement error due to the use of PM10-2.5
predictions that have not been validated by monitored
PM10-2.5 concentrations (U.S. EPA, 2019a, sections 3.3.2.3
and 10.3.4). Relatively few experimental studies of PM10-2.5
have been conducted, though available studies indicate that
PM10-2.5 exhibits two key characteristics of carcinogens:
genotoxicity and oxidative stress. While limited, such experimental
studies provide some evidence of biological plausibility for the
findings in a small number of epidemiologic studies (U.S. EPA, 2019a,
section 10.3.4).
Taken together, the small number of epidemiologic and experimental
studies, along with uncertainty with respect to exposure measurement
error, contribute to the determination in the 2019 ISA that, ``the
evidence is suggestive of, but not sufficient to infer, a causal
relationship between long-term PM10-2.5 exposure and
cancer'' (U.S. EPA, 2019a, p. 10-87).
v. Metabolic Effects
The 2009 ISA did not make a causality determination for
PM10-2.5-related metabolic effects. One epidemiologic study
in the 2019 ISA reports an association between long-term
PM10-2.5 exposure and incident diabetes, while additional
cross-sectional studies report associations with effects on glucose or
insulin homeostasis (U.S. EPA, 2019a, section 7.4). As discussed above
for other outcomes, uncertainties with the epidemiologic evidence
include the potential for copollutant confounding and exposure
measurement error due to the different methods used across studies to
estimate PM10-2.5 concentrations (U.S. EPA, 2019a, Tables 7-
14 and 7-15). The evidence base to support the biological plausibility
of metabolic effects following PM10-2.5 exposures is
limited, but a cross-sectional study that investigated biomarkers of
insulin resistance and systemic and peripheral inflammation may support
a pathway leading to type 2 diabetes (U.S. EPA, 2019a, sections 7.4.1
and 7.4.3). Based on the expanded, though still limited evidence base,
the 2019 ISA concludes that, ``[o]verall, the evidence is suggestive
of, but not sufficient to infer, a causal relationship between [long]-
term PM10-2.5 exposure and metabolic effects'' (U.S. EPA,
2019a, p. 7-56).
vi. Nervous System Effects
The 2009 ISA did not make a causality determination for
PM10-2.5-related nervous system effects. In the 2019 ISA,
available epidemiologic studies report associations between
PM10-2.5 and impaired cognition and anxiety in adults in
longitudinal analyses (U.S. EPA, 2019a, Table 8-25, section 8.4.5).
Associations of long-term exposure with neurodevelopmental effects are
not consistently reported in children (U.S. EPA, 2019a, sections 8.4.4
and 8.4.5). Uncertainties in these studies include the potential for
copollutant confounding, as no studies examined copollutants models
(U.S. EPA, 2019a, section 8.4.5), and for exposure measurement error,
given the use of various methods to estimate PM10-2.5
concentrations (U.S. EPA, 2019a, Table 8-25). In addition, there is
limited animal toxicological evidence supporting the biological
plausibility of nervous system effects (U.S. EPA, 2019a, sections 8.4.1
and 8.4.5). Overall, the 2019 ISA concludes that, ``the evidence is
suggestive of, but not sufficient to infer, a causal relationship''
between long-term PM10-2.5 exposure and nervous system
effects (U.S. EPA, 2019a, p. 8-75).
[[Page 16296]]
B. Conclusions on the Primary PM10 Standard
In drawing conclusions on the adequacy of the current primary
PM10 standard, in view of the advances in scientific
knowledge and additional information now available, the Administrator
has considered the evidence base, information, and policy judgments
that were the foundation of the 2020 review and reflects upon the body
of information and evidence available in this reconsideration. In so
doing, the Administrator has taken into account both evidence-based and
quantitative information-based considerations, as well as advice from
the CASAC and public comments. Evidence-based considerations draw upon
the EPA's integrated synthesis of the scientific evidence from animal
toxicologic, controlled human exposure, and epidemiologic studies
evaluating health effects related to exposures to PM10-2.5
as presented in the 2019 ISA and discussed in section III.A.2. In
addition to the evidence, the Administrator has weighed a range of
policy-relevant considerations as discussed in the 2022 PA and
summarized in sections III.B and III.C of the proposal and summarized
in section III.B.2 below. These considerations, along with the advice
from the CASAC (section III.B.1) and public comments (section III.B.3),
are discussed below. A more detailed summary of all significant
comments, along with the EPA's responses in the Response to Comments
document, can be found in the docket for this rulemaking (Docket No.
EPA-HQ-OAR-2015-00072). This document is available for review in the
docket for this rulemaking and through EPA's NAAQS website (link). The
Administrator's conclusions in this reconsideration regarding the
adequacy of the current primary PM10 standard and whether
any revisions are appropriate are described in section III.B.4.
1. CASAC Advice
As described in section I.X, the EPA decided to prepare a revised
PA for the reconsideration of the 2020 final decision. The CASAC's
advice on the 2019 draft PA and the 2021 draft PA was documented in
letters to the prior and current Administrators (Cox, 2019b; Sheppard,
2022a) and is summarized below. In reviewing both the 2019 draft PA and
the 2021 draft PA, the CASAC agreed with the EPA's preliminary
conclusion that the available scientific evidence, including its
uncertainties and limitations, does not call into question the adequacy
of the current primary PM10 standard and that the standard
should be retained, without revision.
In its review of the 2019 draft PA, the CASAC concurred with the
overall preliminary conclusion that it is appropriate to consider
retaining the current primary PM10 standard, without
revision. In their agreement with the conclusions in the 2019 draft PA,
the CASAC stated that ``that key uncertainties identified in the last
review remain'' (Cox, 2019b) and that ``none of the identified health
outcomes linked to PM10-2.5'' were judged to be causal or
likely to be causal (Cox, 2019b, p. 12 of consensus responses).
Moreover, to reduce these uncertainties in future reviews, the CASAC
recommended improvements to PM10-2.5 exposure assessment,
including a more extensive network for direct monitoring of the
PM10-2.5 fraction (Cox, 2019b, p. 13 of consensus
responses). The CASAC also recommended additional controlled human
exposure and animal toxicological studies of the PM10-2.5
fraction to improve the understanding of biological mechanisms and
pathways (Cox, 2019b, p. 13 of consensus responses). Overall, the CASAC
agreed with the EPA's preliminary conclusion in the 2019 draft PA that
``. . . the available evidence does not call into question the adequacy
of the public health protection afforded by the current primary
PM10 standard and that evidence supports consideration of
retaining the current standard in this review'' (Cox, 2019b, p. 3 of
letter).
In its review of the 2021 draft PA, the CASAC provided advice on
the adequacy of the current primary PM10 standard in the
context of its review of the revised PA for this reconsideration
(Sheppard, 2022a) \129\.) \130\. In this context, the CASAC supported
the preliminary conclusion in the 2021 draft PA that the evidence
reviewed in the 2019 ISA does not call into question the public health
protection provided by the current primary PM10 standard
against PM10-2.5 exposures and concurs with the 2021 draft
PA's overall preliminary conclusion that it is appropriate to consider
retaining the current primary PM10 standard (Sheppard,
2022a, p. 4 of consensus letter). Additionally, the CASAC concurred
that ``. . . at this time, PM10 is an appropriate choice as
the indicator for PM10-2.5'' and ``that it is important to
retain the level of protection afforded by the current PM10
standard'' (Sheppard, 2022a, p. 4 of consensus letter). The CASAC also
recognized uncertainties associated with the scientific evidence,
including ``compared to PM2.5 studies, the more limited
number of epidemiology studies with positive statistically significant
findings, and the difficulty in extracting the sole contribution of
coarse PM to observed adverse health effects'' (Sheppard, 2022a, p. 19
of consensus responses).
---------------------------------------------------------------------------
\129\ As described in section I.C.5.b above, the scope of the
ISA Supplement did not include consideration of studies of health
effects associated with exposure to PM10-2.5. Therefore,
the information and conclusions presented in the 2022 PA are very
similar to those in the 2020 PA.
\130\ As described in section I.C.5.b above, the scope of the
ISA Supplement did not include consideration of studies of health
effects associated with exposure to PM10-2.5. Therefore,
the information and conclusions presented in the 2022 PA are very
similar to those in the 2020 PA.
---------------------------------------------------------------------------
The CASAC recommended several areas for additional research to
reduce uncertainties in the PM10-2.5 exposure estimates used
in the epidemiologic studies, to evaluate the independence of
PM10-2.5 health effect associations, to evaluate the
biological plausibility of PM10-2.5-related effects, and to
increase the number of studies examining PM10-2.5-related
health effects in at-risk populations (Sheppard, 2022a, p. 20 of
consensus responses). Furthermore, the CASAC ``recognizes a need for,
and supports investment in research and deployment of measurement
systems to better characterize PM10-2.5'' and to ``provide
information that can improve public health'' (Sheppard, 2022a, p. 20 of
consensus responses).
2. Basis for the Proposed Decision
At the time of the proposal, the Administrator carefully considered
the assessment of the current evidence and conclusions reached in the
2019 ISA, considerations and staff conclusions and associated
rationales presented in the 2020 PA and 2022 PA, and advice and
recommendations of the CASAC (88 FR 5634, January 27, 2023). Consistent
with previous reviews, the Administrator first considered the available
scientific evidence for PM10-2.5-related exposures and
health effects, as evaluated in the 2019 ISA. As an initial matter, the
Administrator recognized that the scientific evidence for
PM10-2.5-related effects available in this reconsideration
is the same body of evidence that was available at the time of the 2020
review, as evaluated in the 2019 ISA and summarized in section III.A.2
above. The 2019 ISA concludes that the evidence supports ``suggestive
of, but not sufficient to infer'' causal relationships between short-
and long-term exposures to PM10-2.5 and cardiovascular
effects, cancer, and mortality and long-term PM10-2.5
exposures and metabolic effects and nervous system effects (U.S. EPA,
2019a). The Administrator noted that
[[Page 16297]]
the evidence for several PM10-2.5-related health effects has
expanded since the completion of the 2009 ISA, but important
uncertainties remain. The uncertainties in the epidemiologic studies
contribute to the determinations in the 2019 ISA that the evidence for
short and long-term PM10-2.5 exposures and mortality,
cardiovascular effects, metabolic effects, nervous system effects, and
cancer is ``suggestive of, but not sufficient to infer'' causal
relationships (U.S. EPA, 2019a; U.S. EPA, 2022b, section 4.3.1).
Drawing from the evidence evaluated in the 2019 ISA and consideration
of the scientific evidence in the 2022 PA, the Administrator noted
that, consistent with previous reviews, the 2019 ISA and the 2022 PA
highlight a number of uncertainties associated with the evidence,
including: (1) PM10-2.5 exposure estimates used in
epidemiologic studies, (2) independence of PM10-2.5 health
effect associations, and (3) biological plausibility of the
PM10-2.5-related effects. These uncertainties contribute to
the determinations in the 2019 ISA that the evidence for short-term
PM10-2.5 exposures and key health effects is ``suggestive
of, but not sufficient to infer'' causal relationships. In considering
the available scientific evidence, consistent with approaches employed
in past NAAQS reviews, the Administrator placed the most weight on
evidence supporting ``causal'' and ``likely to be causal''
relationships. In so doing, he noted that the available evidence for
short- and long-term PM10-2.5 exposures and health effects
does not support causality determinations of a ``causal relationship''
or ``likely to be causal relationship.'' Furthermore, the Administrator
recognized that, because of the uncertainties and limitations in the
evidence base, the 2022 PA does not include a quantitative assessment
of PM10-2.5 exposures and risk that might further inform
decisions regarding the adequacy of the current 24-hour primary
PM10 standard. Therefore, in light of the 2019 ISA
conclusions that the evidence supports ``suggestive of, but not
sufficient to infer'' causal relationships. The Administrator judged
that there are substantial uncertainties that raise questions regarding
the degree to which additional public health improvements would be
achieved by revising the existing PM10 standard. In
considering the available evidence for long-term PM10-2.5
exposures, the Administrator noted that there is limited evidence that
would support consideration of an annual standard to provide protection
against such effects, in conjunction with the current primary 24-hour
PM10 standard. He preliminarily concluded that the current
primary 24-hour PM2.5 standard that reduces 24-hour
exposures also likely reduces long-term average exposures, and
therefore provides some margin of safety against the health effects
associated with long-term PM10-2.5 exposures.
In reaching his proposed decision on the adequacy of the current
primary 24-hour PM10 standard, the Administrator also
considered advice from the CASAC. As noted above in section III.B.1,
the CASAC recognized uncertainties associated with the scientific
evidence and agreed with the 2019 draft PA and 2021 draft PA
conclusions that the scientific evidence does not call into question
the adequacy of the primary PM10 standard and supports
consideration of retaining the current standard.
When considering the above information together, the Administrator
proposed to conclude that the available scientific evidence continues
to support a PM10 standard to provide some measure of
protection against PM10-2.5 exposures. Additionally, he
recognized that there are important uncertainties and limitations
associated with the available evidence for PM10-2.5-related
health effects, for both short and long-term exposure, as evaluated in
the 2019 ISA. Consistent with the decisions in the previous reviews,
the Administrator proposed to conclude that these limitations lead to
considerable uncertainty regarding the potential public health
implications of revising the level of the current primary 24-hour
PM10 standard. Thus, based on his consideration of the
evidence and associated uncertainties and limitations for
PM10-2.5-related health effects and his consideration of
CASAC advice on the primary PM10 standard, the Administrator
proposed to retain the current primary PM10 standard,
without revision.
3. Comments on the Proposed Decision
Of the public comments received on the proposal, very few
commenters provided comments on the primary PM10 standard.
Of those commenters who did provide comments on the primary
PM10 standard, the majority agree with the EPA's proposed
decision to retain the primary PM10 standard. In so doing,
these commenters agree with the EPA's rationale regarding the available
scientific information, including uncertainties and limitations, for
informing decisions on the standard. These commenters state that no new
scientific evidence or quantitative information has emerged since the
2020 decision to retain the current standard. Furthermore, these
commenters note that the EPA did not evaluate any new scientific
evidence related to PM10-2.5 exposures and health effects as
a part of the 2022 ISA Supplement developed for this reconsideration,
nor did the revised 2022 PA consider any new or different information
from the 2020 PA, and therefore, the EPA reached the same conclusion as
is the 2020 PA that the current standard is adequate and should be
retained. This group includes industries and industry groups, as well
as some State and local governments. All of these commenters generally
note their agreements with the rationale provided in the proposal and
the CASAC concurrence with the 2021 draft PA conclusion that the
available information does not call into question the adequacy of the
current standard, and therefore, does not support revision and that the
current standard should be retained.
Some commenters, including those from environmental and public
health organizations and groups, some states, and individuals,
disagreed with the Administrator's proposed decision to retain the
current primary PM10 standard. These commenters recommend
that the EPA revise the primary PM10 standard to a lower
level to provide increased public health protection, citing to the
available scientific evidence, as well as the proposed revision to the
primary PM2.5 standard.
Commenters who disagreed with the proposal to retain the current
standard state that revision to the primary PM10 standard is
necessary to protect public health with an adequate margin of safety.
In their recommendations for revising the standard, some commenters
contend that the current standard, with its indicator of
PM10 to target exposures to PM10-2.5, has become
less protective as ambient concentrations of PM2.5 have been
reduced with revisions to that standard. These commenters assert that
the current primary PM10 standard allows increased exposure
to PM10-2.5 in ambient air because retaining the primary
PM10 would allow proportionately more PM10-2.5
mass as the PM2.5 standard has been revised downward.
Moreover, in support of their recommendations, the commenters note that
the available evidence of PM10-2.5-related health effects
has been expanded and strengthened since the time of the last review.
Taken together, the commenters contend that the primary PM10
standard should be revised and failure to do so would be arbitrary and
capricious. Some of these
[[Page 16298]]
commenters assert that the level of the primary PM10
standard should be revised to 140 or 145 [micro]g/m\3\, concurrent with
a strengthened primary 24-hour PM2.5 standard, while other
commenters recommend revising the level of the standard to within the
range of 65-75 [micro]g/m\3\, to provide increased public health
protection.
We disagree with the commenters that the primary PM10
standard should be revised because of reductions in ambient
concentrations of PM2.5. As an initial matter, we note that
overall, ambient concentrations of both PM10 and
PM2.5 have declined significantly over time. Ambient
concentrations of PM10 have declined by 46% across the U.S.
from 2000 to 2019,\131\ while PM2.5 concentrations in
ambient air have declined by 43% during this same time period.\132\ As
noted in the 2022 PA (p. 2-41), the majority of PM10-2.5
sites have generally remained steady and do not exhibit a trend of
increasing or decreasing concentrations during this time period,
reflecting the relatively consistent level of dust emission across the
U.S. from 2000 to 2019 (U.S. EPA, 2022b).
---------------------------------------------------------------------------
\131\ PM10 concentrations presented as the annual
second maximum 24-hour concentration (in [micro]g/m\3\) at 262 sites
in the U.S. For more information, see: https://www.epa.gov/air-
trends/particulate-matter-pm10-trends
\132\ PM2.5 concentrations presented as the
seasonally-weighted annual average concentration (in [micro]g/m\3\)
at 406 sites in the U.S. For more information, see: https://www.epa.gov/air-trends/particulate-matter-pm25-trends
---------------------------------------------------------------------------
The 2019 ISA provides a comparison of the relative contribution of
PM2.5 and PM10-2.5 to PM10
concentrations by region and season using the more comprehensive
monitoring data from the NCore network available in this
reconsideration (U.S. EPA, 2019, section 2.5.1.1.4). The data indicate
that, for urban areas, there are roughly equivalent amounts of
PM2.5 and PM10-2.5 contributing to
PM10 in ambient air, while rural locations have a slightly
higher contribution of PM10-2.5 contributing to
PM10 concentrations than PM2.5 (U.S. EPA, 2019,
section 2.5.1.1.4, Table 2-7). There is generally a greater
contribution from the PM2.5 fraction in the East and a
greater contribution from the PM10-2.5 fraction in the West
and Midwest.
The EPA recognizes that when the primary annual PM2.5
standard was revised from 15.0 [micro]g/m\3\ to 12.0 [micro]g/m\3\
while leaving the 24-hour PM2.5 standards unchanged at 35
[micro]g/m\3\ and the 24-hour PM10 standard unchanged at 150
[micro]g/m\3\, the PM10-2.5 fraction of PM10
could increase in some areas as the PM2.5 fraction decreases
(78 FR 3085, March 03, 2013). As described in the 2019 ISA,
PM10 has become considerably coarser across the U.S.
compared to similar observations in the 2009 ISA such that, in urban
areas, the mass of the coarse fraction of PM is similar to or greater
than the mass of the fine fraction of PM (U.S. EPA, 2019, section
2.5.1.1.4; U.S. EPA, 2009c). However, in considering recent air quality
data, the EPA notes that in most areas of the country PM2.5
and PM10 concentrations have declined and are well below
their respective 24-hour standards. While the contribution of fine and
coarse PM to PM10 mass concentrations may vary spatially and
temporally, based on the trends in recent air quality data, the
Administrator concludes that the current primary 24-hour
PM10 standard is maintaining air quality at level that
provides requisite protection against PM10-2.5. That is,
recent air quality data does not suggest that PM10-2.5
concentrations have been increasing as PM2.5 concentrations
have been decreasing. In considering the available PM10-2.5
health effects evidence in this reconsideration, there continue to be
significant uncertainties and limitations, specifically with respect to
the exposure assessment methods used to estimate PM10-2.5
concentrations, that make it difficult to fully assess the public
health implications of revising the primary PM10 standard
even considering the possibility for additional variability in the
relative ratio of PM2.5 to PM10-2.5 in current
PM10 air quality across the U.S. As described in detail
above in section III.A.2 and in the proposal (85 FR 5558, January 27,
2023), the uncertainties and limitations in the health effects evidence
for PM10-2.5 contributed to the determinations in the 2019
ISA that the evidence for key PM10-2.5 health effects is
``suggestive of, but not sufficient to infer, a causal relationship''
or ``inadequate to infer the presence, or absence of a causal
relationship'' (U.S. EPA, 2019a). While the evidence base for
PM10-2.5-related health effects has somewhat expanded since
the 2009 ISA, the Administrator concludes that the evidence remains too
limited to inform judgments regarding whether a more protective primary
PM10 standard is warranted at this time.
Beyond the uncertainties and limitations associated with the
available scientific evidence, the EPA also notes that, while the NCore
monitoring network has been expanded since the time of the last review,
epidemiologic studies available in this review do not use
PM10-2.5 NCore data in evaluating associations between
PM10-2.5 in ambient air and long- or short-term exposures.
In the absence of such evidence, the public health implications of
changes in ambient PM10-2.5 concentrations as
PM2.5 concentrations decrease remain unclear. Therefore, the
EPA continues to recognize this as an area for future research, to
address the existing uncertainties (U.S. EPA, 2022b, section 4.6), and
inform future reviews of the PM NAAQS. Taken together, as at the time
of proposal, the Administrator concludes that these and other
limitations in the PM10-2.5 evidence raised questions as to
whether additional public health improvements would be achieved by
revising the existing PM10 standard, particularly when
considering such judgments along with his decision to retain the
current primary 24-hour PM2.5 standard. Therefore, the EPA
does not agree with the commenters that the currently available air
quality information or scientific evidence support revisions to the
primary PM10 standard in this reconsideration.
Consistent with their comments on the 2020 proposal, some
commenters disagreed with the Administrator's proposed conclusion to
retain the current primary PM10 standard, primarily focusing
their comments on the need for revisions to the form of the standard or
the level of the standard. With regard to comments on the form of the
standard, some commenters assert that the EPA should revise the
standard by adopting a separate form (or a ``compliance threshold'' in
their words)--the 99th percentile, averaged over three years--for the
primary PM10 standard for continuous monitors, which provide
data every day, while maintaining the current form of the standard (one
exceedance, averaged over three years) for 1-in-6 samplers, given the
increased use of continuous monitoring and to ease the burden of
demonstrating exceptional events. These commenters, in support of their
comment, contend that the 99th percentile would effectively change the
form from the 2nd highest to the 4th highest and would allow no more
than three exceedances per year, averaged over three years. These
commenters additionally highlight the EPA's decision in the 1997 review
to adopt a 99th percentile form, averaged over three years, citing to
advantages of a percentile-based form in the Administrator's rationale
in that review. The comments further assert that a 99th percentile form
for the primary PM10 standard is still more conservative
than the form for other short-term NAAQS (e.g., PM2.5 and
NO2).
[[Page 16299]]
First, the EPA has long recognized that the form is an integral
part of the NAAQS and must be selected together with the other elements
(i.e., indicator, averaging time, level) of the NAAQS to ensure the
appropriate stringency and requisite degree of public health
protection. Thus, if the EPA were to change the form according to the
monitoring method it would be establishing two different NAAQS, varying
based on the monitoring method. The EPA has not done this to date, did
not propose such an approach, and declines to adopt it for the final
rule, as we believe such a decision in this final rule is beyond the
scope of the proposal, and that each PM standard should have a single
form, indicator, level and averaging time, chosen by the Administrator
as necessary and appropriate. While certain continuous monitors may be
established and approved as a Federal Equivalent Method (FEM) for
PM10, as an alternative to a Federal Reference Method (FRM),
the use of an FEM is intended as an alternative means of determining
compliance with the NAAQS, not as authorizing a different NAAQS.
Even if the commenters had asked that the change in form be made
without regard to monitoring method, the EPA does not believe such a
change would be warranted. The change in form for continuous monitors
suggested by the commenters, without also lowering the level of such a
standard, would allow more exceedances and thereby reduce the public
health protection provided against exposures to PM10-2.5 in
ambient air, resulting in a less stringent primary PM10
standard than the current standard. These commenters have not provided
new evidence or analyses to support their conclusion that an
appropriate degree of public health protection could be achieved by
allowing the use of an alternative form (i.e., 99th percentile), while
retaining the other elements of the standard.
With regard to the commenters' assertion that an alternate form of
the standard would ease the burden of demonstrating exceptional events,
the EPA recognizes, consistent with the CAA, that it may be appropriate
to exclude monitoring data influenced by ``exceptional'' events when
making certain regulatory determinations. However, the EPA notes that
the cost of implementation of the standards may not be considered by
the EPA in reviewing the standards. The EPA continues to update and
develop documentation and tools to facilitate the implementation of the
2016 Exceptional Events Rule, including new PM2.5
implementation focused products under development that are intended to
assist air agencies with the development of demonstrations for specific
types of exceptional events. With regard to the commenters' specific
concerns for wildfires or high winds, the EPA released updated guidance
documents on the preparation of exceptional event demonstrations
related to wildfires in September 2016, high wind dust events in April
2019, and prescribed fires in August 2019. These guidance documents
outline the regulatory requirements and provide examples for air
agencies preparing demonstrations for wildfires, high wind dust, and
prescribed fire events. For all of the reasons discussed above, the EPA
does not agree with the commenters that the form of the primary
PM10 standard should be revised to a 99th percentile for
continuous monitors.
4. Administrator's Conclusions
This section summarizes the Administrator's considerations and
conclusions related to the current primary PM10 standard. In
establishing primary standards under the Act that are ``requisite'' to
protect the public health with an adequate margin of safety, the
Administrator is seeking to establish standards that are neither more
nor less stringent than necessary for this purpose. In so doing, the
Administrator notes that his final decision in this reconsideration is
a public health policy judgment that draws upon scientific information,
as well as judgments about how to consider the range and magnitude of
uncertainties that are inherent in the information. Accordingly, he
recognizes that his decision requires judgments based on the
interpretation of the evidence that neither overstates nor understates
the strength or limitations of the evidence nor the appropriate
inferences to be drawn. He recognizes, as described in section I.A
above, that the Act does not require that primary standards be set at a
zero-risk level; rather, the NAAQS must be sufficient but not more
stringent than necessary to protect public health, including the health
of sensitive groups with an adequate margin of safety.
Given these requirements, and consistent with the primary
PM2.5 standards discussed above (section II.C.3), the
Administrator's final decision in this reconsideration of the current
primary PM10 standard will be a public health policy
judgment that draws upon the scientific information examining the
health effects of PM10-2.5 exposures, including how to
consider the range and magnitude of uncertainties inherent in that
information. The Administrator's final decision is based on an
interpretation of the scientific evidence that neither overstates nor
understates its strengths and limitations, nor the appropriate
inferences to be drawn.
Having carefully considered advice from the CASAC and public
comments, as discussed above, the Administrator notes that the
fundamental scientific conclusions on health effects of
PM10-2.5 in ambient air that were reached in the 2019 ISA
and summarized in the 2020 PA and 2022 PA remain valid. Additionally,
the Administrator believes the judgments he proposed (85 FR 5558,
January 27, 2023) with regard to the evidence remain appropriate.
Further, in considering the adequacy of the current primary
PM10 standard in this reconsideration, the Administrator has
carefully considered the policy-relevant evidence and conclusions
contained in the 2019 ISA; the rationale and conclusions presented in
the 2020 PA and 2022 PA; the advice and recommendations from the CASAC
in their reviews of the 2019 draft PA and 2021 draft PA; and public
comments, as addressed in section III.B.3 above and in the RTC
document. In the discussion below, the Administrator gives weight to
the conclusions in the 2020 PA and 2022 PA, with which the CASAC has
concurred, as summarized in section III.C of the proposal and takes
note of the key aspects of the rationale for those conclusions that
contribute to his decision in this review. In considering this
information, the Administrator concludes that the preliminary
conclusions and policy judgments supporting his proposed decision
remain valid, and that the current primary PM10 standard
provides requisite protection of public health with an adequate margin
of safety and should be retained. In considering the 2020 PA and 2022
PA evaluations and conclusions, the Administrator notes that, while the
health effects evidence is somewhat expanded since the 2009 ISA as
described in section III.A.2 above, the overall conclusions are
generally consistent with those reached in the 2009 ISA (U.S. EPA,
2020b, section 4.4). In so doing, he additionally notes that the CASAC
supported the preliminary conclusion in the 2019 draft PA and 2021
draft PA that the evidence reviewed in the 2019 ISA does not call into
question the public health protection provided by the current primary
PM10 standard against PM10-2.5 exposures and
concurs that it is appropriate to consider retaining the current
primary PM10 standard (Cox,
[[Page 16300]]
2019b, p. 13 of consensus responses; Sheppard, 2022a, p. 4 of consensus
letter).
As noted below, the scientific evidence for PM10-2.5-
related health effects has expanded somewhat since the 2012 review, in
particular for long-term exposures. The Administrator recognizes,
however, that there are a number of uncertainties and limitations
associated with the available information, as described in the proposal
(85 FR 5558, January 27, 2023) and below. With regard to the current
evidence on PM10-2.5-related health effects, the
Administrator takes note of recent epidemiologic studies that continue
to report positive associations with mortality and morbidity in cities
across North America, Europe, and Asia, where PM10-2.5
sources and composition are expected to vary widely. While significant
uncertainties remain, as described below, the Administrator recognizes
that this expanded body of evidence has broadened the range of effects
that have been linked with PM10-2.5 exposures. These studies
provide an important part of the scientific foundation supporting the
2019 ISA's revised causality determinations (and new determinations)
for long-term PM10-2.5 exposures and mortality,
cardiovascular effects, metabolic effects, nervous system effects, and
cancer (U.S. EPA, 2019a; U.S. EPA, 2022b, section 4.2). Drawing from
his consideration of this evidence, the Administrator concludes that
the available scientific information supports a decision to maintain a
primary PM10 standard to provide public health protection
against PM10-2.5 exposures, regardless of location, source
of origin, or particle composition. With regard to uncertainties in the
evidence, the Administrator first notes that a number of limitations
were identified in the 2012 review related to: (1) Estimates of ambient
PM10-2.5 concentrations used in epidemiologic studies; (2)
limited evaluation of copollutant models to address the potential for
confounding; and (3) limited experimental studies supporting biological
plausibility for PM10-2.5-related effects. Despite the
expanded body of evidence for PM10-2.5 exposures and health
effects assessed in the 2019 ISA, the Administrator recognizes that
uncertainties remain, similar to those in the 2012 review. As
summarized in section III.A.2 above and in responding to public
comments, uncertainties in the available scientific evidence continue
to include those associated with the exposure estimates used in
epidemiologic studies, the independence of the PM10-2.5
health effect associations, and the biologically plausible pathways for
PM10-2.5 health effects (U.S. EPA, 2022b, section 4.3).
These uncertainties contribute to the 2019 ISA determinations that the
evidence is ``suggestive of, but not sufficient to infer'' causal
relationships (U.S. EPA, 2019a). The Administrator recognizes that the
NAAQS must allow for a margin of safety but also places emphasis on
evidence supporting ``causal'' or ``likely to be causal'' relationships
(as described in sections II.A.2 and III.A.2 above). Finding that there
is too much uncertainty that a more stringent standard would improve
public health, the Administrator judges that the available evidence
provides support for his conclusion that the current standard provides
the requisite level of protection from the effects of
PM10-2.5. In making this judgment, the Administrator
considers whether this level of protection is more than what is
requisite and whether a less stringent standard would be appropriate to
consider. He notes that there continues to be uncertainty associated
with the evidence, as reflected by the ``suggestive of, but not
sufficient to infer'' causal determinations. The Administrator
recognizes that the CAA requirement that primary standards provide an
adequate margin of safety, as summarized in section I.A above, is
intended to address uncertainties associated with inconclusive
scientific evidence and technical information, as well as to provide a
reasonable degree of protection against hazards that research has not
yet identified. In light of these considerations and the current body
of evidence, including uncertainties and limitations, the Administrator
concludes that a less stringent standard would not provide the
requisite protection of public health, including an adequate margin of
safety. The Administrator also considers whether the level of
protection associated with the current standard is less than what is
requisite and whether a more stringent standard would be appropriate to
consider. In so doing, the Administrator considers, as discussed above,
the level of protection offered from exposures for which public health
implications are less clear. In so doing, he again notes the
significant uncertainties and limitations that persist in the
scientific evidence. In particular, he notes limitations in the
approaches used to estimate ambient PM10-2.5 concentrations
in epidemiologic studies, limited examination of the potential for
confounding by co-occurring pollutants, and limited support for the
biological plausibility of the serious effects reported in many
epidemiologic studies that are reflected by the ``suggestive of, but
not sufficient to infer'' causal determinations. Thus, in light of the
currently available information, including the uncertainties and
limitations of the evidence base available to inform his judgments
regarding protection against PM10-2.5-related effects, the
Administrator does not find it appropriate to increase the stringency
of the standard in order to provide the requisite public health
protection. Rather, he judges it appropriate to maintain the level of
protection provided by the current primary PM10 standard for
PM10-2.5 exposures and he does not judge that the available
information and the associated uncertainties indicate the need for a
greater level of public health protection.
In reaching his conclusions on the primary PM10
standard, the Administrator also considers advice from the CASAC. In
their comments, the CASAC noted that uncertainties that were identified
in the 2012 review persist in the evidence for PM10-2.5-
related health effects (Cox, 2019b, p. 13 of consensus responses;
Sheppard, 2022a, p. 4 of consensus letter) In considering these
comments, the Administrator takes note of the CASAC consideration of
the evidence, and associated uncertainties, and its conclusion that the
evidence reviewed in the 2019 ISA does not call into question the
adequacy of the public health protection afforded by the current
primary PM10 standard (Cox, 2019b, p. 3 of letter; Sheppard,
2022a, p. 4 of consensus letter). The Administrator further notes the
unanimous conclusions of the CASAC that evidence supports consideration
of retaining the current primary PM10 standard (Cox, 2019b,
p. 3 of consensus letter; Sheppard, 2022a, p. 4 of consensus letter).
In addition to the CASAC's advice, the Administrator also considers
public comments, the majority of which supported retaining the primary
PM10 standard, citing to and agreeing with the
Administrator's rationale for his proposed decision. The Administrator
also recognizes that a few public commenters supported revising the
primary PM10 standard in order to provide increased
protection against PM10-2.5-related health effects.
The Administrator also notes that the scientific record for his
decision on the primary PM10 standard is the same as the
record before the then-Administrator in 2020, as the scope of the ISA
Supplement focused on health effect categories where the 2019 ISA
concluded a causal relationship (i.e.,
[[Page 16301]]
short- and long-term PM2.5 exposure and cardiovascular
effects and mortality). Therefore, because no health outcome categories
for short- or long-term PM10-2.5 exposure in the 2019 ISA
were greater than ``suggestive of, but not sufficient to infer, a
causal relationship'', the ISA Supplement did not evaluate studies
published after the literature cutoff date of the 2019 ISA related to
PM10-2.5 exposures and health effects. The Administrator
further notes his decision is consistent with the decision of the prior
Administrator in 2020 to retain the primary PM10 standard.
With regard to the indicator for the primary PM10
standard, the Administrator recognizes that the 2022 PA notes that the
evidence continues to support retaining the PM10 indicator
to provide public health protection against PM10-2.5-related
effects. He notes that, consistent with the approaches in previous
reviews, a standard with a PM10 mass-based indicator, in
conjunction with a PM2.5 mass-based standard, will result in
controlling allowable concentrations of PM10-2.5. The
Administrator also takes note of the 2019 ISA comparison that showed
that the relative contribution of PM2.5 and
PM10-2.5 to PM10 concentrations can vary across
the U.S. by region and season, with urban locations having a somewhat
higher contribution of PM2.5 contributing to PM10
concentrations than PM10-2.5 (U.S. EPA, 2019a, section
2.5.1.1.4, Table 2-7). In these urban locations, where PM2.5
concentrations are somewhat higher than in rural locations, the
toxicity of the PM10 may be higher due to contaminating
PM2.5. Further, although uncertainties with the evidence
persist, the strongest health effects evidence associated with
PM10-2.5 comes from epidemiologic studies conducted in urban
areas. He also notes that the CASAC agreed with the EPA's conclusions
that a PM10 indicator remained appropriate (Cox, 2019b, p.
13 of consensus responses; Sheppard, 2022a, p. 4 of letter). In light
of this information, the Administrator concludes that the
PM10 indicator remains appropriate and provides protection
from exposure to all coarse PM, regardless of location, source of
origin, or particle composition.
Similarly, with regard to averaging time, form, and level of the
standard, the Administrator takes note of uncertainties in the
available evidence and information and continues to find that the
current standard, as defined by in all of its elements, is requisite.
As an initial matter, the Administrator notes that the current primary
PM10 standard, with its level of 150 [micro]g/m\3\, 24-hour
averaging time, not to be exceeded more than once per year on average
over three years, is intended to protect against short-term peak
PM10-2.5 exposures. In so doing, while the Administrator
notes that changes in PM2.5 concentrations in ambient air
can influence the contribution of the fine and coarse fractions to
PM10 mass, such that reductions in PM2.5
concentrations can lead to more allowable PM10-2.5 under the
current primary PM10 standard, he recognizes that there is
no new information available in this reconsideration to suggest that
the public health protection provided by the current standard is not
requisite or that a more stringent standard is warranted at this time.
The Administrator concludes that, particularly in light of his decision
to retain the primary 24-hour PM2.5 standard with its level
of 35 [micro]g/m\3\ as described in section II.B.4 above, the primary
PM10 standard would be expected to maintain
PM10-2.5 concentrations in ambient air below those that have
been considered to be associated with serious health effects in past
NAAQS reviews. The Administrator also notes that while the scientific
evidence available in the 2019 ISA has expanded since the completion of
the 2009 ISA, he concludes that this information does not provide
support for the causal or likely to be causal relationships upon which
he places the greatest weight in considering the adequacy of the
current standards. He further concludes that the uncertainties and
limitations of the scientific evidence, along with the absence of
information to inform a quantitative exposure or risk assessment, make
it difficult to reach decisions regarding whether a more protective
standard is warranted at this time. He has additionally considered the
public comments regarding revisions to these elements of the standard
and continues to judge that the existing level and the existing form,
in all its aspects, together with the other elements of the existing
standard provide an appropriate level of public health protection. For
all of the reasons discussed above and recognizing the CASAC's
conclusion that the current evidence provides support for retaining the
current standard, the Administrator concludes that the current primary
PM10 standard (in all of its elements) is requisite to
protect public health with an adequate margin of safety from effects of
PM10-2.5 in ambient air and should be retained without
revision.
C. Decision on the Primary PM10 Standard
For the reasons discussed above and considering information and
assessments presented in the 2019 ISA and the 2022 PA, the advice from
the CASAC, and public comments, the Administrator concludes that the
current primary PM10 standard is requisite to protect public
health with an adequate margin of safety, including the health of at-
risk populations, and is retaining the current standard without
revision.
IV. Communication of Public Health
A. Air Quality Index Overview
Information about the public health implications of ambient
concentrations of criteria pollutants is communicated to the public
using the Air Quality Index (AQI) reported on the EPA's AirNow
website.\133\ The current AQI has been in use since its inception in
1999.\134\ It provides useful, timely, and easily understandable
information about the daily degree of pollution. The goal of the AQI is
to establish a nationally uniform system of indexing pollution
concentrations for ozone, carbon monoxide, nitrogen dioxide, PM, and
sulfur dioxide. The AQI is recognized internationally as a proven tool
to effectively communicate air quality information to the public as
demonstrated by the fact that many countries have created similar
indices based on the AQI.
---------------------------------------------------------------------------
\133\ See https://www.airnow.gov/.
\134\ In 1976, the EPA established a nationally uniform air
quality index, then called the Pollutant Standard Index (PSI), for
use by State and local agencies on a voluntary basis (41 FR 37660,
September 7, 1976; 52 FR 24634, July 1,1987). In August 1999, the
EPA adopted revisions to this air quality index (64 FR 42530, August
4, 1999) and renamed the index the AQI.
---------------------------------------------------------------------------
The AQI converts an individual pollutant concentration in a
community's air to a number on a scale from 0 to 500. Reported AQI
values for specific pollutants enable the public to know whether air
pollution levels in a particular location are characterized as good (0-
50), moderate (51-100), unhealthy for sensitive groups (101-150),
unhealthy (151-200), very unhealthy (201-300), or hazardous (301+).
Across criteria pollutants, the AQI value of 100 typically corresponds
to the level of the short-term (e.g., 24-hour, 8-hour, or 1-hour
standard) NAAQS for each pollutant. Below an index value of 100, an
intermediate value of 50 is defined either as the level of the annual
standard if an annual standard has been established (e.g.,
PM2.5, nitrogen dioxide), a
[[Page 16302]]
concentration equal to one-half the value of the 24-hour standard used
to define an index value of 100 (e.g., carbon monoxide), or a
concentration based directly on health effects evidence (e.g., ozone).
An AQI value greater than 100 means that a pollutant is in one of the
unhealthy categories (i.e., unhealthy for sensitive groups, unhealthy,
very unhealthy, or hazardous). An AQI value at or below 100 means that
a pollutant concentration is in one of the satisfactory categories
(i.e., moderate or good). The scientific evidence on pollutant-related
health effects for each NAAQS review support decisions related to
pollutant concentrations at which to set the various AQI breakpoints,
which delineate the AQI categories for each individual pollutant (i.e.,
the pollutant concentrations corresponding to index values of 150, 200,
300, and 500). The AQI is reported three ways by the EPA and State,
local and Tribal agencies, all of which are useful and complementary.
The daily AQI is reported for the previous day and used to observe
trends in community air quality, the AQI forecast helps people plan
their outdoor activities for the next day, and the near-real-time AQI,
or NowCast AQI, tells people whether it is a good time for outdoor
activity.
Historically, State and local agencies have primarily used the AQI
to provide general information to the public about air quality and its
relationship to public health. For more than two decades, many State
and local agencies, as well as the EPA and other Federal agencies, have
been developing new and innovative programs and initiatives to provide
more information related to air quality and health messaging to the
public in a more timely way. These initiatives, including air quality
forecasting, near real-time data reporting through the AirNow website,
use of data from air quality sensors on the EPA and U.S. Forest
Service's (USFS) Fire and Smoke Map, and air quality action day
programs, provide useful, up-to-date, and timely information to the
public about air pollution and its health effects. Such information can
help the public learn when their well-being may be compromised, so they
can take actions to avoid or to reduce exposures to ambient air
pollution at concentrations of concern. This information can also
encourage the public to take actions that will reduce air pollution on
days when concentrations are projected to be of concern to local
communities (e.g., air quality action day programs can encourage
individuals to drive less or carpool).
B. Air Quality Index Category Breakpoints for PM2.5
Recognizing the scientific information available and current AQI
reporting practices, the EPA proposed several revisions to the AQI
PM2.5 breakpoints. EPA solicited and received comments on
these proposed revisions. Upon reviewing the information in the
proposal and considering the comments received EPA is making final
revisions to the AQI category breakpoints for PM2.5. This
section summarizes the proposed revisions, which can be read in full in
the proposal (88 FR 5638, January 27, 2023), significant comments, and
final revisions.
1. Summary of Proposed Revisions
One purpose of the AQI is to communicate to the public when air
quality is poor and thus when they should consider taking actions to
reduce their exposures. The higher the AQI value, the higher the level
of air pollution and the greater the health concern. In recognition of
the scientific information available that is informing the
reconsideration of the 2020 final decision on the primary
PM2.5 standards, including a number of new controlled human
exposure and epidemiologic studies published since the completion of
the 2009 ISA, as well as additional epidemiologic studies from other
peer reviewed documents that evaluate the health effects of wildfire
smoke exposure and that can inform the selection of AQI breakpoints at
higher PM2.5 concentrations,\135\ the EPA proposed to make
two sets of changes to the PM2.5 sub-index of the AQI.
First, the EPA proposed to continue to use the approach used in the
revisions to the AQI in 2012 (77 FR 38890, June 29, 2012) of setting
the lower breakpoints (50, 100 and 150) to be based on the levels of
the primary PM2.5 annual and 24-hour standards and proposed
to revise the lower breakpoints to be consistent with changes to the
primary PM2.5 standards that are part of this
reconsideration. Second, the EPA proposed to revise the upper AQI
breakpoints (200 and above) and to replace the linear-relationship
approach used in 1999 to set these breakpoints, with an approach that
more fully considers the PM2.5 health effects evidence from
controlled human exposure and epidemiologic studies that have become
available in the last 20 years (64 FR 42530, August 4, 1999).
---------------------------------------------------------------------------
\135\ In evaluating the scientific evidence available to inform
decisions regarding the AQI breakpoints, the EPA considered studies
that were included as a part of the 2019 ISA and ISA Supplement, but
also considered other studies that were not included as a part of
the review of the air quality criteria. The ISAs have specific
criteria for study inclusion and consideration in reaching
conclusions regarding causal relationships, and some studies that
may not have met those criteria (e.g., epidemiologic studies that
evaluate the health effects of wildfire smoke exposure that would
have higher PM2.5 concentrations, which are outside of
the scope of the ISA) were identified as studies that could be used
to inform decisions on the AQI, particularly for the upper
breakpoints.
---------------------------------------------------------------------------
a. Air Quality Index Values of 50, 100 and 150
With respect to the lower AQI breakpoints in the proposal (88 FR
5638, January 27, 2023), the EPA proposed to conclude that it is
appropriate to continue setting these breakpoints to be consistent with
the primary annual and 24-hour PM2.5 standard levels. The
lowest AQI value of 50 provides the breakpoint between the ``good'' and
``moderate'' categories. At and below this concentration, air quality
is considered ``good'' for everyone. Above this concentration, in the
``moderate'' category, the AQI contains advisories for unusually
sensitive individuals. The EPA has historically set this breakpoint at
the level of the primary annual PM2.5 standard. In doing so,
the EPA has recognized that: (1) The annual standard is set to provide
protection to the public, including at-risk populations, from
PM2.5 concentrations, which, when experienced on average for
a year, have the potential to result in adverse health effects; and (2)
the AQI exposure period represents a shorter exposure period (e.g., 24-
hour (or less)) while focusing on the most sensitive individuals. The
EPA saw no basis for deviating from this approach in this
reconsideration. Thus, the EPA proposed to set the AQI value of 50 at a
daily (i.e., 24-hour) average concentration equal to the level of the
primary annual PM2.5 standard that is promulgated.
The historical approach to setting an AQI value of 100, which is
the breakpoint between the ``moderate'' and ``unhealthy for sensitive
groups'' categories, and above which advisories are generated for
sensitive groups, is to set it at the same level as the primary 24-hour
PM2.5 standard. In so doing, the EPA has recognized that the
primary 24-hour PM2.5 standard is set to provide protection
to the public, including at-risk populations, from short-term exposures
to PM2.5 concentrations that have the potential to result in
adverse health effects. Given this, it is appropriate to generate
advisories for sensitive groups at concentrations above this level. In
the past, State, local, and Tribal air quality agencies have expressed
strong support for this approach (78 FR 3086, January 15, 2013). The
EPA saw no basis to deviate
[[Page 16303]]
from this approach in this reconsideration. In the proposal (88 FR
5638, January 27, 2023), the EPA proposed to retain the current primary
24-hour PM2.5 standard with its level of 35 [mu]g/m\3\ but
took comment on revising the level of that standard to 25 [mu]g/m\3\
(section II.D.3.b). Thus, the EPA proposed to retain the AQI value of
100 set at the level of the current primary 24-hour PM2.5
standard concentration of 35 [mu]g/m\3\ (i.e., 24-hour average).
With respect to an AQI value of 150, which is the breakpoint
between the ``unhealthy for sensitive groups'' and ``unhealthy
categories,'' this breakpoint concentration in this reconsideration is
based upon the considering the same health effects information, as
assessed in the 2019 ISA and ISA Supplement and described in section II
above, that informs the proposed decisions on the level of the 24-hour
standard and the AQI value of 100. Previously, the Agency has used a
proportional adjustment in which the AQI value of 150 was set
proportionally to the AQI value of 100. This proportional adjustment
inherently recognizes that the available epidemiologic studies provide
no evidence of discernible thresholds, below which effects do not occur
in either sensitive groups or in the general population, that could
inform conclusions regarding concentrations at which to set this
breakpoint. Given that the epidemiologic evidence continues to be the
most relevant health effects evidence for informing this range of AQI
values, the EPA saw no basis to deviate from this approach in this
reconsideration. Therefore, the EPA proposed to set an AQI value of 150
proportionally, depending on the breakpoint concentration of the AQI
value of 100 (i.e., 55.4 for a 24-hour standard of 35 [mu]g/m\3\).
b. Air Quality Index Values of 200 and Above
In the proposal (88 FR 5639, January 27, 2023), the EPA summarized
the history of setting the AQI values of 300 and above in the 1999 rule
(64 FR 42530, August 4, 1999) and established breakpoints for
PM2.5 in that range. In general, the AQI values between 100
and 500 were based on PM2.5 concentrations that generally
reflected a linear relationship between increasing index values and
increasing PM2.5 concentrations.\136\ It was found that this
linear relationship was generally consistent with the health effects
evidence, which suggested that as PM2.5 concentrations
increase, increasingly larger numbers of people are likely to
experience serious health effects in this range of PM2.5
concentrations (64 FR 42536, August 4, 1999). For the AQI breakpoint of
500, the concentration was based on the method used to establish a
previously existing PM10 breakpoint that was informed by
studies conducted in London using the British Smoke method, which uses
a different particle size cutpoint as noted in the proposal (88 FR
5639, January 27, 2023). Due to limited ambient PM2.5
monitoring data available at that time, the decision on the 500 value
concentration for PM2.5 was based on the stated assumption
that PM concentrations measured by the British Smoke method were
approximately equivalent to PM2.5 concentrations (64 FR
42530, August 4, 1999). Given that the British Smoke method has a
larger particle size cutpoint than the current PM2.5
monitoring method, which has a cutpoint of 2.5 microns, a concentration
of 500 [mu]g/m\3\ based on the British Smoke method would be equivalent
to a lower PM2.5 concentration. With respect to the upper
breakpoints of the AQI, the EPA has historically been concerned about
establishing these upper breakpoints using evidence based on larger
size fractions of PM, given that PM2.5 is the indicator for
the AQI. While monitoring data for higher PM2.5
concentrations in ambient air has been available for many years, the
health effects evidence has only recently become available for
consideration in informing decisions on the upper breakpoints of the
AQI.
---------------------------------------------------------------------------
\136\ The AQI breakpoint at 150 was originally set in 1999 to be
linearly related to the concentrations at the 100 and 500
breakpoints but then revised in 2012 to be proportional to the AQI
breakpoint concentration at 100 (78 FR 3181, January 15, 2013).
---------------------------------------------------------------------------
As part of this reconsideration, the EPA recognized that the health
effects evidence associated with PM2.5 exposure has greatly
expanded in recent years. Multiple controlled human exposure studies
have become available that provide information about health effects
across a range of concentrations. While many of the new studies
evaluated in the 2019 ISA focused on examining health effects
associated with exposure to lower PM2.5 concentrations,
there are also several new controlled human exposure studies that
provide information about the health effects observed in study
participants at concentrations well above the standard levels.
Additionally, there are also epidemiologic studies now available and
evaluated in other Agency peer-reviewed documents that can inform
health effects associated with higher PM2.5 concentrations
(U.S. EPA, 2021b).\137\ Thus, the EPA concluded that it is appropriate
to reevaluate the upper AQI breakpoints, taking into account the
expanded body of scientific evidence, particularly given several new
epidemiologic studies conducted during high pollution events like
wildfires and multiple controlled human exposure studies. While it
remains unclear the exact PM2.5 concentrations at which
specific health effects occur, the more recent studies do provide more
refined information about the concentration range in which these
effects might occur in some populations. These studies provide support
for coherence of effects across scientific disciplines and potentially
biologically plausible pathways for the overt population-level health
effects observed in epidemiologic studies. Therefore, taking into
account the short exposure time period in these studies (e.g., 1-6
hours) and that the studies generally do not include at-risk (or
sensitive) populations, but rather young, healthy adults, these
studies, in conjunction with information from epidemiologic studies,
the EPA preliminarily concluded it would be appropriate to be more
cautionary and offer advisories to the public for reducing exposures at
lower concentrations than recommended with the current AQI breakpoints.
---------------------------------------------------------------------------
\137\ In this reconsideration, the controlled human exposure
studies were evaluated in the 2019 ISA, whereas the epidemiologic
studies of wildfire smoke exposures were included in the EPA
Comparative Assessment of the Impacts of Prescribed Fire Versus
Wildfire (CAIF): A Case Study in the Western U.S. (U.S. EPA 2021b).
---------------------------------------------------------------------------
The AQI value of 200 is the breakpoint between the ``unhealthy''
and ``very unhealthy'' categories. At AQI values above 200, the AQI
would be providing a health warning that the risk of anyone
experiencing a health effect following short-term exposures to these
PM2.5 concentrations has increased. To inform proposed
decisions on this breakpoint, the EPA takes note of studies indicating
the potential for respiratory or cardiovascular effects that are on
their own representative of or are on the biologically plausible
pathway to more serious health outcomes (e.g., emergency department
visits, hospital admissions). The controlled human exposure studies
evaluated in the 2009 and 2019 ISAs provide evidence of inflammation as
well as cardiovascular effects in healthy subjects at and above 120
[micro]g/m\3\. For example, Ramanathan et al. (2016) observed a
transient reduction in antioxidant/anti-inflammatory function after
exposing healthy young subjects to a mean concentration of 150
[micro]g/m\3\ of PM2.5 for 2 hours. Urch et al.
[[Page 16304]]
(2010) also reported increased markers of inflammation when exposing
both asthmatic and non-asthmatic subjects to a mean concentration of
140 [micro]g/m\3\ of PM2.5 for 3 hours. In studies
specifically examining cardiovascular effects, Ghio et al. (2000) and
Ghio et al. (2003) exposed healthy subjects to a mean concentration of
120 [micro]g/m\3\ for 2 hours and reported significantly increased
levels of fibrinogen, a marker of coagulation that increases during
inflammation. Sivagangabalan et al. (2011) exposed healthy subjects to
a mean concentration of 150 [micro]g/m\3\ of PM2.5 for 2
hours and noted an increased QT interval (3.4 1.4)
indicating some evidence for conduction abnormalities, an indicator of
possible arrhythmias. Lastly, Brook et al. (2009) reported a transient
increase of 2.9 mm Hg in diastolic blood pressure in healthy subjects
during the 2-hour exposure to a mean concentration of 148 [micro]g/m\3\
of PM2.5.
In addition to epidemiologic studies evaluated in the 2019 ISA that
analyzed exposures at ambient PM2.5 concentrations, there
are a number of recent epidemiologic studies focusing on wildfire smoke
that have become available that were evaluated in the EPA's recently
released peer-reviewed assessment on wildland fire (U.S. EPA, 2021b).
One of these studies, Hutchinson et al. (2018), conducted a
bidirectional case-crossover analysis to examine associations between
wildfire-specific PM2.5 exposure and respiratory-related
healthcare encounters (i.e., ED visits, inpatient hospital admissions,
and outpatient visits) prior and during the 2007 San Diego wildfires.
This study found positive and significant associations to
PM2.5 exposures and respiratory-related healthcare
encounters. Further, during the initial 5-day period of the wildfire
event, the study observed that there was evidence of increases in a
number of respiratory-related outcomes particularly ED visits for
asthma, upper respiratory infection, respiratory symptoms, acute
bronchitis, and all respiratory-related visits (Hutchinson et al.,
2018). When examining the air quality during the wildfire event,
PM2.5 concentrations were highest during the initial five
days of the wildfire, with 24-hour average PM2.5
concentrations of 89.1 [micro]g/m\3\ across all zip codes and with the
highest 24-hour average of 160 [micro]g/m\3\ on the first day
(Hutchinson et al., 2018).
When considering this collective body of evidence from controlled
human exposure and epidemiologic studies, the Agency proposed to set an
AQI value of 200 at a daily (i.e., 24-hour average) concentration of
PM2.5 of 125 [mu]g/m\3\. As discussed above and in the
proposal (88 FR 5640, January 27, 2023), this concentration is at the
lower end of the concentrations consistently shown to be associated
with respiratory and cardiovascular effects in controlled human
exposure studies following short-term exposures (e.g., 2-3 hours) and
in young, healthy adults (Ghio et al., 2000; Ghio et al., 2003; Urch et
al., 2010; Ramanathan et al., 2016; Sivagangabalan et al., 2011; and
Brook et al., 2009) and also within the range of 5-day average and
maximum concentrations observed to be associated with respiratory-
related outcomes following exposure to wildfire smoke (Hutchinson et
al., 2018).
The AQI value of 300 denotes the breakpoint between the ``very
unhealthy'' and ``hazardous'' categories, and thus marks the beginning
of the ``hazardous'' AQI category. At AQI values above 300, the AQI
provides a health warning that everyone is likely to experience effects
following short-term exposures to these PM2.5
concentrations. To inform decisions on this AQI breakpoint, the EPA
takes note of controlled human exposure studies that consistently show
subclinical effects which are often associated with more severe
cardiovascular outcomes. As discussed above, Brook et al. (2009)
reported a transient increase of 2.9 mm Hg in diastolic blood pressure
in healthy subjects during the 2-hour exposure to a mean concentration
of 148 [micro]g/m\3\ of PM2.5. Bellavia et al. (2013)
exposed healthy subjects to an average PM2.5 concentration
of 242 [micro]g/m\3\ for 2 hours and reported increased systolic blood
pressure (2.53 mm Hg). Tong et al. (2015) exposed healthy subjects to
an average PM2.5 concentration of 253 [micro]g/m\3\ for 2
hours and observed a significant increase in diastolic blood pressure
(2.1 mm Hg) and a nonsignificant increase in systolic blood pressure
(2.5 mm Hg). Lucking et al. (2011) reported impaired vascular function
and increased potential for coagulation when exposing healthy subjects
to diesel exhaust (DE) with an average PM2.5 concentration
of 320 [micro]g/m\3\ for a duration of 1 hour.\138\ These studies all
provided evidence of impaired vascular function, including
vasodilatation impairment and increased thrombus formation, with Tong
et al. (2015), Bellavia et al. (2013), Brook et al. (2009) all
reporting increases in blood pressure. Additionally, Behbod et al.
(2013) reported increased inflammatory markers following a 2-hour
exposure to an average PM2.5 concentration of 250 [micro]g/
m\3\ in healthy subjects.
---------------------------------------------------------------------------
\138\ Although participants in Lucking et al. (2011) were
exposed to diesel exhaust (DE), the authors also conducted analyses
using a particle trap, and as noted in the 2019 ISA, this type of
study design allows for the assessment of the role of
PM2.5 on the health effects observed by removing PM from
the DE mixture.
---------------------------------------------------------------------------
In addition to the controlled human exposure studies discussed
above, the epidemiologic study conducted by DeFlorio-Barker et al.
(2019) examined the relationship between wildfire smoke and
cardiopulmonary hospitalizations among adults 65 years of age and older
from 2008-2010 in 692 U.S. counties. The authors reported a 2.22%
increase in all-cause respiratory hospitalizations on wildfire smoke
days for a 10 [micro]g/m\3\ increase in 24-hour average
PM2.5 concentrations (DeFlorio-Barker et al., 2019). The
maximum 24-hour average concentration in this study on wildfire smoke
days was 212.5 [micro]g/m\3\ (DeFlorio-Barker et al., 2019). In
considering this study, the EPA notes the increased probability that
even healthy adults experience effects at this maximum exposure
concentration, particularly given that this maximum concentration is
near the exposure concentrations in controlled human exposure studies
that consistently reported evidence of impaired vascular function and
several that reported increases in blood pressure in healthy adults
following 2-hour exposures.
Based on the information discussed above and in the proposal (88 FR
5640, January 27, 2023), the EPA proposed to revise the 300 level of
the AQI, which marks the beginning of the ``hazardous'' AQI category,
to a concentration that is consistent with the PM2.5
concentrations associated with health effects as reported in the
controlled human exposure (Brook et al., 2009; Bellavia et al., 2013;
Tong et al., 2015; Behbod et al., 2013) and epidemiologic studies
(DeFlorio-Barker et al. (2019). Specifically, the Agency proposed to
set an AQI value of 300 at a daily (i.e., 24-hour average)
PM2.5 concentration of 225 [mu]g/m\3\. This concentration
falls between the 2-hour average concentrations reported in controlled
human exposure studies found to be consistently associated, in healthy
adults, with impaired vascular function and/or increases in blood
pressure, which could both be a precursor to more severe cardiovascular
effects following short-term (1- to 2-hour) exposures, and the maximum
24-hour average PM2.5 concentrations on wildfire smoke days
reported in the epidemiologic study conducted by DeFlorio-Barker et al.
(2019).
[[Page 16305]]
c. Air Quality Index Value of 500
Lastly, the EPA also proposed revisions to the 500 value of the
AQI. The 500 value of the AQI is within the ``hazardous'' category but
is specified and used to calculate the slope of the AQI values in the
``hazardous category'' above and below AQI values of 500. In the past,
this breakpoint had a very prominent role in determining the current
upper AQI values given that it was used as part of the linear
relationship with the concentration at the AQI value of 100 to
determine the AQI values of 200 and 300 in 1999 (64 FR 42530, August 4,
1999).
As discussed above and in the proposal (88 FR 5641, January 27,
2023), the current breakpoint concentration for the 500 value of the
AQI was set in 1999 at a 24-hour average PM2.5 concentration
of 500 [mu]g/m\3\ and was based on studies conducted in London using
the British Smoke method, which used a different particle size cutpoint
and likely overestimated the PM2.5 concentration. In looking
to improve upon that approach, the EPA considered several recent
controlled human exposure studies that observe health effects that are
on the biologically plausible pathway to more severe cardiovascular
outcomes and note that these seem to follow exposures to high
PM2.5 concentrations that are well above those typically
observed in ambient air. More specifically, in controlled human
exposure studies, Vieira et al. (2016a) and Vieira et al. (2016b)
exposed healthy subjects and subjects with heart failure to diesel
exhaust (DE) with a mean PM2.5 concentration of 325
[micro]g/m\3\ for 21 minutes and reported decreased stroke volume, and
increased arterial stiffness (an indicator of endothelial dysfunction)
in both healthy and heart failure subjects.\139\ Also as summarized
above and discussed in the proposal (88 FR 5641, January 27, 2023),
Lucking et al. (2011) exposed healthy subjects to DE with a mean
PM2.5 concentration of 320 [micro]g/m\3\ for 1 hour.\140\
Epidemiologic studies have linked the types of cardiovascular effects
observed in these controlled human exposure studies with the
exacerbation of ischemic heart disease (IHD) and heart failure as well
as myocardial infarction (MI) and stroke.
---------------------------------------------------------------------------
\139\ These effects were attenuated when the DE was filtered, to
reduce PM2.5 concentrations, indicating the effects were
likely associated with PM2.5 exposure.
\140\ When applying a particle trap, PM2.5
concentrations were reduced, and effects associated with
cardiovascular function including impaired vascular function, as
measured by vasodilatation and thrombus formation were attenuated
indicating associations with PM2.5.
---------------------------------------------------------------------------
In addition to the controlled human exposure studies discussed in
the proposal (88 FR 5641, January 27, 2023) and summarized above,
recent epidemiologic studies examining the relationship between
concentrations of PM2.5 during wildfires and respiratory
health also informed the proposed decisions on the concentration for
the AQI value of 500. As discussed in the proposal (88 FR 5641, January
27, 2023) and summarized earlier in this section, Hutchinson et al.
(2018) reported increases in a number of respiratory-related ED visits
for asthma, upper respiratory infection, respiratory symptoms, acute
bronchitis, and all combined respiratory-related visits based on data
from Medi-Cal claims for emergency department presentations, inpatient
hospitalizations, and outpatient visits during the initial 5-day period
of the 2007 San Diego fire. During the initial 5-day window,
PM2.5 concentrations were found to be at their highest with
the 95th percentile of 24-hour average concentrations of 333 [micro]g/
m\3\.
Although studies of short-term (i.e., daily) exposures to wildfire
smoke are more informative in considering alternative level for the AQI
value of 500 since they mirror the 24-hour exposure timeframe,
additional information from epidemiologic studies of longer-term
exposures (i.e., over many weeks) during wildfire events can provide
supporting information. As discussed in the proposal (88 FR 5641,
January 27, 2023) and summarized here, Orr et al. (2020) conducted a
longitudinal study that reported exposure to wildfire smoke from a
multi-month fire resulted in reduced lung function in subsequent years
and concluded that exposure to high PM2.5 concentrations
during a multi-week fire event may lead to health consequences, such as
declines in lung function. During the 2017 wildfire event (August 1 to
September 19, 2017), Orr et al. (2020) reported that many days during
the multi-month fire had PM2.5 concentrations above 300
[micro]g/m\3\, resulting in a daily average PM2.5
concentration of 220.9 [mu]g/m\3\ with a maximum PM2.5
concentration of 638 [micro]g/m\3\.
The controlled human exposure studies provide biological
plausibility for results of epidemiologic studies that document
increases in respiratory-related health care events during the
wildfires. The collective evidence from controlled human exposure and
epidemiologic studies, which includes decreases in stroke volume,
increased arterial stiffness, impaired vascular function and
respiratory-related healthcare encounters provide health-based evidence
that informed the proposed decisions on the level of the AQI value of
500. Given the concentrations observed in these studies, the Agency
proposed to revise the AQI value of 500 to a level set at a daily
(i.e., 24-hour average) PM2.5 concentration of 325 [mu]g/
m\3\. This concentration is at or below the lowest concentrations
observed in the controlled human exposure studies associated with more
severe effects discussed above and also at the low end of the daily
concentrations observed in the epidemiologic studies conducted by
Hutchinson et al. (2018) and Orr et al. (2020).
Table 1 below summarizes the proposed breakpoints for the
PM2.5 sub-index.
Table 1--Proposed Breakpoints for PM2.5 Sub-Index
----------------------------------------------------------------------------------------------------------------
Proposed breakpoints ([mu]g/
AQI category Index values m\3\, 24-hour average)
----------------------------------------------------------------------------------------------------------------
Good............................................................... 0-50 0.0-(9.0-10.0)
Moderate........................................................... 51-100 (9.1-10.1)-35.4
Unhealthy for Sensitive Groups..................................... 101-150 35.5-55.4
Unhealthy.......................................................... 151-200 55.5-125.4
Very Unhealthy..................................................... 201-300 125.5-225.4
[[Page 16306]]
Hazardous \1\...................................................... 301+ 225.5
----------------------------------------------------------------------------------------------------------------
\1\ AQI values between breakpoints are calculated using equation 1 in appendix G. For AQI values in the
hazardous category, AQI values greater than 500 should be calculated using equation 1 and the PM2.5
concentration specified for the AQI value of 500.
2. Summary of Significant Comments on Proposed Revisions
The EPA received many comments on the proposed changes to the
PM2.5 AQI breakpoints. Many commenters generally supported
all the proposed revisions to the AQI breakpoints based on the
revisions to the primary annual and daily PM2.5 standards
and recent scientific evidence discussed in the proposal (88 FR 5558,
January 27, 2023). However, we received specific comments on proposed
revisions to the breakpoints in the lower end of the AQI, related to
their linkage to the annual and daily PM2.5 standards, and
proposed revisions to the breakpoints at the upper end of the AQI,
based on EPA's interpretation of available health effects evidence.
a. Air Quality Index Values of 50, 100, and 150
Some commenters agreed with using the historical approach of
setting the 50, 100 and 150 breakpoints of the AQI to be consistent
with the primary PM2.5 standards. Some cited the reason that
this approach creates consistent communication with respect to air
quality and the standards, and this is how the other AQI sub-indices
are set. A few commenters disagreed with the historical approach and
suggested instead that the 50 breakpoint of the AQI should not be
revised at all, or that the 50 and 100 breakpoints of the AQI should be
supported directly by health data similar to the basis for the proposed
200, 300 and 500 breakpoints.
The few commenters that disagreed with the historical approach of
the 50 breakpoint of the AQI noted that setting a short-term breakpoint
to annual standard was not logical since it is a long-term standard and
not meant to be interpreted for short-term messaging with the AQI, in
particular when reported hourly via the NowCast. These commenters also
noted that additional studies are needed to identify the health impacts
of short-term exposures at low concentrations. They also noted that
lowering the 50 breakpoint of the AQI in conjunction with the annual
standard may cause confusion with the public because some State
programs and policy decisions are connected to the AQI while others are
based on PM concentrations, which could lead to inconsistent messaging
reducing the public's trust. These comments were supported by noting
that revised breakpoints could lead to more moderate days than in the
past, but the monitor values would be the same as before when the
commenters considered it ``healthy,'' possibly eroding trust in air
agencies' messaging. Commenters also noted if the breakpoints are
revised, the public will not visually be able to detect the difference
between what was considered a good AQI day versus a now moderate AQI
day.
The EPA disagrees with these commenters. With respect to setting a
short-term breakpoint to the level of a much longer-term (annual)
standard, setting the lower AQI breakpoints at the level of the annual
and daily PM2.5 standards for communication purposes was
discussed in the proposed reconsideration (88 FR 5558, January 27,
2023) and previously supported by State organizations in the 2012 PM
Final Rule (77 FR 38890, June 29, 2012). Both the AQI and the Pollutant
Standards Index, which came before it, have historically been
normalized across pollutants by defining an index value of 50 and 100
as the numerical level of the annual (when defined) and short-term
(i.e., averaging time of 24-hours or less) primary NAAQS for each
pollutant. This approach clearly communicates the air quality to the
public. The EPA considers this approach to be appropriate given the
available evidence and structure of the standard. As discussed in
section II.B above and in the notice of final rulemaking for the 2012
review (77 FR 38890, June 29, 2012), the primary annual and 24-hour
PM2.5 standards work together in concert to provide public
health protection. The annual PM2.5 standard is generally
viewed as the principal means of providing public health protection
against ``typical'' daily and annual PM2.5 exposures, while
the 24-hour PM2.5 standard is generally viewed as a means of
providing protection against short-term exposures to ``peak''
PM2.5 concentrations, such as can occur in areas with strong
contributions from local or seasonal sources, even when annual average
PM2.5 concentrations remain relatively low. Because the
annual standard provides public health protection for typical daily
PM2.5 exposures, the EPA thinks it is appropriate to use
that level for the 50 breakpoint of the AQI and describe daily air
quality at and below the level of the annual standard ``Good.'' Since
an annual standard allows for days with air quality above that level,
it is appropriate to call days just above it ``Moderate.'' If the 50
breakpoint of the AQI was set at a level above the annual standard, it
would be possible for the majority of days to be called ``good'' in a
year when an area exceeds the annual standard. This could cause
confusion with the public about air quality if the general perception
is that local air quality is ``good,'' but the area fails to meet the
annual standard. In addition, the EPA continues to find it appropriate
to use the NowCast with the PM2.5 AQI index to provide more
real-time information to the public. As discussed in the AQI Technical
Assistance Document, while the NowCast algorithm is approximating a 24-
hour average exposure, it can reflect concentrations observed over
shorter averaging times when air quality is changing rapidly (U.S. EPA,
2018a). The EPA continues to consider the use of the primary annual
standard level suitable in the NowCast given the health evidence
supporting the standard and given that the reported concentrations are
an approximation of ``typical'' daily exposure. Additionally, the EPA
reflects the nature of the NowCast in the associated health messaging.
With regard to the commenter stating the public may not be able to
visually detect a difference in the air quality, the EPA notes that the
AQI is intended to be a communication tool for public awareness
precisely because it is generally difficult for the public to visually
judge air quality risks when air pollution is ``moderate.'' Moreover,
since the establishment of the AQI, the EPA and State and local air
agencies and organizations have developed experience in educating the
public about changes in the standards and,
[[Page 16307]]
concurrently, related changes to AQI breakpoints and advisories. When
the standards change, the EPA and State and local agencies have sought
to help the public understand that air quality is not getting worse,
it's that the health evidence underlying the standards and the AQI has
changed. The EPA's Air Quality System (AQS), the primary repository for
air quality monitoring data, is also adjusted to reflect the revised
breakpoints. Specifically, all historical AQI values in AQS are
recomputed with the revised breakpoints, so that all data queries and
reports downstream of AQS will show appropriate trends in AQI values
over time. If any State, local or Tribal air agency is concerned that
people are or will be confused on a moderate AQI day, then they could
use the communication information that has been developed with this
rulemaking.
Some commenters stated that the AQI should not necessarily be
linked to the primary PM2.5 standards. One example is the
comment that if the annual standard is not lowered to 8 [micro]g/m\3\,
the EPA should lower the 50 breakpoint of the AQI to that level to
better inform the public of the need for behavioral modifications to
reduce the harm to health from PM2.5 exposure. Similar to
the reasons discussed above, the EPA concludes that setting the 50
breakpoint of the AQI at the level of the annual PM2.5
standard is appropriate from a health perspective and for communication
purposes. The Administrator has judged the primary annual standard (in
conjunction with the other primary standards) as revised in this final
action to be requisite to protect public health with an adequate margin
of safety, based on the health evidence discussed in section II.A.2.
Setting the 50 breakpoint lower than the annual standard also has the
potential to cause confusion with the public since it does not reflect
the standards and the Administrator's judgments about the standards as
well.
With regard to the 100 breakpoint of the AQI, several commenters
expressed the view that the level of the 24-hour PM2.5
standard and an AQI value of 100 should be set at 25 [mu]g/m\3\ based
on the body of evidence and lower end of the range recommended by
CASAC. These commenters noted that if the current 24-hour standard and
AQI value of 100 is retained at 35 [mu]g/m\3\ then the public will not
be able to make informed decisions about actions to take to protect
their health. Many of these commenters further recommended that the AQI
value of 100 should be lowered to 25 [mu]g/m\3\ even if the standard is
retained. Commenters expressed the view that this would more adequately
allow the public to take health-protective actions.
The EPA disagrees with these commenters and notes that many State,
Tribal and local air agencies have expressed strong support for
aligning the 100 breakpoint of the AQI with the short-term 24-hour
primary PM2.5 standards as discussed in the proposal (88 FR
5558, January 27, 2023). The EPA agrees with the view, expressed by
State, local and Tribal entities, that aligning the lower breakpoints
with the standards enables clear communication of the standards. This
alignment approach is also utilized in the other AQI sub-indices lower
breakpoints and taking a different approach with the PM2.5
AQI could cause confusion. Additionally, the Administrator has judged
that it is appropriate to retain the 24-hour standard at a level of 35
[mu]g/m\3\ (in conjunction with the other primary standards) to protect
public health with an adequate margin of safety, based on the health
evidence discussed in section II.A.2. Thus, EPA disagrees that it is
necessary or appropriate to set the 100 breakpoint at a lower
concentration to provide further information to the public. The 50
breakpoint, which is set at a level below 25 [mu]g/m\3\, will continue
to provide information to members of the public particularly concerned
about exposures to PM2.5. As with the 50 breakpoint,
aligning the breakpoint with the standard both reflects the
Administrator's judgment about the health risks and eliminates the
potential to cause confusion in the public about those risks.
b. Air Quality Index Values of 200 and Above
Some commenters supported the proposed revisions to the 200, 300
and 500 breakpoints that recognize the expanded body of scientific
evidence, particularly several new epidemiologic studies conducted
during high pollution events such as wildfires and multiple controlled
human exposure studies. A few commenters agreed with incorporating the
expanded body of scientific evidence into the 200, 300 and 500
breakpoints, but suggested a modified linear approach between 200 (115
[mu]g/m\3\) and 500 (312 [mu]g/m\3\, setting the 300 breakpoint to 187
[mu]g/m\3\) based on recent epidemiologic wildfire smoke studies.
Other commenters disagreed with the proposed revisions and
suggested the EPA should continue using the previous breakpoints that
follow the 1999 linear approach (64 FR 42530, August 4, 1999), because
not changing the breakpoints would simplify communications. A few
commenters stated the proposed revisions to the AQI upper breakpoints
are not justified because the scientific evidence supporting the
revisions is inadequate. To support this view, the commenters suggest
that only three epidemiologic studies were used in determining the
upper breakpoints and none of them were representative of potential
effects in the general public; of the 13 studies cited only three were
near the proposed revised breakpoints; four of the studies involved
exposure to PM from diesel and traffic pollution, which is different
than PM from wildfire smoke; and the data supporting the revisions only
indicated ``mild'' health effects that were mostly in sensitive
populations.
The EPA agrees with the majority of commenters that supported
utilizing the expanded body of scientific evidence to revise the 200,
300 and 500 breakpoints of the AQI. The EPA appreciates the suggestion
of using a revised linear approach from 200 to 500. But rather than
using the available evidence to only set the breakpoint of 500, the EPA
finds it appropriate to set the breakpoints for 200, 300 and 500 using
an evidence-based approach, by relying on information presented in both
controlled human exposure studies and epidemiologic studies that
examine relationships between high PM2.5 exposure episodes
(i.e., periods of wildfire smoke) and various health outcomes. Setting
these breakpoints based directly on health effects evidence, which can
be communicated, is more useful and appropriate than using a linear
approach, because it can better describe the potential health effects
and symptoms which also helps the public better understand why more
health protective actions are needed. By its nature, a linear approach
does not evaluate and identify associated health effects and risk
factors.
The EPA disagrees with the commenters that expressed the view that
these upper breakpoints should not be revised based largely on the
numerous peer-reviewed studies published since the 200, 300 and 500
breakpoints were originally established in 1999 (64 FR 42530, August 4,
1999). As discussed in the proposal (88 FR 5641, January 27, 2023), the
rationale behind the proposed revisions is rooted in the fact the upper
AQI breakpoints are based on outdated scientific evidence.
Specifically, the traditional linear approach was predicated on the 500
value of the AQI, which was estimated using health studies that used
the British Smoke Method. The British Smoke Method is based on a
particle size fraction (4.5 microns) that is larger
[[Page 16308]]
than PM2.5. Given that the British Smoke method has a larger
particle size cutpoint than the current PM2.5 monitoring
method, which has a cutpoint of 2.5 microns, a concentration of 500
[mu]g/m\3\ based on the British Smoke method would be equivalent to a
lower PM2.5 concentration (88 FR 5641, January 27, 2023).
The combination of a larger particle size fraction informing previous
decisions around upper AQI breakpoints and more recent scientific
evidence than the London Fog Episode, on the potential health
consequences of what we currently consider to be high PM2.5
exposures, provides the underlying basis for revising the upper
breakpoints to better inform the public about air quality to allow the
public to take health protective actions as appropriate. Moreover, as
discussed above, until recently there was limited information upon
which to base the breakpoints between 150 and 500, so the linear
approach was a reasonable substitute. While not changing the
breakpoints may be easier because there is no change to communicate,
using a health-based approach is more appropriate, because it helps the
public better understand that more health protective actions are
needed.
The Agency disagrees that the scientific evidence discussed in the
proposal is inadequate to revise the 200, 300 and 500 breakpoints of
the AQI (88 FR 5640, January 27, 2023). The EPA disagrees that these
studies should not be considered because they ``indicated mild health
effects in sensitive populations.'' The EPA notes that many of the
subclinical effects discussed in the proposal (88 FR 5640, January 27,
2023) that informed the breakpoints are on the biologically plausible
pathway (see 2019 ISA, section 6.1.1 and Figure 6-1) to more severe
cardiovascular outcomes, such as ED visits, hospital admissions, and
death as depicted in the large number of epidemiologic studies
evaluated in the 2019 ISA and ISA Supplement. From a public health
perspective, the purpose of the AQI is to inform the public when air
quality could adversely affect their health. The scientific evidence
informed revisions to the breakpoints at the upper end of the AQI allow
it to better reflect the risk of experiencing health effects at higher
PM2.5 concentrations. In addition, the EPA disagrees with
the commenter that the effects reported at these higher concentrations
were observed only in sensitive populations as these effects were also
reported in healthy populations (Ghio et al., 2000; Ghio et al., 2003;
Urch et al., 2010; Ramanathan et al., 2016; Sivagangabalan et al.,
2011; Brook et al., 2009; Bellavia et al. (2013); Tong et al. (2015);
Behbod et al. (2013); Vieira et al. (2016a) Vieira et al. (2016b); and
Lucking et al. (2011)).
c. Other Comments
The EPA received a few additional comments on elements of the
PM2.5 AQI, including the averaging time. Some commenters
expressed the view that the 24-hour averaging time was not useful when
informing the public how to protect their health, particularly during
rapidly changing conditions such as wildfire smoke events. Instead,
they suggested a subdaily averaging time of 1-3 hours would be more
effective because it more closely aligns with how people breathe.
A few of these commenters suggested that instead of changing the
AQI averaging time, which aligns with the short-term standard, the EPA
could create a public health warning system for unhealthy
PM2.5 levels. The commenters noted that aligning the AQI
averaging time with the short-term standard could be useful for
consistent communication with the standards and attainment but
suggested that a subdaily warning system could better allow the public
to take health protective actions.
The EPA disagrees that a shorter averaging period for the
PM2.5 AQI sub-index would be better. The health effects
evidence supporting a subdaily metric is limited and inconsistent. As
part of its review of the health effects evidence, the 2019 ISA
evaluated whether a subdaily metric would be more closely related to
health effects. Most epidemiologic studies that examined the
relationship between short-term PM2.5 exposures and health
effects evaluated an exposure metric averaged over 24-hours. Some
recent studies, focusing on respiratory and cardiovascular effects and
mortality, have examined whether there is evidence that subdaily
exposure metrics are more closely related to health effects than a
traditional 24-hour average metric. After evaluating this limited newer
evidence, the 2019 ISA concluded that ``collectively, the available
evidence does not indicate that subdaily averaging periods for
PM2.5 are more closely associated with health effects than
the 24-hour avg exposure metric,'' (2019 ISA, chapter 1, section
1.5.2.1, pp. 146-147; U.S. EPA, 2022a).
In addition, there are communication benefits to aligning the
averaging time of the AQI with the daily standard, as some of these
commenters note, such as providing consistent messages about when it
may be beneficial for people to take actions to reduce PM2.5
exposures. Furthermore, with regard to an additional warning system,
the EPA is concerned that having two air quality communication systems
operating at the same time would likely be confusing to the public and
reduce the effectiveness of the systems.
At the same time, the EPA recognizes that when air quality is
rapidly changing, such as during wildfire smoke events, reporting
information based on a 24-hour metric may not be as useful for the
public as reporting more frequently would be. The EPA has balanced
concerns about being able to provide timely communication of air
quality hazards when conditions are changing quickly with the goal of
limiting the number of air quality communications systems and its
judgment that the evidence supports a 24-hour-based metric linked to
the daily standard by establishing the NowCast, which takes into
consideration subdaily PM2.5 concentrations and provides a
near real-time AQI value based on the AQI colors and scale.
Specifically, the NowCast shows air quality conditions for the most
current hour of PM2.5 data available by using a calculation
that involves multiple hours of past data. As noted in the AQI
Technical Assistance Document, the NowCast currently uses longer
averages during periods of stable air quality and shorter averages
(down to a 3-hour average) when air quality is changing rapidly, such
as during a wildfire (U.S. EPA, 2018a). As discussed further in section
IV.D.2 of this notice, the EPA uses the NowCast to approximate the
complete daily AQI (24-hour average) during any given hour. This means
the subdaily NowCast is approximating a 24-hour average exposure, which
aligns with the health evidence and the existing AQI communications
network, while also being capable of communicating rapidly changing
conditions to the public.
3. Summary of Final Revisions
Upon reviewing and considering the comments on the proposed
revisions (summarized above in Section IV.C) along with the scientific
evidence outlined in the proposal (88 FR 5639, January 27, 2023) and
summarized above in section IV.A, the EPA is finalizing the proposed
changes to the AQI.
Thus, as discussed in section IV of the preamble (88 FR 5639,
January 27, 2023) to the proposed rule, the EPA is taking final action
to revise the AQI value of 50 to 9.0 [mu]g/m\3\, 24-hour average,
consistent with the final decision on the primary annual
PM2.5 standard level as summarized in section II.C of the
[[Page 16309]]
preamble to the final rule; retain the AQI value of 100 at 35 [mu]g/
m\3\, 24-hour average, consistent with the final decision on the
primary 24-hour PM2.5 standard level as summarized in
section II.C of the preamble to the final rule; and retain the AQI
value of 150 at 55 [mu]g/m\3\, 24-hour average. The EPA is also taking
action to revise the AQI value of 200 to 125 [mu]g/m\3\, 24-hour
average; 300 to 225 [mu]g/m\3\, 24-hour average; and 500 to 325 [mu]g/
m\3\, 24-hour average, consistent with the rationale discussed above
and the health evidence discussed in section IV of the preamble (88 FR
5639, January 27, 2023) to the proposed rule. The EPA has prepared
communications materials to assist States with adjusting to the revised
AQI and looks forward to working with, and learning from the
experiences of, State, local, and Tribal governments in implementing
these changes.
C. Air Quality Index Category Breakpoints for PM10
The EPA proposed to retain the PM10 sub-index of the AQI
consistent with the proposed decision to retain the primary
PM10 standard, and consistent with the health effects
information that supports this proposed decision, as discussed in
section III.D of the proposal (88 FR 5632, January 27, 2023). EPA did
not receive comments on this and is taking final action to retain the
PM10 sub-index of the AQI for the reasons stated in the
preamble to the proposed rule (88 FR 5642, January 27, 2023).
D. Air Quality Index Reporting
With respect to the reporting requirements for the AQI and as noted
in the proposal (88 FR 5642, January 27, 2023) there have been many
technological advances in air quality monitoring and data reporting
since the appendix G to 40 CFR part 58 was last revised in 1999.
Federal, State, local, and Tribal agencies have used these changes to
make health information and air quality data more readily available and
easier to access. Given this, it is useful to update the reporting
requirements and recommendations to match current practices and ensure
the public has the most useful and timely information to take health-
protective behaviors.
1. Summary of Proposed Revisions
Currently, appendix G defines daily reporting as five days per
week. When this reporting requirement was originated in 1999 the
technology available at that time was not sufficient to calculate and
report the AQI more than five days per week without requiring
additional staffing on the weekends. Since that time, advances in
technology have allowed for reporting seven days per week automatically
without expending additional resources on weekends. As a result, most
State, local, and Tribal air agencies now report the AQI seven days per
a week. Given these technological advances and noting that reporting
agencies currently report the AQI seven days per week, the EPA proposed
that State, local, and Tribal agencies that report the AQI be required
to report it seven days a week, ensuring that the members of the public
continue to have access to daily air quality and health information
that they can use to take steps to protect their health.
Improvements in monitoring networks and modeling capabilities have
also enabled the ability to report the AQI in near real-time. This
allows State, local, and Tribal air agencies to provide timely air
quality information to the public for making health-protective
decisions and to help satisfy AQI reporting requirements. The
availability of near real-time AQI data also allows for more timely
responses by the public when air quality conditions are changing
rapidly, such as during wildfire smoke events. Subdaily reporting of
the AQI can be critical when there are rapidly change conditions and/or
high pollution events so that the public is able to make informed
decisions to protect their health. Many State, local, and Tribal air
agencies currently report the AQI hourly to ensure that the public has
access to accurate and timely information. In recognition of these
advances, and to continue to provide for near-real time AQI reporting
that the public has come to rely on, the EPA proposed to recommend that
State, local, and Tribal agencies report the AQI in near-real time.
In lieu of or along with reporting the near-real-time AQI directly
to the public, most State/local and Tribal agencies submit hourly air
quality data to the EPA. The EPA and some State, local and Tribal air
quality agencies use this near-real-time data to create products for
use by the public, weather service providers and the media as discussed
in the proposal (88 FR 5643, January 27, 2023). To continue to ensure
the availability of the products that the public and many stakeholders
rely upon, the EPA proposed to recommend that State, local, and Tribal
air quality agencies submit hourly data to the EPA's air quality
database. Submitting hourly data to the EPA for use on the AirNow
website and in other products also enables State, local, and Tribal air
quality agencies to meet the recommendation to report the AQI in near-
real-time.
In addition to the proposed updates to the reporting requirements
and recommendations for near-real-time reporting and data submission
recommendations, the Agency also proposed reformatting the question-
and-answer format used in appendix G to align with the current standard
formatting used in the Code of Federal Regulations. In proposing to
update the format, the EPA did not reopen the language that has merely
been moved or rearranged as there are no substantive changes.
Another change the EPA proposed to make to appendix G is with
regard to Table 2--Breakpoints for the AQI for purposes of clarity. As
discussed in the proposal (88 FR 5642, January 27, 2023) and summarized
here, the EPA proposed to collapse the two rows presented for the
Hazardous Category into one. The two rows in the current table specify
pollutant concentrations for two AQI ranges within the Hazardous
category (301-400 and 401-500), with an intermediate break at 400. The
400 breakpoint for all criteria pollutants in the current Table 2 is
set at the proportional pollutant concentration approximately halfway
between the Index values of 300 and 500. In proposing updated AQI
breakpoints for PM2.5, the EPA considered adjusting the 400
breakpoint similarly. However, the EPA concluded that collapsing the
two rows into a single range (301-500) would provide a more transparent
and easy-to-follow presentation of the pollutant concentrations
corresponding to the AQI range for the Hazardous category. Moreover,
collapsing the Hazardous category into a single row in Table 2 has no
substantive effect on the Emergency Episode program in 40 CFR part 51,
appendix L. Thus, the EPA proposed to remove the breakpoint of 400 from
the table in appendix G but this change would not substantively affect
the derivation of the AQI for any pollutant.
In addition, the EPA proposed to move some information currently in
appendix G into the Technical Assistance Document for the Reporting of
Daily Air Quality, or TAD (U.S. EPA, 2018a), so that it can be updated
in a more timely manner to reflect current scientific and health
effects evidence and current communication methods, thereby assisting
State, local, and Tribal agencies in providing accurate and timely
information to the public. Information that was proposed to be moved
from appendix G to the TAD included the definitions of the sensitive
(at-risk) populations for each pollutant.
[[Page 16310]]
This definition is typically evaluated and updated, as warranted, in
most NAAQS reviews, even if the standard is not revised. Generally, if
the standard is not revised in a review of the NAAQS, then appendix G
is also not revised. Moving the definitions of sensitive groups to the
TAD allows them to be updated even when a NAAQS is not revised to be
consistent with the definitions of the sensitive (at-risk) populations
identified in the ISA for that NAAQS review. Also, the proposal (88 FR
5642, January 27, 2023) recognized that the ways that air quality and
health information is supplied to the news media and public changes
regularly and thus proposed that information about suggested approaches
for public communication be taken out of appendix G and discussed in
the TAD.
2. Summary of Significant Comments on the Proposed Revisions
The EPA received many comments on the proposed changes to AQI
reporting, many of which supported the proposed revisions. EPA
discusses several of the topics that received the most attention from
commenters below. Discussion of other comments received on the proposed
changes to the AQI can be found in section IV of the Responses to
Significant Comments on the 2023 Proposed Reconsideration of the
National Ambient Air Quality Standards for Particulate Matter.
Most commenters expressed support for revising the definition of
``daily reporting'' from five days a week to seven days a week. A
commenter did not support this change and recommended the EPA maintain
the definition of daily as five days per week, noting that State and
local air agencies do not routinely work seven days per week and would
not be available to perform quality control of this data and report it
reliably on weekends.
The EPA appreciates the support for this proposed revision and
disagrees that the proposed change would require personnel to perform
quality control of AQI data on weekends. 40 CFR part 58 Appendix D
defines continuous monitoring requirements for agencies participating
in the State/Local Air Monitoring Stations (SLAMS) network, and
Appendix G states that agencies `` . . . must use concentration data
from State/Local Air Monitoring Stations (SLAMS) required by 40 CFR
58.10'' when reporting the AQI. Therefore, as noted in Appendix D and
G, Agencies are required to report the AQI using monitors within SLAMS,
which are not subject to daily quality control/validation.
A few commenters noted that the proposal preamble language
mentioned AQI is reported three ways (88 FR 5637, 5638, January 27,
2023): ``The AQI is reported three ways all of which are useful and
complementary. The daily AQI is reported for the previous day and used
to observe trends in community air quality, the AQI forecast helps
people plan their outdoor activities for the next day, and the near-
real-time AQI, or NowCast AQI, tells people whether it is a good time
for outdoor activity.'' These commenters suggested that the NowCast is
being codified in 40 CFR part 58 Appendix G as a method of calculating
the AQI, which they oppose, saying that codifying its use is
inappropriate given the shortest averaging period of the
PM2.5 NAAQS remains at 24-hours. Some stated that NowCast
values have no direct correlation to the AQI calculation methodology
codified in 40 CFR part 58 Appendix G. These commenters say that
codifying the NowCast would impose a significant burden on States'
forecasting staff.
However, some other commenters noted they appreciate the public-
friendly format and near real-time data the NowCast provides and use it
in their clinical encounters with patients. One air agency recognized
the importance of the NowCast near real-time AQI during high pollution
events and suggested the EPA should provide more ``concrete'' health
messaging for these short-term spikes.
The EPA disagrees that the preamble language proposed to codify the
NowCast or to impose a burden on reporting agencies. The preamble to
the proposed rule references the AQI being reported in three ways and
it does so because the EPA and many State, local and Tribal air quality
agencies already report it these three ways. However, text included in
the preamble is generally explanatory and does not alter regulatory
provisions. Comments that State that EPA is codifying the NowCast into
Appendix G are incorrect. Further, in proposed revisions to 40 CFR part
58 Appendix G, the EPA recommended, but did not propose to require, the
use of air quality forecasts and a subdaily AQI. Consistent with the
proposal, the EPA is therefore not finalizing any additional
requirement or burden on States' forecasting staff relative to
forecasts or a subdaily AQI.
The EPA disagrees with the comment that the NowCast values have no
direct correlation to the AQI calculation methodology codified in 40
CFR part 58 Appendix G. As noted in the AQI Technical Assistance
Document (Technical Assistance Document for the Reporting of Daily Air
Quality--the Air Quality Index (AQI)), the NowCast algorithm is based
on the AQI methodology but provides more real-time information to the
public (U.S. EPA, 2018a). While the NowCast algorithm is approximating
a 24-hour average exposure, it can reflect concentrations observed over
shorter averaging times when air quality is changing rapidly (U.S. EPA,
2018a). The EPA reflects the nature of the NowCast in the health
messaging provided there.
As noted in the above discussion of the AQI, air quality can change
quickly during the day. A central purpose of the AQI is to help the
public know when it is prudent to take action to reduce their exposure
to pollution. Accordingly, the EPA developed the NowCast to estimate
the 24-hour AQI for the current hour to give people information and
tools to reduce their exposures to protect their health, particularly
when air quality may be changing. The NowCast gives people the
knowledge and ability to take timely action. They can use this
information to reduce their exposure--reducing exposures if
PM2.5 is high only during a few hours a day will help reduce
a person's 24-hour exposure--or be active when air quality is better.
The first NowCast method was developed in 2003 and was designed so
``current conditions'' represent the 24-hour PM2.5 standard
as closely as possible. This method proved to be slow to respond during
rapid air quality changes. In 2013, the EPA developed an updated
NowCast method for PM2.5 \141\ that responds more quickly to
rapidly changing air quality conditions, such as those we see during
wildfires, to make air quality alerts more timely. We analyzed millions
of data points in developing this NowCast method and presented this
information to State, local and Tribal air agencies. The updated
NowCast, which is still in use, was launched August 1, 2013, on
AirNow.gov. It was designed to represent a shorter average (target 3-
hour) when air quality is changing rapidly, in part because 3-hour
averages from some continuous monitors are more stable than 1-hour
averages. The NowCast reflects a longer-term (12-hour) average when air
quality is stable.
---------------------------------------------------------------------------
\141\ U.S. EPA. (2013). Transitioning to a New NowCast Method.
Presentation available in the Rulemaking Docket for the Review of
the National Ambient Air Quality Standards for Particulate Matter
(EPA-HQ-OAR-2015-0072), at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
After evaluating the 2013 NowCast method, the EPA concluded that it
matched the desired characteristics. The NowCast method responds to
rapid changes in air quality yet still reflects a
[[Page 16311]]
longer-term average when air quality is stable; will work in any
location with adequate air quality data and for any air quality
situation; gives people the best possible estimate of a 24-hour
exposure; allows the EPA to caution people in time for them to take
protective action and reduce their 24-hour exposure; and ensures that
AQI maps on AirNow more closely match what people see.
The AQI is designed to allow people to reduce their exposure when
pollution levels are higher and be active outdoors when pollution
levels are lower. Since air quality almost always changes during the
day, that level of granularity is not possible with a 24-hour forecast.
If the public has only the 24-hour forecast, they may miss the times to
be active outdoors when air quality is better and may be active
outdoors when air quality is worse.
Also as noted above, many entities appreciate the near real-time
reporting of the AQI that the NowCast provides and suggested more
specific messaging is needed. The EPA appreciates this insight and will
continue to consider ways to communicate air quality information most
effectively to the public. For example, in light of recent wildfire
events, the EPA worked with the USFS to pilot the AirNow Fire and Smoke
Map.
3. Summary of Final Revisions
Upon reviewing and considering the comments on the proposed
revisions (summarized above in Section IV.C) along with the rationale
outlined in the proposal (88 FR 5638, January 27, 2023) and summarized
above in section IV.C, the EPA is finalizing the proposed changes to
the AQI reporting requirements. Thus, as discussed in section IV of the
preamble to the proposed rule, the EPA is taking final action to
require the AQI be reported seven days a week; recommend that State,
local, and Tribal agencies report the AQI in near-real time; recommend
that State, local, and Tribal air quality agencies submit hourly data
to the EPA's air quality database; reformat appendix G to align with
the current standard formatting used in the Code of Federal
Regulations; collapse the two rows in Table 2 presented for the
Hazardous Category into one by removing the 400 breakpoint; and move
some information currently in appendix G into the Technical Assistance
Document for the Reporting of Daily Air Quality, or TAD (U.S. EPA,
2018a) such as including the definitions of the sensitive (at-risk)
populations for each pollutant and suggested approaches for public
communication as stated in the revised Appendix G.
Table 2 below summarizes the breakpoints for the PM2.5
sub-index.
Table 2--Breakpoints for PM2.5 Sub-Index
----------------------------------------------------------------------------------------------------------------
Breakpoints ([mu]g/m\3\,
AQI category Index values 24-hour average)
----------------------------------------------------------------------------------------------------------------
Good............................................................... 0-50 0.0-9.0
Moderate........................................................... 51-100 9.1-35.4
Unhealthy for Sensitive Groups..................................... 101-150 35.5-55.4
Unhealthy.......................................................... 151-200 55.5-125.4
Very Unhealthy..................................................... 201-300 125.5-225.4
Hazardous \1\...................................................... 301+ 225.5
----------------------------------------------------------------------------------------------------------------
\1\ AQI values between breakpoints are calculated using equation 1 in appendix G. For AQI values in the
hazardous category, AQI values greater than 500 should be calculated using equation 1 and the PM2.5
concentration specified for the AQI value of 500.
V. Rationale for Decisions on the Secondary PM Standards
This section presents the rationale for the Administrator's
decision that no change to the current secondary PM standards is
required at this time to provide requisite protection against the
public welfare effects of PM within the scope of this reconsideration
(i.e., visibility, climate, and materials effects).\142\ This decision
is based on a thorough review of the scientific evidence generally
published through December 2017,\143\ as presented in the 2019 ISA
(U.S. EPA, 2019a), on the non-ecological public welfare effects of PM
pertaining to the presence of PM in ambient air, specifically
visibility, climate, and materials effects. Additionally, this decision
is based on a thorough evaluation of some studies that became available
after the literature cutoff date of the 2019 ISA that could either
further inform the adequacy of the current PM NAAQS or address key
scientific topics that have evolved since the literature cutoff date
for the 2019 ISA, generally through March 2021, as presented in the ISA
Supplement \144\ (U.S. EPA, 2022a). The selection of welfare effects
evaluated within the ISA Supplement was based on the causality
determinations reported in the 2019 ISA and the subsequent use of
scientific evidence in the 2020 PA.\145\
[[Page 16312]]
Specifically, for welfare effects, the focus within the ISA Supplement
is on visibility effects. The ISA Supplement does not include an
evaluation of studies on climate or materials effects. The
Administrator's decision also takes into account the 2022 PA evaluation
of the policy-relevant information in the 2019 ISA and ISA Supplement
and presentation of quantitative analysis of air quality related to
visibility impairment; CASAC advice and recommendations, as reflected
in discussions of the drafts of the ISA Supplement and 2022 PA at
public meetings and in the CASAC's letters to the Administrator; and
public comments received on the proposal.
---------------------------------------------------------------------------
\142\ Consistent with the 2016 Integrated Review Plan (U.S. EPA,
2016), other welfare effects of PM, including ecological effects,
are being considered in the separate, on-going review of the
secondary NAAQS for oxides of nitrogen, oxides of sulfur and PM.
Accordingly, the public welfare protection provided by the secondary
PM standards against ecological effects such as those related to
deposition of nitrogen- and sulfur-containing compounds in
vulnerable ecosystems is being considered in that separate review.
Thus, the Administrator's decision in this reconsideration will be
focused only and specifically on the adequacy of public welfare
protection provided by the secondary PM standards from effects
related to visibility, climate, and materials and hereafter
``welfare effects'' refers to non-ecological welfare effects (i.e.,
visibility, climate, and materials effects).
\143\ In addition to the 2020 review's opening ``call for
information'' (79 FR 71764, December 3, 2014), the 2019 ISA
identified and evaluated studies and reports that have undergone
scientific peer review and were published or accepted for
publication between January 1, 2009 through approximately January
2018 (U.S. EPA, 2019a, p. ES-2). References that are cited in the
2019 ISA, the references that were considered for inclusion but not
cited, and electronic links to bibliographic information and
abstracts can be found at: https://hero.epa.gov/hero/particulate-matter.
\144\ As described in more detail in the ISA Supplement, ``the
scope of this Supplement provides specific criteria for the types of
studies considered for inclusion within the Supplement.
Specifically, studies must be peer reviewed and published between
approximately January 2018 and March 2021'' (U.S. EPA, 2022a,
section 1.2.2).
\145\ As described in section 1.2.1 of the ISA Supplement, ``the
selection of welfare effects to evaluate within this Supplement is
based on the causality determinations reported in the 2019 PM ISA
and the subsequent use of scientific evidence in the 2020 PM PA. The
2019 PM ISA concluded a causal relationship for each of the welfare
effects categories evaluated (i.e., visibility, climate effects, and
materials effects). While the 2020 PM PA considered the broader set
of evidence for these effects, for climate effects and material
effects, it concluded that there remained `substantial uncertainties
with regard to the quantitative relationships with PM concentrations
and concentration patterns that limit[ed] [the] ability to
quantitatively assess the public welfare protection provided by the
standards from these effects (U.S. EPA, 2020b). Given these
uncertainties and limitations, the basis of the discussion on
conclusions regarding the secondary standards in the 2020 PM PA
primarily focused on visibility effects. Therefore, this Supplement
focuses only on visibility effects in evaluating newly available
scientific information and is limited to studies conducted in the
U.S. and Canada'' (U.S. EPA, 2022a, section 1.2.1).
---------------------------------------------------------------------------
In presenting the rationale for the Administrator's final decision
and its foundations, section V.A provides background on the 2020 final
decision to retain the secondary PM standards (section V.A.1), and also
provides brief summaries of key aspects of the currently available
welfare effects evidence (section V.A.2) and quantitative information
(section V.A.3). Section V.B summarizes the CASAC's advice (section
V.B.1) and the proposed conclusions (section V.B.2), addresses public
comments received on the proposal (section V.B.3), and presents the
Administrator's conclusions on the adequacy of the current standards
(section V.B.4), drawing on consideration of the available scientific
and quantitative information, advice from the CASAC, and comments from
the public. Section V.C summarizes the Administrator's decision on the
secondary PM standards.
A. Introduction
The general approach for this reconsideration of the 2020 final
decision on the secondary PM standards relies on the EPA's assessments
of the current scientific evidence and associated quantitative analyses
to inform the Administrator's judgments regarding secondary standards
that are requisite to protect the public welfare from known or
anticipated adverse effects associated with the pollutant's presence in
the ambient air. The EPA's assessments are primarily documented in the
2019 ISA, ISA Supplement, and 2022 PA, which builds on the 2020 PA, all
of which have received CASAC review and public comment (83 FR 53471,
October 23, 2018; 83 FR 55529, November 6, 2018; 85 FR 4655, January
27, 2020; 86 FR 52673, September 22, 2021; 86 FR 54186, September 30,
2021; 86 FR 56263, October 8, 2021; 87 FR 958, January 7, 2022; 87 FR
22207, April 14, 2022; 87 FR 31965, May 26, 2022). In bridging the gap
between the scientific assessments of the 2019 ISA and ISA Supplement
and the judgments required of the Administrator in determining whether
the current standards provide the requisite public welfare protection,
the 2022 PA evaluates policy implications of the evaluation of the
current evidence in the 2019 ISA and ISA Supplement, and the
quantitative information documented in the 2022 PA. In evaluating the
public welfare protection afforded by the current standards against PM-
related effects within the scope of this reconsideration, the four
basic elements of the NAAQS (indicator, averaging time, level, and
form) are considered collectively.
The final decision on the adequacy of the current secondary
standards is a public welfare policy judgment to be made by the
Administrator. In reaching conclusions with regard to the standard, the
decision draws on the scientific information and analyses about welfare
effects, and associated public welfare significance, as well as
judgments about how to consider the range and magnitude of
uncertainties that are inherent in the scientific evidence and
analyses. This approach is based on the recognition that the available
evidence generally reflects a continuum that includes ambient air
exposures at which scientists agree that effects are likely to occur
through lower levels at which the likelihood and magnitude of responses
become increasingly uncertain. This approach is consistent with the
requirements of the provisions of the Clean Air Act related to the
review of NAAQS and with how the EPA and the courts have historically
interpreted the Act. These provisions require the Administrator to
establish secondary standards that, in the judgment of the
Administrator, are requisite to protect public welfare from known or
anticipated adverse effects associated with the presence of the
pollutant in the ambient air. In so doing, the Administrator seeks to
establish standards that are neither more nor less stringent than
necessary for this purpose. The Act does not require that standards be
set at a zero-risk level, but rather at a level that reduces risk
sufficiently so as to protect the public welfare from known or
anticipated adverse effects.
1. Background on the Current Standards
The current secondary PM standards were retained in 2020 based on
the scientific and technical information available at that time, as
well as the then-Administrator's judgments regarding the available
welfare effects evidence, the appropriate degree of public welfare
protection for the existing standards, and available air quality
information on visibility impairment that may be allowed by such a
standard (85 FR 82684, December 18, 2020). With the 2020 decision, the
then-Administrator retained the secondary 24-hour PM2.5
standard, with its level of 35 [micro]g/m\3\, the annual
PM2.5 standard, with its level of 15.0 [micro]g/m\3\, and
the 24-hour PM10 standard, with its level of 150 [micro]g/
m\3\. The subsections below focus on the key considerations and the
then-Administrator's conclusions in the 2020 final decision for climate
and materials effects (section V.A.1.a) and visibility effects (section
V.A.2.b).
a. Non-Visibility Effects
In light of the robust evidence base, the 2019 ISA concluded there
to be causal relationships between PM and climate effects and materials
effects (U.S. EPA, 2019a, sections 13.3.9 and 13.4.2). The 2020 final
decision was based on a thorough review in the 2019 ISA of the
scientific information on PM-induced climate and materials effects. The
decision also took into account: (1) Assessments in the 2020 PA of the
most policy-relevant information in the 2019 ISA regarding evidence of
adverse effects of PM to climate and materials, (2) uncertainties in
the available evidence to inform a quantitative assessment of PM-
related climate and materials effects, (3) CASAC advice and
recommendations, and (4) public comments received during the
development of these documents and on the proposal document.
In considering non-visibility welfare effects in the 2020 decision,
the then-Administrator concluded that, while it is important to
maintain an appropriate degree of control of fine and coarse particles
to address non-visibility welfare effects, ``it is generally
appropriate to retain the existing standards and that there is
insufficient information to establish any distinct secondary PM
standards to address climate and materials effects of PM'' (85 FR
82744, December 18, 2020).
With regard to climate, the then-Administrator recognized that
there were a number of improvements and refinements to climate models
since the 2012 review. However, while the evidence continued to support
a causal relationship between PM and climate effects, the then-
Administrator noted that significant limitations continued to exist
related to quantifying the contributions of direct and indirect
[[Page 16313]]
effects of PM and PM components on climate forcing (U.S. EPA, 2020b,
sections 5.2.2.1.1 and 5.4). He also recognized that the models
continued to exhibit considerable variability in estimates of PM-
related climate impacts at regional scales (e.g., ~100 km) as compared
to simulations at global scales. Therefore, the resulting uncertainty
led the then-Administrator to conclude in the 2020 decision that the
available scientific information remained insufficient to quantify
climate impacts associated with particular concentrations of PM in
ambient air (U.S. EPA, 2020b, section 5.2.2.2.1) or to evaluate or
consider a level of PM air quality in the U.S. to protect against
climate effects and that there was insufficient information available
to base a national ambient standard on climate impacts (85 FR 82744,
December 18, 2020).
With regard to materials effects, the then-Administrator noted that
the evidence available in the 2019 ISA continued to support a causal
relationship between materials effects and PM deposition (U.S. EPA,
2019a, section 13.4). He recognized that the deposition of fine and
coarse particles to materials can lead to physical damage and/or
impaired aesthetic qualities. Particles can contribute to materials
damage by adding to the natural weathering processes and by promoting
the corrosion of metals, the degradation of building materials, and the
weakening of material components. While some new information was
available in the 2019 ISA, the information was from studies primarily
conducted outside of the U.S. in areas where PM concentrations in
ambient air are higher than those observed in the U.S. (U.S. EPA,
2020b, section 13.4). Additionally, the information assessed in the
2019 ISA did not support quantitative analyses of PM-related materials
effects in the 2020 PA (U.S. EPA, 2020b, section 5.2.2.2.2). Given the
limited amount of information available and its inherent uncertainties
and limitations, the Administrator concluded that he was unable to
relate soiling or damage to specific levels of PM in ambient air or to
evaluate or consider a level of air quality to protect against such
materials effects, and that there was insufficient information
available to support a distinct national ambient standard based on
materials effects (85 FR 82744, December 18, 2020).
In reviewing the 2019 draft PA, the CASAC agreed with staff
conclusions that, while these effects are important, ``the available
evidence does not call into question the protection afforded by the
current secondary PM standards'' and recommended that the secondary
standards ``should be retained'' (Cox, 2019b, p. 3 of letter). In
reaching a final decision in 2020, for all of the reasons discussed
above and recognizing the CASAC conclusion that the evidence provided
support for retaining the current secondary PM standards, the then-
Administrator concluded that it was appropriate to retain the existing
secondary PM standards, without revision. For climate and materials
effects, this conclusion reflected his judgment that, although it
remains important to maintain secondary PM2.5 and
PM10 standards to provide some degree of control over long-
and short-term concentrations of both fine and coarse particles, there
was insufficient information to establish distinct secondary PM
standards to address non-visibility PM-related welfare effects (85 FR
82744, December 18, 2020).
b. Visibility Effects
The 2019 ISA concluded that, ``the evidence is sufficient to
conclude that a causal relationship exists between PM and visibility
impairment'' (U.S. EPA, 2019a, section 13.2.6). The 2020 decision on
the adequacy of the secondary standards with regard to visibility
effects was a public welfare policy judgment made by the then-
Administrator, which drew upon the available scientific evidence for
PM-related visibility effects and on analyses of visibility impairment,
as well as judgments about the appropriate weight to place on the range
of uncertainties inherent in the evidence and analyses. The 2020 final
decision was based on a thorough review in the 2019 ISA of the
scientific information on PM-related visibility effects. The decision
also took into account: (1) Assessments in the 2020 PA of the most
policy-relevant information in the 2019 ISA regarding evidence of
adverse effects of PM on visibility; (2) air quality analyses of the
PM2.5 visibility index and design values based on the form
and averaging time of the existing secondary 24-hour PM2.5
standard; (3) CASAC advice and recommendations; and (4) public comments
received during the development of these documents and on the 2020
proposal document.
In considering the visibility effects in the 2020 review, the then-
Administrator noted the long-standing body of evidence for PM-related
visibility impairment. This evidence, which is based on the fundamental
relationship between light extinction and PM mass, demonstrated that
ambient PM can impair visibility in both urban and remote areas, and
had changed very little since the 2012 review (U.S. EPA, 2019a, section
13.1; U.S. EPA, 2009a, section 9.2.5). The evidence related to public
perception of visibility impairment was from studies from four areas in
North America.\146\ These studies provided information to inform our
understanding of levels of visibility impairment that the public judged
to be ``acceptable'' (U.S. EPA, 2010b; 85 FR 24131, April 30, 2020). In
considering these public preference studies, the then-Administrator
noted that no new visibility studies conducted in the U.S. were
discussed in the 2019 ISA, and there was little newly available
information with regard to acceptable levels of visibility impairment
in the U.S. The Administrator recognized that visibility impairment can
have implications for people's enjoyment of daily activities and their
overall well-being, and therefore, considered the degree to which the
current secondary standards protect against PM-related visibility
impairment.
---------------------------------------------------------------------------
\146\ Preference studies were available in four urban areas.
Three western preference studies were available, including one in
Denver, Colorado (Ely et al., 1991), one in the lower Fraser River
valley near Vancouver, British Columbia, Canada (Pryor, 1996), and
one in Phoenix, Arizona (BBC Research & Consulting, 2003). A pilot
focus group study was also conducted for Washington, DC (Abt
Associates, 2001), and a replicate study with 26 participants was
also conducted for Washington, DC (Smith and Howell, 2009). More
details about these studies are available in Appendix D of the 2022
PA (U.S. EPA, 2022b).
---------------------------------------------------------------------------
Consistent with the 2012 review, in the 2020 review, the then-
Administrator first concluded that a target level of protection for a
secondary PM standard is most appropriately defined in terms of a
visibility index that directly takes into account the factors (i.e.,
species composition and relative humidity) that influence the
relationship between PM2.5 in ambient air and PM-related
visibility impairment. In defining a target level of protection, the
then-Administrator considered the specific aspects of such an index,
including the appropriate indicator, averaging time, form and level (78
FR 82742-82744, December 18, 2020).
First, with regard to indicator, the then-Administrator noted that
in the 2012 review, the EPA used an index based on estimates of light
extinction by PM2.5 components calculated using an adjusted
version of the IMPROVE algorithm, which allows the estimation of the
light extinction using routinely monitored components of
PM2.5, PM10-2.5 mass, and estimates of relative
humidity. The then-Administrator recognized that, while there have been
some revisions to the IMPROVE algorithm since the time of the 2012
[[Page 16314]]
review, our fundamental understanding of the relationship between PM in
ambient air and light extinction had changed little and the various
IMPROVE algorithms appropriately reflected this relationship across the
U.S. In the absence of a monitoring network for direct measurement of
light extinction, he concluded that a calculated light extinction
indicator that utilizes the IMPROVE algorithms continued to provide a
reasonable basis for defining a target level of protection against PM-
related visibility impairment (78 FR 82742-82744, December 18, 2020).
In further defining the characteristics of a visibility index, the
then-Administrator next considered the appropriate averaging time,
form, and level of the index. Given the available scientific
information the review, and in considering the CASAC's advice and
public comments, the then-Administrator concluded that, consistent with
the decision in the 2012 review, a visibility index with a 24-hour
averaging time and a form based on the 3-year average of annual 90th
percentile values remained reasonable. With regard to the averaging
time and form of such an index, the Administrator noted analyses
conducted in the last review that demonstrated relatively strong
correlations between 24-hour and subdaily (i.e., 4-hour average)
PM2.5 light extinction (78 FR 3226, January 15, 2013),
indicating that a 24-hour averaging time is an appropriate surrogate
for the subdaily time periods of the perception of PM-related
visibility impairment and the relevant exposure periods for segments of
the viewing public. This decision in the 2020 review also recognized
that a 24-hour averaging time may be less influenced by atypical
conditions and/or atypical instrument performance (78 FR 3226, January
15, 2013). The then-Administrator recognized that there was no new
information to support updated analyses of this nature, and therefore,
he believed these analyses continued to provide support for
consideration of a 24-hour averaging time for a visibility index in
this review. With regard to the statistical form of the index, the
Administrator noted that, consistent with the 2012 review: (1) A multi-
year percentile form offers greater stability from the occasional
effect of interannual meteorological variability (78 FR 3198, January
15, 2013; U.S. EPA, 2011, p. 4-58); (2) a 90th percentile represents
the median of the distribution of the 20 percent worst visibility days,
which are targeted in Federal Class I areas by the Regional Haze
Program; and (3) public preference studies did not provide information
to identify a different target than that identified for Federal Class I
areas (U.S. EPA, 2011, p. 4-59). Therefore, the then-Administrator
judged that a visibility index based on estimates of light extinction,
with a 24-hour averaging time and a 90th percentile form, averaged over
three years, remained appropriate (78 FR 82742-82744, December 18,
2020).
With regard to the level of a visibility index, consistent with the
2012 review, the then-Administrator judged that it was appropriate to
establish a target level of protection of 30 deciviews
(dv),147 148 reflecting the upper end of the range of
visibility impairment judged to be acceptable by at least 50% of study
participants in the available public preference studies (78 FR 3226,
January 15, 2013). The 2011 PA identified a range of levels from 20 to
30 dv based on the responses in the public preference studies available
at that time (U.S. EPA, 2011, section 4.3.4). At the time of the 2012
review, the then-Administrator noted a number of uncertainties and
limitations in public preference studies, including the small number of
stated preference studies available, the relatively small number of
study participants, the extent to which the study participants may not
be representative of the broader study area population in some of the
studies, and the variations in the specific materials and methods used
in each study. In considering the available preference studies in 2012,
with their inherent uncertainties and limitations, the then-
Administrator concluded that the substantial degree of variability and
uncertainty in the public preference studies should be reflected in a
target level of protection based on the upper end of the range of
candidate protection levels (CPLs).
---------------------------------------------------------------------------
\147\ Deciview (dv) refers to a scale for characterizing
visibility that is defined directly in terms of light extinction.
The deciview scale is frequently used in the scientific and
regulatory literature on visibility.
\148\ For comparison, 20 dv, 25 dv, and 30 dv are equivalent to
64, 112, and 191 megameters (Mm-1), respectively.
---------------------------------------------------------------------------
Given that there were no new preference studies in the 2019 ISA,
the then-Administrator's judgments in 2020 were based on the same
studies, with the same range of levels, available in the 2012 review.
The 2020 PA (U.S. EPA, 2020b, section 5.5), discussed a number of
limitations and uncertainties associated with these studies. In
considering the scientific information, with its uncertainties and
limitations, as well as public comments on the level of the target
level of protection against visibility impairment, the then-
Administrator concluded that it was appropriate to again use a level of
30 dv for the visibility index (78 FR 82742-82744, December 18, 2020).
Having concluded that the protection provided by a standard defined
in terms of a PM2.5 visibility index, with a 24-hour
averaging time, and a 90th percentile form, averaged over 3 years, set
at a level of 30 dv, was requisite to protect public welfare with
regard to visual air quality, the Administrator next considered the
degree of protection from visibility impairment afforded by the
existing suite of secondary PM standards.
In this context, the then-Administrator considered the updated
analyses of visibility impairment presented in the 2020 PA (U.S. EPA,
2020b, section 5.2.1.2), which reflected a number of improvements since
the 2012 review. Specifically, the updated analyses examined multiple
versions of the IMPROVE equation, including the version incorporating
revisions since the time of the 2012 review. These updated analyses
provided a further understanding of how variation in the inputs to the
algorithms affect the estimates of light extinction (U.S. EPA, 2020b,
Appendix D). Additionally, for a subset of monitoring sites with
available PM10-2.5 data, the updated analyses better
characterized the influence of coarse PM on light extinction than in
the 2012 review (U.S. EPA, 2020b, section 5.2.1.2).
The results of the updated analyses in the 2020 PA were consistent
with those from the 2012 review. Regardless of which version of the
IMPROVE equation was used, the analyses demonstrated that, based on
2015-2017 data, the 3-year visibility metric was at or below about 30
dv in all areas meeting the current 24-hour PM2.5 standard,
and below 25 dv in most of those areas. In locations with available
PM10-2.5 monitoring, which met both the current 24-hour
secondary PM2.5 and PM10 standards, 3-year
visibility index metrics were at or below 30 dv regardless of whether
the coarse fraction was included as an input to the algorithm for
estimating light extinction (U.S. EPA, 2020b, section 5.2.1.2). While
the inclusion of the coarse fraction had a relatively modest impact on
the estimates of light extinction, the then-Administrator recognized
the continued importance of the PM10 standard given the
potential for larger impacts on light extinction in areas with higher
coarse particle concentrations, which were not included in the analyses
in the 2020 PA due to a lack of available data (U.S. EPA, 2019a,
section 13.2.4.1; U.S. EPA, 2020b, section 5.2.1.2). He
[[Page 16315]]
noted that the air quality analyses showed that all areas meeting the
existing 24-hour PM2.5 standard, with its level of 35
[micro]g/m\3\, had visual air quality at least as good as 30 dv, based
on the visibility index. Thus, the secondary 24-hour PM2.5
standard would likely be controlling relative to a 24-hour visibility
index set at a level of 30 dv. Additionally, areas would be unlikely to
exceed the target level of protection for visibility of 30 dv without
also exceeding the existing secondary 24-hour PM2.5
standard. Thus, the then-Administrator judged that the 24-hour
PM2.5 standard provided sufficient protection in all areas
against the effects of visibility impairment, i.e., that the existing
24-hour PM2.5 standard would provide at least the target
level of protection for visual air quality of 30 dv which he judged
appropriate (78 FR 82742-82744, December 18, 2020).
2. Overview of Welfare Effects Evidence
The information summarized here is based on the scientific
assessment of the welfare effects evidence available in this
reconsideration; this assessment is documented in the 2019 ISA and ISA
Supplement and its policy implications are further discussed in the
2022 PA. While the 2019 ISA provides the broad scientific foundation
for this reconsideration, additional literature has become available
since the cutoff date of the 2019 ISA that expands the body of evidence
related to visibility effects that can inform the Administrator's
judgment on the adequacy of the current secondary PM standards. As
such, the ISA Supplement builds on the information in the 2019 ISA with
a targeted identification and evaluation of new scientific information
regarding visibility effects. As described in the ISA Supplement and
the 2022 PA, the selection of welfare effects to evaluate within the
ISA Supplement were based on the causality determinations reported in
the 2019 ISA and the subsequent use of scientific evidence in the 2020
PA (U.S. EPA, 2019a, section 1.2; U.S. EPA, 2022a, section 1.4.2). The
ISA Supplement focuses on U.S. and Canadian studies that provide new
information on public preferences for visibility impairment and/or
developed new methodologies or conducted quantitative analyses of light
extinction (U.S. EPA, 2022a, section 1.2). Such studies of visibility
effects and quantitative relationships between visibility impairment
and PM in ambient air were considered to be of greatest utility in
informing the Administrator's conclusions on the adequacy of the
current secondary PM standards. The visibility effects evidence
presented within the 2019 ISA, along with the targeted identification
and evaluation of new scientific information in the ISA Supplement,
provides the scientific basis for the reconsideration of the 2020 final
decision on the secondary PM standards for visibility effects. For
climate and materials effects, the 2020 PA concluded that there were
substantial uncertainties associated with the quantitative
relationships with PM concentrations and the concentration patterns
that limited the ability to quantitatively assess the public welfare
protection provided by the standards from these effects. Therefore, the
evaluation of the information related to these effects draws heavily
from the 2019 ISA and 2020 PA. The subsections below briefly summarize
the nature of PM-related visibility (section V.B.1.a), climate (section
V.B.1.b), and materials (section V.B.1.c) effects.
a. Nature of Effects
Visibility impairment can have implications for people's enjoyment
of daily activities and for their overall sense of well-being (U.S.
EPA, 2009a, section 9.2). The strongest evidence for PM-related
visibility impairment comes from the fundamental relationship between
light extinction and PM mass (U.S. EPA, 2009a), which confirms a well-
established ``causal relationship exists between PM and visibility
impairment'' (U.S. EPA, 2009a, p. 2-28). Beyond its effects on
visibility, the 2009 ISA also identified a causal relationship
``between PM and climate effects, including both direct effects of
radiative forcing and indirect effects that involve cloud and feedbacks
that influence precipitation formation and cloud lifetimes'' (U.S. EPA,
2009a, p. 2-29). The evidence also supports a causal relationship
between PM and effects on materials, including soiling effects and
materials damage (U.S. EPA, 2009a, p. 2-31).
The evidence available in this reconsideration is consistent with
the evidence available at the time of the 2012 and 2020 reviews and
supports the conclusions of causal relationships between PM and
visibility, climate, and materials effects (U.S. EPA, 2019a, chapter
13). Evidence newly available in this reconsideration augments the
previously available evidence of the relationship between PM and
visibility impairment (U.S. EPA, 2019a, section 13.2; U.S. EPA, 2022a,
section 4), climate effects (U.S. EPA, 2019a, section 13.3), and
materials effects (U.S. EPA, 2019a, section 13.4).
i. Visibility
The fundamental relationship between light extinction and PM mass,
and the EPA's understanding of this relationship, has changed little
since the 2009 ISA (U.S. EPA, 2009a). The combined effect of light
scattering and absorption by particles and gases is characterized as
light extinction, i.e., the fraction of light that is scattered or
absorbed per unit of distance in the atmosphere.\149\ Light extinction
is measured in units of 1/distance, which is often expressed in the
technical literature as visibility per megameter (abbreviated
Mm-1). Higher values of light extinction (usually given in
units of Mm-1 or dv) correspond to lower visibility. When PM
is present in the air, its contribution to light extinction is
typically much greater than that of gases (U.S. EPA, 2019a, section
13.2.1). The impact of PM on light scattering depends on particle size
and composition, as well as relative humidity. All particles scatter
light, as described by the Mie theory, which relates light scattering
to particle size, shape, and index of refraction (U.S. EPA, 2019a,
section 13.2.3; Mie, 1908, Van de Hulst, 1981). Fine particles scatter
more light than coarse particles on a per unit mass basis and include
sulfates, nitrates, organics, light-absorbing carbon, and soil (Malm et
al., 1994). Hygroscopic particles like ammonium sulfate, ammonium
nitrate, and sea salt increase in size as relative humidity increases,
leading to increased light scattering (U.S. EPA, 2019a, section
13.2.3).
---------------------------------------------------------------------------
\149\ All particles scatter light and, although a larger
particle scatters more light than a similarly shaped smaller
particle of the same composition, the light scattered per unit of
mass is greatest for particles with diameters from ~0.3-1.0 [micro]m
(U.S. EPA, 2009a, section 2.5.1; U.S. EPA, 2019a, section 13.2.1).
Particles with hygroscopic components (e.g., particulate sulfate and
nitrate) contribute more to light extinction at higher relative
humidity than at lower relative humidity because they change size in
the atmosphere in response to relative humidity.
---------------------------------------------------------------------------
As at the time of the 2012 and 2020 reviews, direct measurements of
PM light extinction, scattering, and absorption continue to be
considered more accurate for quantifying visibility than PM mass-based
estimates because measurements do not depend on assumptions about
particle characteristics (e.g., size, shape, density, component
mixture, etc.) (U.S. EPA, 2019a, section 13.2.2.2). Measurements of
light extinction can be made with high time resolution, allowing for
characterization of subdaily temporal patterns of visibility
impairment. A number of measurement methods have been used for
visibility impairment (e.g.,
[[Page 16316]]
transmissometers, integrating nephelometers, teleradiometers,
telephotometers, and photography and photographic modeling), although
each of these methods has its own strengths and limitations (U.S. EPA,
2019a, Table 13-1). While some recent research confirms and adds to the
body of knowledge regarding direct measurements as is described in the
2019 ISA and ISA Supplement, no major new developments have been made
with these measurement methods since prior reviews (U.S. EPA, 2019a,
section 13.2.2.2; U.S. EPA, 2022a, section 4.2).
In the absence of a robust monitoring network for the routine
measurement of light extinction across the U.S., estimation of light
extinction based on existing PM monitoring can be used. The theoretical
relationship between light extinction and PM characteristics, as
derived from Mie theory (U.S. EPA, 2019a, Equation 13.5), can be used
to estimate light extinction by combining mass scattering efficiencies
of particles with particle concentrations (U.S. EPA, 2019a, section
13.2.3; U.S. EPA, 2009a, sections 9.2.2.2 and 9.2.3.1). This estimation
of light extinction is consistent with the method used in previous
reviews. The algorithm used to estimate light extinction, known as the
IMPROVE algorithm,\150\ provides for the estimation of light extinction
(bext), in units of Mm\-1\, using routinely monitored
components of fine (PM2.5) and coarse (PM10-2.5)
PM. Relative humidity data are also needed to estimate the contribution
by liquid water that is in solution with the hygroscopic components of
PM. To estimate each component's contribution to light extinction,
their concentrations are multiplied by extinction coefficients and are
additionally multiplied by a water growth factor that accounts for
their expansion with moisture. Both the extinction efficiency
coefficients and water growth factors of the IMPROVE algorithm have
been developed by a combination of empirical assessment and theoretical
calculation using particle size distributions associated with each of
the major aerosol components (U.S. EPA, 2019a, sections 13.2.3.1 and
13.2.3.3).
---------------------------------------------------------------------------
\150\ The algorithm is referred to as the IMPROVE algorithm as
it was developed specifically to use monitoring data generated at
IMPROVE network sites and with equipment specifically designed to
support the IMPROVE program and was evaluated using IMPROVE optical
measurements at the subset of monitoring sites that make those
measurements (Malm et al., 1994).
---------------------------------------------------------------------------
At the time of the 2012 review, two versions of the IMPROVE
algorithm were available in the literature--the original IMPROVE
algorithm (Lowenthal and Kumar, 2004, Malm and Hand, 2007, Ryan et al.,
2005) and the revised IMPROVE algorithm (Pitchford et al., 2007). As
described in detail in the 2022 PA (U.S. EPA, 2022b, section 5.3.1.1)
and the 2019 ISA (U.S. EPA, 2019a, section 13.2.3), the algorithm has
been further evaluated and refined since the time of the 2012 review
(Lowenthal and Kumar, 2016), particularly for PM characteristics and
relative humidity in remote areas. All three versions of the IMPROVE
algorithm were considered in evaluating visibility impairment in this
reconsideration.
Consistent with the evidence available at the time of the 2012 and
2020 reviews, our understanding of public perception of visibility
impairment comes from visibility preference studies conducted in four
areas in North America.\151\ The detailed methodology for these studies
are described in the 2022 PA (U.S. EPA, 2022b, section 5.3.1.1), the
2019 ISA (U.S. EPA, 2019a), and the 2009 ISA (U.S. EPA, 2019a). In
summary, the study participants were queried regarding multiple images
that were either photographs of the same location and scenery that had
been taken on different days on which measured extinction data were
available or digitized photographs onto which a uniform ``haze'' had
been superimposed. Results of the studies indicated a wide range of
judgments on what study participants considered to be acceptable
visibility across the different study areas, depending on the setting
depicted in each photograph. Based on the results of the four cities, a
range encompassing the PM2.5 visibility index values from
images that were judged to be acceptable by at least 50 percent of
study participants across all four of the urban preference studies was
identified (U.S. EPA, 2010b, p. 4-24; U.S. EPA, 2020b, Figure 5-2).
Much lower visibility (considerably more haze resulting in higher
values of light extinction) was considered acceptable in Washington,
DC, than was in Denver, and 30 dv reflected the level of impairment
that was determined to be ``acceptable'' by at least 50 percent of
study participants (78 FR 3226-3227, January 15, 2013).
---------------------------------------------------------------------------
\151\ Preference studies were available in four urban areas in
the last review: Denver, Colorado (Ely et al., 1991), Vancouver,
British Columbia, Canada (Pryor, 1996), Phoenix, Arizona (BBC
Research & Consulting, 2003), and Washington, DC (Abt Associates,
2001; Smith and Howell, 2009).
---------------------------------------------------------------------------
Since the completion of the 2009 and 2019 ISAs, there has been only
one public preference study that has become available in the U.S. This
study uses images of the Grand Canyon, AZ, described in the ISA
Supplement (U.S. EPA, 2022a). The Grand Canyon study, conducted by Malm
et al. (2019), has a similar study design to that used in the public
preference studies discussed above; however, there are several
important differences that make it difficult to directly compare the
results of the Malm et al. (2019) study with other public preference
studies. As an initial matter, the Grand Canyon study was conducted in
a Federal Class I area, as opposed to in an urban area, with a scene
depicted in the photographs that did not include urban features.\152\
We recognize that public preferences with respect to visibility in
Federal Class 1 areas may well differ from visibility preferences in
urban areas and other contexts, although there is currently a lack of
information to on such questions. Further, the Malm et al. (2019) study
also used a much lower range of superimposed ``haze'' than the
preference studies discussed above.\153\ It is unclear whether the
participant preferences are a function in part of the range of
potential values presented, such that the participant preferences for
the Grand Canyon were generally lower \154\ than the other preference
studies in part because of the lower range of superimposed ``haze'' for
the images in that study, or if their preferences would vary if
presented with images with a range of superimposed ``haze'' more
comparable to the levels used in the other studies (i.e., more ``haze''
superimposed on the images).
---------------------------------------------------------------------------
\152\ The Grand Canyon study used a single scene looking west
down the canyon with a small landscape feature of a 100-km-distant
mountain (Mount Trumbull), along with other closer landscape
features. The scenes presented in the previously available
visibility preference studies are presented in more detail in Table
D-9 in the 2022 PA (U.S. EPA, 2022b, Appendix D).
\153\ The Grand Canyon study superimposed light extinction
ranging from 3 dv to 20 dv on the image slides shown to participants
compared to the previously available preference studies. In those
studies, the visibility ranges presented were as low as 9 dv and as
high as 45 dv. The visibility ranges presented in the previously
available visibility preference studies are described in more detail
in Table D-9 in the 2022 PA (U.S. EPA, 2022b, Appendix D).
\154\ In the Grand Canyon study, the level of impairment that
was determined to be ``acceptable'' by at least 50 percent of study
participants was 7 dv (Malm et al., 2019).
---------------------------------------------------------------------------
The Malm et al. (2019) study also explored alternate methods for
evaluating ``acceptable'' levels of visual air quality from the
preference studies, including the use of scene-specific visibility
indices as potential indicators of visibility levels as perceived by
the observer (Malm et al., 2019). In addition to measures of
atmospheric haze, such
[[Page 16317]]
as atmospheric extinction, used in previously available preference
studies, other indices for visual air quality include color and
achromatic contrast of single landscape figures, average and equivalent
contrast of an entire scene, edge detection algorithms such as the
Sobel index, and just-noticeable difference or change indexes. The
results reported by Malm et al. (2019) suggest that scene-dependent
metrics, such as contrast, may be useful alternate predictors of
preference levels compared to universal metrics like light extinction
(U.S. EPA, 2022a, section 4.2.1). This is because extinction alone is
not a measure of ``haze,'' but of light attenuation per unit distance,
and visible ``haze'' is dependent on both light extinction and distance
to a landscape feature (U.S. EPA, 2022a, section 4.2.1). However, there
are very few studies available that use scene-dependent metrics (i.e.,
contrast) to evaluate public preference information, which makes it
difficult to evaluate them as an alternative to the light extinction
approach.
ii. Climate
The available evidence continues to support the conclusion of a
causal relationship between PM and climate effects (U.S. EPA, 2019a,
section 13.3.9). Since the 2012 review, climate impacts have been
extensively studied and recent research reinforces and strengthens the
evidence evaluated in the 2009 ISA. Recent evidence provides greater
specificity about the details of radiative forcing effects \155\ and
increases the understanding of additional climate impacts driven by PM
radiative effects. The Intergovernmental Panel on Climate Change (IPCC)
assesses the role of anthropogenic activity in past and future climate
change, and since the completion of the 2009 ISA, has issued the Fifth
IPCC Assessment Report (AR5; IPCC, 2013), which summarizes any key
scientific advances in understanding the climate effects of PM since
the previous report. As in the 2009 ISA, the 2019 ISA draws
substantially on the IPCC report to summarize climate effects. As
discussed in more detail in the 2022 PA (U.S. EPA, 2022b, section
5.3.2.1.1), the general conclusions are similar between the IPCC AR4
and AR5 reports with regard to effects of PM on global climate.
Consistent with the evidence available in the 2012 review, the key
components, including sulfate, nitrate, organic carbon (OC), black
carbon (BC), and dust, that contribute to climate processes vary in
their reflectivity, forcing efficiencies, and direction of forcing.
Since the completion of the 2009 ISA, the evidence base has expanded
with respect to the mechanisms of climate responses and feedbacks to PM
radiative forcing; however, the recently published literature assessed
in the 2019 ISA does not reduce the considerable uncertainties that
continue to exist related these mechanisms.
---------------------------------------------------------------------------
\155\ Radiative forcing (RF) for a given atmospheric constituent
is defined as the perturbation in net radiative flux, at the
tropopause (or the top of the atmosphere) caused by that
constituent, in watts per square meter (Wm\-2\), after allowing for
temperatures in the stratosphere to adjust to the perturbation but
holding all other climate responses constant, including surface and
tropospheric temperatures (Fiore et al., 2015; Myhre et al., 2013).
A positive forcing indicates net energy trapped in the Earth system
and suggests warming of the Earth's surface, whereas a negative
forcing indicates net loss of energy and suggests cooling (U.S. EPA,
2019a, section 13.3.2.2).
---------------------------------------------------------------------------
As described in the proposal (88 FR 5650, January 27, 2023), PM has
a very heterogeneous distribution globally and patterns of forcing tend
to correlate with PM loading, with the greatest forcings centralized
over continental regions. The climate response to this PM forcing,
however, is more complicated since the perturbation to one climate
variable (e.g., temperature, cloud cover, precipitation) can lead to a
cascade of effects on other variables. While the initial PM radiative
forcing may be concentrated regionally, the eventual climate response
can be much broader spatially or be concentrated in remote regions, and
may be quite complex, affecting multiple climate variables with
possible differences in the direction of the forcing in different
regions or for different variables (U.S. EPA, 2019a, section 13.3.6).
The complex climate system interactions lead to variation among climate
models, which have suggested a range of factors that can influence
large-scale meteorological processes and may affect temperature,
including local feedback effects involving soil moisture and cloud
cover, changes in the hygroscopicity of the PM, and interactions with
clouds (U.S. EPA, 2019a, section 13.3.7). As a result, there remains
insufficient evidence to related climate effects to specific PM levels
in ambient air or to establish a quantitative relationship between PM
and climate effects, particularly at a regional scale. Further research
is needed to better characterize the effects of PM on regional climate
in the U.S. before PM climate effects can be quantified.
iii. Materials
Consistent with the evidence assessed in the 2009 ISA, the
available evidence continues to support the conclusion that there is a
causal relationship between PM deposition and materials effects.
Effects of deposited PM, particularly sulfates and nitrates, to
materials include both physical damage and impaired aesthetic
qualities, generally involving soiling and/or corrosion (U.S. EPA,
2019a, section 13.4.2). Because of their electrolytic, hygroscopic, and
acidic properties and their ability to sorb corrosive gases, particles
contribute to materials damage by adding to the effects of natural
weathering processes, by potentially promoting or accelerating the
corrosion of metals, degradation of painted surfaces, deterioration of
building materials, and weakening of material components.\156\ There is
a limited amount of recently available data for consideration in this
review from studies primarily conducted outside of the U.S. on
buildings and other items of cultural heritage. However, these studies
involved concentrations of PM in ambient air greater than those
typically observed in the U.S. (U.S. EPA, 2019a, section 13.4).
---------------------------------------------------------------------------
\156\ As discussed in the 2019 ISA (U.S. EPA, 2019a, section
13.4.1), corrosion typically involves reactions of acidic PM (i.e.,
acidic sulfate or nitrate) with material surfaces, but gases like
SO2 and nitric acid (HNO3) also contribute.
Because ``the impacts of gaseous and particulate N and S wet
deposition cannot be clearly distinguished'' (U.S. EPA, 2019a, p.
13-1), the assessment of the evidence in the 2019 ISA considers the
combined impacts.
---------------------------------------------------------------------------
Building on the evidence available in the 2009 ISA, and as
described in detail in the proposal (88 FR 5650, January 27, 2023) and
in the 2019 ISA (U.S. EPA, 2019a, section 13.4), research has
progressed on (1) the theoretical understanding of soiling of items of
cultural heritage; (2) the quantification of degradation rates and
further characterization of factors that influence damage of stone
materials; (3) materials damage from PM components besides sulfate and
black carbon and atmospheric gases besides SO2; (4) methods
for evaluating soiling of materials by PM mixtures; (5) PM-attributable
damage to other materials, including glass and photovoltaic panels; (6)
development of dose-response relationships for soiling of building
materials; and (7) damage functions to quantify material decay as a
function of pollutant type and load. While the evidence of PM-related
materials effects has expanded somewhat since the completion of the
2009 ISA, there remains insufficient evidence to relate soiling or
damage to specific PM levels in ambient air or to establish a
quantitative relationship between PM and materials degradation. The
recent evidence assessed in the 2019 ISA is generally similar to the
evidence available in the 2009 ISA, including
[[Page 16318]]
associated limitations and uncertainties and a lack of evidence to
inform quantitative relationships between PM and materials effects,
therefore leading to similar conclusions about the PM-related effects
on materials.
3. Summary of Air Quality and Quantitative Information
Beyond the consideration of the scientific evidence, as discussed
in section V.A.2 above, quantitative analyses of PM air quality, when
available, can also inform conclusions on the adequacy of the public
welfare protection provided by the current secondary PM standards.
a. Visibility Effects
In the 2012 and 2020 reviews, quantitative analyses for PM-related
visibility effects focused on daily visibility impairment, given the
short-term nature of PM-related visibility effects. The evidence and
information available in this reconsideration continues to provide
support for the short-term (i.e., hourly or daily) nature of PM-related
visibility impairment. As such, the quantitative analyses presented in
the 2022 PA continue to focus on daily visibility impairment and
utilize a two-phase assessment approach for visibility impairment,
consistent with the approaches taken in past reviews. First, the 2022
PA considers the appropriateness of the elements (indicator, averaging
time, form, and level) of the visibility index for providing protection
against PM-related visibility effects. Second, recent air quality was
used to evaluate the relationship between the current secondary 24-hour
PM2.5 standard and the visibility index. The information
available since the 2012 review includes an updated equation for
estimating light extinction, summarized in the 2022 PA (U.S. EPA,
2022b, section 5.3.1.1) and described in the 2019 ISA (U.S. EPA, 2019a,
section 13.2.3.3), as well as more recent air monitoring data, that
together allow for development of an updated assessment of PM-related
visibility impairment in study locations in the U.S.
i. Target Level of Protection in Terms of a PM2.5
Visibility Index
In evaluating the adequacy of the current secondary PM standards,
the 2022 PA first evaluates the appropriateness of the elements
(indicator, averaging time, form, and level) identified for a
visibility index to protect against visibility effects. In previous
reviews, the visibility index as set at a level of 30 dv, with
estimated light extinction as the indicator, a 24-hour averaging time,
and a 90th percentile form, averaged over three years.
With regard to an indicator for the visibility index, the 2022 PA
recognizes the lack of availability of methods and an established
network for directly measuring light extinction (U.S. EPA, 2022b,
section 5.3.1.1). Therefore, consistent with previous reviews, the 2022
PA concludes that a visibility index based on estimates of light
extinction by PM2.5 components derived from an adjusted
version of the original IMPROVE algorithm to be the most appropriate
indicator for the visibility index in this reconsideration. As
described in section 5.3.1.1 of the 2022 PA, the IMPROVE algorithm
estimates light extinction using routinely monitored components of
PM2.5 and PM10-2.5, along with estimates of
relative humidity (U.S. EPA, 2022b, section 5.3.1.1).
With regard to averaging time, the 2022 PA notes that the evidence
continues to provide support for the short-term nature of PM-related
visibility effects. Given that there is no new information available
regarding the time periods during which visibility impairment occurs or
public preferences related to specific time periods for visibility
impairment, the 2022 PA concludes that it is appropriate to continue to
focus on daily visibility impairment. In so doing, the 2022 PA relies
on analyses that were conducted in the 2012 review that showed
relatively strong correlations between 24-hour and subdaily (i.e., 4-
hour average) PM2.5 light extinction that indicated that a
24-hour averaging time is an appropriate surrogate for the subdaily
time periods relevant for visual perception (U.S. EPA, 2011, Figures G-
4 and G-5; Frank, 2012). These analyses continue to provide support for
a 24-hour averaging time for the visibility index in this
reconsideration. Consistent with previous reviews, the 2022 PA also
notes that the 24-hour averaging time may be less influenced by
atypical conditions and/or atypical instrument performance than a
subdaily averaging time (85 FR 82740, December 18, 2020; 78 FR 3226,
January 15, 2013).
With regard to the form for the visibility index, the available
information continues to provide support for a 3-year average of annual
90th percentile values. Given that there is no new information to
inform selection of an alternate form, as in previous reviews, the 2022
PA notes that the 3-year average form provides stability from the
occasional effect of inter-annual meteorological variability that can
result in unusually high pollution levels for a particular year (85 FR
82741, December 18, 2020; 78 FR 3198, January 15, 2013; U.S. EPA, 2011,
p. 4-58). In so doing, the 2022 PA considers the evaluation in the 2010
Urban-Focused Visibility Assessment (UFVA) of three different
statistical forms: 90th, 95th, and 98th percentiles (U.S. EPA, 2010b,
Chapter 4).). In considering this evaluation of statistical forms from
the 2010 UFVA, consistent with the 2011 PA, the 2022 PA notes that the
Regional Haze Program targets the 20 percent most impaired days for
visibility improvements in visual air quality in Federal Class I areas
and that the median of the distribution of these 20 percent most
impaired days would be the 90th percentile. The 2011 PA also noted that
strategies that are implemented so that 90 percent of days would have
visual air quality that is at or below the level of the visibility
index would reasonably be expected to lead to improvements in visual
air quality for the 20 percent most impaired days. Additionally, as in
the 2011 PA, the 2022 PA recognizes that the available public
preference studies do not address frequency of occurrence of different
levels of visibility (U.S. EPA, 2022b, section 5.3.1.2). Therefore, the
analyses and consideration for the form of a visibility index from the
2011 PA continue to provide support for a 90th percentile form,
averaged across three years, in defining the characteristics of a
visibility index in this reconsideration.
With regard to the level for the visibility index, the 2022 PA
recognizes that there is an additional public preference study (Malm et
al., 2019) available in this reconsideration. As noted above, however,
this study differs from the previously available public preference
studies in several ways, which makes it difficult to integrate this
newly available study with the previously available studies. Most
significantly, this study was evaluated public preferences for
visibility in the Grand Canyon, perhaps the most notable Class I area
in the country for visibility purposes. Therefore, the 2022 PA
concludes that the Grand Canyon study is not directly comparable to the
other available preferences studies and public preferences of
visibility impairment in the Malm et al. (2019) study are not
appropriate to consider in identifying a range of levels for the target
level of protection against visibility impairment for this
reconsideration of the secondary PM NAAQS.
[[Page 16319]]
Therefore, the 2022 PA continues to rely on the same studies \157\
and the range of 20 to 30 dv identified from those studies in previous
reviews. With regard to selecting the appropriate target level of
protection for visibility impairment within this range, the 2022 PA
notes that in previous reviews, a level at the upper end of the range
(i.e., 30 dv) was selected given the uncertainties and limitations
associated with the public preference studies (U.S. EPA, 2022b, section
5.3.1.1). However, the 2022 PA also recognizes that (1) the degree of
protection provided by a secondary PM NAAQS is not determined solely by
any one element of the standard but by all elements (i.e., indicator,
averaging time, form, and level) being considered together, and (2)
decisions regarding the adequacy of the current secondary standards is
a public welfare policy judgment to be made by the Administrator. As
such, the Administrator may judge that a target level of protection
below the upper end of the range (i.e., less than 30 dv) is
appropriate, depending on his public welfare policy judgments, which
draw upon the available scientific evidence for PM-related visibility
effects and on analyses of visibility impairment, as well as judgments
about the appropriate weight to place on the range of uncertainties
inherent in the evidence and analyses.
---------------------------------------------------------------------------
\157\ As noted above, the available public preference studies
include those conducted in Denver, Colorado (Ely et al., 1991),
Vancouver, British Columbia, Canada (Pryor, 1996), Phoenix, Arizona
(BBC Research & Consulting, 2003), and Washington, DC (Abt
Associates, 2001; Smith and Howell, 2009).
---------------------------------------------------------------------------
In considering the available public preference studies, consistent
with past reviews, the 2022 PA concludes that it is reasonable to
consider a range of 20 to 30 dv for selecting a target level of
protection, including a high value of 30 dv, a midpoint value of 25 dv,
and a low value of 20 dv. A target level of protection at or in the
upper end of the range would focus on the Washington, DC, preference
study results (Abt Associates, 2001; Smith and Howell, 2009), which
identified 30 dv as the level of impairment that was determined to be
``acceptable'' by at least 50 percent of study participants. The public
preferences of visibility impairment in the Washington, DC, study are
likely to be generally representative of urban areas that do not have
valued scenic elements (e.g., mountains) in the distant background.
This would be more representative of areas in the middle of the country
and many areas in the eastern U.S., as well as possibly some areas in
the western U.S.
A target level of protection in the middle of the range would be
most closely associated with the level of impairment that was
determined to be ``acceptable'' by at least 50 percent of study
participants in the Phoenix, AZ, study (BBC Research & Consulting,
2003), which was 24 dv. This study, while methodologically similar to
the other public preference studies, included participants that were
selected as a representative sample of the Phoenix area population
\158\ and used computer-generated images to depict specific uniform
visibility impairment conditions. This study yielded the best results
of the four public preference studies in terms of the least noisy
preference results and the most representative selection of
participants. Therefore, based on this study, the use of 25 dv to
represent a midpoint within the range of target levels protection is
well supported.
---------------------------------------------------------------------------
\158\ The other preference studies did not include populations
that were necessarily representative of the population in the area
for which the images being judged. For example, in the Denver, CO,
study, participants were from intact groups (i.e., those who were
meeting for other reasons) and were asked to provide a period of
time during a regularly scheduled meeting to participate in the
study (Ely et al., 1991). As another example, in the British
Columbia, Canada, study, participants were recruited from
undergraduate and graduate students enrolled in classes at the
University of British Columbia's Department of Geography (Pryor,
1996).
---------------------------------------------------------------------------
A target level of protection at or just above the lower end of the
range would focus on the Denver, CO, study, but may not be as strongly
supported as higher levels within the range (Ely et al., 1991). Older
studies, such as those conducted in Denver, CO (Ely et al., 1991), and
British Columbia, Canada (Pryor, 1996), used photographs that were
taken at different times of the day and on different days to capture a
range of light extinction levels needed for the preference studies.
Compared to studies that used computer-generated images (i.e., those in
Phoenix, AZ, and Washington, DC) there was more variability in scene
appearance in these older studies that could affect preference rating
and includes uncertainties associated with using ambient measurements
to represent sight path-averaged light extinction values rather than
superimposing a computer-generated amount of haze onto the images. When
using photographs, the intrinsic appearance of the scene can change due
to meteorological conditions (i.e., shadow patterns and cloud
conditions) and spatial variations in ambient air quality that can
result in ambient light extinction measurement not being representative
of the sight-path-averaged light extinction. Computer-generated images,
such as those generated with WinHaze, do not introduce such
uncertainties, as the same base photograph is used (i.e., there is no
intrinsic change in scene appearance) and the modeled haze that is
superimposed on the photograph is determined based on uniform light
extinction throughout the scene.
In addition to differences in preferences that may arise from
photographs versus computer-generated images, urban visibility
preference may differ by location, and such differences may arise from
differences in the cityscape scene that is depicted in the images.
These differences are related to the perceived value of objects and
scenes that are included in the image, as objects at a greater distance
have a greater sensitivity to perceived visibility changes as light
extinction is changed compared to similar scenes with objects at
shorter distances. For example, a person (regardless of their location)
evaluating visibility in an image with more scenic elements such as
mountains or natural views may value better visibility conditions in
these images compared to the same level of visibility impairment in an
image that only depicts urban features such as buildings and roads.
That is, if a person was shown the same level of visibility impairment
in two images depicting different scenes--one with mountains in the
background and urban features in the foreground and one with no
mountains in the background and nearby buildings in the image without
mountains in the distance--may find the amount of haze to be
unacceptable in the image with the mountains in the distance because of
a greater perceived value of viewing the mountains, while finding the
amount of haze to be acceptable in the image with the buildings because
of a lesser value of viewing the cityscape or an expectation that such
urban areas may generally have higher levels of haze in general. This
is consistent when comparing the differences between the Denver, CO,
study results (which found the 50% acceptance criteria occurred at the
best visual air quality levels among the four cities) and the
Washington, DC, results (which found the 50% acceptability criteria
occurred at the worst visual air quality levels among the four cities).
These results may occur because the most prominent and picturesque
feature of the cityscape of Denver is the visible snow-covered
mountains in the distance, while the prominent and
[[Page 16320]]
picturesque features of the Washington, DC, cityscape are buildings
relatively nearby without prominent and/or valued scenic features that
are more distant. Given these variabilities in preferences it is
unclear to what extent, the available evidence provides strong support
for a target level of protection at the lower end of the range. Future
studies that reduce sources of noisiness and uncertainty in the results
could provide more information that would support selection of a target
level of protection at or just above the lower end of the range.
Taken together, the 2022 PA concludes that available information
continues to support a visibility index with estimated light extinction
as the indicator, a 24-hour averaging time, and a 90th percentile form,
averaged over three years, with a level within the range of 20 to 30
dv.
ii. Relationship Between the PM2.5 Visibility Index and the
Current Secondary 24-Hour PM2.5 Standard
The 2022 PA presents quantitative analyses based on recent air
quality that evaluate the relationship between recent air quality and
calculated light extinction. As in previous reviews, these analyses
explored this relationship as an estimate of visibility impairment in
terms of the 24-hour PM2.5 standard and the visibility
index. Generally, the results of the updated analyses are similar to
those based on the data available at the time of the 2012 and 2020
reviews (U.S. EPA, 2022b, section 5.3.1.2). As discussed in section
V.C.1.a above, the 2022 PA concludes that the available evidence
continues to support a visibility index with estimated light extinction
as the indicator, a 24-hour averaging time, and a 90th percentile form,
averaged over three years, with a level within the range of 20 to 30
dv. These analyses evaluate visibility impairment in the U.S. under
recent air quality conditions, particularly those conditions that meet
the current standards, and the relative influence of various factors on
light extinction. Given the relationship of visibility with short-term
PM, we focus particularly on the short-term PM standards.\159\ Compared
to the 2012 review, updated analyses incorporate several refinements,
including (1) the evaluation of three versions of the IMPROVE equation
to calculate light extinction (U.S. EPA, 2022b, Appendix D, Equations
D-1 through D-3) in order to better understand the influence of
variability in equation inputs; \160\ (2) the use of 24-hour relative
humidity data, rather than monthly average relative humidity as was
used in the 2012 review (U.S. EPA, 2022b, section 5.3.1.2, Appendix D);
and (3) the inclusion of the coarse fraction in the estimation of light
extinction (U.S. EPA, 2022b, section 5.3.1.2, Appendix D). The analyses
in the reconsideration are updated from the 2012 and 2020 reviews and
include 60 monitoring sites that measure PM2.5 and
PM10 and are geographically distributed across the U.S. in
both urban and rural areas (U.S. EPA, 2022b, Appendix D, Figure D-1).
---------------------------------------------------------------------------
\159\ The analyses presented in the 2022 PA focus on the
visibility index and the current secondary 24-hour PM2.5
standard with a level of 35 [micro]g/m\3\. However, we recognize
that all three secondary PM standards influence the PM
concentrations associated with the air quality distribution. As
noted in section V.A.1 above, the current secondary PM standards
include the 24-hour PM2.5 standard, with its level of 35
[micro]g/m\3\, the annual PM2.5 standard, with its level
of 15.0 [micro]g/m\3\, and the 24-hour PM10 standard,
with its level of 150 [micro]g/m\3\. With regard to the annual
PM2.5 standard, we note that all 60 areas included in the
analyses meet the current secondary annual PM standard (U.S. EPA,
2022b, Table D-7).
\160\ While the PM2.5 monitoring network has an
increasing number of continuous FEM monitors reporting hourly
PM2.5 mass concentrations, there continue to be data
quality uncertainties associated with providing hourly
PM2.5 mass and component measurements that could be input
into IMPROVE equation calculations for subdaily visibility
impairment estimates. As detailed in the 2022 PA, there are
uncertainties associated with the precision and bias of 24-hour
PM2.5 measurements (U.S. EPA, 2022b, p. 2-18), as well as
to the fractional uncertainty associated with 24-hour PM component
measurements (U.S. EPA, 2022b, p. 2-21). Given the uncertainties
present when evaluating data quality on a 24-hour basis, the
uncertainty associated with subdaily measurements may be even
greater. Therefore, the inputs to these light extinction
calculations are based on 24-hour average measurements of
PM2.5 mass and components, rather than subdaily
information.
---------------------------------------------------------------------------
When light extinction was calculated using the revised IMPROVE
equation, in areas that meet the current 24-hour PM2.5
standard for the 2017-2019 time period, all sites have light extinction
estimates at or below 26 dv (U.S. EPA, 2022b, Figure 5-3). For the four
locations that exceed the current 24-hour PM2.5 standard,
light extinction estimates range from 22 dv to 27 dv (U.S. EPA, 2022b,
Figure 5-3). These findings are consistent with the findings of the
analyses using the same IMPROVE equation in the 2012 review with data
from 102 sites with data from 2008-2010 and in the 2020 review with
data from 67 sites with data from 2015-2017. The analyses presented in
the 2022 PA indicate similar findings to those from the analyses in the
2012 and 2020 reviews, i.e., the updated quantitative analysis shows
that the 3-year visibility metric was no higher than 30 dv \161\ at
sites meeting the current secondary PM standards, and at most such
sites the 3-year visibility index values are much lower (e.g., an
average of 20 dv across the 60 sites).\162\
---------------------------------------------------------------------------
\161\ A 3-year visibility metric with a level of 30 dv would be
at the upper end of the range of levels identified from the public
preference studies.
\162\ When light extinction is calculated using the original
IMPROVE equation, all 60 sites have 3-year visibility metrics below
30 dv, 58 sites are at or below 25 dv, and 26 sites are at or below
20 dv (see U.S. EPA, 2022b, Appendix D, Table D-3).
---------------------------------------------------------------------------
When light extinction was calculated using the revised IMPROVE
equation,\163\ the resulting 3-year visibility metrics are nearly
identical to light extinction estimates calculated using the original
IMPROVE equation (U.S. EPA, 2022b, Figure 5-4), but some sites are just
slightly higher. Using the revised IMPROVE equation, for those sites
that meet the current 24-hour PM2.5 standard, the 3-year
visibility metric is at or below 26 dv. For the four locations that
exceed the current 24-hour PM2.5 standard, light extinction
estimates range from 22 dv to 29 dv (U.S. EPA, 2022b, Figure 5-4).
These results are similar to those for light extinction calculated
using the original IMPROVE equation,\164\ and those from previous
reviews.
---------------------------------------------------------------------------
\163\ As described in more detail in the 2022 PA, the revised
IMPROVE equation divides PM components into smaller and larger sizes
of particles in PM2.5, with separate mass scattering
efficiencies and hygroscopic growth functions for each size category
(U.S. EPA, 2022b, section 5.3.1.1).
\164\ When light extinction is calculated using the revised
IMPROVE equation, all 60 sites have 3-year visibility metrics below
30 dv, 56 sites are at or below 25 dv, and 26 sites are at or below
20 dv (see U.S. EPA, 2022b, Appendix D, Table D-3).
---------------------------------------------------------------------------
When light extinction was calculated using the refined equation
from Lowenthal and Kumar (2016), the resulting 3-year visibility
metrics are slightly higher at all sites compared to light extinction
estimates calculated using the original IMPROVE equation (U.S. EPA,
2022b, Figure 5-5).\165\ These higher estimates are to be expected,
given the higher OC multiplier included in the IMPROVE equation from
Lowenthal and Kumar (2016), which reflects the use of data from remote
areas with higher concentrations of organic PM when validating the
equation. As such, it is important to note that the Lowenthal and Kumar
(2016) version of the equation may overestimate light extinction in
non-remote areas, including the urban areas in the updated analyses in
this reconsideration.
---------------------------------------------------------------------------
\165\ When light extinction is calculated using the Lowenthal
and Kumar IMPROVE equation, 59 sites have 3-year visibility metrics
below 30 dv, 45 sites are at or below 25 dv, and 15 sites are at or
below 20 dv. The one site with a 3-year visibility metric of 32 dv
exceeds the secondary 24-hour PM2.5 standard, with a
design value of 56 [micro]g/m\3\ (see U.S. EPA, 2022b, Appendix D,
Table D-3).
---------------------------------------------------------------------------
Nevertheless, when light extinction is calculated using the
Lowenthal and
[[Page 16321]]
Kumar (2016) equation for those sites that meet the current 24-hour
PM2.5 standard, the 3-year visibility metric is generally at
or below 28 dv. For those sites that exceed the current 24-hour
PM2.5 standard, three of these sites have a 3-year
visibility metric ranging between 26 dv and 30 dv, while one site in
Fresno, California that exceeds the current 24-hour PM2.5
standard and has a 3-year visibility index value of 32 dv (compared to
29 dv when light extinction is calculated with the original IMPROVE
equation) (see U.S. EPA, 2022b, Appendix D, Table D-3). At this site,
it is likely that the 3-year visibility metric using the Lowenthal and
Kumar (2016) equation would be below 30 dv if PM2.5
concentrations were reduced such that the 24-hour PM2.5
level of 35 [micro]g/m\3\ was attained.
In considering visibility impairment under recent air quality
conditions, the 2022 PA recognizes that the differences in the inputs
to equations estimating light extinction can influence the resulting
values. For example, given the varying chemical composition of
emissions from different sources, the 2.1 multiplier for converting OC
to organic matter (OM) in the Lowenthal and Kumar (2016) equation may
not be appropriate for all source types. At the time of the 2012
review, the EPA judged that a 1.6 multiplier was more appropriate, for
the purposes of estimating visibility index at sites across the U.S.,
than the 1.4 or 1.8 multipliers used in the original and revised
IMPROVE equations, respectively. A multiplier of 1.8 or 2.1 would
account for the more aged and oxygenated organic PM that tends to be
found in more remote regions than in urban regions, whereas a
multiplier of 1.4 may underestimate the contribution of organic PM
found in remote regions when estimating light extinction (78 FR 3206,
January 15, 2013; U.S. EPA, 2012, p. IV-5). The available scientific
information and results of the air quality analyses indicate that it
may be appropriate to select inputs to the IMPROVE equation (e.g., the
multiplier for OC to OM) on a regional basis rather than a national
basis when calculating light extinction. This is especially true when
comparing sites with localized PM sources (such as sites in urban or
industrial areas) to sites with PM derived largely from biogenic
precursor emissions (that contribute to widespread secondary organic
aerosol formation), such as those in the southeastern U.S. The 2022 PA
notes, however, that conditions involving PM from such different
sources have not been well studied in the context of applying a
multiplier to estimate light extinction, contributing uncertainty to
estimates of light extinction for such conditions.
At the time of the 2012 review, the EPA noted that PM2.5
is the size fraction of PM responsible for most of the visibility
impairment in urban areas (77 FR 38980, June 29, 2012). Data available
at the time of the 2012 review suggested that, generally,
PM10-2.5 was a minor contributor to visibility impairment
most of the time (U.S. EPA, 2010b) although the coarse fraction may be
a major contributor in some areas in the desert southwestern region of
the U.S. Moreover, at the time of the 2012 review, there were few data
available from PM10-2.5 monitors to quantify the
contribution of coarse PM to calculated light extinction. Since that
time, an expansion in PM10-2.5 monitoring efforts has
increased the availability of data for use in estimating light
extinction with both PM2.5 and PM10-2.5
concentrations included as inputs in the equations. The analysis in the
2020 PA addressed light extinction at 20 of the 67 PM2.5
sites where collocated PM10-2.5 monitoring data were
available. Since that time, PM10-2.5 monitoring data are
available at more locations and the analyses presented in the 2022 PA
include those for light extinction estimated with coarse and fine PM at
all 60 sites. Generally, the contribution of the coarse fraction to
light extinction at these sites is minimal, contributing less than 1 dv
to the 3-year visibility metric (U.S. EPA, 2020b, section 5.2.1.2).
However, the 2022 PA notes that in the updated quantitative analyses,
only a few sites were in locations that would be expected to have high
concentrations of coarse PM, such as the Southwest. These results are
consistent with those in the analyses in the 2019 ISA, which found that
mass scattering from PM10-2.5 was relatively small (less
than 10%) in the eastern and northwestern U.S., whereas mass scattering
was much larger in the Southwest (more than 20%) particularly in
southern Arizona and New Mexico (U.S. EPA, 2019a, section 13.2.4.1, p.
13-36).
Overall, the findings of these updated quantitative analyses are
generally consistent with those in the 2012 and 2020 reviews. The 3-
year visibility metric was generally below 26 dv in most areas that
meet the current 24-hour PM2.5 standard. Small differences
in the 3-year visibility metric were observed between the variations of
the IMPROVE equation, which may suggest that it may be more appropriate
to use one version over another in different regions of the U.S. based
on PM characteristics such as particle size and composition to more
accurately estimate light extinction.
b. Non-Visibility Effects
Consistent with the evidence available at the time of the 2012 and
2020 reviews, and as described in detail in the 2022 PA (U.S. EPA,
2022b, section 5.3.2.2), the data remain insufficient to conduct
quantitative analyses for PM effects on climate and materials. For PM-
related climate effects, as explained in more detail in the proposal
(88 FR 5654, January 27, 2023), our understanding of PM-related climate
effects is still limited by significant key uncertainties. The recently
available evidence does not appreciably improve our understanding of
the spatial and temporal heterogeneity of PM components that contribute
to climate forcing (U.S. EPA, 2022b, sections 5.3.2.1.1 and 5.5).
Significant uncertainties also persist related to quantifying the
contributions of PM and PM components to the direct and indirect
effects on climate forcing, such as changes to the pattern of rainfall,
changes to wind patterns, and effects on vertical mixing in the
atmosphere (U.S. EPA, 2022b, sections 5.3.2.1.1 and 5.5). Additionally,
while improvements have been made to climate models since the
completion of the 2009 ISA, the models continue to exhibit variability
in estimates of the PM-related climate effects on regional scales
(e.g., ~100 km) compared to simulations at the global scale (U.S. EPA,
2022b, sections 5.3.2.1.1 and 5.5). While our understanding of climate
forcing on a global scale is somewhat expanded since the 2012 review,
significant limitations remain to quantifying potential adverse PM-
related climate effects in the U.S. and how they would vary in response
to incremental changes in PM concentrations across the U.S. As such,
while recent research is available on climate forcing on a global
scale, the remaining limitations and uncertainties are significant, and
the recent global scale research does not translate directly for use at
regional spatial scales. Therefore, the evidence does not provide a
clear understanding at the necessary spatial scales for quantifying the
relationship between PM mass in ambient air and the associated climate-
related effects in the U.S. that would be necessary to evaluate or
consider a level of air quality to protect against such effects and for
informing consideration of a national PM standard on climate in this
reconsideration (U.S. EPA, 2022b, section 5.3.2.2.1; U.S. EPA, 2019a,
section 13.3).
[[Page 16322]]
For PM-related materials effects, as explained in more detail in
the 2022 PA (U.S. EPA, 2022b, section 5.3.2.2), the available evidence
has been somewhat expanded to include additional information about the
soiling process and the types of materials impacted by PM. This
evidence provides some limited information to inform dose-response
relationships and damage functions associated with PM, although most of
these studies were conducted outside of the U.S. where PM
concentrations in ambient air are typically above those observed in the
U.S. (U.S. EPA, 2022b, section 5.3.2.1.2; U.S. EPA, 2019a, section
13.4). The evidence on materials effects characterized in the 2019 ISA
also includes studies examining effects of PM on the energy efficiency
of solar panels and passive cooling building materials, although the
evidence remains insufficient to establish quantitative relationships
between PM in ambient air and these or other materials effects (U.S.
EPA, 2022b, section 5.3.2.1.2). While the available evidence assessed
in the 2019 ISA is somewhat expanded since the time of the 2012 review,
quantitative relationships have not been established for PM-related
soiling and corrosion and frequency of cleaning or repair that further
the understanding of the public welfare implications of materials
effects (U.S. EPA, 2022b, section 5.3.2.2.2; U.S. EPA, 2019a, section
13.4). Therefore, there is insufficient information to inform
quantitative analyses assessing materials effects to inform
consideration of a national PM standard on materials in this
reconsideration (U.S. EPA, 2022b, section 5.3.2.2.2; U.S. EPA, 2019a,
section 13.4).
B. Conclusions on the Secondary PM Standards
In drawing conclusions on the adequacy of the current secondary PM
standards, in view of the advances in scientific knowledge and
additional information now available, the Administrator has considered
the evidence base, information, and policy judgments that were the
foundation of the 2020 decision and reflects upon the body of
information and evidence available in this reconsideration. In so
doing, the Administrator has taken into account both evidence-based and
quantitative information-based considerations, as well as advice from
the CASAC and public comments. Evidence-based considerations draw upon
the EPA's assessment and integrated synthesis of the scientific
evidence from studies evaluating welfare effects related to visibility,
climate, and materials associated with PM in ambient air as discussed
in the 2022 PA (summarized in sections V.B and V.D.2 of the proposal,
section V.A.2 above). The quantitative information-based considerations
draw from the results of the quantitative analyses of visibility
impairment presented in the 2022 PA (as summarized in section V.C of
the proposal and V.A.3 above) and consideration of these results in the
2022 PA.
Consideration of the scientific evidence and quantitative
information in the 2022 PA and by the Administrator is framed by
consideration of a series of policy-relevant questions. Section V.B.2
below summarizes the rationale for the Administrators proposed
decision, drawing from section V.D.3 of the proposal. The advice and
recommendations of the CASAC and public comments on the proposed
decision are addressed below in sections V.B.1 and V.B.3, respectively.
The Administrator's conclusions in this reconsideration regarding the
adequacy of the secondary PM standards and whether any revisions are
appropriate are described in section V.D.4.
1. CASAC Advice
In comments on the 2019 draft PA, the CASAC concurred with the
staff's overall preliminary conclusions that it is appropriate to
consider retaining the current secondary standards without revision
(Cox, 2019b). The CASAC ``finds much of the information . . . on
visibility and materials effects of PM2.5 to be useful,
while recognizing that uncertainties and controversies remain about the
best ways to evaluate these effects'' (Cox, 2019b, p. 13 of consensus
responses). Regarding climate, while the CASAC agreed that research on
PM-related effects has expanded since the 2012 review, it also
concluded that ``there are still significant uncertainties associated
with the accurate measurement of PM to the direct and indirect effects
of PM on climate'' (Cox, 2019b, pp. 13-14 of consensus responses). The
committee recommended that the EPA summarize the ``current scientific
knowledge and quantitative modeling results for effects of reducing
PM2.5'' on several climate-related outcomes (Cox, 2019b, p.
14 of consensus responses), while also recognizing that ``it is
appropriate to acknowledge uncertainties in climate change impacts and
resulting welfare impacts in the United States of reductions in
PM2.5 levels'' (Cox, 2019b, p. 14 of consensus responses).
When considering the overall body of scientific evidence and technical
information for PM-related effects on visibility, climate, and
materials, the CASAC agreed with the EPA's preliminary conclusions in
the 2019 draft PA, stating that ``the available evidence does not call
into question the protection afforded by the current secondary PM
standards and concurs that they should be retained'' (Cox, 2019b, p. 3
of letter).
In this reconsideration, the CASAC provided its advice regarding
the current secondary PM standards in the context of its review of the
2021 draft PA (Sheppard, 2022a). In its comments on the 2021 draft PA,
the CASAC first recognized that the scientific evidence is sufficient
to support a causal relationship between PM and visibility effects,
climate effects and materials effects.
With regard to visibility effects, the CASAC recognized that the
identification of a target level of protection for the visibility index
is based on a limited number of studies and suggested that ``additional
region- and view-specific visibility preference studies and data
analyses are needed to support a more refined visibility target''
(Sheppard, 2022a, p. 21 of consensus responses). While the CASAC did
not recommend revising either the target level of protection for the
visibility index or the level of the current 24-hour PM2.5
standard, they did state that a visibility index of 30 deciviews
``needs to be justified'' and ``[i]f a value of 20-25 deciviews is
deemed to be an appropriate visibility target level of protection, then
a secondary 24-hour PM2.5 standard in the range of 25-35
[micro]g/m\3\ should be considered'' (Sheppard, 2022a, p. 21 of
consensus responses).
The CASAC also recognized the limited availability of monitoring
methods and networks for directly measuring light extinction. As such,
they suggest that ``[a] more extensive technical evaluation of the
alternatives for visibility indicators and practical measurement
methods (including the necessity for a visibility FRM) is need for
future reviews'' (Sheppard, 2022a, p. 22 of consensus letter). The
majority of the CASAC ``recommend[ed] that an FRM for a directly
measured PM2.5 light extinction indicator be developed'' to
inform the consideration of the protection afforded by the secondary PM
standards against visibility impairment, the minority of the CASAC
``believe that a light extinction FRM is not necessary to set a
secondary standard protective of visibility'' (Sheppard, 2022a, p. 22
of consensus responses).
[[Page 16323]]
With regard to climate, the CASAC noted that ``there is a causal
relationship between PM and climate change, but large uncertainties
remain'' and recommended additional research (Sheppard, 2022a, p. 22 of
consensus responses). With respect to materials damage, the CASAC noted
that ``[q]uantitative information on the relationship between PM and
material damage is lacking'' and suggested some additional studies and
research approaches that could provide additional information on the
effects of PM on materials and the quantitative assessment of the
relationship between materials effects and PM in ambient air (Sheppard,
2022a, p. 23 of consensus responses).
2. Basis for the Proposed Decision
In reaching his proposed conclusions, the Administrator first
recognized that, consistent with the scope of this reconsideration, his
decision in this reconsideration will be focused only and specifically
on the adequacy of public welfare protection provided by the secondary
PM standards from effects related to visibility, climate, and
materials. He then considered the assessment of the current evidence
and conclusions reached in the 2019 ISA and ISA Supplement; the
currently available quantitative information, including associated
limitations and uncertainties, described in detail and characterized in
the 2022 PA; considerations and staff conclusions and associated
rationales presented in the 2022 PA; and the advice and recommendations
from the CASAC (88 FR 5655, January 27, 2023).
With respect to visibility, the Administrator noted the
longstanding body of evidence that demonstrates a causal relationship
between ambient PM and effects on visibility (U.S. EPA, 2019a, section
13.2). and that visibility impairment can have implications for
people's enjoyment of daily activities and for their overall sense of
well-being. Therefore, as in previous reviews, he considered the degree
to which the current secondary standards protect against PM-related
visibility impairment. In so doing, and consistent with previous
reviews, the Administrator considered the protection provided by the
current secondary standards against PM-related visibility impairment in
conjunction with the Regional Haze Program \166\ for protecting
visibility in Class I areas,\167\ which together would be expected to
achieve appropriate visual air quality across all areas (88 FR 5658,
January 27, 2023). The Administrator proposed to conclude that
addressing visibility impairment in Class I areas is beyond the scope
of the secondary PM NAAQS and that setting the secondary PM NAAQS at a
level that would remedy visibility impairment in Class I areas would
result in standards that are more stringent than is requisite.
---------------------------------------------------------------------------
\166\ The Regional Haze Program was established by Congress
specifically to achieve ``the prevention of any future, and the
remedying of existing, impairment of visibility in mandatory Class I
areas, which impairment results from man-made air pollution,'' and
that Congress established a long-term program to achieve that goal
(CAA section 169A).
\167\ In adopting section 169A, Congress set a goal of
eliminating anthropogenic visibility impairment at Class I areas, as
well as a framework for achieving that goal which extends well
beyond the planning process and timeframe for attaining secondary
NAAQS. Thus, the Regional Haze Program will continue to contribute
to reductions in visibility impairment in Class I areas.
---------------------------------------------------------------------------
In further considering what standards are requisite to protect
against adverse public welfare effects from visibility impairment, the
Administrator adopted an approach consistent with the approach used in
previous reviews (88 FR 5645, January 27, 2023). That is, he first
identified an appropriate target level of protection in terms of a PM
visibility index that accounts for the factors that influence the
relationship between particles in the ambient air and visibility (i.e.,
size fraction, species composition, and relative humidity). He then
considered air quality analyses examining the relationship between this
PM visibility index and the current secondary 24-hour PM2.5
standard in locations meeting the current 24-hour PM2.5 and
PM10 standards (U.S. EPA, 2022b, section 5.3.1.2; 88 FR
5650, January 27, 2023).
To identify a target level of protection, the Administrator first
considered the characteristics of the visibility index and defines its
elements (indicator, averaging time, form, and level). With regard to
the indicator for the visibility index, the Administrator recognized
that there is a lack of availability of methods and an established
network for directly measuring light extinction, consistent with the
conclusions reached in the 2022 PA (U.S. EPA, 2022b, section 5.3.1.1)
and with the CASAC's recommendation for additional research on direct
measurement methods for light extinction in their review of the 2021
draft PA (Sheppard, 2022a, p. 22 of consensus responses). Consistent
with the approaches used in reaching decisions in 2012 and 2020, given
the lack of such monitoring data, the Administrator preliminarily
judged that estimated light extinction, as calculated using one or more
versions of the IMPROVE algorithms, continues to be the most
appropriate indicator for the visibility index in this reconsideration
(88 FR 5659, January 27, 2023).
In further defining the characteristics of a visibility index based
on estimates of light extinction, the Administrator considered the
appropriate averaging time, form, and level of the index. With regard
to the averaging time and form, the Administrator noted that in
previous reviews, a 24-hour averaging time was selected and the form
was defined as the 3-year average of annual 90th percentile values. The
Administrator recognized that the evidence available in this
reconsideration and described in the 2022 PA continue to provide
support for the short-term nature of PM-related visibility effects.
Considering the available analyses of 24-hour and subdaily
PM2.5 light extinction, and noting that the CASAC did not
provide advice or recommendations with regard to the averaging time of
the visibility index, the Administrator preliminarily judged that the
24-hour averaging time continues to be appropriate for the visibility
index (88 FR 5659, January 27, 2023).
With regard to the form of the visibility index, the Administrator
noted that, consistent with the approach taken in other NAAQS,
including the current secondary 24-hour PM2.5 NAAQS, a
multi-year percentile form offers greater stability to the air quality
management process by reducing the possibility that statistically
unusual indicator values will lead to transient violations of the
standard. Using a 3-year average provides stability from the occasional
effects of inter-annual meteorological variability that can result in
unusually high pollution levels for a particular year (88 FR 5659,
January 27, 2023). In considering the percentile that would be
appropriate with the 3-year average, the Administrator first noted that
the Regional Haze Program targets the 20% most impaired days for
improvements in visual air quality in Class I areas.\168\ Based on
analyses examining 90th, 95th, and 98th percentile forms, the
Administrator preliminarily judged that a focus similar to the Regional
Haze Program focused on improving the 20% most impaired days suggest
that the 90th percentile, which represents the median of the 20% most
impaired days, such that 90% of days have visual air quality that is at
or below the target level of protection of the visibility
[[Page 16324]]
index, would be reasonably expected to lead to improvements in visual
air quality for the 20% most impaired days (88 FR 5659, January 27,
2023). In the analyses of percentiles, the results suggest that a
higher percentile value could have the effect of limiting the
occurrence of days with peak PM-related light extinction in areas
outside of Federal Class I areas to a greater degree. However, the
Administrator preliminarily concluded that it is appropriate to balance
concerns about focusing on the group of most impaired days with
concerns about focusing on the days with peak visibility impairment.
Additionally, the Administrator noted that the CASAC did not provide
advice or recommendations related to the form of the visibility index.
Therefore, the Administrator preliminarily judged that it remains
appropriate to define a visibility index in terms of a 24-hour
averaging time and a form based on the 3-year average of annual 90th
percentile values (88 FR 5659, January 27, 2023).
---------------------------------------------------------------------------
\168\ As noted above, the Administrator viewed the Regional Haze
Program as a complement to the secondary PM NAAQS, and thus took
into consideration its approach to improving visibility in
considering how to address visibility outside of Class I areas.
---------------------------------------------------------------------------
With regard to the level of the visibility index, the Administrator
first noted that the scientific evidence that is available to inform
the level of the visibility index is largely the same as in previous
reviews, and continues to provide support for a level within the range
of 20 to 30 dv (88 FR 5659-5660, January 27, 2023). The Administrator
recognized that significant uncertainties and limitations remained, in
particular those related to the public preference studies, including
methodological differences between the studies, and that the available
studies may not capture the full range of visibility preferences in the
U.S. population (88 FR 5659-5660, January 27, 2023). The Administrator
also noted that, in their review of the 2021 draft PA, the CASAC
recognized that a judgment regarding the appropriate target level of
protection for the visibility index is based on a limited number of
visibility preference studies, with studies conducted in the western
U.S. reporting public preferences for visibility impairment associated
with the lower end of the range of levels, while studies conducted in
the eastern U.S. reporting public preferences associated with the upper
end of the range (Sheppard, 2022a, p. 21 of consensus responses). The
Administrator noted that there have long been significant questions
about how to set a national standard for visibility that is not
overprotective for some areas of the U.S. In establishing the Regional
Haze Program to improve visibility in Class I areas, Congress noted
that ``as a matter of equity, the national ambient air quality
standards cannot be revised to adequately protect visibility in all
areas of the country.'' H.R. Rep. 95-294 at 205. Thus, in reaching his
proposed conclusion, the Administrator recognized that there are
substantial uncertainties and limitations in the public preference
studies that should be considered when selecting a target level of
protection for the visibility index and took the uncertainties and
variability inherent in the public preference studies into account. In
so doing, the Administrator first preliminarily judged that, consistent
with similar judgments in past reviews, it is appropriate to recognize
that the secondary 24-hour PM2.5 standard is intended to
address visibility impairment across a wide range of regions and
circumstances, and that the current standard works in conjunction with
the Regional Haze Program to improve visibility, and therefore, it is
appropriate to establish a target level of protection based on the
upper end of the range of levels. In considering the information
available in this reconsideration and the CASAC's advice, the
Administrator proposed to conclude that the protection provided by a
visibility index based on estimated light extinction, a 24-hour
averaging time, and a 90th percentile form, averaged over 3 years, set
at a level of 30 dv (the upper end of the range of levels) would be
requisite to protect public welfare with regard to visibility
impairment (88 FR 5660, January 27, 2023).
In preliminarily concluding that it remains appropriate in this
reconsideration to define the target level of protection in terms of a
visibility index based on estimated light extinction as described above
(i.e., with a 24-hour averaging time; a 3-year, 90th percentile form;
and a level of 30 dv), the Administrator next considered the degree of
protection from visibility impairment afforded by the existing
secondary standards. He considered the updated analyses of PM-related
visibility impairment presented in the 2022 PA (U.S. EPA, 2022b,
section 5.3.1.2), which reflect several improvements over the analyses
conducted in the 2012 review. Specifically, the updated analyses
examine multiple versions of the IMPROVE algorithm, including the
version incorporating revisions since the 2012 review (section
V.B.1.a), which provides an improved understanding of how variation in
equation inputs impacts calculated light extinction (U.S. EPA, 2022b,
Appendix D). In addition, unlike the analyses in the 2012 review and
the 2020 PA, all of the sites included in the analyses had
PM10-2.5 data available, which allows for better
characterization of the influence of the coarse fraction on light
extinction (U.S. EPA, 2022b, section 5.3.1.2).
The Administrator noted that the results of these updated analyses
are consistent with the results from the 2012 and 2020 reviews (88 FR
5660, January 27, 2023). Regardless of the IMPROVE equation used, these
analyses demonstrate that the 3-year visibility metric is at or below
28 dv in all areas meeting the current 24-hour PM2.5
standard (section V.C.1.b). Given the results of these analyses, the
Administrator preliminarily concluded that the updated scientific
evidence and technical information support the adequacy of the current
secondary PM2.5 and PM10 standards to protect
against PM-related visibility impairment. While the inclusion of the
coarse fraction had a relatively modest impact on calculated light
extinction in the analyses presented in the 2022 PA, he nevertheless
recognized the continued importance of the PM10 standard
given the potential for larger impacts in locations with higher coarse
particle concentrations, such as in the southwestern U.S., for which
only a few sites met the criteria for inclusion in the analyses in the
2022 PA (U.S. EPA, 2019a, section 13.2.4.1; U.S. EPA, 2022b, section
5.3.1.2).
With regard to the adequacy of the secondary 24-hour
PM2.5 standard, the Administrator noted that the CASAC
stated that ``[i]f a value of 20-25 deciviews is deemed to be an
appropriate visibility target level of protection, then a secondary 24-
hour PM2.5 standard in the range of 25-35 [micro]g/m\3\
should be considered'' (Sheppard, 2022a, p. 21 of consensus responses).
The Administrator recognized that the CASAC recommended that the
Administrator provide additional justification for a visibility index
target of 30 dv but did not specifically recommend that he choose an
alternative level for the visibility index. The Administrator
considered the CASAC's advice, together with the available scientific
evidence and quantitative information, in reaching his proposed
conclusions. He recognized conclusions regarding the appropriate weight
to place on the scientific and technical information examining PM-
related visibility impairment including how to consider the range and
magnitude of uncertainties inherent in that information is a public
welfare policy judgment left to the Administrator. As such, the
Administrator noted his conclusion on
[[Page 16325]]
the appropriate visibility index (i.e., with a 24-hour averaging time;
a 3-year, 90th percentile form; and a level of 30 dv) and his
conclusions regarding the quantitative analyses of the relationship
between the visibility index and the current secondary 24-hour
PM2.5 standard. In so doing, he proposed to conclude that
the current secondary standards provide requisite protection against
PM-related visibility effects (88 FR 5661, January 27, 2023).
In reaching his proposed conclusions, the Administrator also
recognized that the available evidence on visibility impairment
generally reflects a continuum and that the public preference studies
did not identify a specific level of visibility impairment that would
be perceived as ``acceptable'' or ``unacceptable'' across the whole
U.S. population. However, he noted that a judgment regarding the
appropriate target level of protection would take into consideration
the appropriate weight to place on the individual public preference
studies. In so doing, he noted that placing more weight on the public
preference study from Washington, DC, could provide support for a
target level of protection at or near 30 dv, whereas placing more
weight on the public preference study performed in the Phoenix, AZ,
study could provide support for a target level of protection below 30
dv and down to 25 dv. While the Administrator noted that, in their
review of the 2021 draft PA, the CASAC did not recommend revising the
level of the current 24-hour PM2.5 standard, the
Administrator recognized that they did recommend greater justification
for a target level of protection of 30 dv, and noted that if a target
level of protection of 20-25 dv was identified, then a secondary 24-
hour PM2.5 standard in the range of 25-35 [mu]g/m\3\ should
be considered (Sheppard, 2022a, p. 21 of consensus responses). For
these reasons, the Administrator solicited comment on his proposed
decision to retain the current secondary 24-hour PM2.5
standard, as well as the appropriateness of a target level of
protection for visibility below 30 dv and as low as 25 dv, and on
revising the level of the current secondary 24-hour PM2.5
standard to a level as low as 25 [micro]g/m\3\.
With respect to climate effects, the Administrator recognized that
a number of improvements and refinements have been made to climate
models since the time of the 2012 review. However, despite continuing
research and the strong evidence supporting a causal relationship with
climate effects (U.S. EPA, 2019a, section 13.3.9), the Administrator
noted that there are still significant limitations in quantifying the
contributions of the direct and indirect effects of PM and PM
components on climate forcing (U.S. EPA, 2022b, sections 5.3.2.1.1 and
5.5). He also recognized that models continue to exhibit considerable
variability in estimates of PM-related climate impacts at regional
scales (e.g., ~100 km), compared to simulations at the global scale
(U.S. EPA, 2022b, sections 5.3.2.1.1 and 5.5). As noted above, the
CASAC recognized a causal relationship between PM and climate effects
but also the large uncertainties associated with quantitatively
assessing such effects, particularly on a national level in the context
of a U.S.-based standard. These uncertainties led the Administrator to
preliminarily conclude that the scientific information available in
this reconsideration remains insufficient to quantify, with confidence,
the impacts of ambient PM on climate in the U.S. (U.S. EPA, 2022b,
section 5.3.2.2.1) and that there is insufficient information at this
time to revise the current secondary PM standards or to promulgate a
distinct secondary standard to address PM-related climate effects (88
FR 5661, January 27, 2023).
With respect to materials effects, the Administrator noted that the
available evidence continues to support the conclusion that there is a
causal relationship with PM deposition (U.S. EPA, 2019a, section 13.4).
He recognized that deposition of particles in the fine or coarse
fractions can result in physical damage and/or impaired aesthetic
qualities. Particles can contribute to materials damage by adding to
the effects of natural weathering processes and by promoting the
corrosion of metals, the degradation of painted surfaces, the
deterioration of building materials, and the weakening of material
components. While some recent evidence on materials effects of PM is
available in the 2019 ISA, the Administrator noted that this evidence
is primarily from studies conducted outside of the U.S. in areas where
PM concentrations in ambient air are higher than those observed in the
U.S. (U.S. EPA, 2019a, section 13.4). The CASAC also noted the lack of
quantitative information relating PM and material effects. Given the
limited amount of information on the quantitative relationships between
PM and materials effects in the U.S., and uncertainties in the degree
to which those effects could be adverse to the public welfare, the
Administrator preliminarily judged that the scientific information
available in this reconsideration remains insufficient to quantify,
with confidence, the public welfare impacts of ambient PM on materials
and that there is insufficient information at this time to revise the
current secondary PM standards or to promulgate a distinct secondary
standard to address PM-related materials effects (88 FR 5661, January
27, 2023).
Taken together, the Administrator proposed to conclude that the
scientific and technical information for PM-related visibility
impairment, climate impacts, and materials effects, with its attendant
uncertainties and limitations, supports the current level of protection
provided by the secondary PM standards as being requisite to protect
against known and anticipated adverse effects on public welfare. For
visibility impairment, this proposed conclusion reflected his
consideration of the evidence for PM-related light extinction, together
with his consideration of updated analyses of the protection provided
by the current secondary PM2.5 and PM10
standards. For climate and materials effects, this conclusion reflected
his preliminary judgment that, although it remains important to
maintain secondary PM2.5 and PM10 standards to
provide some degree of control over long- and short-term concentrations
of both fine and coarse particles, it is generally appropriate not to
change the existing secondary standards at this time and that it is not
appropriate to establish any distinct secondary PM standards to address
PM-related climate and materials effects at this time. As such, the
Administrator recognized that current suite of secondary standards
(i.e., the 24-hour PM2.5, 24-hour PM10, and
annual PM2.5 standards) together provide such control for
both fine and coarse particles and long- and short-term visibility and
non-visibility (e.g., climate and materials) \169\ effects related to
PM in ambient air. His proposed conclusions on the secondary standards
were consistent with advice from the CASAC, which noted substantial
uncertainties remain in the scientific evidence for climate and
materials effects. Thus, based on his consideration of the evidence and
analyses for PM-related welfare effects, as described above, and his
consideration of CASAC advice on the secondary standards, the
Administrator proposed not to change those standards (i.e., the current
24-hour and annual PM2.5 standards, 24-hour PM10
standard) at this time (88 FR 5662, January 27, 2023).
---------------------------------------------------------------------------
\169\ As noted earlier, other welfare effects of PM, such as
ecological effects, are being considered in the separate, on-going
review of the secondary NAAQS for oxides of nitrogen, oxides of
sulfur and PM.
---------------------------------------------------------------------------
[[Page 16326]]
3. Comments on the Proposed Decision
Of the public comments received on the proposal, very few were
specific to the secondary PM standards. Of those commenters who did
provide comments on the secondary PM standards, the majority support
the Administrator's proposed decision to retain the current standards.
Some commenters disagree with the Administrator's proposed conclusion
to retain the current secondary standards, primarily focusing their
comments on the need for a revised standard to protect against
visibility impairment. In addition to the comments addressed in this
notice, the EPA has prepared a Response to Comments document that
addresses other specific comments related to setting the secondary PM
standards. This document is available for review in the docket for this
rulemaking and through the EPA's NAAQS website (https://www.epa.gov/naaqs/particulate-matter-pm-air-quality-standards).
We first note that some commenters raise questions about the
protection provided by the secondary PM standards for ecological
effects (e.g., effects on ecosystems, ecosystem services, or species).
However, consistent with the 2016 IRP and as described in the proposal
(88 FR 5643, January 27, 2023), other welfare effects of PM, such as
the ecological effects identified by commenters, are being considered
as part of the separate, ongoing review of the secondary standards for
oxides of sulfur, oxides of nitrogen and PM, and thus, those comments
are beyond the scope of this action.
Of the comments addressing the proposed decision for the secondary
PM standards, many of the commenters support the Administrator's
proposed decision to retain the current secondary PM standards, without
revision. This group includes industries and industry groups and State
and local governments and organizations. All of these commenters
generally note their agreement with the rationale provided in the
proposal, with a focus on the strength of the available scientific
evidence for PM-related welfare effects. Most also recognize that the
scientific evidence and quantitative information available in this
reconsideration have not substantially altered our previous
understanding of PM-related effects on non-ecological welfare effects
(i.e., visibility, climate, and materials) and do not call into
question the adequacy of the current secondary standards. They find the
proposed decision not to change the standards at this time to be well
supported and a reasonable exercise of the Administrator's public
welfare policy judgment under the CAA. The EPA agrees with these
comments regarding the adequacy of the current secondary PM standards
and the lack of support for revision of these standards at this time.
The EPA received relatively few comments on the proposed decision
that it is not appropriate to establish any distinct secondary PM
standards to address PM-related climate effects. Several commenters
agree that the available scientific evidence provides support for the
2019 conclusion that there is a causal relationship between PM and
climate effects, and the commenters also agree with the EPA that the
currently available information is not sufficient for supporting
quantitative analyses for the climate effects of PM in ambient air.
These commenters support the Administrator's proposed decision not to
set a distinct standard for climate.
There were also very few commenters who commented on the proposed
decision that it is not appropriate to establish any distinct secondary
PM standards to address PM-related materials effects. As with comments
on climate effects, commenters generally agree with the EPA that the
evidence is not sufficient to support quantitative analyses for PM-
related materials effects. However, some commenters contend that EPA
failed to explain in the proposal how the current standard is
appropriate to protect materials from the effects of PM. These
commenters disagree with the EPA's conclusion that quantitative
relationships have not been established for PM-related soiling and
corrosion and frequency of cleaning or repair of materials, and cite to
several studies conducted outside the U.S. that they contend that the
EPA should consider since the same materials are present in the U.S.
They further contend that, in discussing the available scientific
evidence in the 2019 ISA for studies conducted outside of the U.S., the
EPA did not provide references to these studies and, therefore, the
public is unable to comment on these studies. They further State that
EPA failed to consider the following information: (1) Recent work
related to soiling of photovoltaic modules and other surfaces, and; (2)
damage and degradation resulting from oxidant concentrations and solar
radiation for a number of materials, including polymeric materials,
plastic, paint, and rubber. These commenters further assert that the
EPA failed to propose a standard that provides requisite protection
against materials effects attributable to PM.
As an initial matter, we note that the commenters submitted the
same comments related to materials effects during the 2020 review.
Consistent with our response in the 2020 notice of final rulemaking (85
FR 82737, December 18, 2020), we disagree with the commenters that the
EPA failed to consider the relevant scientific information about
materials effects available in this reconsideration. The 2019 ISA
considered and included studies related to materials effects of PM,
including studies conducted in and outside of the U.S., on newly
studied materials including photovoltaic modules that were published
prior to the cutoff date for the literature search.\170\ These include
the Besson et al. (2017) study referenced by the commenters (U.S. EPA,
2019a, section 13.4.2). The Gr[oslash]ntoft et al. (2019) study
referenced by the same commenters was published after the cutoff date
for the literature search for the 2019 ISA. However, the EPA
provisionally considered new studies in responding to comments in the
2020 review, including the new studies highlighted by the commenters in
their comments on the 2020 notice of proposed rulemaking, in the
context of the findings of the 2019 ISA (see Appendix in U.S. EPA,
2020a).\171\ Based on the provisional consideration, the EPA concluded
in the 2020 review that the new studies are not sufficient to alter the
conclusions reached in the 2019 ISA regarding PM and materials effects.
For example, the Gr[oslash]ntoft et al. (2019) study was based on
European air pollution which as the EPA has noted has higher
concentrations (as well as diversity in sources, such as light duty
diesel engines) compared to the U.S.. Thus, the EPA did not find it
necessary or appropriate to reopen the air quality criteria to consider
this study because it would not have been an adequate basis on which to
set a NAAQS. As discussed in section I, when the EPA decided to
reconsider the standards, it also decided to reopen the air quality
criteria to a limited degree, based on its judgment that certain new
studies were likely to be useful in reconsidering the standards.
[[Page 16327]]
Based on the provisional consideration in the 2020 review and the
significant data gaps that existed at that time, the EPA did not
include these studies within the scope of the 2022 ISA Supplement
because, although these studies provide additional support for PM-
related materials, the studies would not support quantitative analyses
or alternative conclusions regarding these effects. As described in
section I.C.5.b above, the ISA Supplement focuses on a thorough
evaluation of some studies that became available after the literature
cutoff date of the 2019 ISA that could either further inform the
adequacy of the current PM NAAQS or address key scientific topics that
have evolved since the literature cutoff date for the 2019 ISA. In
developing the ISA Supplement, the EPA focused on the non-ecological
welfare effects for which the evidence supported a ``causal
relationship'' and for which quantitative analyses could be supported
by the evidence because those were the welfare effects that were most
useful in informing conclusions in the 2020 PA. While the 2020 PA
considered the broader set of evidence for materials effects, it
concluded that there remained `substantial uncertainties with regard to
the quantitative relationships with PM concentrations and concentration
patterns that limit[ed] [the] ability to quantitatively assess the
public welfare protection provided by the standards from these effects'
(U.S. EPA, 2020b).'' Therefore, the ISA Supplement did not include an
evaluation of scientific evidence for PM-related materials effects.
However, the EPA has once again provisionally considered new studies in
this reconsideration, including the studies highlighted by the
commenters, in the context of the 2019 ISA and concludes that, as in
the 2020 review, these studies are not sufficient to alter the
conclusions reached in the 2019 ISA regarding PM and materials effects
or to provide sufficient information on which to base a secondary
NAAQS. The EPA agrees there is a causal relationship between the
presence of PM in the ambient air and materials effects, but to set a
standard, the EPA needs not only to understand at what point materials
effects become adverse to public welfare but to be able to relate
specific concentrations of ambient PM to those levels of materials
effects. Given the significant gaps in the evidence, particularly given
that the majority of the recent evidence has been conducted outside of
the U.S., establishing any quantitative relationships between particle
size, concentration, chemical components, and specific measures of
materials damage, such as frequency of painting or repair of materials,
the EPA finds the evidence is insufficient to support a secondary NAAQS
to protect against materials effects.
---------------------------------------------------------------------------
\170\ As noted earlier in section V, the 2019 ISA ``identified
and evaluated studies and reports that that have undergone
scientific peer review and were published or accepted for
publication between January 1, 2009, and March 31, 2017. A limited
literature update identified some additional studies that were
published before December 31, 2017'' (U.S. EPA, 2019a, Appendix, p.
A-3).
\171\ As discussed in section I.D, the EPA has provisionally
considered studies that were highlighted by commenters and that were
published after the 2019 ISA. These studies are generally consistent
with the evidence assessed in the 2019 ISA, and they do not
materially alter our understanding of the scientific evidence or the
Agency's conclusions based on that evidence.
---------------------------------------------------------------------------
With regard to studies conducted outside of the U.S., including
those referenced by the commenters, as described in the proposal, in
reaching his proposed conclusion, the Administrator recognized that
while there was some newly available information related to materials
effects of PM included in the 2019 ISA, ``this evidence is primarily
from studies conducted outside of the U.S. in areas where PM
concentrations in ambient air are higher than those observed in the
U.S. (U.S. EPA, 2019a, section 13.4)'' (88 FR 5661, January 27, 2023).
We disagree with the commenters that EPA did not provide references for
these studies, nor that the lack of references inhibited the public's
ability to provide comment on this proposed conclusion. First, the
reference to section 13.4 in the 2019 ISA is a direct citation to the
evaluation of newly available studies on PM-related materials effects,
which includes citations for all materials effects evidence considered
in the 2020 review and in this reconsideration. Second, section
5.3.2.1.2 of the 2022 PA considers the available scientific evidence
for PM-related materials effects--including citations to the studies
newly available in the 2019 ISA--and how that evidence informs
conclusions regarding the adequacy of the standard (U.S. EPA, 2022b,
section 5.3.2.1.2). Therefore, the EPA disagrees that the proposal
failed to provide the proper references to the studies conducted
outside of the U.S., and that the public was not provided the
opportunity to provide comment on these studies.
Moreover, we disagree with the commenters that the EPA failed to
consider quantitative information from studies available in this
reconsideration. As detailed in sections 5.3.2.1.2 and 5.3.2.2 of the
2022 PA, and consistent with the information available in the 2020
review, a number of new studies are available that apply new methods to
characterize PM-related effects on previously studied materials;
however, the evidence remains insufficient to relate soiling or damage
to specific levels of PM in ambient air or to establish quantitative
relationships between PM and materials degradation. The uncertainties
in the evidence identified in the 2012 review persist in the evidence
in the 2020 review and in this reconsideration, with significant
uncertainties and limitations to establishing quantitative
relationships between particle size, concentration, chemical
components, and frequency of painting or repair of materials. While
some new evidence is available in the 2019 ISA, overall, the data are
insufficient to conduct quantitative analyses for PM-related materials
effects. Quantitative relationships have not been established between
characteristics of PM and frequency of repainting or cleaning of
materials, including photovoltaic panels and other energy-efficient
materials, that would help inform our understanding of the public
welfare implications of soiling in the U.S. (U.S. EPA, 2022b, section
5.3.2.2.2; U.S. EPA, 2019a, section 13.4). Similarly, the information
does not support quantitative analyses between microbial deterioration
of surfaces and the contribution of carbonaceous PM to the formation of
black crusts that contribute to soiling (U.S. EPA, 2022b, section
5.3.2.2.2; U.S. EPA, 2019a, section 13.4). We also note that
quantitative relationships are difficult to assess, in particular those
characterized using damage functions as these approaches depend on
human perception of the level of soiling deemed to be acceptable and
evidence in this area remains limited in this reconsideration (U.S.
EPA, 2022b, section 5.3.2.1.2). Additionally, we note the CASAC's
concurrence with conclusions in the 2020 PA (Cox, 2019b, p. 13 of
consensus responses) and the 2022 PA (Sheppard, 2022a, p. 23 of
consensus responses) that uncertainties remain about the best way to
evaluate materials effects of PM in ambient air. Further, no new
studies are available in this reconsideration to link human perception
of reduced aesthetic appeal of buildings and other objects to materials
effects and PM in ambient air. Finally, uncertainties remain about
deposition rates of PM in ambient air to surfaces and the interaction
of PM with copollutants on these surfaces (U.S. EPA, 2022b, section
5.6).
With respect to the commenters' assertion that the EPA failed to
consider information related to materials damage and degradation from
oxidant concentrations and solar radiation for a variety of materials,
we first note that, even assuming these sources of materials damage are
within the scope of this review of the PM NAAQS, the commenter did not
provide any references to the scientific studies that they suggest that
the EPA did not consider. Despite the lack of a list of specific
references from the commenter, we note that the 2019 ISA considered a
number of studies that examined the
[[Page 16328]]
relationships between PM and several of the materials listed by the
commenters (e.g., paint, plastic, rubber). However, as described in the
2022 PA, these studies did not provide additional information regarding
quantitative relationships between PM and materials that could inform
quantitative analyses (U.S. EPA, 2022b, sections 5.3.2.1.2 and
5.3.2.2.2), nor did they alter conclusions regarding the adequacy of
the current standard (U.S. EPA, 2022b, section 5.5).
As summarized above and in the proposal, the evidence in the 2020
review and in this reconsideration for PM-related effects on materials
is not substantively changed from that in the 2012 review. There
continues to be a lack of evidence related to materials effects that
establishes quantitative relationships and supports quantitative
analyses of PM-related materials soiling or damage. While the
information available in the 2020 review and in this reconsideration
continues to support a causal relationship between PM in ambient air
and materials effects (U.S. EPA, 2019a, section 13.4), the EPA is
unable to relate soiling or damage to specific levels of PM in ambient
air and is unable to evaluate or consider a level of air quality to
protect against such materials effects. Although the EPA did not
propose a distinct level of air quality or a national standard based on
air quality impacts (88 FR 5662, January 27, 2023), we did identify
data gaps that prevented us from doing so. The EPA identified a number
of key uncertainties and areas of future research (U.S. EPA, 2022b,
section 5.6) that may inform consideration of the materials effects of
PM in ambient air in future reviews of the PM NAAQS. The EPA notes that
one commenter objected to the Administrator's proposed conclusion in
the proposal (88 FR 5661, January 27, 2023) that in light of the
available evidence for PM-related impacts on climate and on materials
that it is appropriate not to change the existing secondary standards
at this time. The EPA has explained, in both the proposal and this
final action, the basis for its conclusion that there is insufficient
evidence to identify any particular secondary standard or standards
that would provide requisite protection against climate effects or
materials damage. The EPA acknowledges that, as a result, the adoption
of any distinct secondary PM standards for those effects would be
inconsistent with the requirements of the CAA. The EPA is clarifying
that it is not basing its decisions on secondary standards in this
reconsideration to address these welfare effects because it has
concluded that the available scientific evidence is insufficient to
allow the Administrator to make a reasoned judgment about what specific
standard(s) would be requisite to protect against known or anticipated
adverse effects to public welfare from PM-related materials damage or
climate effects.
Some commenters agree with the Administrator's proposed conclusion
that a target level of protection for visibility of 30 dv and the level
of the secondary 24-hour PM2.5 standard of 35 [micro]g/m\3\
continues to be adequate to protect visibility, highlighting
improvements in visibility in the U.S. Other commenters who disagree
with the proposed decision indicated support for a more stringent
standard for visibility impairment, although some of these commenters
did not necessarily specify the alternative standard that would, in
their judgment, address their concerns related to various aspects of
the EPA's proposal, including the available public preference studies,
specific aspects of the visibility index, and the target level of
protection identified by the Administrator. Rather, most commenters
focused on particular aspects of the visibility metric underlying the
current secondary 24-hour PM2.5 standard, including the
form, averaging time, and target level of protection necessary to
protect against visibility impairment.
With regard to the commenters' assertion that the current secondary
standards are inadequate to protect the public welfare from PM-related
visibility impairment, the EPA disagrees that the currently available
information is sufficient to suggest that a more stringent standard is
warranted. The EPA identified and addressed in great detail the
limitations and uncertainties associated with the public preference
studies as a part of the 2012 review (78 FR 3210, January 15, 2013).
Given that the evidence related to public preferences has not
substantially changed since the 2012 review, the EPA reiterated the
limitations and uncertainties inherent in the evidence as a part of the
2020 PA (U.S. EPA, 2020b, section 5.5), as well as in the 2022 PA for
this reconsideration (U.S. EPA, 2022b, section 5.6). The 2022 PA
highlights key uncertainties associated with public perception of
visibility impairment and identifies areas for future research to
inform future PM NAAQS reviews, including those raised by the
commenters (U.S. EPA, 2022b, section 5.6). Specifically, the EPA agrees
with commenters that there are several areas where additional
information would reduce uncertainty in our interpretation of the
available information for purposes of characterizing visibility
impairment. As described in more detail in the 2020 PA (U.S. EPA,
2020b, p. 5-41) and the 2022 PA (U.S. EPA, 2022b, p. 5-53), briefly,
these areas include: (1) Expanding the number and geographic coverage
of preference studies in urban, rural, and Class I areas; (2)
evaluating visibility preferences of the U.S. population today, given
that the preference studies were conducted more than 15 years ago,
during which time air quality in the U.S. has improved; (3) accounting
for the influence of varying study methods may have on an individual's
response as to what level of visibility impairment is acceptable, and;
(4) information on people's judgments on acceptable visibility based on
factors that can influence their perception of visibility (e.g.,
duration of impairment experiences, time of day, frequency of
impairment).
However, the EPA disagrees with the commenters that the current
secondary PM standards are inadequate and should be made more stringent
because of the limitations and uncertainties associated with the
available public preference studies. The EPA does not view the
limitations of the preference studies and other available evidence as
so significant as to render the EPA unable to identify a secondary
standard to protect against the adverse effects of PM on visibility,
but the EPA also does not believe that the limitations themselves mean
that the standards are inadequate. In fact, there is a limited amount
of recently available scientific evidence to further inform our
understanding of public preferences and visibility impairment is
recognized by the Administrator in reaching his proposed decision not
to change the current secondary PM standards at this time, given that
the evidence base is largely the same as at the time of the 2012 and
2020 reviews.
These same commenters further contend that the EPA failed to use
the latest science to develop a visibility index, stating that the EPA
failed to consider the contrast of distance methodology employed in a
recent meta-analysis of available preference studies (Malm et al.,
2019). Commenters claim that the EPA draws conclusions from the Malm et
al. (2019) study about how to relate contrast to acceptable visibility
preferences in the 2022 ISA Supplement, yet ignores the findings of the
study and fails to consider the ``contrast of distance'' methodology in
the 2022 PA and the proposal, thereby, in their view, departing from
the CASAC's advice to consider this evidence in setting the secondary
[[Page 16329]]
standard. Finally, the commenters assert that the EPA did not explain
why the available public preference studies are adequate for analysis
using a light extinction approach but not using the contrast of
distance approach, and that such differential treatment is arbitrary.
We disagree with the commenters that the EPA did not use the latest
science in evaluating the visibility index, and that the EPA failed to
consider the contrast of distance methodology used in Malm et al.
(2019). As the commenters state, the Malm et al. (2019) study was
included in the ISA Supplement (U.S. EPA, 2022a, section 4.2.1).
However, the EPA disagrees with the assertion that the ISA Supplement
reached conclusions about how to relate contrast to acceptable
visibility preferences. The ISA Supplement provided an overview of the
Malm et al. (2019) study, stating that ``[t]he main conclusion of this
study was that the level of acceptable visual air quality is more
consistent across studies using metrics that evaluate the distinction
of an object from a background than using metrics that evaluate the
greatest distance at which an object can be observed.'' Furthermore,
the statements that the commenters are referencing in support of this
statement (i.e., U.S. EPA, 2022b, pp. 4-5-4-6) are in fact the
conclusions of the study itself, rather than conclusions of the EPA.
For example, the ISA Supplement notes that ``Malm et al. (2019)
suggested that scene-dependent metrics like contrast, which integrate
the effects of bext along the sight paths between observers
and landscape features, are better predictors of preference levels than
universal metrics like light extinction.'' The suggestion that the
contrast of distance methodology is a better predictor than light
extinction is one of the study authors, not the EPA. The EPA has not
reached a conclusion on whether contrast of distance methodology would
be a more appropriate indicator for a visibility index than estimated
light extinction because the EPA finds that there is insufficient
information in the record at this time to support that it is practical
to evaluate, much less adopt, the contrast of distance methodology on a
national basis. Specifically, the Malm et al. (2019) study does not
provide as a part of their publication the specific input values to the
equation to calculate the contrast of distance associated with the
available public preference studies (e.g., sight paths from the
images), nor do the preference studies present or make publicly
available these data in their publications. In the absence of
additional studies or publicly available data to further evaluate the
contrast of distance methodology, the EPA is unable to consider
contrast of distance as an alternative to estimated light extinction in
this reconsideration, although we note that it may be appropriate to
evaluate it more closely in future reviews.
In reaching conclusions regarding the appropriate indicator for the
visibility index, the 2022 PA specifically notes ``that limited new
research is available on methods of characterizing visibility or on how
visibility is valued by the public, such as visibility preference
studies. Thus, while limited new research has further informed our
understanding of the influence of atmospheric components of
PM2.5 on light extinction, the available evidence to inform
consideration of the public welfare implications of PM-related
visibility impairment remains relatively unchanged'' (U.S. EPA, 2022b,
p. 5-50). The EPA again notes in the proposal that ``there are very few
studies available that use scene-dependent metrics (i.e., contrast) to
evaluate public preference information, which makes it difficult to
evaluate them as an alternative to the light extinction approach'' (88
FR 5649-5650, January 27, 2023). To further expand on this statement,
the Malm et al. (2019) study does not provide enough information to
replicate the results of their contrast of distance approach to allow
for a comprehensive evaluation of the potential use of this methodology
in considering the results of the public preference studies for
determining the target level of protection for visibility.
Some commenters suggests that the methodology could be approximated
by simply ensuring that people could always see distant scenic
elements, and that characterizing typical average and/or maximal
viewing distances cross different geographical areas and regions would
be a straightforward Geographical Information Systems (GIS) exercise.
The EPA disagrees that this assessment would be straightforward, given
the lack of data establishing viewing distances in the available
scientific record and the diversity of distance to scenic elements
across different areas and regions of the U.S., and finds that this
approach is also not practical to adopt in this reconsideration.
Finally, while the Malm et al. (2019) study is using an alternative
approach for evaluating public preferences and acceptability, we note
that this study is evaluating the same public preference studies that
have been available for the past several decades. For these reasons,
the EPA disagrees with the commenters' allegation that the EPA ignored
the findings of the Malm et al. (2019) study and failed to consider the
contrast of distance methodology in the 2022 PA and the proposal, and
ignored the CASAC's advice to consider this study. The ISA Supplement
and the 2022 PA considered the Malm et al. (2019) study, along with the
full body of available scientific evidence, and took into account the
uncertainties and limitations associated with the evidence for
visibility preferences, in reaching conclusions regarding the adequacy
of the secondary 24-hour PM2.5 standard (U.S. EPA, 2022b,
pp. 5-24-5.25, 5-50).
Several comments in support of revising the secondary 24-hour
PM2.5 standard to protect against visibility generally
recommend revisions to the elements of the standard and visibility
index (indicator, averaging time, form, and level) consistent with
those supported by the CASAC and public comments in previous PM NAAQS
reviews. Some commenters assert that the EPA's approach in the 2022 PA
and in the proposal for this reconsideration did not evaluate options
for alternative secondary PM standards and thereby is flawed. We
address comments on the elements of a visibility index and a revised
standard for visibility effects below.
As an initial matter, the EPA disagrees to the extent commenters
are suggesting that the PA is legally required to analyze options for
alternative standards. The PA is a document developed by the EPA in
order to assist the Administrator and the CASAC in reaching conclusions
regarding the adequacy of the current standards, and its scope is
determined by the EPA. Moreover, the 2022 PA did assess a wide range of
information relevant to the Administrator's decision and considered a
range of potential standards.
First, in developing the 2022 PA and in responding to CASAC's
advice and recommendations during its review of the 2021 draft PA, the
EPA expanded upon its discussion of determining the target level of
protection for the visibility index and considered the extent to which
the available scientific information would alter regarding the
visibility index and the appropriate target level of protection against
PM-related visibility effects (U.S. EPA, 2022b, pp. 5-27-5-29). This
detailed discussion expands the consideration of the target level of
protection for the visibility index presented in the 2020 PA (U.S. EPA,
2020b) and the 2021 draft PA (U.S. EPA, 2021c), neither of which
specifically considered the elements of the visibility index in
determining the appropriate target level of protection. In
[[Page 16330]]
considering the available information in the 2022 PA, the EPA concluded
that the available information continued to provide support for a
visibility index with a level of 30 dv, with estimated light extinction
as the indicator, a 24-hour averaging time, and a 90th percentile form,
averaged over three years.
Additionally, in summarizing the air quality and quantitative
information in the proposal for this reconsideration, the EPA further
expands upon the discussion added to the 2022 PA related to the target
level of protection in terms of a PM2.5 visibility index. In
so doing, the EPA considers even more extensively the available public
preference studies and quantitative analyses (88 FR 5651-5652, January
27, 2023). In particular, there is a more detailed discussion of the
public preference studies, including the levels of impairment
determined to be ``acceptable'' by at least 50 percent of study
participants and the methodologies used in the studies, including
uncertainties and limitations associated with the methodologies (88 FR
5652, January 27, 2023). In reaching a proposed decision regarding the
adequacy of the secondary PM standards, as well as the appropriate
target level of protection for the visibility index, the Administrator
considered the available scientific evidence and quantitative analyses,
as well as judgments about how to consider the range and magnitude of
uncertainties that are inherent in the scientific evidence and
analyses. In so doing, the Administrator proposed to conclude that the
protection provided by a visibility index based on estimated light
extinction, a 24-hour averaging time, and a 90th percentile form,
averaged over 3 years, set at a level of 30 dv would be requisite to
protect public welfare with regard to visibility impairment (88 FR
5660, January 27, 2023).
Having provisionally concluded that it was appropriate to define
the target level of protection in terms of a visibility index based on
estimated light extinction as described above (i.e., with a 24-hour
averaging time; a 3-year, 90th percentile form; and a level of 30 dv),
the Administrator next considered the degree of protection from
visibility afforded by the current secondary PM standards. In so doing,
he considered the updated analyses of PM-related visibility impairment
presented in the 2022 PA (U.S. EPA, 2022b, section 5.3.1.2) and
described in more detail in the proposal (88 FR 5656, January 27,
2023), which included estimating light extinction using multiple
versions of the IMPROVE algorithm and inclusion of PM10-2.5
data at all sites to allow for better characterization of the influence
of the coarse fraction of PM on light extinction. The Administrator
noted that the results of the analyses in the 2022 PA were consistent
with those from the 2012 and 2020 reviews. He also recognized that,
regardless of the IMPROVE equation that was used, the analyses
demonstrated that the 3-year visibility metric is at or below 28 dv in
all areas meting the current 24-hour PM2.5 standard (88 FR
5657, January 27, 2023). The Administrator also noted that, in their
review of the 2021 draft PA, the CASAC stated that ``[i]f a value of
20-25 deciviews is deemed to be an appropriate visibility target level
of protection, then a secondary 24-hour standard in the range of 25-35
[micro]g/m\3\ should be considered (Sheppard, 2022a, p. 21 of consensus
responses). The Administrator recognized that while the CASAC
recommended that additional justification be provided for a visibility
index target level of protection of 30 dv, they did not specifically
recommend that he choose an alternative level for the visibility index.
Therefore, the Administrator considered the available scientific
evidence, quantitative information, and the CASAC's advice in reaching
his proposed conclusions. The Administrator recognized conclusions
regarding the appropriate weight to place on the scientific and
technical information, including how to consider the range and
magnitude of uncertainties inherent in that information, is a public
welfare policy judgment left to the Administrator. As such, the
Administrator noted his preliminary conclusion on the appropriate
visibility index (i.e., with a 24-hour averaging time; a 3-year, 90th
percentile form; and a level of 30 dv) and his preliminary conclusions
regarding the quantitative analyses of the relationship between the
visibility index and the current secondary 24-hour PM2.5
standard. In so doing, he proposed to conclude that the current
secondary standards provide requisite protection against PM-related
visibility effects (88 FR 5661, January 27, 2023).
However, the Administrator additionally recognized that the
available evidence on visibility impairment generally reflects a
continuum and that the public preference studies did not identify a
specific level of visibility impairment that would be perceived as
``acceptable'' or ``unacceptable'' across the whole U.S. population. He
noted a judgment of a target level of protection, below 30 dv and down
to 25 dv, could be supported if more weight was put on the public
preference study performed in the Phoenix, AZ, study (BBC Research &
Consulting, 2003). As described above, while the Administrator noted
that the CASAC did not recommend revising the level of the current 24-
hour PM2.5 standard in their review of the 2021 draft PA,
they did state that, should an alternative level be considered for the
visibility index, revisions to the secondary 24-hour PM2.5
standard should also be considered (Sheppard, 2022a, p. 21 of consensus
responses). Thus, the Administrator solicited comment on the
appropriateness of a target level of protection for visibility below 30
dv and down as low as 25 dv, and of revising the level of the current
secondary 24-hour PM2.5 standard to a level as low as 25
[micro]g/m\3\ (88 FR 5662, January 27, 2023), and the Administrator
considered these public comments in reaching his final decision on the
secondary standards. Thus, the EPA disagrees that the 2022 PA and the
proposal did not adequately consider options for revising the secondary
PM NAAQS.
With regard to the elements of the visibility index, in considering
the adequacy of the current secondary 24-hour PM2.5 standard
to protect against visibility impairment, as described in the proposal
(88 FR 5658-5660, January 27, 2023), the Administrator first defined an
appropriate target level of protection in terms of a PM visibility
index. In considering the information available in this reconsideration
and the CASAC's advice, the Administrator proposed to conclude that the
protection provided by a visibility index based on estimated light
extinction, a 24-hour averaging time, and 90th percentile form,
averaged over 3 years, set at a level of 30 dv, would be requisite to
protect public welfare with regard to visibility impairment (88 FR
5660, January 27, 2023).
In defining this target level of protection, the Administrator
first considered the indicator of such an index. He noted that, given
the lack of availability of methods and an established network for
directly measuring light extinctions, a visibility index based on
estimates of light extinction by PM2.5 components derived
from an adjusted version of the original IMPROVE algorithm would be
most appropriate, consistent with the 2012 and 2020 reviews. As
described in the proposal (88 FR 5649, January 27, 2023) and above
(section V.A.2), the IMPROVE algorithm estimates light extinction using
routinely monitored components of PM2.5 and
PM10-2.5, along with estimates of relative humidity. The
[[Page 16331]]
Administrator, while recognizing that some revisions to the IMRPOVE
algorithm were newly available in the 2020 review, noted that the
fundamental relationship between ambient PM and light extinction has
changed very little and the different versions of the IMPROVE
algorithms can appropriately reflect this relationship across the U.S.
(88 FR 5658-5659, January 27, 2023). As such, he judged that defining a
target level of protection in terms of estimated light extinction
continues to be a reasonable approach in this reconsideration.
Some commenters who criticized the EPA's interpretation and
application of the Malm et al. (2019) study also contend that an
indicator based on the contrast of distance would be a significant
improvement over the current indicator for the visibility index and
would more accurately evaluate public preferences. However, as
described in the 2022 PA (U.S. EPA, 2022b, section 5.3.1.1), while
scene-dependent metrics, such as contrast, may be useful alternative
predictors of preferences compared to universal metrics like light
extinction, there are a very limited number of studies that use such
metrics to evaluate public preferences of visibility impairment and
there is a lack of scientific evidence that supports one metric over
another. Moreover, the EPA finds that even if the Administrator agreed
that the contrast of distance methodology was an improvement over light
extinction, there is insufficient information available to evaluate and
adopt contrast of distance as an indicator for a national visibility
target at this time. While, in its review of the 2021 draft PA the
CASAC suggested that the EPA consider this method in developing the
secondary PM standards, the CASAC also noted that ``more extensive
technical evaluation of the alternatives for visibility indicators and
practical measurement methods'' is needed to inform future reviews of
the secondary PM standards (Sheppard, 2022a, p. 22 of consensus
responses). The CASAC did not recommend using a different indicator for
this reconsideration, with the majority of CASAC members reiterated
past advice recommending development of a visibility FRM for a directly
measured PM2.5 light extinction indicator (Sheppard, 2022a,
p. 22 of consensus responses), a recommendation that was supported by
other public commenters as well, and the minority of the CASAC
suggested that such an FRM is not necessary. For these reasons, the EPA
does not consider it feasible or appropriate to define the visibility
index in terms of a contrast of distance indicator at this time.
With regard to averaging time, some commenters suggested to the EPA
that a secondary standard with a different form than the primary
standard may be a more relevant for welfare effects. While they do not
recommend a specific alternative form, the commenters point to CASAC
advice in past reviews where the CASAC stated that a subdaily standard
based on daylight hours better reflects visibility impairment.
In defining the characteristics of a visibility index, the EPA
continues to believe that a 24-hour averaging time is reasonable. This
is in part based on analyses conducted in the 2012 review that showed
relatively strong correlations between 24-hour and subdaily (i.e., 4-
hour average) PM2.5 light extinction (88 FR 5659, January
27, 2023; 85 FR 82740, December 18, 2020; 78 FR 3226, January 15,
2013), indicating that a 24-hour averaging time is an appropriate
surrogate for the subdaily time periods relevant for visual perception.
The EPA believes that these analyses continue to provide support for a
24-hour averaging time for the visibility index in this
reconsideration. The EPA also recognizes that the longer averaging time
may be less influenced by atypical conditions and/or atypical
instrument performance (88 FR 5659, January 27, 2023; 85 FR 82740,
December 18, 2020; 78 FR 3226, January 15, 2013). When taken together,
the available scientific information and updated analyses of calculated
light extinction available in this reconsideration continue to support
that a 24-hour averaging time is appropriate when defining a target
level of protection against visibility impairment in terms of a
visibility index.
Moreover, the EPA disagrees with commenters that a secondary
PM2.5 standard with a 24-hour averaging time does not
provide requisite protection against the public welfare impacts of
visibility impairment. At the time of the 2012 review, the EPA
recognized that hourly or subdaily (i.e., 4- to 6-hour) averaging
times, within daylight hours and excluding hours with high relative
humidity, are more directly related to the short-term nature of
visibility impairment and the relevant viewing periods for segments of
the viewing public than a 24-hour averaging time. At the time of the
2012 review, the EPA agreed that a subdaily averaging time would
generally be preferable. However, the Agency noted significant data
quality uncertainties associated with the instruments that would
provide hourly PM2.5 mass concentrations necessary to inform
a subdaily averaging time. These uncertainties, as described in the
2012 review, included short-term variability in hourly data from
available continuous monitoring methods, which would prohibit
establishing a subdaily averaging time (78 FR 3209, January 15, 2013).
For all of these reasons, and consistent with the 2020 review, the EPA
continues to believe that a subdaily averaging time is not supported by
the information available in this reconsideration.
With regard to the form of the visibility index, some commenters
contend that the form used in evaluating visibility impairment is not
appropriate. First, commenters contend that the EPA incorrectly stated
that the CASAC did not provide advice on the 3-year, 90th percentile
form of the visibility index and that the CASAC specifically
recommended that the EPA further justify the metric and form, and by
not doing so, the proposal arbitrarily departs from the CASAC's
recommendations. The commenters also contend that the EPA fails to
explain how averaging the form over three years is protective given
that the public does not perceive visibility in three-year averages.
We disagree with the commenters that the EPA departed from the
CASAC's recommendations that ``[t]he final PA should provide a robust
justification for the daily light extinction percentile used in the
analysis'' (Sheppard, 2022a, p. 22 of consensus responses). In this
statement, the CASAC did not make explicit recommendations for
revisions to the form of the visibility index, as the commenters
assert, but rather requested additional justification for the
percentile selected for the visibility index in the 2022 PA. In
response to the CASAC's recommendation after reviewing the 2021 draft
PA, the EPA included a new section in the 2022 PA that explicitly
discusses the elements (i.e., indicator, averaging time, form, and
level) of the visibility index, including additional justification for
the conclusions regarding the appropriate elements for the index (U.S.
EPA, 2022b, pp. 5-27-5-29). In so doing, the 2022 PA recognizes that
there is no new information available in this reconsideration to inform
selection of an alternative form of the visibility index, and
therefore, relied on the analyses presented in the 2010 UFVA that
evaluated the different statistical forms of the visibility index. The
2022 PA also discusses the approach to improving visual air quality in
Federal Class I areas as a part of the Regional Haze Program (U.S. EPA,
2022b, p. 5-28). Furthermore, as reflected in responding to public
comments below, and in
[[Page 16332]]
reaching his final conclusions in section V.B.4 below, the
Administrator further considers the available scientific and
quantitative information, the CASAC's advice, and public comments in
informing his final conclusions regarding the appropriate target level
of protection for the visibility index. With regard to the commenters'
assertion that the EPA did not justify why averaging the form over
three years is protective, we agree with the commenters that people do
not perceive visibility impairment in three year averages. As described
in the 2022 PA, visibility-related effects and perceived impairment are
often associated with short-term PM concentrations, and therefore, the
focus of the visibility analyses is centered on the adequacy of the 24-
hour PM2.5 standard (U.S. EPA, 2022b, p. 5-29). However, as
described in the 2022 PA, the 3-year average form provides stability
from the occasional effect of inter-annual meteorological variability
that can result in unusually high pollution levels for a particular
year (U.S. EPA, 2022b, p. 5-28). Occasional meteorological variability
is of particular concern for the visibility index, which can be
impacted by not only PM concentrations in ambient air but also relative
humidity. The D.C. Circuit has previously recognized that it is
legitimate for the EPA to consider overall stability of the standard
and its resulting promotion of overall effectiveness of NAAQS control
programs in setting a standard. See American Trucking Ass'ns v.
Whitman, 283 F.3d 355, 375-76 (D.C. Cir. 2002). The 2022 PA concluded
that the available information continues to provide support for a 90th
percentile form, averaged over three years, and the inclusion of
additional justification for the elements of the visibility index
responds to the CASAC's recommendation (U.S. EPA, 2022b, section
5.3.1.2).
Some commenters suggest that the 90th percentile form is too low
and would result in 36 days being excluded annually, presuming that the
public only finds it objectionable when visibility is worse than the
standard on 37 or more days per year. The commenters also contend that
the EPA's approach of using a 90th percentile form for the visibility
index is inconsistent with the goals of the Regional Haze Program. In
so doing, the commenters note that the Regional Haze Rule focuses on
improving conditions on the worst days, while they argue that a 90th
percentile form for the visibility index would ignore the 36 worst
visibility days, rather than identifying them and reducing pollution on
those days.
In reaching conclusions regarding the appropriate form of the
visibility index, the EPA is following the same approach employed in
past reviews of the secondary PM NAAQS, including those in the 2012 and
2020 rulemakings. In reaching conclusions regarding the appropriate
form of the visibility index in the 2011 PA, the EPA considered the
percentile forms of the visibility index assessed in the 2010 PA (i.e.,
90th, 95th, 98th) along with the approach for improving visual air
quality under the Regional Haze Program. In so doing, the 2011 PA notes
that the Regional Haze Program targets the 20% most impaired days for
improvements in visual air quality in Federal Class I areas (i.e., the
days more impaired than the 80th percentile). The 2011 PA recognized
that to increase the likelihood of improving visual air quality on the
worst days, the form of the visibility index should be set well above
the 80th percentile. The 2011 PA further concluded that a 90th
percentile form would represent the median of the distribution of the
20% most impaired days, and meeting a visibility index with a 90th
percentile form would mean that 90% of the days have visual air quality
that is at or below the level of the visibility index and would
reasonably expected to lead to improvements in visual air quality for
the 20% most impaired days (U.S. EPA, 2011, p. 4-59). The 2022 PA noted
that there is no new information from public preference studies that
would inform the Administrator's consideration of the appropriate form
for the visibility target index, and reached conclusions consistent
with those of 2011 PA. However, as discussed below, the EPA disagrees
that a focus on the 90th percentile ``ignores'' any days with worse
visibility. It is possible to examine past patterns of air quality to
judge the relationship between the 90th percentile and higher
percentiles, and to assess whether achieving a 90th percentile
visibility target will also result in air quality improvements, where
necessary, at higher percentiles. Based on its assessment of past air
quality and potential alternative percentiles for the form, the EPA
judged that a 90th percentile would appropriately achieve improved air
quality both above and below that percentile.
Some commenters suggest that the analyses conducted in the 2010
UFVA are based on a different metric than the 24-hour average being
considered in the reconsideration, that the analyses are outdated and
irrelevant. Therefore, the commenters assert that relying on the
analyses in the 2010 UFVA is not a rational justification for the use
of a 90th percentile for the visibility index in this reconsideration.
Moreover, these commenters state that, in past reviews, both the EPA
and the CASAC have considered and recommended a 98th percentile form,
but the proposal does not consider the 98th percentile.
[[Page 16333]]
These commenters assert that the 2010 UFVA was not considering the
same metric under consideration here. However, the EPA was citing to
the 2010 UFVA for the conclusion that there are correlations between
different statistical forms of the visibility index. To confirm whether
these correlations occur under recent air quality, we conducted
additional air quality analyses evaluating the visibility index using
the current percentile form (i.e., 90th) and two alternative forms
(i.e., 95th and 98th).\172\ While a higher percentile form would
further limit the number of days with peak PM-related light extinction,
the analyses confirm that a 90th percentile form is effective in
limiting visibility impairment at higher percentiles. Based on these
analyses, depending on which version of the IMPROVE equation is used to
estimate light extinction, the differences in the 3-year averages of
estimated light extinction for the 90th, 95th, and 98th percentile
forms are small. For example, in areas that meet the current 24-hour
PM2.5 standard, for light extinction estimated using the
original IMPROVE equation, all sites have light extinction estimates
for a 90th percentile form at or below 26 dv, for a 95th or 98th
percentile form at or below 29 dv.\173\ In most locations, when
estimating light extinction based on the original IMPROVE equation, the
difference between a 95th or 98th percentile form and a 90th percentile
form is generally less than 3 dv.\174\ As noted in previous reviews, a
change of 1 to 2 dv in light extinction under many viewing conditions
will be perceived as a small, but noticeable, change in the appearance
of a scene, regardless of the initial amount of visibility impairment
(88 FR 5657, January 27, 2023; U.S. EPA, 2004b; U.S. EPA, 2010b). Thus,
differences between a 90th percentile form and a 95th or 98th
percentile form remain small, and for any of these forms of the
visibility index, the estimated light extinction based on the original
IMPROVE equation in areas meeting the current secondary 24-hour
PM2.5 standard is below the upper end of the range of the
levels considered for the visibility index (i.e., below 30 dv).
---------------------------------------------------------------------------
\172\ Gantt, B., and Hagan, N. (2023). Analysis of Percentile
Forms of the Visibility Index. Memorandum to the Rulemaking Docket
for the Review of the National Ambient Air Quality Standards for
Particulate Matter (EPA-HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
\173\ Gantt, B., and Hagan, N. (2023). Analysis of Percentile
Forms of the Visibility Index. Memorandum to the Rulemaking Docket
for the Review of the National Ambient Air Quality Standards for
Particulate Matter (EPA-HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
\174\ Gantt, B., and Hagan, N. (2023). Analysis of Percentile
Forms of the Visibility Index. Memorandum to the Rulemaking Docket
for the Review of the National Ambient Air Quality Standards for
Particulate Matter (EPA-HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
Some commenters disagree with the EPA's proposed conclusion that a
level of 30 dv is appropriate for the visibility index and support a
lower level in order to provide increased protection against visibility
impairment. Commenters who support a revised level for the visibility
index state that a target level of protection of 30 dv would mean that
less than 10% of participants in the public preference studies, other
than the Washington, DC, study, would accept visibility conditions
above 29 dv. These commenters further suggest that a 75% acceptability,
rather than 50% acceptability, is requisite to protect visibility
sources, which would be on average a level of 21 dv when using the
light extinction method or 18 dv when using the contrast of distance
method. These commenters argue that, based on the available
information, a target level of protection for the visibility index of
approximately 20 dv would be more appropriate, and therefore, the level
of the secondary 24-hour PM2.5 standard should be
strengthened to 25 [mu]g/m\3\. Other commenters who support a revised
level for the visibility index suggest that public preference studies
with longer sight paths to distant landscape features or with lower
target levels than those in the Washington, DC study, such as the
Phoenix study, would support a lower level. These commenters support
revising the target level of protection for the visibility index to a
25 dv, and revising the level of the secondary 24-hour PM2.5
standard to a level as low as 25 [mu]g/m\3\, suggesting that in low
relative humidity environments, 25 dv is consistent with
PM2.5 concentrations of less than 25 [mu]g/m\3\.
Some commenters state that EPA's justification for setting a target
level of protection at the upper end of the 20 to 30 dv range is
arbitrary. These commenters state that the EPA's reliance on the
standard operating in many regions and circumstances as support for the
upper end of the range is irrational and illegal. Moreover, these
commenters contend that EPA provided no rational connection between the
Regional Haze Program and the proposed decision to set the target level
of protection at the upper end of the range. They suggest that the EPA
proposed to rely exclusively on the Regional Haze Program to protect
visibility in Class I areas and to give visibility in these areas no
weight in considering the secondary PM standard and that it is not
rational to entirely ignore visibility in Class I areas when setting
the secondary standard. These commenters assert that the Regional Haze
Program provides no rational basis for a target level of protection at
the upper end of the range, nor does the EPA identify one.
Some commenters contend that the EPA failed to justify the adequacy
of the current secondary annual PM2.5 standard, noting that
the secondary 24-hour and annual PM2.5 standards work
together to provide protection against short- and long-term effects of
PM2.5. These commenters point to CASAC comments on the 2021
draft PA and the comments of an individual CASAC member's support for
strengthening the secondary annual PM2.5 standard to provide
increased protection against climate and materials effects over time.
They contend that EPA arbitrarily failed to discuss the secondary
annual PM2.5 standard not only in the proposal, but also in
the 2022 PA and in the 2020 final decision.
The EPA recognizes that the selection of the target level of
protection for the visibility index is fundamentally a public welfare
policy judgment for the Administrator. The Administrator is tasked by
the CAA to judge when visibility impairment becomes an adverse effect
on public welfare. It is clear that visibility impairment can become
adverse to public welfare, but the Administrator does not consider that
every deciview of impairment is adverse to public welfare. In
considering the point at which visibility impairment becomes adverse to
public welfare, such that the attainment of the secondary PM NAAQS
would prevent the adverse effect, the Administrator gives weight to the
public preference studies as to when visibility impairment is
unacceptable. At the same time, the Administrator recognizes the
limitations of these studies, which have been detailed in the proposal
and the 2022 PA. Similarly, the EPA discussed the Regional Haze program
in the proposal to highlight that there is a distinct program to
protect against visibility impairment in Class I areas, and the
existence of that program is relevant to the Administrator's judgment
about the level of visibility impairment that is adverse to public
welfare under CAA 109(d), because in determining what is requisite the
Administrator is primarily considering visibility impairment outside of
Class I areas.
[[Page 16334]]
In considering how to use the results of the public preference
studies, the Administrator concludes that a 50th acceptability
criterion is an appropriate tool. The Administrator's task is to set
standards that are neither more stringent nor less stringent than
necessary, and a 50% acceptability criterion seems most appropriate to
use in judging when visibility impairments become adverse, because it
should more closely represent when the median person would find the
impairment to be adverse. The Administrator notes this conclusion is
consistent with the approach adopted in the Denver study by Ely et al.
(1991) where the 50% acceptability criterion for urban visibility was
first presented. This study discussed the use of the 50% acceptability
criteria as a reasonable basis for setting a standard to protect
visibility in urban areas. In doing so, Ely et al. (1991) noted that
the 50% acceptability criterion divided the slides into two groups--
those judged acceptable and those judged unacceptable by a majority of
people in the study--and therefore, was reasonable since it defines the
point where the majority of the study participants began to judge
levels of visibility impairment as unacceptable (Ely et al., 1991).
In considering the appropriate target level of protection, we next
look to the available public preference studies, noting that the
selecting of the range of 20 to 30 dv for the target level of
protection for the visibility index is informed by the 50%
acceptability values from these studies. The Denver, CO, (Ely et al.,
1991) and British Columbia, Canada, (Pryor, 1996) studies met the 50%
acceptability criteria at 20 dv and 19-23 dv, respectively (U.S. EPA,
2022b, Table D-8). As described in the proposal, these studies used
photographs that were taken at different times of the day and on
different days to capture a range of light extinction levels needed for
the preference studies (88 FR 5652, January 27, 2023). Compared to
studies that used computer-generated images (i.e., those in Phoenix,
AZ, and Washington, DC) there was more variability in scene appearance
in these older studies that could affect preference rating and includes
uncertainties associated with using ambient measurements to represent
sight path-averaged light extinction values rather than superimposing a
computer-generated amount of haze onto the images. When using
photographs, the intrinsic appearance of the scene can change due to
meteorological conditions (i.e., shadow patterns and cloud conditions)
and spatial variations in ambient air quality that can result in
ambient light extinction measurement not being representative of the
sight-path-averaged light extinction. Computer-generated images, such
as those generated with WinHaze, do not introduce such uncertainties,
as the same base photograph is used (i.e., there is no intrinsic change
in scene appearance) and the modeled haze that is superimposed on the
photograph is determined based on uniform light extinction throughout
the scene. Because of the uncertainties and limitations associated with
the Denver, CO, and British Columbia, Canada, the EPA concludes that it
is appropriate to place less weight on these studies, and to instead
focus on the public preference studies that were designed to reduce
these uncertainties and limitations.
In so doing, we focus on the public preference studies that use
computer-generated images (i.e., those in the Phoenix, AZ, and
Washington, DC) studies. As described in the proposal, the use of
computer-generated images have less variability in scene appears than
in those studies that use photographs taken on different days and at
different times of the days (i.e., those in the Denver, CO, study) that
would be likely to influence preference rating and introduces
uncertainties associated with using ambient measurements to present
sight path-averaged light extinction values rather than superimposing a
computer-generated amount of haze onto the images (88 FR 5652, January
27, 2023).
The Phoenix, AZ, public preference study (BBC Research &
Consulting, 2003) had several strengths compared to some of the other
public preference studies. The Phoenix, AZ, study had the largest
number of participants (385 in 27 separate focus group sessions) of all
of the public preference studies, with a sample group designed to be
demographically representative of the Phoenix population at that time.
The age range in the Phoenix study was also more inclusive (18-65+),
with the distribution of the study participants corresponding
reasonably well to the overall age distribution in the 2000 U.S. Census
for the Phoenix area (BBC Research & Consulting, 2003). Furthermore,
the 21 images used in the Phoenix, AZ, study were developed using the
WinHaze software with visual air quality ranging from 15 to 35 dv, and
the view was toward the southwest, including downtown Phoenix, with the
Sierra Estrella Mountains in the background at a distance of 25 miles.
This study had the least noisy preference results, perhaps because a
larger, more representative group of participants combined with the use
of computer-generated images resulted in the smoother distribution of
responses of ``acceptable'' visual air quality. Based on the EPA's
evaluation of the public preference studies in the 2012 review, the 50%
``acceptable'' criteria was met at approximately 24 dv (U.S. EPA, 2010,
Table 2-3).
We also consider the public preferences for the Washington, DC,
studies (Abt Associates, 2001; Smith and Howell, 2009). The 2001
Washington, DC study included nine participants, and the 2009
Washington, DC, study replicated the 2001 study with 26 additional
participants. Similar to the Phoenix study, the Washington, DC, studies
also had the strength of having the 20 images included in the study
generated using WinHaze with visual air quality ranging from 9 to 45
dv. The study depicted a scene of a panoramic view of the Potomac
River, the National Mall, and downtown Washington, DC. All of the
distinct buildings in the scene were within four miles and the higher
elevations in the background were less than 10 miles from where the
image was taken from the Arlington National Cemetary in Virginia. The
50% ``acceptable'' criteria was met at approximately 29 dv (U.S. EPA,
2010, Table 2-3).
As described in more detail in the proposal, visibility preferences
can vary by location, and such differences may arise based on the
differences in the cityscape scene that is depicted in the images (88
FR 5652, January 27, 2023). In considering the geographical differences
between the public preference studies, we recognize that the
methodological differences between the studies may influence the
resulting ``acceptable'' level of visibility impairment. In the
Phoenix, AZ, study, the image depicted mountains in the background and
urban features in the foreground, whereas the Washington, DC, study
depicted nearby buildings in the image without mountains in the
distance. As an initial matter, we note that the object of interest to
the study participant could differ across the studies based on the
scenes included in the images being evaluated--with the mountains being
of greater interest in the images in the Phoenix, AZ, study, despite
also depicting buildings that are similar to those shown and presumed
to be of interest in the images in the Washington, DC, study (88 FR
5652, January 27, 2023). We also agree with the commenters that the
distance between the object of interest and the camera is an important
consideration in
[[Page 16335]]
evaluating the public preference studies. Objects at greater distances
from the camera location (such as those in the Phoenix, AZ, study which
had a maximum distance of 42 km (U.S. EPA, 2022b, Table D-8)) have a
greater sensitivity to light extinction, which alone could explain
differences in preferences but coupled with an object of greater
interest results in lower acceptable levels of visibility impairment.
Conversely, objects at closer distances from the camera location (such
as those in the Washington, DC, study which had a maximum distance of 8
km (U.S. EPA, 2022b, Table D-8)) have less sensitivity to light
extinction, which coupled with objects of interest (compared to the
mountainous views in the Phoenix, AZ, study) result in higher
acceptable levels of visibility impairment. These studies clearly
demonstrate that there are differences in the public preferences across
the studies depending on the images that are used, in particular the
object of interest to the study participant depicted in the image and
the distance of the sight path to the object, and that such differences
can influence preference results.
However, we note that these uncertainties and limitations have
persisted from past reviews, and there is very little new information
to inform conclusions regarding the interpretation of these results
with regard to the target level of protection. In selecting a target
level of protection, and in considering the CASAC's advice in their
review of the 2021 draft PA and public comments, we conclude that it is
appropriate to consider the information from the public preference
studies in Washington, DC, and Phoenix, AZ, and in so doing, that it is
appropriate to place weight on both of these studies in reaching
conclusions on the appropriate target level of protection. The EPA
recognizes that the scenes depicted in these two studies are different
and may influence public preferences of visibility impairment, but
notes these studies can be considered together as providing information
about different areas across the U.S. with variations in the scenes
that people are likely to most commonly encounter. The scene depicted
in the images used in the Washington, DC, study have a mix of
buildings, landmarks, and open space. On the other hand, the scene
depicted in the Phoenix, AZ, study included a mix of buildings in the
foreground and with more distant mountains in the background. The
Administrator considers it appropriate to consider these studies
together because in combination, they provide a greater diversity of
scenes, which is more likely to be representative of scenes people
typically experience around the country (e.g., not only in eastern
metropolitan statistical areas, but also in western areas with
different vistas). In considering these two studies together, the EPA
recognizes that, first, the ``object of interest'' is a subjective
judgment left to the participants of the public preference studies, and
second, the images in these two studies may differ in terms of
sensitivity to changes in light extinction because of the distance
between the object of interest in the scene and the camera. As noted by
the public commenters, the sight path for the images in the public
preference studies is an important consideration in reaching
conclusions regarding the appropriate target level of protection for
the visibility index. In addition, the Administrator judges that giving
weight to multiple studies is a more appropriate approach than focusing
on a single study, particularly where the study design (including the
representativeness of the participants and the scenes depicted in the
images) may be important for interpreting the results of the public
preference studies for informing conclusions regarding the visibility
index. Given these considerations and taking into consideration public
comments on the target level of protection for the visibility index,
the Administrator recognizes that it is more appropriate to consider a
broader range of public preferences, reflecting a broader range of
scenes, by putting significant weight on both the Washington, DC, and
Phoenix, AZ, studies. In so doing, he reaches the conclusion that it
would be appropriate to identify secondary PM standards that generally
limit visibility impairment to a level between the two studies.
The Administrator next considers what target level of protection
would be appropriate based on the available information from these
public preference studies. He first recognizes that, in the 2012 and
2020 final decisions, the then-Administrators selected a target level
of protection of 30 dv, based on the upper end of the range. In so
doing, the then-Administrators judged that it was appropriate to place
more weight on the uncertainties associated with the public preference
studies in reaching their conclusions. However, in this
reconsideration, the current Administrator, while continuing to
recognize that substantial uncertainties remain and that there is
relatively limited new information regarding public preferences of
visibility impairment, judges that it is important to balance the
weight placed on uncertainties with the strength of the scientific
evidence. As such, the Administrator concludes that it is appropriate
to consider a target level of protection within the range of 20 to 30
dv. He further concludes that in selecting a target level within that
range it is appropriate to place weight on both the mid-point of the
range, as supported by the study in Phoenix, AZ, as well as the upper
end, as supported by the Washington, DC, study. The Administrator notes
that these two studies both employ similar methodologies that are
subject to fewer uncertainties than older public preference studies
(including their use of WinHaze to reduce uncertainties in the
preference solicitations) although he notes that the Phoenix, AZ, study
yielded the best results of the four public preference studies in terms
of the least noisy preference results and the most representative
selection of participants. Furthermore, he notes the differences
between the scenes used for each study and finds that consideration of
these studies together is more appropriate in selecting a national
target for visibility protection than considering either study alone.
Thus, in considering this information, along with the uncertainties and
limitations of the public preference studies, the Administrator judges
that it would be appropriate to select a target level of protection
based on placing equal weight on the upper end of the range (i.e., 30
dv) and the middle of the range (i.e., 24 dv based on the Phoenix, AZ,
study) in order to identify a nationwide target for protection against
visibility impairment. In so doing, the Administrator concludes that a
visibility index with a target level of protection of 27, defined in
terms of estimated light extinction, with a 24-hour averaging time and
a 3-year, 90th percentile form, would provide adequate protection
against PM-related visibility effects on public welfare. Such a target
level of protection balances the information from two key studies
reflecting different participant preferences for different vistas in
different parts of the country, appropriately weighting both near-field
and more distant landscape features that may be of importance to public
perceptions of visibility.
The Administrator notes that the available evidence indicates that
the relationship between PM and light extinction is complex, depending
on factors such as PM composition, size fraction, and age of the
particles in ambient air, as well as relative
[[Page 16336]]
humidity. These factors can vary across the country based on
differences in regional influences, as well as meteorological
conditions that can vary spatially and temporally in different areas.
The Administrator also recognizes that this variability, coupled with
the age of the PM depending on the distance from the source to the
monitor location, also complicates the selection of which IMPROVE
equation is most appropriate in different areas, although he notes that
different IMPROVE equations will yield similar, but not identical,
results. In so doing, the Administrator takes note of the figures
presented in the 2022 PA, which depict the comparisons using the
original IMPROVE equation (Figure 5-3), the revised IMPROVE equation
(Figure 5-4), and the Lowenthal & Kumar equation (Figure 5-6), as well
as the estimated light extinction values for the three different
equations presented in Table D-7.
The Administrator notes that when light extinction is calculated
using the original IMPROVE equation, all 60 sites have 3-year
visibility metrics below 28 dv, 58 sites are at or below 25 dv, 26
sites are at or below 20 dv, and of the two sites above 25 dv one is at
26 dv and the other has a 24-hour PM2.5 design value of 56
[mu]g/m\3\ (i.e., well above the current 24-hour standard). Results are
similar for other IMPROVE equations.\175\ Based on these analyses, and
consistent with the results of similar analyses in the 2012 review and
the 2020 PA, the Administrator concludes that the current secondary 24-
hour PM2.5 standard, with its level of 35 [mu]g/m\3\,
maintains the visibility index below 27 dv, and in fact, the current
standard maintains air quality such that many areas have visibility
index values that range between 15 and 25 dv for all three IMPROVE
equations. In the areas that meet the secondary 24-hour
PM2.5 standard, all locations were below 27 dv when using
the original and revised IMPROVE equation and all but three locations
were at or below 27 dv when using the Lowenthal & Kumar IMPROVE
equation. Three locations (two in California and one in Utah) had air
quality that was at 28 dv when the Lowenthal & Kumar IMPROVE equation
was used. As described in more detail in section V.A.1.3, we recognize
that there are differences in the inputs for the three IMPROVE
equations that can influence the resulting estimated light extinction
values. The higher multiplier for converting OC to OM in the Lowenthal
& Kumar IMPROVE equation (i.e., a multiplier of 2.1) may be more
appropriate in more remote locations where there is more aged and
oxygenated organic PM than in urban locations. The three locations with
air quality at 28 dv are all in urban areas (downtown Los Angeles, CA;
Rubidoux, CA; Salt Lake City, UT) and tend to have higher levels of
nitrate and OC, especially during the wintertime when peak
PM2.5 concentrations typically occur. In these locations, it
may be more appropriate to use either the original or revised IMPROVE
equation, which have multipliers of 1.4 and 1.8, respectively, in order
to refine the inputs such that estimated light extinction in these
locations is more accurately characterized based on site-specific
characteristics.
---------------------------------------------------------------------------
\175\ When light extinction is calculated using the revised
IMPROVE equation, all 60 sites have 3-year visibility metrics below
28 dv, 56 sites are at or below 25 dv, and 26 sites are at or below
20 dv. When light extinction is calculated using the Lowenthal and
Kumar IMPROVE equation, 59 sites have 3-year visibility metrics
below 28 dv, 45 sites are at or below 25 dv, and 15 sites are at or
below 20 dv. The one site with a 3-year visibility metric of 32 dv
exceeds the secondary 24-hour PM2.5 standard, with a
design value of 56 [mu]g/m\3\ (see U.S. EPA, 2022b, Appendix D,
Table D-3).
---------------------------------------------------------------------------
We also note that the four areas that exceed the secondary 24-hour
PM2.5 standard also generally had air quality that was below
27 dv in terms of the visibility index, with only two locations
experiencing a visibility index above 27 dv. One location that exceeds
the secondary 24-hour PM2.5 standard had a visibility index
of 29 dv using the original IMPROVE equation, while two locations were
30 and 32 dv using the Lowenthal & Kumar IMPROVE equation. We believe
attainment and maintenance of the secondary 24-hour PM2.5
standard will result in improved air quality in these areas, such that
the visibility index values for these areas will decrease even further.
The Administrator recognizes that in concluding that it is
appropriate to identify secondary PM standards that generally limit
visibility impairment to as low as 27 dv in terms of the visibility
index, the current secondary PM standards continue to provide
protection against visibility impairment associated with a visibility
index as low as, or even lower than, 27 dv. In so doing, he notes that
when meeting the current 24-hour PM2.5 standard, all sites
have a visibility index at or below 27 dv with the original and revised
IMPROVE equations, and all but three sites at or below 27 dv with the
Lowenthal and Kumar IMPROVE equation. Furthermore, the Administrator
notes that this conclusion is consistent with the CASAC's advice who,
in their review the 2021 draft PA, stated that ``[i]f a value of 20-25
deciviews is deemed to be an appropriate visibility target level of
protection, then a secondary 24-hour PM2.5 standard in the
range of 25-35 [mu]g/m\3\ should be considered'' (Sheppard, 2022a, p.
21 of consensus responses).
Thus, the Administrator concludes that weight on both the upper end
of the range of target levels of protection for the visibility index
identified in previous reviews and the mid-point of the range, as
presented by the Phoenix, AZ, public preference study, and focusing on
a target level of protection of 27 dv, he still judges the current
secondary 24-hour PM2.5 standard requisite to achieve that
target because the standard generally maintains the visibility index at
or below 27 dv such that more stringent standards are not warranted.
The EPA agrees with the commenters that the secondary PM standards
work together to provide protection against short- and long-term
effects of both fine and coarse particles (U.S. EPA, 2022b, section
5.5; 88 FR 5661, January 27, 2023). However, the EPA disagrees with
commenters that we failed to discuss the secondary annual
PM2.5 standard in the proposal, 2022 PA, and the 2020 final
notice and that we failed to justify the adequacy of the secondary
annual PM2.5 standard. As described in the 2022 PA and the
proposal, we recognize that PM2.5 is the size fraction of PM
responsible for most of the visibility impairment in urban areas (U.S.
EPA, 2022b, section 5.3.1.2; 88 FR 5654, January 27, 2023). Analyses in
the 2019 ISA found that mass scattering from PM10-2.5 was
relatively small (less than 10%) in the eastern and northwestern U.S.,
whereas mass scattering was much larger in the Southwest (more than
20%), particularly in southern Arizona and New Mexico (U.S. EPA, 2019,
section 13.2.4.1, p. 13-36). Given the relationship between visibility
and PM2.5 along with the short-term nature of visibility
effects, we focus more on the adequacy of the secondary 24-hour
PM2.5 standard for providing protection against visibility
impairment (U.S. EPA, 2022b, section 5.3.1.2; 88 FR 5653, January 27,
2023). In reaching his proposed conclusions, the Administrator clearly
states that he ``recognizes that the current suite of secondary
standards (i.e., the 24-hour PM2.5, 24-hour PM10,
and annual PM2.5 standards) together provide . . . control
for both fine and coarse particulates and long- and short-term
visibility and non-visibility (e.g., climate and materials) effects
related to PM in ambient air'' (88 FR 5661, January 27, 2023). Thus, by
explaining how the secondary standards work together to provide
protection from adverse effects, why we focus on the secondary 24-hour
PM2.5 standard as
[[Page 16337]]
most relevant to visibility impairment, and how the Administrator
selected the target level of protection for the visibility index, we
have addressed the CASAC's request to support the proposed decision to
revise the secondary 24-hour PM2.5 standard while retaining
the secondary annual PM2.5 standard. The commenters also
cite to an individual CASAC member's comments for the review of the
2021 draft PA who stated ``[f]or the limited scope of this
reconsideration review, I see no reason to not simply set the Secondary
equal to the Primary PM Standards, whatever they may be'' (Sheppard,
2022a, p. A-3). This CASAC member did not provide a supporting
rationale for revising the secondary standards to levels equal to the
primary standards. Although areas across the country are required to
attain both the primary and secondary PM2.5 standards so air
quality is unaffected by the Administrator's decision not to revise the
secondary standards to be equal to the primary standards, as described
in responding to comments above, the CAA provisions require the
Administrator to establish secondary standards that, in the judgment of
the Administrator, are requisite to protect public welfare from known
or anticipated adverse effects associated with the presence of the
pollutant in ambient air. In so doing, the Administrator seeks to
establish standards that are neither more nor less stringent than
necessary for this purpose. The Act does not require that standards be
set at a zero-risk level, but rather at a level that reduces risk
sufficiently so as to protect the public welfare from known or
anticipated adverse effects. The final decision on the adequacy of the
current secondary standards is a public welfare policy judgment to be
made by the Administrator. In reaching his proposed and final decisions
regarding the adequacy of the current secondary PM standards, the
Administrator considered the available scientific information and
analyses about welfare effects, and associated public welfare
significance, as well as judgments about how to consider the range and
magnitude of uncertainties that are inherent in the scientific evidence
and analyses. In so doing, the Administrator concluded that the
currently available scientific evidence and quantitative analyses,
including uncertainties and limitations, do not call into question the
adequacy of the current secondary PM standards and that the current
secondary PM standards should be retained, without revision. The
Administrator's judgments and decisions on the primary and secondary
standards are independent and consider different aspects of the
available scientific evidence and information in reaching conclusions
regarding the adequacy of the standards in protecting against PM-
related health and welfare effects.
4. Administrator's Conclusions
This section summarizes the Administrator's considerations and
conclusions related to the current secondary PM2.5 and
PM10 standards and presents the rationale for his decision
that no change is required for those standards at this time. The CAA
provisions require the Administrator to establish secondary standards
that, in the judgment of the Administrator, are requisite to protect
public welfare from known or anticipated adverse effects associated
with the presence of the pollutant in the ambient air. In so doing, the
Administrator seeks to establish standards that are neither more nor
less stringent than necessary for this purpose. The Act does not
require that standards be set at a zero-risk level, but rather at a
level that reduces risk sufficiently so as to protect the public
welfare from known or anticipated adverse effects. The final decision
on the adequacy of the current secondary standards is a public welfare
policy judgment to be made by the Administrator. The decision should
draw on the scientific information and analyses about welfare effects,
and associated public welfare significance, as well as judgments about
how to consider the range and magnitude of uncertainties that are
inherent in the scientific evidence and analyses. This approach is
based on the recognition that the available evidence generally reflects
a continuum that includes ambient air exposures at which scientists
agree that effects are likely to occur through lower levels at which
the likelihood and magnitude of responses become increasingly
uncertain. This approach is consistent with the requirements of the
provisions of the Clean Air Act related to the review of NAAQS and with
how the EPA and the courts have historically interpreted the Act.
Given these requirements, the Administrator's final decision in
this reconsideration is a public welfare policy judgment that draws
upon the scientific and technical information examining PM-related
visibility impairment, climate effects and materials effects, including
how to consider the range and magnitude of uncertainties inherent in
that information. The Administrator recognizes that his final decision
is based on an interpretation of the scientific evidence and technical
analyses that neither overstates nor understates their strengths and
limitations, or the appropriate inferences to be drawn. In particular,
the Administrator notes that the assessment of when visibility
impairment is adverse to public welfare requires a public welfare
policy judgment informed by available scientific and quantitative
information.
In considering the adequacy of the current secondary PM standards
in this reconsideration, the Administrator has carefully considered
the: (1) Policy-relevant evidence and conclusions contained in the 2019
ISA and 2022 ISA Supplement; (2) the quantitative information presented
and assessed in the 2022 PA; (3) the evaluation of this evidence, the
quantitative information, and the rationale and conclusions presented
in the 2022 PA; (4) the advice and recommendations from the CASAC; and
(5) public comments. In the discussion below, the Administrator gives
weight to the 2022 PA conclusions, with which the CASAC generally
concurred during their review of the 2019 draft PA and 2021 draft PA,
as summarized in section IV.B.1 of the 2020 final notice and section
V.D.1 of the 2022 proposal, and takes note of key aspects of the
rationale for those conclusions that contribute to his decision in this
reconsideration. After giving careful consideration to all of this
information, the Administrator judges that no change is required for
the secondary PM standards at this time.
In considering the 2022 PA evaluations and conclusions, the
Administrator takes note of the overall conclusions that the non-
ecological welfare effects evidence and quantitative information are
generally consistent with what was considered in the 2020 final
decision and in the 2012 review (U.S. EPA, 2022b, section 5.5). The
scientific evidence for non-ecological welfare effects in this
reconsideration is largely the same as that available in the 2019 ISA
and 2020 PA. As described in section I.C.5.b above, the 2022 ISA
Supplement included a limited number of newly available studies on PM-
related visibility effects. This newly available evidence on visibility
effects, along with the full body of non-ecological welfare effects
evidence assessed in the 2019 ISA, reaffirms conclusions on the
visibility, climate, and materials effects recognized in the 2020 final
decision and in the 2012 review, including key conclusions on which the
standards are based. Further, as discussed in more
[[Page 16338]]
detail above, the updated quantitative analyses of visibility
impairment for areas meeting the current standards in the 2022 PA
support the adequacy of the current secondary PM standards to protect
against PM-related visibility impairment. The Administrator also
recognizes that uncertainties and limitations continue to be associated
with the available scientific evidence and quantitative information.
With regard to the current evidence on visibility effects, as
summarized in the 2022 PA and discussed in detail in the 2019 ISA and
ISA Supplement, the Administrator notes the long-standing body of
evidence for PM-related visibility impairment. As in previous reviews,
this evidence continues to demonstrate a causal relationship between PM
in ambient air and effects on visibility (U.S. EPA, 2019a, section
13.2). The Administrator recognizes that visibility impairment can have
implications for people's enjoyment of daily activities and for their
overall sense of well-being. Therefore, as in previous reviews, he
considers the degree to which the current secondary standards protect
against PM-related visibility impairment and the degree to which PM-
related visibility impairment is adverse to public welfare. In
particular, in recognizing the short-term nature of visibility
impairment along with the fact that PM2.5 is the size
fraction that contributes most to light extinction, the Administrator
especially focuses on the adequacy of the current secondary 24-hour
PM2.5 standard in providing protection against PM-related
visibility effects judged to be adverse. The Administrator also
considers the protection provided by the current secondary 24-hour
PM2.5 standard against PM-related visibility impairment in
conjunction with the Regional Haze Program as a means of achieving
appropriate levels of protection against PM-related visibility
impairment in urban, suburban, rural, and Federal Class I areas across
the U.S. Programs implemented to meet the secondary PM standards, along
with the requirements of the Regional Haze Program established for
protecting against visibility impairment in Class I areas, would be
expected to improve visual air quality across all areas of the country.
As described in the proposal (88 FR 5658, January 27, 2023), the
Administrator recognizes that the Regional Haze Program was established
by Congress specifically to achieve ``the prevention of any future, and
the remedying of existing, impairment of visibility in mandatory Class
I areas, which impairment results from man-made air pollution,'' and
that Congress established a long-term program to achieve that goal (CAA
section 169A). In adopting section 169, Congress set a goal of
eliminating anthropogenic visibility impairment at Class I areas, as
well as a framework for achieving that goal which extends well beyond
the planning process and timeframe for attaining the secondary PM
NAAQS. Recognizing that the Regional Haze Program will continue to
contribute to reductions in visibility impairment in Class I areas,
consistent with his proposed conclusions, the Administrator concludes
that addressing visibility impairment in Class I areas is largely
beyond the scope of the secondary PM standards and that setting the
secondary 24-hour PM2.5 standard at a level that would
remedy visibility impairment in Class I areas would result in standards
that are more stringent than is requisite.
In further considering what standards are requisite to protect
against adverse public welfare effects from visibility impairment, the
Administrator concludes that it is appropriate to use an approach
consistent with the approach used past reviews (88 FR 5650, January 27,
2023). He first identifies an appropriate target level of protection in
terms of a PM visibility index that takes into account the factors that
influence the relationship between PM in ambient air and visibility
(i.e., size fraction, species composition, and relative humidity). He
then considers the air quality analyses conducted in the 2022 PA that
examine the relationship between the PM visibility index and the
current secondary 24-hour PM2.5 standard in locations that
meet the current 24-hour PM2.5 and PM10 standards
(U.S. EPA, 2022b, section 5.3.1.2).
In reaching conclusions regarding the target level of protection,
the Administrator first considers the characteristics of the visibility
index and defines its elements (indicator, averaging time, form, and
level). With regard to the indicator for the visibility index, the
Administrator continues to recognize that, consistent with the
conclusions of the 2022 PA and the CASAC's advice in their review of
the 2021 draft PA, there is a lack of availability of methods and an
established network for directly measuring light extinction. Therefore,
the Administrator concludes that it continues to be appropriate to
using an index based on estimates of light extinction by
PM2.5 components based on the IMPROVE algorithm. In so
doing, the Administrator recognizes that the fundamental understanding
of the relationship between ambient PM and light extinction has
generally changed very little over time; however, several versions of
the IMPROVE equation have been developed and evaluated that could be
used to estimate light extinction. As at the time of the proposal, the
Administrator recognizes that the results of the quantitative analyses
in the 2022 PA that examined three versions of the IMPROVE equation
indicate that there are very small differences in estimates of light
extinction between the equations, and that it is not always clear that
one version of the IMPROVE equation is more appropriate for estimating
light extinction across the U.S. than other versions of the IMPROVE
algorithm (88 FR 5659, January 27, 2023). He also recognizes that the
selection of inputs to the IMPROVE equation (e.g., the multiplier for
OC to OM) may be more appropriate on a regional basis rather than a
national basis when calculating light extinction, and notes the CASAC's
advice that PM-visibility relationships are region specific (Sheppard,
2022a, p. 21 of consensus responses). The Administrator further notes
that neither the CASAC nor public commenters recommended a specific
IMPROVE equation or an approach for using different IMPROVE equations
across the U.S. Therefore, given the absence of a robust monitoring
network to directly measure light extinction, the Administrator
concludes that light estimated light extinction, as calculated using
one or more versions of the IMPROVE algorithms, continues to be the
most appropriate indicator for the visibility index.
Having reached the conclusion that estimated light extinction is
the appropriate indicator for the visibility index, the Administrator
next considers the appropriate averaging time and form of the index.
With regard to the averaging time and form, the Administrator notes
that in previous reviews, a 24-hour averaging time was selected and the
form was defined as the 3-year average of annual 90th percentile
values. As at the time of proposal, the Administrator recognizes that
the available information continues to provide support for the short-
term nature of visibility effects. He further recognizes that no new
information is available in this reconsideration to inform his
conclusions regarding averaging time, and therefore, he considers past
analyses of 24-hour and subdaily PM2.5 light extinction to
inform his conclusions on averaging time. As described in the proposal
(88 FR 5659, January 27, 2023) and in responding to comments in section
V.B.3 above, prior
[[Page 16339]]
analyses demonstrated that there are strong correlations between 24-
hour and subdaily (i.e., 4-hour average) PM2.5 light
extinction, indicating that a 24-hour averaging time is an appropriate
surrogate for the subdaily time periods associated with when
individuals experience visibility impairment and that a longer
averaging time may also be less influenced by atypical conditions and/
or atypical instrument performance. The Administrator also notes that
the CASAC did not provide advice or recommendations with regard to the
averaging time of the visibility index, although some public commenters
referenced CASAC advice in past reviews that a subdaily standard based
on daylight hours would better reflect the public welfare effects of
public perceptions of visibility impairment than a 24-hour standard.
However, in considering the available scientific and quantitative
information, as well as the CASAC's advice in their reviews of the 2019
draft PA and 2021 draft PA, the Administrator concludes that the 24-
hour averaging time continues to be appropriate for the visibility
index because it is an appropriate surrogate for subdaily time periods
and results in a more stable target.
With regard to the form of the visibility index, the Administrator
notes the approach in other NAAQS that a multi-year percentile form
offers greater stability to the air quality management process by
reducing the possibility that statistically unusual indicator values
will lead to transient violations of the standard. He recognizes that
using a 3-year average provides stability from the occasional effects
of inter-annual meteorological variability (including relative
humidity) that can result in unusually high pollution levels for a
particular year (88 FR 5659, January 27, 2023) and recognizes that a
stable standard contributes to the benefits of the NAAQS by ensuring
that attainment strategies are designed to address non-transient
problems and achieve durable air quality improvements. For these
reasons, he concludes that a 3-year average continues to be
appropriate.
In considering the percentile that would be appropriate with the 3-
year average, the Administrator recognizes that there is very little
new information available in this reconsideration to inform selection
of an alternative form of the visibility index and that the appropriate
form requires the exercise of public welfare policy judgment. In
selecting the appropriate target level of protection for the visibility
index, the Administrator is required to assess when visibility
impairment becomes adverse to public welfare, weighing both the degree
of visibility impairment (in dv) and the frequency of such impairment
(through the form). As with the mass-based PM air quality standard, the
target level of protection for the visibility index must be selected in
conjunction with the form to determine the appropriate stringency. In
so doing, consistent with approaches in past reviews, the Administrator
first notes that the Regional Haze Program targets the 20% most
impaired days for improvements in visual air quality in Class I areas,
which are the days above the 80th percentile form of the visibility
index. The Administrator concludes that a percentile form set at the
80th percentile would not be likely to sufficiently improve visual air
quality on the worst days based on the visibility index. In considering
the information available in past reviews regarding the form of the
visibility index, as well as the analysis of alternative forms based on
recent air quality discussed above, the Administrator notes that a 90th
percentile form would represent the median of the distribution of the
20% most impaired days, and meeting a visibility index with a 90th
percentile form would reasonably be expected to lead to improvements in
visual air quality for days both above and below the 90th percentile
(88 FR 5660, January 27, 2023). In reaching his conclusion that a 90th
percentile would appropriately achieve improved air quality both above
and below that percentile, the Administrator took into consideration
assessments of air quality data and potential alternative percentiles
for the form. The Administrator further notes that, consistent with the
conclusions in the 2011 PA and 2020 PA, the 2022 PA concluded that
there is no new information from public preference studies that would
suggest that a 90th percentile form is not appropriate. The
Administrator also considers air quality analyses described above in
responding to public comments regarding the percentile form of the
visibility index. In particular, the Administrator notes that while a
higher percentile form (i.e., 95th or 98th) would somewhat further
limit the number of days with peak PM-related light extinction, the
differences in the 3-year averages of estimated light extinction for
the 90th, 95th, and 98th percentile forms are small. For example, he
notes that for the original IMPROVE equation, in areas that meet the
current 24-hour PM2.5 standard, all sites have light
extinction estimates for a 90th percentile form at or below 26 dv, and
for a 95th or 98th percentile form light extinction estimates are at or
below 29 dv.\176\ He further notes that, in most locations when
estimating light extinction based on the original IMPROVE equation, the
difference between a 95th or 98th percentile form and a 90th percentile
form is generally less than 3 dv.\177\ Moreover, the Administrator
concludes that a 90th percentile form achieves a very high degree of
control but appropriately targets the group of worst days, rather than
the few very worst days. Based on the available information and these
analyses, the Administrator concludes that the information does not
indicate that it would be appropriate to consider limiting the
occurrence of days with peak PM-related light extinction to a greater
degree, nor did the CASAC provide advice or recommendations related to
the form of the visibility index. Therefore, the Administrator judges
that it remains appropriate to define a visibility index in terms of a
24-hour averaging time and form based on the 3-year average of annual
90th percentile values.
---------------------------------------------------------------------------
\176\ Gantt, B., and Hagan, N. (2023). Analysis of Percentile
Forms of the Visibility Index. Memorandum to the Rulemaking Docket
for the Review of the National Ambient Air Quality Standards for
Particulate Matter (EPA-HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
\177\ Gantt, B., and Hagan, N. (2023). Analysis of Percentile
Forms of the Visibility Index. Memorandum to the Rulemaking Docket
for the Review of the National Ambient Air Quality Standards for
Particulate Matter (EPA-HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
With regard to the level of the visibility index, as at the time of
proposal, the Administrator continues to recognize that there is very
little new information available to inform his judgment regarding the
range of levels of visibility impairment judged to be acceptable by at
least 50% of study participants in the visibility preference
studies,\178\ and therefore, the range of 20 to 30 dv identified in the
2022 PA remains appropriate for considering the level of the visibility
index. The Administrator also recognizes that the uncertainties and
limitations associated with the public preferences identified in the
2012 and 2020 reviews continue to persist, and that these limitations
and uncertainties contributed to the decisions in 2012 and 2020 that a
level at the upper end of the range (i.e., 30 dv) was selected. The
Administrator specifically notes that, while the studies
[[Page 16340]]
are methodologically similar, there are a number of factors that can
influence comparability across the studies and that the available
studies may not capture the full range of visibility preferences in the
U.S. population, as described in more detail in section V.D.3 of the
2022 proposal (88 FR 5659-5660, January 27, 2023). The Administrator
also notes the CASAC's advice in their review of the 2021 draft PA that
there are a limited number of visibility preference studies available
to inform the Administrator's judgment regarding the appropriate target
level of protection for the visibility index (Sheppard, 2022a, p. 21 of
consensus responses). In considering the available information,
including uncertainties and limitation, and the CASAC's advice, the
Administrator proposed to conclude that it is appropriate to consider a
target level of protection for the visibility index within the range of
20 to 30 dv, and that establishing a target level of protection at the
upper end of the range was appropriate. In so doing, the Administrator
proposed to conclude that the protection provided by a visibility index
based on estimated light extinction, a 24-hour averaging time, and a
90th percentile form, averaged over 3 years, set to a level of 30 dv
would be requisite to protect public welfare with regard to visibility
impairment.
---------------------------------------------------------------------------
\178\ For reasons stated above and described in the 2022 PA and
proposal, the Administrator does not find it appropriate to use the
most recent preference study based on the Grand Canyon study area
(Malm et al., 2019) for purposes of identifying a target level of
protection for the visibility index.
---------------------------------------------------------------------------
However, at the time of proposal, the Administrator recognized that
the available evidence on visibility impairment generally reflects a
continuum and that the public preference studies do not provide
information about the specific level for which visibility impairment
would be ``acceptable'' or ``unacceptable'' across the country, and
that alternative target levels of protection could be supported. At
that time, in soliciting public comments, the Administrator recognized
that other interpretations, assessments, and judgments based on the
available welfare effects evidence for this reconsideration could be
possible (88 FR 5662, January 27, 2023).
With regard to the appropriate target level of protection for the
visibility index, the Administrator first notes that while the public
preference studies were conducted in several geographical areas across
the U.S., and they provide insight into regional preferences for
visibility impairment, none of the studies identify a specific level of
visibility impairment that would be perceived as ``acceptable'' or
``unacceptable'' across the whole U.S. population. He also noted that
there have been significant questions about how to set a standard for
visibility that is neither overprotective nor underprotective for some
areas of the U.S. As described in the proposal (88 FR 5660, January 27,
2023), in establishing the Regional Haze Program to improve visibility
in Class I areas, Congress noted that ``as a matter of equity, the
national ambient air quality standards cannot be revised to adequately
protect visibility in all areas of the country.'' H.R. Rep. 95-294 at
205. For the reasons noted above, in reaching his proposed decision
regarding visibility impairment, the Administrator recognized that he
is not seeking to set a standard that would eliminate visibility
impairment in Class I areas, but significant uncertainties remain
regarding how to judge when visibility impairment becomes adverse to
public welfare across the range of daily outdoor activities for
Americans across the country.
In reaching final conclusions regarding the available information,
along with the CASAC's advice and public comments, the Administrator
again considers what constitutes an appropriate target level of
protection, and in particular considers whether a target level of
protection below 30 dv is warranted. In so doing, he first notes the
variability in public preferences of visibility impairment as
demonstrated by the available public preferences, which support a range
of potential target levels of protection for the visibility index from
20 to 30 dv. He also notes that this range informed the 2012 and 2020
then-Administrators final decisions that a target level of protection
at the upper end of the range (i.e., 30 dv) would be most appropriate,
given the uncertainties and limitations associated with the public
preference studies. As described in in section V.B.3 above in
responding to public comments, the Administrator recognizes that a
number of factors can influence public preferences across studies, in
particular due to the types of scenes depicted in the images as well as
the distances at which the objects of interest are located from the
camera. Furthermore, the Administrator recognizes the small number of
public preference studies currently available makes precise
interpretations of their results challenging for determining a
nationally appropriate target level of visibility protection. The
Administrator also recognizes that the CASAC, in their review of 2021
draft PA, reiterated that PM-visibility relationships are region-
specific based on aerosol composition, and that several public
commenters emphasized the importance of the sight path distance in the
images when considering how to interpret the public preference studies.
In this reconsideration, the Administrator judges that in
determining when visibility impairment becomes adverse to public
welfare for purposes of the secondary NAAQS, while continuing to
recognize that substantial uncertainties remain and that there is
relatively limited new information regarding public preferences of
visibility impairment, it is important to balance the weight placed on
uncertainties with the strength of the scientific evidence. In so
doing, the Administrator first concludes that, consistent with previous
reviews and his proposed decision, it remains appropriate to consider a
target level of protection within the range of 20 to 30 dv. However, in
further considering the available scientific and quantitative
information, CASAC advice, and public comments, he further concludes
that in selecting a target level within that range it is appropriate to
place weight on both the middle of the range, as supported by the study
in Phoenix, AZ, as well as the upper end, as supported by the
Washington, DC, study. In so doing, he notes that the Washington, DC,
and Phoenix, AZ, studies employ similar methodologies that are subject
to fewer uncertainties than older public preference studies (including
their use of WinHaze to reduce uncertainties in the preference
solicitations) although he does note that the Phoenix, AZ, study
yielded the best results of the four public preference studies in terms
of the least noisy preference results and the most representative
selection of participants. Further, the Administrator judges that this
approach would take into account scenes that are similar to both the
Washington, DC, study and Phoenix, AZ, study, which would be more
representative of the ``typical'' scenes encountered across more areas
of the U.S. than an approach that places weight on just one study or on
studies conducted in certain geographical areas of the country. In
considering this information, along with the uncertainties and
limitations of the public preference studies, the Administrator judges
that it would be appropriate to select a target level of protection
based on placing equal weight on the upper end of the range (i.e., 30
dv) and the middle of the range (i.e., 24 dv based on the Phoenix, AZ,
study) in order to provide protection against visibility impairment in
different geographical areas of the U.S. For these reasons, the
Administrator concludes that a visibility index with a target level of
protection of 27 dv,
[[Page 16341]]
defined in terms of estimated light extinction, with a 24-hour
averaging time and a 3-year, 90th percentile form, would provide
adequate protection against PM-related visibility effects. In reaching
this conclusion, the Administrator judges that such a target level of
protection balances the information from these two key public
preference studies in such a way appropriately weighs both near-field
and more distant landscape features that may be of importance to public
perceptions of visibility.
In further considering the appropriate target level of protection
for the visibility index, the Administrator again recognizes the
complexity of the relationship between PM and light extinction which is
dependent on a number of factors, including PM composition, size
fraction, and age of the particles in ambient air, as well as relative
humidity. As noted in responding to comments above, these factors can
vary geographically across the U.S. and local or regional
meteorological conditions can also vary spatially and temporally. These
factors are critical inputs to the IMPROVE equation and can influence
the resulting estimated light extinction such that it is not a
straightforward comparison between estimated light extinction in one
area of the country versus another. Moreover, the Administrator
recognizes that there is variability in estimated light extinction
depending on the version of the IMPROVE equation that is used. As
described in more detail in the 2022 PA and the proposal, and in
reaching his decisions on the indicator of the visibility index above,
the Administrator notes that the 2022 PA concluded that one version of
the IMPROVE equation is not more accurate or precise in estimating
light extinction, and that difference in locations may support the
selection of inputs into the IMPROVE equation or of the appropriate
IMPROVE equation to estimate light extinction on a regional basis
rather than on a national basis.
In considering the available information, including variations in
both public preferences of visibility impairment and estimates of light
extinction using one or more IMPROVE equation, as well as the CASAC's
advice in their review of the 2019 draft PA and 2021 draft PA and
public comments, the Administrator judges that a target level of
protection of 27 dv would be appropriate. In so doing, he concludes
that a target level of protection above 27 dv would not provide
adequate protection against PM-related visibility impairment based on
the 50% acceptability values when both the Washington, DC, and Phoenix,
AZ, studies are considered. However, he also notes that when
considering the 50% acceptability values from studies conducted in
different areas of the U.S. and with different scenes and images
depicted, the available public preference studies do not provide a
``bright line'' at and above which visibility impairment is considered
adverse to public welfare. He further recognizes that, as discussed
just above, there are a number of region-specific factors that can
influence light extinction, and thereby influence visibility
impairment, as well as variations in public preferences of visibility
impairment based on the available studies, that complicate selection of
a single target level of protection that would be appropriate for a
national visibility index. While the Administrator recognizes that the
uncertainties and limitations associated with public preferences of
visibility and estimating light extinction have persisted over the last
several PM NAAQS reviews, he also recognizes that in reaching
conclusions regarding the appropriate target level of protection for
the visibility index also involves public welfare policy judgments
regarding how to appropriately consider the particular uncertainties
around identifying when visibility impairment becomes adverse to public
welfare, and the limitations on relying on the public preference
studies.
The Administrator also places weight on the high degree of spatial
and temporal variability in PM composition and relative humidity across
the U.S. in considering a target level of protection. This approach of
establishing a target level of protection that takes into account 50%
acceptability values from both eastern and western sites is a more
appropriate basis for determining the requisite level of protection
against known or anticipated adverse effects on public welfare across
diverse locations, i.e., a standard that is neither more nor less
stringent than necessary nationwide. Specifically, the Administrator
judges that a target level of protection for the visibility index
focused on maintaining estimated light extinction between the upper end
of the range of the target levels of protection (i.e., 30 dv based on
the Washington, DC, study) and the middle of the range (i.e., 24 dv
based on the Phoenix, AZ, study) to be more appropriate for a
nationwide standard to protect against visibility impairment compared
to a value derived from one location or one type of scene alone. For
these reasons, in selecting a target level of protection, the
Administrator concludes that a target level of protection somewhere
between the upper end and middle of the range is appropriate because he
judges that this approach, in conjunction with the Regional Haze
program, is sufficient, but not more stringent than necessary, to
protect against adverse effects on public welfare. Thus, he concludes a
secondary 24-hour PM2.5 NAAQS should be evaluated based on
its ability to provide protection against visibility impairment
associated with estimated light extinction of 27 dv based on estimated
light extinction, a 24-hour averaging time, and a 90th percentile form,
averaged over 3 years.
Having concluded that it is appropriate to identify a target level
of protection in terms of a visibility index based on estimated light
extinction as described above, the Administrator next considers the
degree of protection from visibility impairment afforded by the current
secondary PM standards. He considers the updated analyses of PM-related
visibility impairment presented in the 2022 PA (U.S. EPA, 2022b,
section 5.3.1.2) and described in section V.B.1.a of the proposal, and
notes that the results of the analyses are consistent with the results
from the 2012 and 2020 reviews.
Taking into consideration the full body of scientific evidence and
technical information concerning the known and anticipated effects of
PM on visibility impairment, the Administrator concludes that the
current secondary PM2.5 and PM10 standards are
requisite to protect against PM-related visibility impairment. While
the inclusion of the coarse fraction had a relatively modest impact on
calculated light extinction in the analyses presented in the 2022 PA,
he recognizes the continued importance of the PM10 standard
given the potential for larger impacts in locations with higher coarse
particle concentrations, such as in the southwestern U.S., for which
only a few sites met the criteria for inclusion in the analyses in the
2022 PA (U.S. EPA, 2019a, section 13.2.4.1; U.S. EPA, 2022b, section
5.3.1.2).
With regard to the adequacy of the secondary 24-hour
PM2.5 standard, the Administrator notes that, in their
review of the 2021 draft PA, the CASAC stated that ``[i]f a value of
20-25 deciviews is deemed to be an appropriate visibility target level
of protection, then a secondary 24-hour PM2.5 standard in
the range of 25-35 [micro]g/m\3\ should be considered'' (Sheppard,
2022a, p. 21 of consensus responses). The Administrator recognizes that
the CASAC recommended that the Administrator provide additional
justification for a visibility index target
[[Page 16342]]
of 30 dv but did not specifically recommend that he choose an
alternative level for the visibility index. The Administrator carefully
considered the advice of CASAC and the public comments and concluded
that a lower target level of visibility was appropriate in order to
properly reflect both a broader set of studies and a broader range of
vistas that were the subject of those studies. However, in their review
of the 2021 draft PA, the CASAC recognized that even a visibility index
target in the range of 20-25 dv could still warrant retention of the
current secondary 24-hour PM2.5 standard. The Administrator
also considers the advice from the CASAC in their review of the 2019
draft PA, who ``recogniz[ed] that uncertainties. . .remain about the
best way to evaluate'' PM-related visibility effects (Cox, 2019b, p. 13
consensus responses). The Administrator considered the CASAC's advice,
together with the available scientific evidence and quantitative
information, in reaching his conclusions.
The Administrator recognizes that conclusions regarding the
appropriate weight to place on the scientific and technical information
examining PM-related visibility impairment, including how to consider
the range and magnitude of uncertainties inherent in that information,
is a public welfare policy judgment left to the Administrator. In
reaching his final decision in 2020, the then-Administrator noted that
the available evidence regarding visibility effects had changed very
little since the 2012 review, specifically recognizing that, as
evaluated in the 2019 ISA, there were no new visibility studies that
were conducted in the U.S. and there was little new information
available with regard to acceptable levels of visibility impairment in
the U.S. (85 FR 82742, December 18, 2020). As such, the then-
Administrator concluded that the protection provided by a standard
defined in terms of a PM2.5 visibility index, with a 24-hour
averaging time, a 90th percentiles form averaged over three years, set
at a level of 30 dv, was requisite to protect public welfare against
visibility impairment (85 FR 82743, December 18, 2020). He also
recognized that there was some new information to inform quantitative
analyses of light extinction, but that the results of the analyses
conducted in the 2020 PA were consistent with those from the 2012
review. The then-Administrator recognized that the analyses
demonstrated that the 3-year visibility metric was at or below about 30
dv in all areas that met the current secondary 24-hour PM2.5
standard, and was below 25 dv in most of those areas (85 FR 82743,
December 18, 2020). Therefore, the Administrator judged that the
secondary 24-hour PM2.5 standard provided sufficient
protection for visual air quality of 30 dv, which he judged appropriate
(88 FR 82744, December 18, 2020). In this reconsideration, the ISA
Supplement evaluated newly available studies on public preferences for
visibility impairment and/or development methodologies or conducted
quantitative analyses of light extinction. In considering the available
scientific and quantitative information, including that newly available
in this reconsideration, the current Administrator reached the same
preliminary conclusions in the notice of proposed rulemaking regarding
the 3-year visibility index and the current secondary PM standards as
the then-Administrator in the 2020 final decision. However, in light of
public comments on the proposal, the Administrator has further
considered the available scientific evidence and information, as well
as the CASAC's advice regarding visibility effects in their review of
the 2021 draft PA. In so doing, the Administrator judges that it is
appropriate to place more weight on certain aspects of the evidence
that he had placed less weight on in reaching his proposed conclusions
(i.e., he focused on the both the middle and the upper end of the range
of the 50% acceptability values from the available public preference
studies). As such, the Administrator notes his conclusion on the
appropriate visibility index (i.e., with a 24-hour averaging time; a 3-
year, 90th percentile form; and a level of 27 dv), which takes into
account the regional variations in public preferences and equations for
estimating light extinction, and his conclusions regarding the
quantitative analyses of the relationship between the visibility index
and the current secondary 24-hour PM2.5 standard. In so
doing, the Administrator concludes that the current secondary standards
provide requisite protection against PM-related visibility effects.
With respect to climate effects, as at the time of proposal, the
Administrator recognizes that a number of improvements and refinements
have been made to climate models since the time of the 2012 review.
However, despite continuing research and the strong evidence supporting
a causal relationship with climate effects (U.S. EPA, 2019a, section
13.3.9), the Administrator notes that there are still significant
limitations in quantifying the contributions of the direct and indirect
effects of PM and PM components on climate forcing (U.S. EPA, 2022b,
sections 5.3.2.1.1 and 5.5). He also recognizes that models continue to
exhibit considerable variability in estimates of PM-related climate
impacts at regional scales (e.g., ~100 km), compared to simulations at
the global scale (U.S. EPA, 2022b, sections 5.3.2.1.1 and 5.5).
Moreover, the effects of PM on climate are diverse as well as
uncertain. Depending on the circumstances, the radiative forcing
effects of PM in the atmosphere can vary, such that positive forcing
could result in warming of the Earth's surface, whereas a negative
forcing could result in cooling (U.S. EPA, 2019a, section 13.3.2.2).
The resulting uncertainty leads the Administrator to conclude that the
scientific information available in this reconsideration remains
insufficient to quantify, with confidence, the impacts of ambient PM on
climate in the U.S. (U.S. EPA, 2022b, section 5.3.2.2.1) and that there
is not an adequate scientific basis to link attainment of any
particular PM concentration in ambient air in the U.S. to specific
climate effects. Consequently, the Administrator judges that there is
insufficient information at this time to revise the current secondary
PM standards or to promulgate a distinct secondary standard to address
PM-related climate effects.
With respect to materials effects, the Administrator notes that the
available evidence continues to support the conclusion that there is a
causal relationship with PM deposition (U.S. EPA, 2019a, section 13.4).
He recognizes that deposition of particles in the fine or coarse
fractions can result in physical damage and/or impaired aesthetic
qualities. Particles can contribute to materials damage by adding to
the effects of natural weathering processes and by promoting the
corrosion of metals, the degradation of painted surfaces, the
deterioration of building materials, and the weakening of material
components. While some recent evidence on materials effects of PM is
available in the 2019 ISA, the Administrator notes that this evidence
is primarily from studies conducted outside of the U.S. in areas where
PM concentrations in ambient air are higher than those observed in the
U.S. (U.S. EPA, 2019a, section 13.4). Given the limited amount of
information on the quantitative relationships between PM and materials
effects in the U.S., and uncertainties in the degree to which those
effects could be adverse to the public welfare, the Administrator
judges that the available scientific information
[[Page 16343]]
remains insufficient to quantify, with confidence, the public welfare
impacts of ambient PM on materials and that there is insufficient
information at this time to revise the current secondary PM standards
or to promulgate a distinct secondary standard to address PM-related
materials effects.
Taken together, the Administrator concludes that the scientific and
quantitative information for PM-related non-ecological welfare effects
(i.e., visibility, climate, and materials),\179\ along with the
uncertainties and limitations, supports the current level of protection
provided by the secondary PM standards as being requisite to protect
against known and anticipated adverse effects on public welfare. For
visibility impairment, this conclusion reflects his consideration of
the evidence for PM-related light extinction, together with his
consideration of updated air quality analyses of the relationship
between the visibility index and the current secondary 24-hour
PM2.5 standard and the protection provided by the current
secondary PM2.5 and PM10 standards. For climate
and materials effects, this conclusion reflects his judgment that,
although it remains important to maintain secondary PM2.5
and PM10 standards to provide some degree of control over
long- and short-term concentrations of both fine and coarse particles,
it is appropriate not to change the existing secondary standards at
this time and that it is not appropriate to establish any distinct
secondary PM standards to address PM-related climate and materials
effects at this time. As such, the Administrator recognizes that
current suite of secondary standards (i.e., the 24-hour
PM2.5, 24-hour PM10, and annual PM2.5
standards) together provide such control for both fine and coarse
particles and long- and short-term visibility and non-visibility (e.g.,
climate and materials) effects related to PM in ambient air. His
conclusions on the secondary standards are consistent with advice from
the CASAC, which noted substantial uncertainties remain in the
scientific evidence for climate and materials effects, as well as the
majority of public comments on the secondary PM standards. Thus, based
on his consideration of the evidence and analyses for PM-related
welfare effects, as described above, and his consideration of CASAC
advice and public comments on the secondary standards, the
Administrator concludes that it is appropriate not to change those
standards (i.e., the current 24-hour and annual PM2.5
standards, 24-hour PM10 standard) at this time.
---------------------------------------------------------------------------
\179\ As noted earlier, other welfare effects of PM, such as
ecological effects, are being considered in the separate, on-going
review of the secondary NAAQS for oxides of nitrogen, oxides of
sulfur and PM.
---------------------------------------------------------------------------
C. Decision on the Secondary PM Standards
For the reasons discussed above and taking into account information
and assessments presented in the 2019 ISA, ISA Supplement, and 2022 PA,
advice from the CASAC, and consideration of public comments, the
Administrator concludes that the current secondary PM standards are
requisite to protect public welfare from known or anticipated adverse
effects and is not changing the standards at this time.
VI. Interpretation of the NAAQS for PM
The EPA is finalizing revisions on data calculations in appendix K
for PM10 and appendix N for PM2.5. Revisions to
appendix K make the PM10 data handling procedures for the
24-hour PM10 standards more consistent with those of other
NAAQS pollutants and codify existing practices. Revisions to appendix N
update references to the revision(s) of the standards and change data
handling provisions related to combining data from nearby monitoring
sites to codify existing practices that are currently being implemented
as the EPA standard operating procedures.
A. Amendments to Appendix K: Interpretation of the NAAQS for
Particulate Matter
The EPA proposed to modify its data handling procedures for the 24-
hour PM10 standard in appendix K to part 50 (88 FR 5662,
January 27, 2023). The proposed modifications include: (1) Revising
design value calculations to be on a site-level basis, (2) codifying
site combinations to maintain a continuous data record, and (3)
clarifying daily validity requirements for continuous monitors. The
purpose of these modifications is to make the data handling procedures
for the 24-hour PM10 standard more consistent with those of
other NAAQS pollutants and codify existing practices that are currently
being implemented as EPA standard operating procedures.
The EPA received few comments on these proposed appendix K
revisions, the majority of which were supportive.
One commenter was not supportive of the proposed appendix K
revision to site-level PM10 design values, asserting that it
would amount to an imposition of a more stringent PM10
standard due to the potential high bias of FEMs. The EPA disagrees with
this assertion because site-level design values would combine data from
any high biased FEM with other monitors at the site rather than
calculate a monitor-level design value with data solely from that high-
biased FEM. The EPA tested the impact of calculating site-level
PM10 design values for the 2019-2021 period by assigning the
lowest parameter occurrence code as the primary monitor and calculating
site-level design values. Most resulting site-level design values were
either identical to or in-between the multiple monitor-level design
values at the site. Combining data from two or more monitors also has
the benefit of increasing the number of valid sample days at many
sites. For the 2019-2021 test period, approximately 10% of the sites
with more than one monitor went from having multiple invalid design
values to a single valid design value.
One commenter was not supportive of a footnote in the preamble of
the NPRM stating that in the absence of a designated primary monitor at
a given site, the default primary monitor would be one with the most
complete data record (88 FR 5662, January 27, 2023). Because the
procedure for calculating PM10 design values on a site-level
basis being finalized here will require monitoring agencies to
designate a primary monitor for each site in their annual network plans
(88 FR 5694, January 27, 2023; App. K, 1.0(b)), the EPA agrees with the
commenter that this footnote was unnecessary.
Therefore, the EPA is finalizing these appendix K revisions as
proposed.
B. Amendments to Appendix N: Interpretation of the NAAQS for PM2.5
The EPA proposed to modify its data handling procedures for the
annual and 24-hour PM2.5 standards in appendix N to part 50
(88 FR 5663, January 27, 2023). These proposed revisions include: (1)
Updating references to the revisions of the standards rather than
stating the specific level, and (2) codifying site combinations to
maintain a continuous data record. The purpose of both modifications is
to codify existing practices that are currently being implemented as
the EPA standard operating procedures.
The EPA received few comments on these revisions in the proposed
rule, with most supportive of the appendix N revisions.
Although the EPA did not propose or request comment on this issue,
one commenter suggested that appendix N be revised to only allow data
from the primary monitor to be used in PM2.5 NAAQS
designations asserting that it would add flexibility. The EPA disagrees
with the commenter's assertion that this would add flexibility because
it could force agencies to run
[[Page 16344]]
their FRMs on a daily schedule or potentially lead to invalid design
values if manual sampling interruptions or laboratory issues impact FRM
data completeness. This change would also be undesirable because it
could reduce by two-thirds the number of days used in calculations for
the annual and 24-hour PM2.5 design values at many sites.
Therefore, the EPA is finalizing these appendix N revisions as
proposed.
VII. Amendments to Ambient Monitoring and Quality Assurance
Requirements
The EPA is finalizing revisions to ambient air monitoring
requirements for PM to improve the usefulness of and appropriateness of
data used in regulatory decision making. These changes focus on ambient
monitoring requirements found in 40 CFR parts 50 (appendix L), 53, and
58 with associated appendices (A, B, C, D, and E). These changes
include addressing updates in the approval of reference and equivalent
methods, updates in quality assurance statistical calculations to
account for lower concentration measurements, updates to support
improvements in PM methods, a revision to the PM2.5 network
design to account for at-risk populations, and updates to the Probe and
Monitoring Path Siting Criteria for NAAQS pollutants. The EPA also took
comment on how to incorporate data from next generation technologies
into Agency efforts. A summary of the comments received is included in
this section.
A. Amendment to 40 CFR Part 50 (Appendix L): Reference Method for the
Determination of Fine Particulate Matter as PM2.5 in the Atmosphere--
Addition of the Tisch Cyclone as an Approved Second Stage Separator
The EPA proposed a change to the FRM for PM2.5 (40 CFR
part 50, appendix L), the addition of an alternative PM2.5
particle size separator to that of the Well Impactor Ninety-Six (WINS)
and the Very Shape Cut Cyclone (VSCC) size separators (88 FR 5663,
January 27, 2023). The new separator is the TE-PM2.5C
cyclone manufactured by Tisch Environmental Inc.,\180\ Cleves Ohio,
which has been shown to have performance equivalent to that of the
originally specified WINS impactor with regards to aerodynamic cutpoint
and PM2.5 concentration measurement. In addition, the new
TE-PM2.5C has a significantly longer service interval than
the WINS and is comparable to that of the VSCC separator. Generally,
the TE-PM2.5C is also physically interchangeable with the
WINS and VSCC where both are manufactured for the same sampler. The
proposed change would allow either the WINS, VSCC, or TE-
PM2.5C to be used in a PM2.5 FRM sampler. As is
the case for the WINS and VSCC, the TE-2.5C is now also an approved
size separator for candidate PM2.5 FEMs. Currently, the EPA
has designated one PM2.5 sampler configured with TE-
PM2.5C separator as a Class II PM2.5 equivalent
method and one as a PM10-2.5 equivalent method. Upon
promulgation of this change to appendix L, these instruments would be
redesignated as PM2.5 and PM10-2.5 FRMs,
respectively. Owners of such samplers should contact the sampler
manufacturer to receive a new reference method label for the samplers.
---------------------------------------------------------------------------
\180\ Mention of commercial names does not constitute EPA
endorsement.
---------------------------------------------------------------------------
The EPA received only one comment regarding this proposed change,
which was supportive. Therefore, the EPA is finalizing this change to
Appendix L as proposed.
B. Issues Related to 40 CFR Part 53 (Reference and Equivalent Methods)
The EPA proposed to clarify the regulations associated with FRM and
FEM applications for review by the EPA (88 FR 5664, January 27, 2023).
Revisions were also proposed in instances where current regulatory
specifications are no longer pertinent and require updating. In
addition, the EPA proposed to correct a compiled a list of noted minor
errors in the regulations associated with the testing requirements and
acceptance criteria for FRMs and FEMs in part 53. These errors are
typically not associated with the content of Federal Register documents
but often relate to transcription errors and typographical errors in
the electronic CFR (eCFR) and printed versions of the CFR.
1. Update to Program Title and Delivery Address for FRM and FEM
Applications
The EPA proposed a change to 40 CFR 53.4(a) to update the delivery
address for FRM and FEM Applications and Modification Requests, as well
as update the name of the program responsible for their review (88 FR
5664, January 27, 2023). These revisions are due solely to
organizational changes and do not affect the structure or role of the
Reference and Equivalent Methods Designation Program in reviewing new
FRM and FEM application requests and requests to modify existing
designated instruments. The EPA received no comments on this revision
and, therefore, the EPA is finalizing this revision as proposed.
2. Requests for Delivery of a Candidate FRM or FEM Instrument
The EPA proposed a change to 40 CFR 53.4(d), which currently allows
the EPA to request only candidate PM2.5 FRMs and Class II or
Class III equivalent methods for testing purposes as part of the
applicant review process (88 FR 5664, January 27, 2023). The EPA
proposed to revise this section to enable requesting any candidate FRM,
FEM, or a designated FRM or FEM associated with a Modification Request,
regardless of NAAQS pollutant type or metric. The EPA received no
comments on these revisions; therefore, the EPA is finalizing this
revision as proposed.
3. Amendments to Requirements for Submission of Materials in 40 CFR
53.4(b)(7) for Language and Format
The EPA proposed a change to 40 CFR 53.4(b)(7) to specify that all
written FRM and FEM application materials must be submitted to the EPA
in English in MS Word format and that submitted data must be submitted
in MS Excel format (88 FR 5664, January 27, 2023). The EPA received no
comments on these revisions; therefore, the EPA is finalizing this
section as proposed.
4. Amendment to Designation of Reference and Equivalent Methods
The EPA proposed a change to 40 CFR 53.8(a) to clarify the terms of
new FRM and FEM methods to ensure that candidate samplers and analyzers
are not publicly announced, marketed, or sold until the EPA's approval
has been formally announced in the Federal Register (88 FR 5664,
January 27, 2023). The EPA received no comments on these revisions;
therefore, the EPA is finalizing this section as proposed.
5. Amendment to One Test Field Campaign Requirement for Class III
PM2.5 FEMs
The EPA proposed a change to 40 CFR 53.35(b)(1)(ii)(D) that
involves field comparability tests for candidate Class III
PM2.5 FEMs, including the requirement that a total of five
field campaigns must be conducted at four separate sites, A, B, C, and
D (88 FR 5664, January 27, 2023). The existing Site D specifications
require that the site ``shall be in a large city east of the
Mississippi River, having characteristically high sulfate
concentrations and high humidity levels.'' However, dramatic decreases
in ambient sulfate concentration make it difficult for applicants to
routinely meet the high sulfate concentration requirement. Therefore,
the EPA proposed to revise the Site D specifications to read ``shall be
in a large
[[Page 16345]]
city east of the Mississippi River, having characteristically high
humidity levels.'' Only one comment was received on this proposed
revision, which was supportive. Therefore, the EPA is finalizing the
revision to 40 CFR 53.35(b)(1)(ii)(D), as proposed.
6. Amendment to Use of Monodisperse Aerosol Generator
The EPA proposed a change to 40 CFR 53.61(g), 53.62(e), and Table
F-1 that involves the wind tunnel evaluation of candidate
PM10 inlets and candidate PM2.5 fractionators
under static conditions, which requires the generation and use of
monodisperse calibration aerosols of specified aerodynamic sizes (88 FR
5664, January 27, 2023). In the current regulations, the TSI
Incorporated Vibrating Orifice Aerosol Generator (VOAG) is the only
monodisperse generator that is approved for this purpose. However, TSI
Incorporated no longer manufacturers nor supports the VOAG. Therefore,
a commercially available monodisperse aerosol generator (Model 1520
Fluidized Monodisperse Aerosol Generator, MSP Corporation, Shoreview,
MN) has been added to list of approved generators for this purpose. No
comments were received on this revision; therefore, the EPA is
finalizing this revision as proposed.
7. Corrections to 40 CFR Part 53 (Reference and Equivalent Methods)
Certain provisions of 40 CFR 53.14, Modification of a reference or
equivalent method, incorrectly state an EPA response deadline of 30
days for receipt of modification materials in response to an EPA
notice. Per a 2015 amendment (80 FR 65460, 65416, Oct. 26, 2015), all
EPA response deadlines for modifications of reference or equivalent
methods are 90 days from day of receipt. Thus, the EPA proposed a
correction to specify the correct 90-day deadline (88 FR 5664, January
27, 2023).
Requirements for Reference and Equivalent Methods for Air
Monitoring of Criteria Pollutants identifies the applicable 40 CFR part
50 appendices and 40 CFR part 53 subparts for each criteria pollutant.
The four rows in the section for PM10-2.5 erroneously do not
include the footnote instruction that the aforementioned pollutant
alternative Class III requirements may be substituted in regard to
Appendix O to Part 50--Reference Method for the Determination of Coarse
Particulate Matter as PM10-2.5 in the Atmosphere.
Table B-1 specifies that the interference equivalent for each
interferent is 0.005 ppm for both the standard-range and
lower-range limits, with the exception of nitric oxide (NO) for the
lower-range limit per note 4. When testing the lower range of
SO2, the limit for NO is 0.003 ppm, therefore,
an incorrect lower limit (0.0003) is currently stated in
note 4 for this exception to the SO2 lower range limit.
Thus, the EPA proposed a correction to Table B-1 to specify the correct
limit in note 4 (88 FR 5664, January 27, 2023).
After the EPA received an inquiry regarding the interaction of NO
and O3, the EPA investigated the interferent testing
requirements stated by 40 CFR part 53, subpart B. The EPA has
determined that during the 2011 SO2 amendment and subsequent
2015 O3 amendment, several typographical errors were
introduced into Table B-3, the most significant of which is the
omission of note 3, which instructs the applicant to not mix the
pollutant with the interferent. Thus, the EPA proposed revisions to
Table B-3 to correct these errors (88 FR 5664, January 27, 2023).
Additionally, appendix A to subpart B of part 53 provides figures
depicting optional forms for reporting test results. Figure B-3 lists
an incorrect formula: the lower detectible limit section is missing the
proper operator in the LDL calculation formula and Figure B-5 lists an
incorrect calculation metric, and there is a typesetting error in the
calculation of the standard deviation. The EPA proposed to correct the
typesetting errors and noted other errors to be corrected in several
formulas provided throughout Sec. 53.43 (88 FR 5664, January 27,
2023).
The EPA proposed a revision to 40 CFR 53.43(a)(2)(xvi),
53.43(b)(2)(iv), and 53.43(b)(2)(iv) to correct typographical errors in
equations.
The EPA proposed a revision to Table C-4 of part 53 Subpart C (88
FR 5700). This change is related to field comparability tests of
candidate PM2.5, PM10-2.5, and PM10
FEMs, which requires testing at wide range of ambient concentrations.
For this reason, Table C-4 specifies a minimum number of valid sample
sets to be conducted at specified high concentrations. However, due to
the dramatic decrease in ambient PM concentrations in the past two
decades, these number of valid test days at high concentrations has
been difficult to achieve. Accordingly, the EPA proposed to revise the
testing specifications for high concentration events in Table C-4 to
reflect current levels of ambient PM for all three PM metrics. In
addition to the revision of the ambient PM concentration specifications
to Table C-4, there are also several entry errors that required
correction.
The EPA received no comments on these proposed revisions;
therefore, the EPA is finalizing the changes as proposed.
C. Changes to 40 CFR Part 58 (Ambient Air Quality Surveillance)
1. Quality Assurance Requirements for Monitors Used in Evaluations for
National Ambient Air Quality Standards
In the proposal, the EPA described how we evaluated the quality
system as part of the PM NAAQS reconsideration (88 FR 5665, January 27,
2023). In this section, the EPA identified several areas for
improvement in steadily declining average ambient PM2.5
concentrations across the country and the final decision to revise
primary annual PM2.5 NAAQS described in section II above. We
assessed PM2.5 concentration data across a range of values
to determine if any changes to the statistical calculations used to
evaluate the data quality in the PM2.5 network were
warranted. This section describes the EPA's assessment, comments
received, and the EPA's final decisions on the proposed changes. Other
changes in this section include clarifications and other improvements
that will facilitate consistency and the operation of quality assurance
programs by State, local, and Tribal (SLT) agencies nationwide.
a. Quality System Requirements
The EPA reconsidered the appendix A, section 2.3.1.1 goal for
acceptable measurement uncertainty (88 FR 5665, January 27, 2023) for
automated and manual PM2.5 methods for total bias. The
existing total bias goal is an upper 90 percent confidence limit for
the coefficient of variation (CV) of 10 percent and 10
percent for total bias. The intent of the proposal was to investigate
if this bias goal is still realistic given updated precision and bias
statistic. The EPA received one comment that bias reevaluation may be
premature, since the final NAAQS standard had not yet been determined
at the time of the proposal. The EPA acknowledges this comment but
clarifies that the proposed new bias statistic was evaluated at a range
of levels including the range of proposed PM2.5 standards in
the technical memorandum, ``Task 16 on PEP/NPAP Task Order: Bias and
Precision DQOs for the PM2.5 Ambient Air Monitoring
Network.'' \181\ Considering the
[[Page 16346]]
justification in the technical memorandum and the lack of adverse
comments regarding this part of the proposal, the EPA is retaining the
appendix A, section 2.3.1.1, goal for acceptable measurement
uncertainty for automated and manual PM2.5 methods for total
bias.
---------------------------------------------------------------------------
\181\ Noah, G. (2023). Task 16 on PEP/NPAP Task Order: Bias and
Precision DQOs for the PM2.5 Ambient Air Monitoring
Network. Memorandum to the Rulemaking Docket for the Review of the
National Ambient Air Quality Standards for Particulate Matter (EPA-
HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
The EPA also proposed to update and clarify ambient air monitoring
requirements found in 40 CFR part 58, appendix A, section 2.6.1
pertaining to EPA Protocol Gas standards used for ambient air
monitoring and the Ambient Air Protocol Gas Verification Program (PGVP)
(88 FR 5665, January 27, 2023). The EPA proposed to revise appendix A
to clarify that in order to participate in the Ambient Air PGVP,
producers of Protocol Gases must adhere to the requirements of 40 CFR
75.21(g), and only regulatory ambient air monitoring programs may
submit cylinders for assay verification to the EPA Ambient Air PGVP.
The EPA received mixed comments in support of and in opposition to this
proposed revision. The sole commenter opposing the proposed revision
indicated that the proposed PGVP requirements would be additional and
is concerned with an increased resource burden. But the EPA responds
that the PGVP requirements that were proposed to be added are
consistent with the existing PGVP requirements in 40 CFR 75.21(g), and
PGVP has been defined as a regulatory requirement since 2016 (81 FR
17263, March 28, 2016), so the proposed part 58 changes are not
``additional'' to existing regulations. After consideration of the
comments, the EPA is finalizing the update and clarification of ambient
air monitoring requirements found in appendix A, section 2.6.1
pertaining to EPA Protocol Gas standards used for ambient air
monitoring and the Ambient Air PGVP as proposed.
b. Measurement Quality Check Requirements
The EPA proposed to remove section 3.1.2.2 from appendix A, which
allows NO2 compressed gas standards to be used to generate
audit standards (88 FR 5665, January 27, 2023). The EPA received one
comment supporting this change. As a result of the comment received and
other general supportive comments regarding quality assurance, the EPA
is finalizing the removal of section 3.1.2.2 from appendix A as
proposed.
The EPA proposed to revise the requirement in Appendix A, section
3.1.3.3 changing the National Performance Audit Program (NPAP)
requirement for annual verification of gaseous standards to the ORD-
recommended certification periods identified in Table 2-3 of the EPA
Traceability Protocol for Assay and Certification of Gaseous
Calibration Standards (appendix A, section 6.0(4)) (88 FR 5665). The
EPA received one comment supporting this change. As a result of the
comment received and other general supportive comments regarding
quality assurance, the EPA is finalizing the updated NPAP gaseous
certification requirement in section 3.1.3.3 as proposed.
The EPA proposed to adjust the minimum value required by appendix
A, section 3.2.4, to be considered valid sample pairs for the
PM2.5 Performance Evaluation Program (PEP) from 3 [mu]g/m\3\
to 2 [mu]g/m\3\ (88 FR 5665, January 27, 2023). The EPA received
comments in support and against the change. In the only opposing
comment, the commenter expressed concern that the method detection
limit (MDL) for PM2.5 is 2 [mu]g/m\3\. The commenter also
indicated that the MDL ``typically has minimal value per the definition
of the MDL.'' 40 CFR part 50, appendix L states, ``The lower detection
limit of the mass concentration measurement range is estimated to be
approximately 2 [mu]g/m\3\, based on noted mass changes in field blanks
in conjunction with the 24 m\3\ nominal total air sample volume
specified for the 24-hour sample.'' The EPA notes that field blanks
currently average less than 10 [mu]g nationally, and when divided by
the 24 m\3\ nominal total air sample volume specified for a 24-hour
sample, the result is 0.4 [mu]g/m\3\. The appendix L MDL referenced by
the commenter was part of the 1997 PM NAAQS rulemaking (62 FR 38652,
July 18, 1997); current data shows that the MDL is substantially lower
than the EPA's original estimate. After review of the comments, and in
consideration of the recently calculated detection limit for the
PM2.5 FRM that is substantially lower than our original
estimate,\182\ the EPA is finalizing the revised minimum value for
valid sample pairs for the PM2.5 Performance Evaluation
Program (PEP) from 3 [mu]g/m\3\ to 2 [mu]g/m\3\ in appendix A, section
3.2.4 as proposed.
---------------------------------------------------------------------------
\182\ See the EPA's PM2.5 Data Quality Dashboard
available at https://sti-r-shiny.shinyapps.io/QVA_Dashboard/.
---------------------------------------------------------------------------
c. Calculations for Data Quality Assessments
The EPA proposed to change Equations 6 and 7 of appendix A, section
4.2.1 that are used to calculate the Collocated Quality Control Sampler
Precision Estimate for PM10, PM2.5 and Pb (88 FR 5666, January 27,
2023). The proposed new statistics are designed to address the high
imprecision values that result from using these calculations to compare
low concentrations that are now more routinely observed in the
networks. The EPA received several comments in support of this change
in general, but some commenters indicated that they believed there was
an error in the new calculation that may result in high imprecision
from the calculation of the equation. The EPA reviewed the technical
memorandum and confirmed that a multiplier of 100 was unintentionally
left in the proposed relative difference equation, Equation 6. Also,
equation 6 was corrected from a normalized percent difference to a
normalized relative percent difference that is appropriate for
comparing collocated pairs at low concentrations. The technical
memorandum titled ``Task 16 on PEP/NPAP Task Order: Bias and Precision
DQOs for the PM2.5 Ambient Air Monitoring Network'' has been
amended to correct the error and is included in the docket for this
action.\183\
---------------------------------------------------------------------------
\183\ Noah, G. (2023). Task 16 on PEP/NPAP Task Order: Bias and
Precision DQOs for the PM2.5 Ambient Air Monitoring
Network. Memorandum to the Rulemaking Docket for the Review of the
National Ambient Air Quality Standards for Particulate Matter (EPA-
HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
Equation 6 as proposed at 88 FR 5666 (January 27, 2023) was:
[[Page 16347]]
[GRAPHIC] [TIFF OMITTED] TR06MR24.031
As a result of the positive comments received and the correction to
the equation made in response to some comments, the EPA is finalizing
the updated Equation 6 as described and is finalizing Equation 7 as
proposed for the calculation of the Collocated Quality Control Sampler
Precision Estimate for PM10, PM2.5, and Pb in section 4.2.1.
The EPA proposed to update the appendix A, section 4.2.5, Equation
8, calculation for the Performance Evaluation Program Bias Estimate for
PM2.5 (88 FR 5666-67, January 27, 2023). Because average
ambient PM concentrations across the nation have steadily declined
since the promulgation of the PM2.5 standard, the EPA
proposed to replace the current percent difference equation with a
relative difference equation. The EPA received several comments in
support of this change in general, but some commenters identified a
potential error in the new calculation that resulted in an artificially
high estimate, which they do not support. The EPA reviewed the
technical memorandum and discovered that a multiplier of 100 was left
in the new relative difference equation used in the bias equation. The
technical memorandum, ``Task 16 on PEP/NPAP Task Order: Bias and
Precision DQOs for the PM2.5 Ambient Air Monitoring
Network'' has been amended to correct the error and is included in the
docket.\184\ The proposed Equation 8 proposed at 88 FR 5667 (January
27, 2023) was:
---------------------------------------------------------------------------
\184\ Noah, G. (2023). Task 16 on PEP/NPAP Task Order: Bias and
Precision DQOs for the PM2.5 Ambient Air Monitoring
Network. Memorandum to the Rulemaking Docket for the Review of the
National Ambient Air Quality Standards for Particulate Matter (EPA-
HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
[GRAPHIC] [TIFF OMITTED] TR06MR24.032
As a result of the supportive comments received and the correction
to the equation in response to some comments, the EPA is updating and
finalizing Equation 8 as described for the calculation for the
Performance Evaluation Program Bias Estimate for PM2.5, in
section 4.2.5.
d. References
The EPA proposed to update the references and hyperlinks in
appendix A, section 6 (88 FR 5667, January 27, 2023) to provide
accuracy in identifying and locating essential supporting documentation
and delete references to historical documents that do not represent
current practices. The EPA received only favorable comments, and as a
result, the EPA is finalizing the updated the references and hyperlinks
in appendix A, section 6, as proposed.
The EPA also proposed to add a footnote to Table A-1 of part 58,
appendix A--Minimum Data Assessment Requirements for NAAQS Related
Criteria Pollutant Monitors (88 FR 5669, January 27, 2023). The
proposed footnote clarifies the allowable time (i.e., every two weeks,
once a month, once a quarter, once every six months, or distributed
over all four quarters depending on the check) between checks and
encourages monitoring organizations to perform data assessments at
regular intervals. The EPA received two comments regarding this
proposed footnote. One commenter indicated that this change is
inconsistent with the QA Handbook for Air Pollution Measurement
Systems: ``Volume II: Ambient Air Quality Monitoring Program QA
Handbook.'' The EPA agrees with the commenter; because the QA Handbook
is guidance,
[[Page 16348]]
the EPA will revise it after this action is finalized to be consistent
with the updated CFR provision. Another commenter does not support the
addition of the footnote due to concerns about limiting flexibility. In
response, the EPA reiterates that the proposed revision is intended to
clarify intent and does not make any changes to the required
frequencies or acceptance criteria for data assessment. A ``weight of
evidence'' narrative is still found in 40 CFR part 58, appendix A,
section 1.2.3. As a result of the comments received and the rationale
discussed above, the EPA is finalizing the addition of the new footnote
to Table A-1 of part 58, appendix A--Minimum Data Assessment
Requirements for NAAQS Related Criteria Pollutant Monitors as proposed.
2. Quality Assurance Requirements for Prevention of Significant
Deterioration (PSD) Air Monitoring
The EPA proposed to revise appendix B, Quality Assurance
Requirements for Prevention of Significant Deterioration (PSD) Air
Monitoring (88 FR 5667, January 27, 2023), in parallel to the proposal
to revise appendix A. Thus, this section of the proposal included
similar detail and proposed revisions related to evaluating quality
system statistical calculations for PM2.5, clarifications
and other improvements that would facilitate consistency and the
operation of quality assurance programs for PSD by SLT agencies
nationwide.
a. Quality System Requirements
The EPA reconsidered the goal in appendix B, section 2.3.1.1 for
acceptable measurement uncertainty for automated and manual
PM2.5 methods for total bias (88 FR 5668, January 27,
2023).\185\ The current total bias goal is an upper 90 percent
confidence limit for the coefficient of variation (CV) of 10 percent
and 10 percent for total bias. The EPA's intent was to
investigate if this goal is still realistic given updated precision and
bias statistics. The EPA received one comment that bias reevaluation
may be premature, since the final NAAQS standard had not yet been
determined at the time of the proposal. The EPA acknowledges this
comment but clarifies that the proposed new bias statistic was
evaluated at a range of levels including the proposed range of
PM2.5 standards in the technical memorandum, ``Task 16 on
PEP/NPAP Task Order: Bias and Precision DQOs for the PM2.5
Ambient Air Monitoring Network.'' \186\ Considering the justification
in the technical memorandum and the lack of adverse comments regarding
the substantive proposal, the EPA is retaining the appendix B, section
2.3.1.1, goal for acceptable measurement uncertainty for automated and
manual PM2.5 methods for total bias.
---------------------------------------------------------------------------
\185\ In the proposal, in section VII.C.2 Quality Assurance
Requirements for Prevention of Significant Deterioration (PSD) Air
Monitoring (88 FR 5667-69), the EPA inadvertently referred to
``appendix A'' in the section rather than the correct ``appendix
B.'' The EPA's intent to have proposed changes to appendix B on
these pages is made clear by the section header, the Table of
Contents on page 5559, and the proposed regulatory text for appendix
B on pages 5707-08. See, e.g., id. at p.5668 (preamble erroneously
states that the EPA proposed to change appendix A, section 2.6.1);
id. at p.5668 (preamble erroneously states that the EPA proposed to
adjust the minimum value required by appendix A, section 3.2.4).
\186\ Noah, G. (2023). Task 16 on PEP/NPAP Task Order: Bias and
Precision DQOs for the PM2.5 Ambient Air Monitoring
Network. Memorandum to the Rulemaking Docket for the Review of the
National Ambient Air Quality Standards for Particulate Matter (EPA-
HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
The EPA also proposed to update and clarify ambient air monitoring
requirements found in 40 CFR part 58, appendix B, section 2.6.1
pertaining to EPA Protocol Gas standards used for ambient air
monitoring and the Ambient Air PGVP (88 FR 5668, January 27, 2023). The
EPA proposed to revise appendix B to clarify that in order to
participate in the Ambient Air PGVP, producers of Protocol Gases must
adhere to the requirements of 40 CFR 75.21(g), and only regulatory
ambient air monitoring programs may submit cylinders for assay
verification to the EPA Ambient Air PGVP. The EPA received comments in
support of and in opposition to this proposed revision. The commenter
opposing the revision indicated that the proposed PGVP requirements
would be additional and is concerned with an increased resource burden.
However, the EPA disagrees with the commenter because that the proposed
PGVP requirements are consistent with the existing PGVP requirements in
40 CFR 75.21(g). PGVP has been defined as a regulatory requirement
since 2016 (81 FR 17263, March 28, 2016), so the proposed part 58
changes are not ``additional'' to existing regulations. After
consideration of the comments, the EPA is finalizing the update and
clarification of ambient air monitoring requirements found in appendix
B, section 2.6.1 pertaining to EPA Protocol Gas standards used for
ambient air monitoring and the Ambient Air PGVP as proposed.
b. Measurement Quality Check Requirements
The EPA proposed to remove section 3.1.2.2 from appendix B, which
allows NO2 compressed gas standards to be used to generate
audit standards (88 FR 5668, January 27, 2023). The EPA received one
comment supporting this change. As a result of the comment received and
other general supportive comments regarding quality assurance, the EPA
is finalizing the removal of section 3.1.2.2 from appendix B as
proposed.
The EPA proposed to revise the requirement in Appendix B, section
3.1.3.3 changing the National Performance Audit Program (NPAP)
requirement for annual verification of gaseous standards to the ORD-
recommended certification periods identified in Table 2-3 of the EPA
Traceability Protocol for Assay and Certification of Gaseous
Calibration Standards (appendix B, section 6.0(4)) (88 FR 5668, January
27, 2023). The EPA received one comment supporting this change. As a
result of the comment received and other general supportive comments
regarding quality assurance, the EPA is finalizing the updated NPAP
gaseous certification requirement in section 3.1.3.3 as proposed.
The EPA proposed to adjust the minimum value required by appendix
B, section 3.2.4, to be considered valid sample pairs for the
PM2.5 Performance Evaluation Program (PEP) from 3 [mu]g/m\3\
to 2 [mu]g/m\3\ (88 FR 5668, January 27, 2023). The EPA received
comments in support and against the change. In the only opposing
comment, the commenter expressed concern that the method detection
limit (MDL) for PM2.5 is 2 [mu]g/m\3\. The commenter also
indicated that the MDL ``typically has minimal value per the definition
of the MDL.'' 40 CFR part 50, appendix L states, ``The lower detection
limit of the mass concentration measurement range is estimated to be
approximately 2 [mu]g/m\3\, based on noted mass changes in field blanks
in conjunction with the 24 m\3\ nominal total air sample volume
specified for the 24-hour sample''. The EPA notes that field blanks
currently average less than 10 [mu]g nationally, and when divided by
the 24 m\3\ nominal total air sample volume specified for a 24-hour
sample, the result is 0.4 [mu]g/m\3\. The appendix L MDL referenced by
the commenter was part of the 1997 PM NAAQS rulemaking more than 20
years ago (62 FR 38652, July 18, 1997); current data shows that the MDL
is substantially lower than EPA's original estimate. After review of
the comments, and in consideration of the recently calculated
[[Page 16349]]
detection limit for the PM2.5 FRM that is substantially
lower than our original estimate, the EPA is revising the minimum value
for valid sample pairs for the PM2.5 Performance Evaluation
Program (PEP) from 3 [mu]g/m\3\ to 2 [mu]g/m\3\ in appendix B, section
3.2.4 as proposed.
c. Calculations for Data Quality Assessments
The EPA proposed to change Equations 6 and 7 of appendix B, section
4.2.1 used for calculating the Collocated Quality Control Sampler
Precision Estimate for PM10, PM2.5 and Pb (88 FR 5707, January 27,
2023). These new statistics are designed to address the high
imprecision values that result from using these calculations to compare
low concentrations that are now more routinely observed in the
networks. The EPA received several comments in support of this change
in general, but a couple commenters indicated that there could be an
error in the new calculation that resulted in high imprecision from the
calculation of the equation. The EPA reviewed the technical memorandum
and discovered that a multiplier of 100 was unintentionally left in the
proposed relative difference equation, Equation 6. Also, equation 6 was
corrected from a normalized percent difference to a normalized relative
percent difference that is appropriate for comparing collocated pairs
at low concentrations. The technical memorandum titled ``Task 16 on
PEP/NPAP Task Order: Bias and Precision DQOs for the PM2.5
Ambient Air Monitoring Network'' was amended to correct the error and
is included in the docket.\187\
---------------------------------------------------------------------------
\187\ Noah, G. (2023). Task 16 on PEP/NPAP Task Order: Bias and
Precision DQOs for the PM2.5 Ambient Air Monitoring
Network. Memorandum to the Rulemaking Docket for the Review of the
National Ambient Air Quality Standards for Particulate Matter (EPA-
HQ-OAR-2015-0072). Available at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
[GRAPHIC] [TIFF OMITTED] TR06MR24.033
As a result of the positive comments received and the correction to
the equation made in response to those comments, the EPA is finalizing
the update to Equation 6 and retaining Equation 7 as proposed for the
calculation of the Collocated Quality Control Sampler Precision
Estimate for PM10, PM2.5 and Pb in section 4.2.1.
The EPA proposed to update the appendix B, section 4.2.5, Equation
8, calculation for the Performance Evaluation Program Bias Estimate for
PM2.5 (88 FR 5668-59, January 27, 2023). Because average
ambient PM concentrations across the nation have steadily declined
since the promulgation of the PM2.5 standard, the EPA
proposed to replace the current percent difference equation with a
relative difference equation. The EPA received several comments in
support of this change in general, but some commenters identified a
potential error in the new calculation that resulted in an artificially
high estimate, which they do not support. The EPA reviewed the
technical memorandum and discovered that a multiplier of 100 was left
in the new relative difference equation used in the bias equation. The
technical memorandum, ``Task 16 on PEP/NPAP Task Order: Bias and
Precision DQOs for the PM2.5 Ambient Air Monitoring
Network'' has been amended to correct the error and is included in the
docket. The proposed Equation 8 (88 FR 5669, January 27, 2023) was:
[[Page 16350]]
[GRAPHIC] [TIFF OMITTED] TR06MR24.034
As a result of the supportive comments received and the correction
to the equation in response to some comments, the EPA is updating and
finalizing Equation 8 as described for the calculation for the
Performance Evaluation Program Bias Estimate for PM2.5, in
section 4.2.5.
d. References
The EPA proposed to update the references and hyperlinks in
appendix B, section 6 (88 FR 5669, January 27, 2023) to provide
accuracy in identifying and locating essential supporting documentation
and delete references to historical documents that do not represent
current practices. The EPA received only favorable comments, and as a
result, the EPA is finalizing the updated the references and hyperlinks
in appendix B, section 6, as proposed.
The EPA also proposed to add a footnote to Table B-1 of part 58,
appendix B--Minimum Data Assessment Requirements for NAAQS Related
Criteria Pollutant PSD Monitors (88 FR 5669, January 27, 2023). The
proposed footnote clarifies the allowable time (i.e., every two weeks,
once a month, once a quarter, once every six months, or distributed
over all four quarters depending on the check) between checks and
encourages monitoring organizations to perform data assessments at
regular intervals. The EPA received two comments regarding this
proposal. One commenter indicated that this change is inconsistent with
the QA Handbook. The EPA agrees with the commenter; because the QA
Handbook is guidance, the EPA will revise it after this action is
finalized to be consistent with the updated CFR provision. Another
commenter does not support the addition of the footnote due to concerns
about limiting flexibility. In response, the EPA reiterates that the
proposed revision is intended to clarify intent and does not make any
changes to the required frequencies or acceptance criteria for data
assessment. A ``weight of evidence'' narrative is still found in 40 CFR
part 58, appendix B, section 1.2.3. As a result of the comments
received and the rationale discussed above, the EPA is adding the new
footnote to Table B-1 of part 58, appendix B--Minimum Data Assessment
Requirements for NAAQS Related Criteria Pollutant PSD Monitors as
proposed.
3. Amendments to PM Ambient Air Quality Methodology
a. Revoking Approved Regional Methods (ARMs)
The EPA proposed to remove provisions for approval and use of
Approved Regional Methods (ARMs) throughout parts 50 and 58 of the CFR
(88 FR 5669, January 27, 2023). ARMs are continuous PM2.5
methods that have been approved specifically within a State or local
air agency monitoring network for purposes of comparison to the NAAQS
and to meet other monitoring objectives. Currently, there are no
approved ARMs. There are, however, more than a dozen approved Federal
Equivalent Methods (FEMs) for PM2.5. These approved FEMs are
eligible for comparison to the NAAQS and to meet other monitoring
objectives.
The EPA received comments from multiple State air programs in
support of the proposal to remove provisions for approval and use of
ARMs. One commenter cites that there are multiple FEMs available for
monitoring agencies to work with and that the agency was never able to
get a candidate ARM to meet the requirements for approval. With the
availability of multiple FEMs that now work in the monitoring agency's
network, the commenting agency does not anticipate the need to ever
pursue an ARM in the future and, therefore, suggests that the ARM
provision is no longer needed. Another commenter strongly supported the
proposed changes to remove the ARM provisions. The EPA also received
comments from a few agencies that supported retaining the ARM
provisions instead. One commenter cited the need to consider the rapid
advancement of various new technologies and that, in some cases,
approved continuous FEMs may have shortcomings, meaning that losing the
ability to propose an ARM in the future may limit useful alternative
options to monitoring agencies. Another commenter suggested that the
removal of the ARM would take away the ability and right to use locally
derived correction factors.
After considering the comments for and against removing the
provisions for ARMs, the EPA believes it is most appropriate to remove
the ARM provisions. As described in the proposal, when the EPA first
proposed the process for approving and using ARMs, there were no
continuous FEMs approved. There are now over a dozen approved
PM2.5 continuous FEMs and no approved ARMs. Therefore, the
EPA is finalizing the removal of ARMs throughout 40 CFR parts 50 and 58
as proposed.
b. Calibration of PM Federal Equivalent Methods (FEMs)
The EPA proposed to modify its specifications for PM FEMs in
appendix C to Part 58 (88 FR 5670-73, January 27, 2023). Specifically,
the EPA proposed that valid State, local, and Tribal (SLT) air
monitoring data from Federal Reference Methods (FRMs) generated in
routine networks and submitted to the EPA may be used to improve the PM
concentration measurement performance of approved FEMs. This approach,
initiated by instrument manufacturers, would be implemented as a
national solution in factory calibrations of approved FEMs through a
firmware update. This could apply to any PM FEM methods (i.e.,
PM10, PM2.5, and PM10-2.5).
The EPA proposed this modification because there are some approved
PM FEMs that are not currently meeting bias measurement quality
objectives (MQOs) when evaluating data nationally as described in the
2022 PA (U.S. EPA, 2022b, section 2.2.3.1), meaning that an update to
factory calibrations may be appropriate; however, there is no clearly
defined process to update the calibration of FEMs. While there are
several types of data available to use as the reference for such
updates (e.g., routinely operated FRMs, audit program FRMs, and
chemical speciation sampler
[[Page 16351]]
data), we proposed to use routinely operated SLT FRMs as the basis of
comparison upon which to calibrate FEMs. The goal of updating factory
calibrations would be to increase the number of routinely operating
FEMs meeting bias MQOs across the networks in which they are operated.
While there are other approaches that could improve data comparability
between PM FEMs and collocated FRMs, the EPA believes that the proposed
modification to calibrate PM FEMs represents the most reliable approach
to update FEM factory calibrations, since the existing FRM network data
that meet MQOs would be used to set updated factory calibrations.
While the Agency proposed to add this language to more expressly
define a process to update factory calibrations of approved PM FEMs,
the EPA believes that the existing rules for updating approved FRMs and
FEMs found at 40 CFR 53.14 may also continue to be utilized for this
purpose, as appropriate. 40 CFR 53.14 allows instrument manufactures to
submit to the EPA a ``Modification of a reference or equivalent
method.'' Submitting a modification request may be appropriate to
ensure an approved FEM continues to meet 40 CFR 53.9, ``Conditions of
designation.'' Specifically, 40 CFR 53.9(c) requires that, ``Any
analyzer, PM10 sampler, PM2.5 sampler, or
PM10-2.5 sampler offered for sale as part of an FRM or FEM
shall function within the limits of the performance specifications
referred to in Sec. 53.20(a), Sec. 53.30(a), Sec. 53.35, Sec.
53.50, or Sec. 53.60, as applicable, for at least 1 year after
delivery and acceptance when maintained and operated in accordance with
the manual referred to in Sec. 53.4(b)(3).'' Thus, instrument
manufacturers are encouraged to seek improvements to their approved FEM
methods as needed to continue to meet data quality needs as operated
across the network.
There are several technical components to EPA's proposed
modification, including: the reference data to be used in the
calibrations; implementing as a national solution in factory
calibrations of approved FEMs through firmware updates; application to
any PM FEM methods (i.e., PM10, PM2.5, and
PM10-2.5); the appropriate range of data to be used to
develop and test new factory calibrations, from just the most
representative concentrations up to all available concentrations; the
representative set of geographic locations that can be used; whether
outliers may be included or not included; that new factory calibrations
should be developed using data from at least 2 years and tested on data
from a separate year or years; that updates to factory calibrations can
occur as often as needed; that calibrations should be evaluated by
monitoring agencies as part of routine data assessments, e.g., during
certification of data and 5-year assessments; the EPA's recognition
that only data from existing operating sites is available; and finally,
that an updated factory calibration does not have to work with the
original field study data submitted that led to the original FEM
designation.
With the proposed modification, the EPA solicited input on these
technical issues as well as the overall approach and any alternatives
that could lead to more sites meeting the bias MQO with automated FEMs,
especially for those sites that are near the level of the primary
annual PM2.5 NAAQS, as proposed to be revised in section II
above. In response, the EPA received comments from about two dozen
entities, most of which were SLT air programs or Multi-Jurisdictional
Organizations (MJOs) comprised of these entities.
Overall, there was broad and strong support from a majority of
commenters for the proposed requirement to use FRM data generated in
routine networks and submitted to the EPA to update factory
calibrations included as part of approved FEMs. There were a smaller
number of critical comments on the proposed process as well as some
commentors that supported the proposed requirement but also provided
additional suggestions for the EPA's consideration. Below, we address
each of the areas on which the EPA requested comment regarding the
calibration of PM FEMs, as well as a few additional areas where
multiple commenters offered input on other areas related to our
proposal.
A majority of the commenters on the proposed PM FEM calibration
process support the process to use valid State, local, and Tribal FRM
data generated in routine networks and submitted to the EPA to improve
the PM concentration measurement performance of approved FEMs. Some
commenters suggested that this action is needed to ensure that data
reported from FRMs and FEMs are comparable and correction methods
applied to data from FEM monitors are defensible across the national PM
monitoring network. Others stated that they agree with the EPA that
this is a critical step in the right direction to account for the
discrepancies between PM2.5 FRM data and PM2.5
FEM data. Some commented that applying corrections includes a
recognition that, while different measurement principles may produce
differences in the resulting data, having an approach that minimizes
bias is extremely important. Finally, some stated their belief that a
correction factor is necessary to preserve data integrity with the FRM.
The EPA also received comments suggesting ways that the PM FEM
correction could be performed, including through detailed analysis of
data; by having PM FEM instrument manufacturers evaluate nationally
available valid FRM data to update factory calibrations; and, by having
the instrument manufacturers implement calibration adjustments at the
factory.
The EPA also received supportive comments on the PM FEMs
calibration relating to comparability to the NAAQS. For example, a
commenter stated that it is important to ensure bias MQOs are met for
FEMs run at sites potentially affected by revised standards as well as
the need to accurately designate areas as attaining or not attaining
the NAAQS. There were comments supporting the correction of PM FEM data
as helping the EPA and SLT monitoring programs continue to evolve
toward more automated methods. For example, one commenter appreciates
the EPA's support for the ongoing move from filter-based
PM2.5 FRMs to use of continuous FEMs, stating that they
concur with the EPA's assessment that there is monitoring bias between
FRMs and FEMs, and commending EPA for recognizing ongoing data quality
issues for FEMs and for taking action to improve these issues in
collaboration with instrument manufacturers and SLT agencies.
A small number of commenters were critical of the proposed FEM
calibration approach. One commenter noted that EPA should further
examine the handling of FEM PM2.5 data when used for
comparison to the NAAQS. In response, we note that monitoring agencies
and the EPA will continue to examine the comparability and use of FEM
data used in comparison to the NAAQS. Another commenter suggested that
the calibration process for a designated PM monitor should not be
altered following Class III designation approval. The EPA disagrees as
we believe it is appropriate for FEMs to be calibrated with routinely
operated FRMs, because doing so is an efficient way to work towards FEM
data meeting the bias MQO across the networks in which the FEMs are
currently being operated. Also, having continuous PM FEMs meeting bias
MQOs allows the use of the data in a variety of other ways that
manually operated FRMs samplers cannot support. Another commenter
stated that, if a particular FEM designated make or model of
[[Page 16352]]
instruments fails to meet MQOs, then that make or model should be
removed from the designations altogether. The EPA agrees and clarifies
that the modification would not prevent removal of FEM designation from
a make or model of instrument under the existing 40 CFR 53.11--
Cancellation of reference or equivalent method designation. This may be
appropriate if there are no other solutions to improve the method such
that it achieves bias MQOs.
A few commenters provided specific recommendations for how the
regulatory language could be improved. These included comments that the
new regulatory language proposed for 40 CFR part 58, appendix C,
section 2.2 must ensure consistency and transparency when requesting
changes to the factory calibration; that the EPA should incorporate
binding regulatory language in 40 CFR part 58, appendix C, section 2.2
(i.e., it currently lacks ``shall'' or ``must'') to ensure the language
is not open to inconsistency and does not provide unique deference to
instrument manufacturers without a mechanism for transparent
communication of the changes being made and the supporting technical
analysis. A commenter also requested that the EPA define the core
requirements needed to ensure all requests for updating factory
calibrations are required to follow the same process, using data of the
same known quality, and evaluating the effectiveness of the resulting
correction factors consistently.
In response to these comments, while the EPA agrees that the
proposed regulatory language for 40 CFR part 58, appendix C, section
2.2 must ensure consistency and transparency when entities request
changes to factory calibrations, the EPA disagrees that the regulations
cannot also provide some flexibility. For example, we believe that a
degree of flexibility is appropriate regarding whether outliers in the
data to be used for factory calibration should or should not be
included, the range of data to be included, and in utilizing collocated
FRM and FEM data for updated calibrations from a representative set of
geographic areas in which it is produced. The EPA believes that the
proposal defined the core requirements needed to ensure all requests
for updating FEM factory calibrations will follow the same process,
using data of the same known quality and evaluating the effectiveness
of the resulting correction factors consistently.
In its proposal, the EPA identified that while there are several
types of data available to use as the reference for FEM calibration
updates, including data from routinely operated FRMs, audit program
FRMs, and PM2.5 chemical speciation samplers, the EPA
proposed to use routinely operated State, local, and Tribal FRMs as the
basis of comparison upon which to calibrate FEMs (88 FR 5670-71,
January 27, 2023). Importantly, routine SLT agency FRM data form the
largest portion of the monitored air quality data used in epidemiologic
studies that are being used to inform proposed decisions regarding the
adequacy of the public health protection afforded by the primary
PM2.5 NAAQS, as discussed in section II above.
Overall, there was broad and strong support for utilizing
collocated FRM data from routine SLT networks to provide calibrations
of the continuous FEMs. For example, several commenters agree that
valid SLT air monitoring data generated in routine networks and
submitted to the EPA will improve the PM concentration measurement
performance of approved FEMs. Another commenter provided support for PM
FEM instrument manufacturers to evaluate nationally available valid FRM
data as well as other data sets such as the performance evaluation
audit program to update factory calibrations. The EPA believes that the
routinely operated PM FRMs represent the best and largest source of
data to calibrate continuous PM FEMs, and that performance evaluation
audit program data should be kept independent of the calibration
process. This will mean that assessments of the routine monitoring
operations, including both the FRM and any future updated PM FEMs, will
appropriately remain independent in evaluating whether updated methods
are meeting bias MQOs. The EPA is, therefore, finalizing its approach
to use routinely operated SLT FRMs as the basis of comparison upon
which to calibrate continuous PM FEMs as proposed.
Regarding the EPA's proposed requirement to utilize factory
calibrations (88 FR 5670-71, January 27, 2023), several commenters
agreed that factory calibrations provide the best option to improve PM
FEMs. For example, one commenter stated that the correction factors are
necessary to preserve data integrity with the FRM, and they support the
proposal that the approach be initiated by instrument manufacturers and
implemented as a national solution through firmware updates.
Regarding the proposed requirement that calibrations be initiated
by instrument manufacturers (88 FR 5671, January 27, 2023), most
commenters were supportive of the proposed approach that recalibration
of FEM PM instruments be initiated by instrument manufacturers. For
example, one commenter stated they support allowing instrument
companies submit improvements to their existing FEMs, as vendors should
be encouraged to improve their methods. Another commenter noted that
having a methodology initiated by the manufacturer will have nationwide
consistency. A few of commenters recommended that SLT air agencies
should have the additional ability to petition the EPA Administrator to
initiate factory calibrations of FEMs to better meet MQOs when data
collected by their agencies indicate disparities, because the
monitoring agencies are responsible for the quality of the data from
the specific makes and models of instrumentation used in their
networks. While the EPA believes that, in most cases, the instrument
companies should be the ones to initiate the process for calibration of
FEMs to routinely operated FRMs, we agree with the commenters who
suggested that other options should be available, including allowing
monitoring agencies or MJOs to work independently or together to pursue
improvements to designated FEMs. However, the EPA believes that any
such improvements initiated by monitoring agencies or MJOs should still
be facilitated through the responsible instrument company. Also, any
such effort to improve data quality should be employed across all the
networks in which the methods are operated and not limited to the
networks operated by the agency(s) pursuing such improvements.
Regarding how frequently factory calibrations should be updated,
our proposal identified that it would be most appropriate to not define
a specific time period for updates; rather, updates should be based on
whether or not quality data is being produced across a given network
(88 FR 5672, January 27, 2023). Regarding this issue, one commenter
recommended that instrument manufacturers be required to evaluate and,
if necessary, adjust PM FEMs factory calibrations on an ongoing basis
at regular intervals. The EPA notes that while it does not have the
authority to require instrument companies to evaluate the quality of
data from operating FEMs under 40 CFR part 58, the EPA does routinely
participate in conferences and workshops and makes assessments of data
quality specific to instrument makes and models publicly available. The
EPA also regularly summarizes relevant FRM and FEM data
[[Page 16353]]
quality in documents such as the 2022 PA (U.S. EPA, 2022b). Therefore,
consistent with the proposal, we are not finalizing any specifics
regarding how frequently factory calibrations should be updated but
commit to continue to routinely provide information to SLT agencies
regarding FEM data quality.
The EPA proposed that the calibration of FEMs could apply to any of
the PM FEM method indicators (i.e., PM10, PM2.5,
and PM10-2.5) (88 FR 5670, January 27, 2023). The EPA
received only supportive comments. All comments that included a
discussion of three PM metrics support their inclusion for calibration
of PM FEMs. Therefore, the EPA is finalizing the inclusion of all three
PM indicators (i.e., PM10, PM2.5, and
PM10-2.5) as proposed.
The EPA proposed that either all data available or a range of data
up to 125% of the 24-hour NAAQS for the PM indicator of interest may be
used to establish new factory calibrations, (88 FR 5671-73, January 27,
2023). The EPA received many comments supportive of the proposal and
one comment offering a different approach on the range of data to use.
One commenter recommends that the EPA should consider using all
``validated'' data because how these instruments behave under normal
operating ranges may be just as important as how they behave when
monitoring conditions are low or elevated, and that the full range of
data should be used when determining the appropriate level of the
standard, just as the full range of data is used in determining if an
area is attaining the standard. In response to this comment, the EPA
believes that making allowances for some flexibilities will increase
the likelihood of instrument companies pursuing such improvements.
Also, even though there is flexibility, the EPA will still be able to
evaluate the appropriateness of a range of concentration data included
as part of each application submitted. Also, the EPA notes that in
certain circumstances, States do petition the EPA to set aside data
under the Exceptional Events Rule (Sec. 50.14, ``Treatment of air
quality monitoring data influenced by exceptional events''). Where
approved, exceptional event data are set aside from use in regulatory
decisions. Thus, there is a process to set aside certain high
concentration data for certain purposes. Therefore, the EPA is
finalizing the provision that factory calibrations may be based on a
range of valid data as proposed.
The EPA solicited comment on the representative set of geographic
locations to use in the calibration of FEMs compared to collocated FEMs
(88 FR 5671, January 27, 2023). Most commenters were supportive of the
approach of using representative sites in SLT networks from across the
country. For example, several commenters provided their support for PM
FEM instrument manufacturers to evaluate nationally available valid FRM
data to update factory calibrations. Commenters disagreeing with a
national geographic approach preferred to allow local solutions to
correct data. For example, one commenter suggested having a local or
regional option because PM instruments are impacted by, and respond
differently to, a variety of local factors, including relative
humidity, temperature, concentration levels, and particle composition.
The EPA agrees that there are challenges in the response of PM FEMs to
a variety of local factors; however, this can be true of many methods
and are not specific to PM FEMs and, therefore, does not provide a
reason to reject this approach in this instance. Another commenter
stated that the proposed national correction factor is a ``flawed
concept,'' suggesting that it is ``widely understood throughout the
monitoring community that monitors perform best with a local correction
factor.'' This commentor offered no record or citation supporting this
point. The EPA counters that while monitoring agencies may
statistically correct data from a PM continuous monitor for AQI
purposes (40 CFR part 58, appendix G), there are both examples of well
performing statistically corrected PM continuous monitors being used
for AQI purposes; however, without proper attention and updates, there
are also examples of poorly performing ones. Finally, another commenter
believes that a national correction factor cannot possibly incorporate
data to represent all the scenarios across the nation that have an
impact on monitor performance and data quality. Although the EPA agrees
that there are a variety of local scenarios that could affect monitor
performance, the overall benefits of having nationally consistent
measurement of PM concentrations and national calibration of data
outweigh the potential advantages of locally specific calibrations.
Several commenters also disagreed with using local and regional
calibrations of data, including some monitoring agencies that asserted
being unable to reinvest in the operation of FRMs that would be
required to locally calibrate their own PM FEMs. Further, every
approved PM FEM method designated today is effectively calibrated
through demonstration of field testing in the areas in which it was
required to be tested (40 CFR 53.35(b)(1)). Moreover, the EPA proposed
to require instrument manufacturers to demonstrate that they can
improve the number of sites meeting bias MQOs by initiating a
recalibration of an FEM. Thus, the use of a national set of sites where
the methods are operated is essentially a fine-tuning of the PM FEMs
performance across all sites where it is used.
After considering all the comments received, the EPA believes it is
appropriate to finalize as proposed with a representative set
geographic locations at SLT sites to calibrate PM FEMs. Identification
of such sites would be made by the applicant of the planned updated
calibration, subject to EPA approval, and submitted to the EPA in
accordance with the requirements and application instructions in 40 CFR
part 58, appendix C, sections 2.2 and 2.7. The EPA encourages early
communication between an applicant seeking a method update and the EPA
to facilitate the most appropriate sites are included in any updated
application of the methods calibration.
The EPA proposed that instrument companies may, but are not
required to, check for and exclude any potential outliers that may
exist in the validated State, local, and Tribal agency network data
available from AQS that would be used to establish new factory
calibrations. The EPA received two comments regarding potential outlier
approaches. One commenter disagreed with the proposed approach and
instead recommended the use of all ``validated'' data, because how
these instruments behave under normal operating ranges may be just as
important as how they behave when monitoring conditions are low or
elevated. The EPA acknowledges this point; however, the proposal on
outliers allows flexibility in using standard outlier tests if needed
to include or exclude such data as part of the calibration process.
Ultimately, the true test of success for an updated method calibration
will be that a higher number of sites are meeting bias MQOs in the
areas in which the method is used, which will include all routine valid
data including any potential outliers. Another commenter asserted
concerns with the ability of instrument manufacturers to analyze data
within individual monitoring agencies. The EPA disagrees with the
commenter because decisions whether to include or exclude outliers
should be flexible and made on a case-by-case basis. Moreover, the
expected substantially larger dataset from routinely operated
collocated FRMs and FEMs compared to what was
[[Page 16354]]
used in the original FEM designation testing (Sec. 53.35 Test
procedure for Class II and Class III methods for PM2.5 and
PM10-2.5) will minimize the effect of any potential
outliers.
In contrast to these two comments, the EPA received many comments
supportive of the proposed outlier approach overall. Therefore, the EPA
is finalizing this part of the proposal that instrument companies may,
but are not required to, check for and exclude any potential outliers
that may exist in the validated State, local, and Tribal agency network
data available from AQS that would be used to establish new factory
calibrations.
Several commenters offered input on statistical criteria and
initial testing requirements for approval of candidate PM FEMs and the
role of instrument manufacturers in this process. The EPA did not
propose any changes related to these issues; however, these comments
have been considered below.
One commenter suggested that data quality objectives, bias, and
precision estimators for different monitoring methods should be based
on averages at both national and regional levels for purposes of
comparison. Another commenter asked to strengthen the criteria for
Class 3 Equivalency standards for candidate PM instrumentation. On
testing requirements, one commenter recommended that the EPA consider
updating the 40 CFR part 53 process for approving FEMs so that the
testing process more closely reflects the regulatory deployment and
data handling that generates NAAQS-comparable data. Another commenter
asked that the results from ``summer'' and ``winter'' field evaluations
not be averaged together because it allows agencies to minimize the
error of biased instruments by averaging poor results with data often
biased in the other direction. The same commenter also recommended that
candidate instruments data sets should not be averaged together as is
done currently where data from triplicate instruments are averaged for
each day. Another commenter asked that the EPA require FEM field
comparability tests in the northwest (e.g., in EPA Region 10) in areas
where particulate derived from biomass predominates to ensure that
certified instruments will perform reliably in regions influenced by
these sources. Related to the different measurement principles and the
instrument companies' role in PM FEMs, one commenter noted that FEMs
may never align perfectly with the FRMs due to the use of different
measurement principles. Another commenter asked that manufacturers of
FEM instruments be held accountable for ensuring that they continue to
meet FEM criteria, whether through calibration updates and/or follow-up
evaluations. Another commenter suggested that instrument manufacturers
should be required to further evaluate the FEM monitoring data at
defined intervals including, but not limited to, the 2-year and 5-year
approval anniversaries.
The EPA did not propose to make modifications to the statistical
criteria or testing requirements; however, we did solicit comment on
any alternatives that would lead to more sites meeting the bias MQO
with automated FEMs, especially for those sites that are near the level
of the primary annual PM2.5 NAAQS as proposed (88 FR 5672-
73, January 27, 2023). While the comments requesting that the
statistical criteria be strengthened may have merit, doing so would not
address the large inventory of already deployed PM FEMs used throughout
the country. Also, without performing a detailed Data Quality Objective
(DQO) design process, it is unclear how changing one or more
statistical criteria would help improve the number of sites meeting the
bias MQO now or in the future. Similarly, while the comments asking for
changes to the locations of testing may also have merit, the EPA
believes this could be a deterrent for instrument manufactures to seek
additional improvements since more testing would be required, at least
for candidate methods. Regarding the comment on the different
measurement principles, the EPA concurs that different measurement
principles may never align perfectly. Also, the EPA notes that the
Agency has longstanding goals for acceptable measurement uncertainty of
automated and manual PM2.5 methods in 40 CFR part 58,
appendix A, section 2.3.1.1. Therefore, while having different
measurement principles align is useful, meeting the goal for acceptable
measurement uncertainty is the objective.
Regarding the comments related to the instrument companies' role in
PM FEMs, the EPA notes that FEMs are already required to meet 40 CFR
53.9, ``Conditions of designation.'' Specifically, 40 CFR 53.9(c)
requires that, ``Any analyzer, PM10 sampler,
PM2.5 sampler, or PM10-2.5 sampler offered for
sale as part of an FRM or FEM shall function within the limits of the
performance specifications referred to in Sec. 53.20(a), Sec.
53.30(a), Sec. 53.35, Sec. 53.50, or Sec. 53.60, as applicable, for
at least 1 year after delivery and acceptance when maintained and
operated in accordance with the manual referred to in Sec.
53.4(b)(3).'' The EPA does not have the authority to require instrument
manufacturers to further evaluate the FEM monitoring data at defined
intervals, including but not limited to the 2-year and 5-year approval
anniversaries, as one commenter suggested.
In addition to these few recommendations, the EPA received many
comments supportive of the proposal that valid State, local, and Tribal
air monitoring data from FRMs generated in routine networks and
submitted to the EPA may be used to improve the PM concentration
measurement performance of approved FEMs; therefore, consistent with
the proposal we are not finalizing any updates to the statistical
criteria, testing requirements, or requirements on instrument
manufactures as proposed.
The EPA proposed that any new factory calibration should be
developed using data from at least 2 years and tested on a separate
year(s) of data (88 FR 5672, January 27, 2023). Comments on this part
of the proposal were generally supportive. One commenter requested that
at least a 3-year dataset, rather than the proposed 2 years, be used
for a representative design value comparison of the FEM and FRM
datasets to be evaluated. Another commenter pointed out that as large a
data set as possible should be used, but EPA should not limit it to
only data collected by instruments that have operated for more than 2
years.
In response to these comments, the EPA notes the broad support for
the proposal as written. Also, the EPA notes that the 2-year period for
using data to develop a factory calibration is a minimum, and that more
years may be used as appropriate. Therefore, the EPA is finalizing its
approach that any new factory calibration should be developed using
data from at least 2 years and tested on a separate year(s) of data as
proposed.
The EPA proposed several aspects of the FEM calibration on which we
did not receive specific comments, including a provision that FEM
methods should be evaluated by monitoring agencies as part of routine
data assessments, such as during certification of data and 5-year
assessments; the fact that the EPA recognizes only data from existing
operating sites are available for use in factory calibrations; and
recognition that an updated factory calibration does not have to work
with the original field study data submitted that led to the
designation as an FEM. With the broad general support from commenters
summarized above, the EPA is finalizing each of these
[[Page 16355]]
individual aspects of the FEM calibration as proposed.
In the proposal, the EPA identified that we should expect a lag
between the date when an already designated method is approved with a
new factory calibration as an updated method by the EPA and when it can
be implemented in the field. The EPA solicited comment on how to
approach the data produced during this lag. Commenters provided input
not only on how to address data during the lag, but also regarding how
to address data already collected prior to a method update that has the
potential to be used in regulatory decision making, particularly where
such collected data do not meet the bias MQO. In response to this
solicitation of comment, there was a consistent recommendation that
calibrations of data associated with method updates should be applied
to all relevant PM data prior to the EPA using it for designations
under a final NAAQS.
While the EPA appreciates these comments and recognizes their
support for retroactive data correction, at this time and following
this final rule, monitoring agencies should continue to report PM FEM
data as measured. This component of this final rule is focused only on
revising 40 CFR part 53, appendix C to implement an updated calibration
for approved PM FEMs. The issue of how prior and future monitoring data
will be used in the implementation of this NAAQS, such as for
designations, and for air quality regulatory programs is outside the
scope of this rulemaking and will therefore be addressed by the EPA in
a subsequent relevant action or actions.
The EPA received comments on whether updates to PM FEM methods
should be required to be implemented or there would flexibility in when
and if a monitoring agency implemented them. The commenters asked that
EPA be flexible in allowing the use of updated method correction
factors intended to improve the data comparability between the FRMs and
FEMs.
In most cases, the EPA expects that updating the FEMs will result
in improved data quality and more sites meeting bias MQOs; however, the
EPA is not finalizing an update requirement in this action. Monitoring
agencies can assess their data and make decisions on an update based on
whether they are meeting the bias MQOs. Such decisions on whether or
not to update a method may efficiently be included in those agencies'
annual monitoring network plans under 40 CFR 58.10, ``Annual monitoring
network plan and periodic assessment,'' which are already subject to
EPA Regional office approval. In some circumstances, it is possible the
original PM FEM may be revised in a manner where only the updated
method has an active approved designation. In these cases, monitoring
agencies would need to address updating their PM FEM in a timely
manner.
The EPA solicited input on any alternative approaches that could
lead to more sites meeting the bias MQO with automated PM FEMs,
especially for those sites that are near the level of the primary
annual PM2.5 NAAQS as proposed to be revised in section II
above. A few commentors provided input on potential options for
alternative approaches and several others offered input on how a local
or regional calibration of an FEM could work. Among alternative
approaches, one commenter suggested that manufacturers of FEMs could
provide settings that would allow for adjustments to make FEM data more
``FRM-like.'' Another commenter suggested working with the
manufacturers of FEM equipment to diagnose the cause of the bias and
then to address it appropriately.
The EPA received several comments on how to implement a local or
regional calibration of FEMs. One commenter suggested that EPA could
allow for SLT agencies to adjust FEM data to be more ``FRM-like'' prior
to submitting data to AQS. Another commenter suggested using a rolling
3-month linear regression based on a comparison of FEM data to
PM2.5 levels measured by a 1-in-6-day FRM. Another commenter
recommended that the EPA allow the application of a correction factor
that is from an area with a similar climate and other conditions.
Another commenter suggested that, for metropolitan statistical areas
(MSAs) where the re-calibrated FEMs still do not meet equivalency
criteria, monitoring agencies should be able to use the rolling linear
regression technique to further calibrate the FEMs within an MSA.
Another commenter suggested that developing a simple linear regression
could establish the relationship between FEM data and FRM data and be
used to adjust the FEM data at each site where they are collocated.
Another commenter suggested that averaging the results within a MSA and
applying it on an MSA basis with the previous 2 years of data could
provide an adjustment method for sites without a collocated FRM.
Another commenter identified that a regional correction factor
potentially could improve instrument accuracy to biomass sources, which
are a large component of PM in many communities.
Among the alternative approaches suggested, having settings that
would allow for adjustments to make FEM data more ``FRM-like'' has
merit, but assuming this was within a PM FEM itself, it would need to
be separately incorporated into each make and model of FEM. If EPA were
to pursue this alternative approach, the suggestion could be
incorporated into a future regulatory action as a potential condition
of designation because, without having the opportunity to thoughtfully
consider how every step of such an approach would need to work,
including what such requirements would look like and how potential
settings adjustments would be made, it is not appropriate for the EPA
to require the availability of such settings now, nor would it address
the inventory of currently available PM FEMs already operating.
Regarding the suggestion that the EPA and SLTs should work with the
manufacturers of FEM equipment to diagnose the cause of any biases and
then to address them appropriately, the EPA supports this
recommendation, but does not believe a regulatory change is required to
allow the monitoring community (EPA and SLTs) to work with instrument
manufacturers in this way.
Regarding the several comments on how to implement a local or
regional calibration of FEMs, the EPA acknowledges the desire for this
flexibility but believes that any such provisions for local or regional
calibration of FEMs would need to be thoroughly thought out and
proposed for consideration across the monitoring community. While
several commenters support such an approach, the EPA also received
adverse comments on the potential for local and regional calibration of
PM FEMs instead of national. Most of the criticism of local and
regional calibration of PM FEMs centered on both the lack of existing
operating PM FRMs in commenters' networks and monitoring agencies'
inability to staff the higher number of operating FRMs that would have
to be collocated with PM FEMs to calibrate. Thus, the commenters that
oppose local and regional calibrations of data prefer to utilize the
national calibration of FEM data as proposed. Acknowledging all of
these viewpoints, the EPA believes that it would not be appropriate to
institute such an approach at this time. As discussed throughout this
section, this final rule, the EPA is embarking on a new national
approach to calibration of FEMs where valid State, local, and
[[Page 16356]]
Tribal air monitoring data from FRMs generated in routine networks and
submitted to the EPA may be used to improve the PM concentration
measurement performance of approved FEMs. The EPA and the community of
SLT monitoring agencies can further consider other solutions to
improving PM FEM methods, including local and regional scale
calibration of FEMs, in a future review of the PM NAAQS.
In summary, the EPA is finalizing its proposal to allow valid
State, local, and Tribal air monitoring data from PM FRMs and FEMs
generated in routine networks and submitted to the EPA to update
factory calibrations included as part of approved FEMs (40 CFR part 58,
appendix C, sections 2.2 and 2.7). This approach, which will typically
be initiated by instrument manufacturers but can also be spurred by
monitoring agencies, MJOs of monitoring agencies, and the EPA itself,
is to be implemented as a national solution in factory calibrations of
approved FEMs through a firmware update, subject to EPA approval. FEM
calibrations can apply to any PM FEM methods (i.e., PM10,
PM2.5, and PM10-2.5). As part of this process,
the EPA is finalizing that a range of data based on the most
representative concentrations up to all available concentrations may be
used in developing and testing a new factory calibration; that a
representative set of geographic locations can be used; that outliers
may be included or not included; that a new factory calibration should
be developed using data from at least 2 years and tested on a separate
year(s) of data; that updates to factory calibrations can occur as
often as needed and should be evaluated by monitoring agencies as part
of routine data assessments such as during certification of data and 5-
year assessments; that the EPA recognizes only data from existing
operating sites is available; and that an updated factory calibration
does not have to work with the original field study data submitted that
led to the designation as an FEM. The EPA is finalizing this approach
as proposed with the intention of having more sites meet the bias MQOs
with automated PM FEMs.
4. Revisions to the PM2.5 Monitoring Network Design Criteria
To Address At-Risk Communities
To enhance protection of air quality in communities subject to
disproportionate air pollution risk, particularly in light of the
proposed range for a revised primary annual PM2.5 standard,
the EPA proposed to modify the PM2.5 monitoring network
design criteria to include an environmental justice (EJ) factor that
accounts for proximity of at-risk populations (i.e., those identified
in the 2019 ISA and ISA Supplement as being at increased risk of
adverse health effects from PM2.5 exposures to sources of
concern), consistent with the statutory requirement that the NAAQS
protect the health of at-risk populations (88 FR 5673, January 27,
2023). Specifically, the EPA proposed to modify the existing
requirement at 40 CFR part 58, appendix D, section 4.7.1(b)(3)): ``For
areas with additional required SLAMS, a monitoring station is to be
sited in an area of poor air quality,'' to additionally address at-risk
communities with a focus on anticipated exposures from local sources of
emissions. The scientific evidence evaluated in the 2019 ISA and ISA
Supplement indicates that sub-populations at potentially greater risk
from PM2.5 exposures include children, lower socioeconomic
status (SES) \188\ populations, minority populations (particularly
Black populations), and people with certain preexisting diseases
(particularly cardiovascular disease and asthma). The EPA proposed that
communities with relatively higher proportions of sub-populations at
greater risk from PM2.5 exposure within the jurisdiction of
a State or local monitoring agency should be considered ``at-risk
communities'' for these purposes.
---------------------------------------------------------------------------
\188\ SES is a composite measure that includes metrics such as
income, occupation, and education, and can play a role in
populations' access to healthy environments and healthcare.
---------------------------------------------------------------------------
The PM2.5 network design criteria have led to a robust
national network of PM2.5 monitoring stations. These
monitoring stations are largely in Core-Based Statistical Areas (CBSAs)
\189\ across the country that include many PM2.5 monitoring
sites in at-risk communities. Many of the epidemiologic studies
evaluated in the 2019 ISA and ISA Supplement, including those that
provide evidence of disparities in PM2.5 exposure and health
risk in minority populations and low-SES populations, often use data
from these existing PM2.5 monitoring sites. However, we
anticipate that with the more protective annual NAAQS finalized in
section II above, characterizing localized air quality issues around
local emission sources may become even more important. The EPA believes
that adding a network design requirement to locate monitors in at-risk
communities will improve our characterization of exposures for at-risk
communities where localized air quality issues may contribute to air
pollution exposures. Requiring that PM2.5 monitoring
stations be sited in at-risk communities will allow other methods to be
operated alongside PM2.5 measurements to support multiple
monitoring objectives per 40 CFR part 58, appendix D, section 1.1. The
EPA believes that it is appropriate to formalize the monitoring
network's characterization of PM2.5 concentrations in
communities at increased risk to provide such areas with the level of
protection intended with the PM2.5 NAAQS. The addition of
this requirement will also lead to enhanced local data that will allow
air quality regulators help communities reduce exposures and inform
future implementation and reviews of the NAAQS.
---------------------------------------------------------------------------
\189\ Metropolitan and Micropolitan Statistical Areas are
collectively referred to as ``Core-Based Statistical Areas.''
Metropolitan statistical areas have at least one urbanized area of
50,000 or more population, plus adjacent territory that has a high
degree of social and economic integration with the core as measured
by commuting ties. Micropolitan statistical areas are a set of
statistical areas that have at least one urban cluster of at least
10,000 but less than 50,000 population, plus adjacent territory that
has a high degree of social and economic integration with the core
as measured by commuting ties.
---------------------------------------------------------------------------
The EPA received comments concerning the proposed requirement to
modify the PM2.5 monitoring network design criteria to
include an EJ factor that accounts for the proximity of populations at
increased risk of adverse health effects from PM2.5
exposures to sources of concern. Commenters included State, local, and
Tribal air agencies and multijurisdictional organizations (MJOs)
comprised of those agencies; industry and industry groups; other
Federal, State, and local government entities; public health, medical,
and environmental nongovernmental organizations (NGOs); and private
citizens. The EPA proposed to require that sites located in at-risk
communities (particularly those whose air quality is potentially
affected by local sources of concern) should nonetheless meet the
requirements to be considered representative of ``areawide'' air
quality as this is consistent with all other minimally required sites.
There were several other technical components of the proposed
requirement for which we asked for comment, including: how to identify
at-risk communities; the PM sources of concern important to consider;
the datasets that can be used to identify communities with high
exposures; the most useful measurement methods to collocate with
PM2.5 in at-risk communities; and the timeline to implement
any new or moved sites.
Overall, most commenters were very supportive of the EPA's proposed
modification to the PM2.5 monitoring
[[Page 16357]]
network design criteria to include an EJ factor that accounts for
proximity of populations at increased risk of adverse health effects
from PM2.5 exposures to sources of concern. A few commenters
offered detailed supporting comments. For example, one commenter
recommended targeting investment in regulatory monitors in EJ
communities, opining that there is presently a lack of equitable
distribution of these monitors in low-income and minority communities.
Another commenter supports the inclusion of an EJ factor in
PM2.5 monitoring network design criteria as a means to
assess whether disparities in exposure are reduced in the future. The
EPA appreciates the support for the proposed requirement and
acknowledges the desirability of a goal to assess if disparities in
exposure are reduced in the future as a result of these monitoring
efforts.
Some commenters were generally supportive of the proposed
requirement but suggested that the EPA should recast the approach in a
more specific way or offered additional examples of sources of concern.
For example, one commenter stated that PM2.5 emissions from
residential and commercial wood burning result in localized hotspots
that are often not revealed by community air monitoring. Another
commenter asked that the EPA adopt a strategy to monitor EJ communities
near both larger well-known point sources of PM2.5 and along
traffic corridors as well as smaller sources that, when taken together,
may create a large amount of emissions and health harms in the area.
Another commenter stated that the national network of monitors operated
by the EPA captures data used for generalized modeling, but overall
monitoring is not as granular as one would expect, especially in urban
areas. For instance, the commenter suggested that EPA could monitor
suspected ``hot spots'' (e.g., residential development adjacent to
highways and active construction sites) to better manage and mitigate
PM2.5 pollution at their sites of origin, and that more
extensive and granular monitoring data would also facilitate essential
research and inform future evaluations and adjustments of the NAAQS.
The EPA acknowledges these comments identifying other sources of
concern, and we address these and other potential sources of concern
below.
Among adverse comments, a few commenters stated that ``at-risk
communities'' is not well defined. The EPA disagrees and directs those
commenters to the numerous places where this definition is covered,
including in Section II.B.2 of the proposal where we explained the term
related to a variety of at-risk populations (88 FR 5591-92, January 27,
2023) as well as section 12.5 of the 2019 ISA (U.S. EPA, 2019a) and
section 3.3.3 of the ISA Supplement (U.S. EPA, 2022a). Other commenters
oppose the addition of the proposed monitoring because they feel it
would reduce flexibility for agencies in deciding where they should
site monitors, advocating that monitoring agencies should be afforded
maximum flexibility to identify where to site monitors for at-risk
areas. Because the EPA recognizes the challenges cited by these
commenters related to establishing new ambient air monitoring stations,
the EPA is finalizing the modified requirement on PM2.5
monitoring network design criteria intended to address at-risk
communities that allows flexibility regarding which EJ communities
should be monitored. Finally, one commenter asked that the EPA clarify
a specific metric to judge how to site monitors in at-risk communities.
Instead, the EPA believes it is appropriate for agencies to recommend
what they believe to be the most important things to consider for their
sites to meet the PM2.5 network design requirements and,
thus, applying a new metric could take away from local priorities for
at-risk communities.
A few commenters asked that the EPA require more monitoring than
proposed. One commenter stated that it would be more beneficial to
overburdened communities if air monitoring were required in all at-risk
communities. A few commenters asked that EPA require additional
monitoring for attainment of PM2.5 NAAQS in EJ communities.
In response to these comments, the EPA supports the SLT agencies'
initiatives to conduct additional monitoring beyond the minimum
monitoring requirements and network design criteria. In addition, the
EPA supports agencies' use of alternative datasets such as sensors and
sensors networks, satellites, and other non-regulatory monitoring where
appropriate for non-regulatory data uses. The EPA notes that many
monitoring agencies already operate more monitoring sites than are
minimally required, and we expect this to continue as agencies consider
siting monitors in at-risk communities.
However, the EPA also received substantial concerns from monitoring
agencies about their resource constraints, including staffing to
support any potential new monitoring. The EPA also notes that the
existing and robust network of almost 1,000 PM2.5 sites
nationally is designed to continue to protect all populations at the
level of the NAAQS discussed in section II of this final action by
always having at least one site in the area of expected maximum
concentration for each CBSA where monitoring is required. As a result
of the revisions to the annual PM2.5 NAAQS being finalized
in this action, a small number of new monitoring sites will also be
required under EPA's current minimum monitoring requirements. With the
monitoring network design changes finalized in this rule, many of these
existing and new sites will form an important sub-component of the
PM2.5 network by better characterizing air quality in at-
risk communities, particularly with respect to sources of concern.
The EPA concludes that the requirements in this final rule for
siting of monitoring in at-risk communities will meaningfully improve
the PM2.5 monitoring network and its characterization of air
quality in at-risk communities, without placing substantial new
resource burdens on States and their monitoring agencies that would be
associated with requirements for additional monitoring sites.
Therefore, the EPA is finalizing this part of the proposed action
without requiring additional monitoring sites beyond what would be
associated with the revised annual PM2.5 NAAQS described in
section II as they pertain to the minimum requirements associated with
Table D-5 of Appendix D to Part 58--PM2.5 Minimum Monitoring
Requirements.
A few commenters asked that the EPA enhance monitoring in smaller
cities and rural areas. One commenter asked for the EPA to extend the
proposed monitoring network to Micropolitan Statistical Areas with
populations of 10,000-50,000 and to rural areas. Another commenter
pointed out that current air quality monitoring networks focus on urban
and densely populated areas; therefore, rural areas are often not
captured in this existing monitoring infrastructure, despite well-
documented examples of high PM concentration in rural communities. The
commenter believes this results in inadequate assessment of air
pollution exposures for a substantial segment of the U.S. population.
The EPA disagrees that there needs to be additional requirements for
small CBSA's and rural areas. Regarding these comments, the EPA points
out that we have a long-standing requirement for each State to monitor
at background and transport sites (40 CFR part 58, appendix D, section
4.7.3--Requirement for PM2.5 Background and Transport
Sites). Also, if an agency deems it appropriate to do so, monitoring
coverage of rural areas can be accomplished with other tools
[[Page 16358]]
such as sensors and sensors networks, satellites, and other non-
regulatory monitoring. Although there may be short-term high exposures
in rural areas, there is no evidence that long-term averages are higher
in rural areas compared to urban areas with significantly higher
density of populations and emissions. For smaller cities or rural areas
that may have concentrations near the level of the PM2.5
NAAQS finalized in section II above, monitoring agencies are encouraged
to monitor and address emissions as appropriate.
Some commenters disagree that the proposed revision to the
PM2.5 monitoring network design criteria to address at-risk
communities is needed. One commenter stated that including an EJ factor
is not necessary because the current network is designed to protect all
citizens. Another commenter stated that EJ factors could be cumbersome
to implement. Another commenter asserted the proposal to add SLAMS in
at-risk communities with higher PM2.5 concentrations might
create more granular data and provide for a greater margin of safety
for those communities and monitors in such a way that data from those
areas could misrepresent the larger area represented by the network. In
response to the comment on the current network protecting all citizens,
the EPA agrees that by measuring in the community with the highest
concentration of PM2.5 we protect other citizens; however,
as stated in the proposal, the EPA believes that adding a requirement
for sites with an EJ factor near sources of concern will enhance the
overall network to the benefit of all citizens. Also, we anticipate
that with the more protective annual NAAQS finalized in section II
above, characterizing localized air quality issues will become even
more important around local emission sources. As for EJ factors being
cumbersome to implement, the EPA disagrees because there are many such
locations already operating successfully in the current network.
Regarding the comment that sites in at-risk communities may
misrepresent the larger area represented by a particular network, the
EPA notes that pursuant to 40 CFR part 58, minimally required sites in
a given network are to represent area-wide air quality; therefore,
sites in at-risk communities, by definition, would be representative of
the communities within the network in which they are sited for the
level of protection intended under the annual PM2.5 NAAQS.
In the proposal, the EPA identified that, in light of the evidence
of increased risk to at-risk communities, it would be appropriate to
better characterize exposures for communities in proximity to local
sources of concern (88 FR 5673-76, January 27, 2023). Thus, the EPA
proposed that enhanced networks should include representation of at-
risk communities living near emission sources of concern (e.g., major
ports, rail yards, airports, industrial areas, or major transportation
corridors). The EPA requested comment on the types of sources of
concern most important to consider. In addition to supporting the types
of sources the EPA identified in the proposal, commenters also
identified several additional localized sources such as railroads,
stationary sources, transportation facilities, and communities with
high numbers of wood stoves.
A few commenters suggested the inclusion of sources that are often
considered line and/or area sources, e.g., traffic corridors and
emissions from federally regulated facilities, military installations,
and national forests. Commenters also identified other sources usually
associated with long-range transport such as smoke from wildfire and
prescribed fires and long-distance transport of PM, for example from
Saharan dust and other international transport. As explained in the
proposal, the site with the highest expected PM2.5 is
already required to have a monitor by our long-standing requirement
that monitors be placed ``. . . in the area of expected maximum
concentration'' (Sec. 58.1 and appendix D, section 4.7.1(b)(1)). The
EPA expects that both sites with the expected maximum concentration and
sites specifically placed in at-risk communities would be impacted by
any long-range transport in the area. Therefore, the EPA believes any
emphasis on the sources of concern should prioritize localized sources,
including point, area, and line sources of concern impacting the at-
risk community of interest. Therefore, based upon the comments, the EPA
is finalizing a broader example list of sources of concern to include
localized sources such as point sources and transportation facilities,
since these are the most commonly expected additional sources of
concern. In response to the other sources of concern suggested by
commenters, the EPA notes that while it has provided examples, the
siting of monitors in EJ communities would not be limited to these
examples. Thus, the revised set of examples would include ``a major
industrial area, point source(s), port, rail yard, airport, or other
transportation facility or corridor.'' In finalizing this modified list
of examples, the EPA is not looking to prioritize one type of source
category over another; rather, we intend to further illustrate the
types of localized sources of pollution that might impact at-risk
communities such that the siting of monitors nearby may be appropriate.
One commenter noted that the proposal may have unintentionally
taken out the requirement related to specific design criteria for
PM2.5 in 40 CFR part 58, appendix D, 4.7.1(b)(3) that, for
an area with a requirement for an additional SLAMS monitor, it should
``be sited in an area of poor air quality.'' Thus, the language as
proposed neither requires that such monitors be sited in areas of poor
air quality, nor does it require that the monitor be sited in an area
that is anticipated to experience poor air quality from unspecified
(and thus potentially relatively insignificant) sources in the area.
The EPA agrees that this was not our intention; the EPA wants to
protect populations in at-risk communities by ensuring they are
protected by the NAAQS when there are sources of concern that may be
impacting them (i.e., not insignificant sources). Thus, the EPA is
reinstating this requirement in the network design language and
combining it with the examples of the types of localized sources of
concern: ``For areas with additional required SLAMS, a monitoring
station is to be sited in an at-risk community with poor air quality,
particularly where there are anticipated effects from sources in the
area (e.g., a major industrial area, point source(s), port, rail yard,
airport, or other transportation facility or corridor).''
To ensure minimally required monitoring sites appropriately
represent exposures in at-risk communities, the EPA proposed that sites
represent ``area-wide'' air quality near local sources of concern (88
FR 5674, January 27, 2023). Sites representing ``area-wide'' air
quality are those monitors sited at neighborhood, urban, and regional
scales, as well as those monitors sited at either micro- or middle-
scale that are identified as being representative of many such
locations in the same Metropolitan Statistical Area (MSA).\190\ Most
existing--as well as new or moved sites--are expected to be
neighborhood-scale, which means that the monitoring stations would
typically represent conditions throughout some reasonably homogeneous
urban sub-region with dimensions of a few kilometers per part 58,
appendix D, section 4.7.1(c)(3). Additionally, as described in Sec.
58.30,
[[Page 16359]]
sites representing ``area-wide'' air quality have a long-standing
applicability to both the annual and 24-hour PM2.5 NAAQS.
Our proposed requirement for siting monitors in communities
representing ``area-wide'' air quality is consistent with other network
design objectives pursuant to which we seek to have monitors located
where people live, work, and play.
---------------------------------------------------------------------------
\190\ MSA means a CBSA associated with at least one urbanized
area of 50,000 population or greater. The central-county, plus
adjacent counties with a high degree of integration, comprise the
area.
---------------------------------------------------------------------------
The EPA received a few comments on its proposed requirement that
minimally required sites represent ``area-wide'' air quality. One
commenter stated that the inclusion of a provision for EJ would narrow
the location of monitors to certain communities that may not best
represent ``areawide'' air quality. Another commenter asked the EPA to
consider removing requirements that sites be area-wide, since 24-hour
and annual averaging times would miss short, elevated pollution events.
A couple commenters had concerns with the difference in the scale of
representation between EJ monitors using small scale and other NAAQS
monitors using area-wide scale, in that area-wide scale would not
protect those most at risk. However, another commenter agreed with the
EPA that sites representing at-risk communities should represent area-
wide air quality. In addition to these comments, the EPA received many
comments with support for its proposed modifications to the network
design criteria as whole.
Regarding whether narrowing the location to certain communities may
not best represent ``area-wide'' air quality, the EPA notes that sites
are either identified as being area-wide or not; the EPA did not
suggest it was seeking a best ``area-wide'' location. In response to
the comment that area-wide site may miss short, elevated pollution
events, the EPA is aware that there can be local, short-term spikes in
PM2.5 concentrations. However, the network design criteria
associated with minimally required sites is applicable to both the
annual and 24-hour PM2.5 NAAQS, and the EPA believes it is
appropriate to continue to ensure all minimally required sites have the
most utility and remain applicable to both forms of the
PM2.5 NAAQS. The identification of unique micro- and middle-
scale sites was directed at discretionary efforts of any monitoring
agency, with the recognition that such sites, (i.e., relatively unique
micro-scale, or localized hot spot, or unique middle-scale impact
sites), are not applicable to the annual NAAQS as described in Sec.
58.30--Special consideration for data comparison to the NAAQS.
After considering all the comments on this topic, the EPA is
finalizing this part of the modification to the network design criteria
to maintain, consistent with our long-standing network design criteria,
that all minimally required sites are to represent area-wide air
quality.
In addition to using data from the robust network of almost 1,000
PM2.5 sites for NAAQS and AQI purposes, having a stable
network of long-term sites is especially valuable to examine trends and
to inform long-term health and epidemiology studies that support
reviews of the PM NAAQS. Therefore, while we proposed to add a
PM2.5 network design criterion to address at-risk
communities, many sites are likely already in valuable locations
meeting one of the existing network design criteria (i.e., being in an
area-wide area of expected maximum concentration or collocated with
near-road sites) and supporting multiple monitoring objectives. Also,
in many communities, there may already be sites meeting the network
design criterion we proposed for at-risk communities. Thus,
acknowledging the value of having long-term data from a consistent set
of network sites, the EPA believes that moving sites should be
minimized, especially in MSAs with a small number of sites. However,
because a small number of new sites are expected to be required due to
the existing minimum monitoring requirements (40 CFR part 58, appendix
D, Table D-5) \191\ and the revised primary annual PM2.5
NAAQS detailed in section II, and because sites occasionally have to be
moved--due to, for example, loss of access to a site or a site no
longer meeting siting criteria--the EPA believes it is appropriate to
prioritize establishing sites in at-risk communities near sources of
concern, whenever new sites are established, whether because it is a
new site or a replacement for a prior site that must be moved. The EPA
accordingly proposed that annual monitoring network plans (40 CFR
58.10(a)(1)) and 5-year assessments (40 CFR 58.10(d)) that include any
of the few new sites that will be required include a commitment to
examine the ability of existing and proposed sites to support air
quality characterization for areas with at-risk populations in the
community and the objective discussed herein.
---------------------------------------------------------------------------
\191\ Gantt, B. (2022). Analyses of Minimally Required
PM2.5 Sites Under Alternative NAAQS. Memorandum to the
Rulemaking Docket for the Review of the National Ambient Air Quality
Standards for Particulate Matter (EPA-HQ-OAR-2015-0072). Available
at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
In the proposal, the EPA identified that assessing and prioritizing
at-risk communities for monitoring can be accomplished through several
approaches (88 FR 5675). The most critical aspect of prioritizing which
communities to monitor is their representation of the at-risk
populations described earlier in this section. The other major
consideration is whether the community is near a source or sources of
concern. While many CBSAs have one or more sources of concern described
above, some CBSAs will not have a quantity of emissions from sources of
concern that result in an elevated level of measured PM2.5
concentrations in surrounding communities. The siting criteria to be
``in the area of expected maximum concentration,'' Sec. 58.1 &
appendix D, section 4.7.1(b)(1) ensures there is a monitoring site in
the community with the highest exposure in each CBSA with a monitoring
requirement. Some CBSAs may also have a requirement to collocate a
PM2.5 monitor at a near-road NO2 station.
Therefore, the EPA believes that for cases where an additional
PM2.5 site is required, we should include a criterion that
the site be in an at-risk community when there are no sources of
concern identified in that CBSA, or such sources do exist but are not
expected to lead to elevated levels of measured PM2.5
concentrations.
In its proposal, the EPA highlighted that tools such as the EPA's
EJSCREEN \192\ are available to identify the at-risk communities
intended for monitoring as part of the proposed revision to the
PM2.5 network design criteria (88 FR 5675-76, January 27,
2023). The EPA solicited comment on other tools and/or datasets that
can be utilized to identify at-risk communities. In addition to support
for using EJSCREEN, commenters identified several other options to
identify at-risk communities intended for monitoring as part of the
proposed revision to the PM2.5 network design criteria.
Among similar tools, one commentor suggesting using
CalEnviroScreen.\193\ Commenters also identified different options for
models including InMAP,\194\ satellite-derived models that can be
employed to help identify EJ communities, and hybrid models. A few
commenters also suggested using sensors and sensor networks such as the
BlueSky \195\ and PurpleAir \196\ sensors.
---------------------------------------------------------------------------
\192\ See: https://www.epa.gov/ejscreen.
\193\ See: https://oehha.ca.gov/calenviroscreen.
\194\ See: https://inmap.run/#home.
\195\ Mention of commercial names does not constitute EPA
endorsement.
\196\ Mention of commercial names does not constitute EPA
endorsement.
---------------------------------------------------------------------------
The EPA supports the use of other State and local tools designed to
help identify the at-risk communities that
[[Page 16360]]
should be monitored to meet the revised network design criteria. The
EPA additionally agrees with commenters that the use of models as well
as sensors and sensor networks may be appropriate and helpful in
identifying the most appropriate at-risk communities in which to locate
monitors.
For at-risk communities, monitoring agencies need data that can
best inform where there may be elevated levels of exposures from
sources of concern. While we use FRMs and FEMs to determine compliance
with the NAAQS, data from these methods will only be available at
existing sites. However, there are several additional datasets
available that may be useful in evaluating the potential for elevated
levels of exposure to communities near sources of concern. In the
proposal, EPA identified potential non-regulatory monitoring datasets
such as CSN, IMPROVE, and AQI non-regulatory PM2.5
continuous monitors; modeling data that utilizes emission inventory and
meteorological data; emerging sensor networks such as those that
comprise EPA and the USFS's Fire and Smoke Map; \197\ and satellites
that measure radiance and, with computational algorithms, can be used
to estimate PM2.5 from aerosol optical depth (AOD) (88 FR
5675-76, January 27, 2023). The EPA solicited comment on datasets most
useful to identify communities with high exposures for PM2.5
NAAQS (i.e., annual or 24-hour). In addition to providing information
about datasets that can inform the NAAQS comparison, commenters
additionally identified several types of datasets that may be useful to
identify where there may be elevated levels of exposures from sources
of concern. These datasets include satellite measurements, sensors, and
sensor network data, which may all be useful to find hot spots in
communities. Commenters also identified EJScreen and CalEnviroScreen,
which are screening and mapping tools that utilize several datasets.
Another commenter stated that to better understand exposure differences
in disadvantaged communities, shorter measurement intervals should be
measured and reported.
---------------------------------------------------------------------------
\197\ See: https://fire.airnow.gov/.
---------------------------------------------------------------------------
In considering the datasets identified in the proposal as well as
the ones commenters provided, the EPA believes all the datasets have
value to help inform where there may be elevated levels of exposures
from sources of concern. However, each of them may also have
limitations and, therefore, users should be careful not to rely solely
on one dataset versus another for all purposes. Fortunately, many of
the available datasets are becoming easier to work with and more
accessible, which will allow interested parties and monitoring agencies
the opportunity to efficiently review the datasets and determine best
applicability. For all of these reasons, the EPA is not finalizing a
requirement to use a specific dataset or tool to identify at risk
communities; however, whatever datasets a monitoring agency elects to
use, its plan to use such data for purposes of meeting the network
design requirements will be subject to EPA approval as part of the 40
CFR 58.10 annual monitoring network plan. Regarding the comment
recommending shorter measurement intervals in measuring and reporting
data to better understand exposure differences in disadvantaged
communities, the EPA agrees and generally supports use of continuous
methods. While we generally support use of continuous methods, approved
filter-based technologies and methods also provide valuable air quality
information. Therefore, the EPA is not requiring the use of automated
continuous methods beyond what is already required in 40 CFR part 58,
appendix D, section 4.7.2--Requirement for Continuous PM2.5
Monitoring.
The monitoring methods appropriate for use at required
PM2.5 sites in at-risk communities are FRMs and automated
continuous FEMs (88 FR 5675-76, January 27, 2023). These are the
methods eligible to compare to the PM2.5 NAAQS, which is the
primary objective for collecting this data. There are several other
monitoring objectives that would benefit from the use of automated
continuous FEMs. For example, having hourly data available from
automated continuous FEMs would allow sites to provide data in near-
real time to support forecasting and near real-time reporting of the
AQI. Automated continuous methods are also useful to support evaluation
of other methods such as low-cost sensors. When used in combination
with on-site wind speed and wind direction measurements, automated FEMs
can provide useful pollution roses, which help in identifying the
origin of emissions that affect a community. Additionally, when
collocated with continuous carbon methods such as an aethalometer,
automated FEMs can help identify potential local carbon sources
contributing to increased exposure in the community. While either FRMs
or automated FEMs may be used at a site for comparison to the
PM2.5 NAAQS, the EPA supports use of automated continuous
FEMs at sites in at-risk communities.
The EPA requested comment on the measurement methods most useful to
collocate with PM2.5 in at-risk communities (88 FR 5675-76,
January 27, 2023), and a few commenters provided input. One commenter
recommended that the EPA should employ supplemental technologies and
systems to increase coverage of the regulatory monitoring network and
obtain more complete data to further protect public health and address
environmental injustice in air pollution exposure. Another commenter
recommended that the EPA invest in community-led monitoring and mobile
air quality monitoring with a goal of recording block-level
variabilities in data. And another commenter cited the value of
community-deployed PM2.5 monitoring.
The EPA appreciates the comments provided on the measurement
methods most useful to collocate with PM2.5 monitoring sites
in at-risk communities. Because the use of methods beyond the required
PM2.5 FRMs or FEMs or other criteria pollutant measurements
meeting a NAAQS monitoring requirement is voluntary, the establishment
of PM2.5 NAAQS comparable sites in at-risk communities will
allow for collaboration at multiple levels. The EPA strongly encourages
such collaboration with impacted communities, and the measurement
methods discussed here should be considered for use as appropriate.
In the proposal, the EPA identified that, to meet the revised
network design criteria, there will be only a few new sites
required,\198\ plus any potentially moved sites in cases where an
existing site lease is lost or otherwise requires relocation (88 FR
5675-76, January 27, 2023). To handle these new or relocated sites, the
EPA proposed to build upon our existing regulatory process for
selecting and approving these sites under 40 CFR 58.10 (88 FR 5676,
January 27, 2023). In the proposal, we stated it would be appropriate
to provide at least 12 months from the effective date of the final rule
to allow monitoring agencies to initiate planning to implement these
measures by seeking input from communities and other interested parties
and considering whether to revise their PM2.5 networks
[[Page 16361]]
or explain how their existing networks meet the objectives of the
proposed modification to the network design criteria. Thus, the EPA
proposed that monitoring agencies should address their approach to the
question of whether any new or moved sites are needed and identify the
potential communities in which the agencies are considering adding
monitoring, if applicable, as well as identifying how they intend to
meet the revised criteria for PM2.5 network design to
address at-risk communities in the agencies' annual monitoring network
plans due to each applicable EPA Regional office no later than July 1,
2024 (see 40 CFR 58.10). Specifics on the resulting new or moved sites
for PM2.5 network design to address at-risk communities were
proposed to be detailed in the annual monitoring network plans due to
each applicable EPA Regional office no later than July 1, 2025 (40 CFR
58.10). The EPA proposed that any new or moved sites would be required
to be implemented and fully operational no later than 24 months from
the date of approval of a plan or January 1, 2027, whichever comes
first, but the EPA solicited comment on whether less time is needed
(e.g., 12 months from plan approval and/or January 1, 2026).
---------------------------------------------------------------------------
\198\ Gantt, B. (2022). Analyses of Minimally Required
PM2.5 Sites Under Alternative NAAQS. Memorandum to the
Rulemaking Docket for the Review of the National Ambient Air Quality
Standards for Particulate Matter (EPA-HQ-OAR-2015-0072). Available
at: https://www.regulations.gov/docket/EPA-HQ-OAR-2015-0072.
---------------------------------------------------------------------------
The EPA received a few comments on its proposed timeline for
monitoring agencies to identify, propose, and ultimately bring any new
or moved sites online. One commenter asked that the timeline give
states more time to start or move sites. A few commenters asked that
the EPA only require meeting a timeline for identifying whether any new
or moved sites are needed after the EPA has provided the monitoring
agencies with guidance on the priority of the potential at-risk
communities. One of those commenters further requests that the EPA
allow at least 24 months from the date of approval of a Sec. 58.10
monitoring plan identifying any relocation of monitoring sites or
establishment of new monitoring sites to implement any changes to the
network, citing the need for more time to work with local officials,
procure monitoring equipment, and contract for services, all of which
can cause significant delays in establishing a monitoring site. Another
commenter asked that the EPA remain attentive to the challenges that
States, and air agencies face regarding recruiting and retaining the
specialized staff needed to support their existing regulatory
monitoring networks and the capital resources needed to implement and
sustain new monitoring stations in areas that are clearly meeting the
existing PM NAAQS or any revised PM NAAQS. Another commenter stated
that the July 1, 2024, timeline for a network evaluation this complex
is insufficient, noting that they submit their draft annual monitoring
network plan for public review and comment in mid-April for 30 days.
Because the final plan is due July 1 and must include all comments and
responses and describe any changes based on those comments, the
timeline does not take these requirements into consideration by
allowing for the more extensive assessment of changes that may be
needed to meet the proposed new monitoring requirements. The commenter
stated that it would be appropriate to provide at least 12 months from
the effective date of this final rule for monitoring agencies to
initiate planning to implement these measures, seek input, consider
revisions to their PM2.5 networks, and explain how their
existing networks meets the objectives of the final rule. The commenter
notes that that SLT agencies should be provided a minimum of 18 months
after the final recommendation is published to add this information to
their Sec. 58.10 annual monitoring network plans. Another commenter
encourages the EPA to retain the proposed deadline for any newly
required monitoring stations in at-risk communities to be operational
(i.e., 24 months after the July 2025 network plan approval or January
1, 2027, whichever is earlier). While the need for this data is urgent,
the commenter stated that the process for procuring instrumentation,
securing leases, and building permits, and other logistics in
constructing new monitoring sites can take a significant amount of
time, some of which are outside of agencies' control.
As stated earlier, the EPA received strong support for our proposal
to modify the PM2.5 monitoring network design criteria to
include an EJ factor that accounts for proximity of populations at
increased risk of adverse health effects from PM2.5
exposures to sources of concern from a wide range of commenters. A few
commenters support the timeline proposed, a few others support starting
any new or moved sites sooner than proposed, while other commenters
asked for more time or offered conditions regarding how to establish an
appropriate timeline.
The EPA disagrees with the commenter that suggested the EPA should
only require agencies to meet a timeline to identify whether any new or
moved sites are needed after the EPA has provided the monitoring
agencies with guidance on the priority of the potential at-risk
communities, because the regulatory text provides all the guidance
required for agencies to begin this process. As we explained above, the
EPA does not anticipate that many new or moved sites will be required
based on the final rule because we think most sites are already in
suitable locations and long-term sites are highly valued. Also,
monitoring agencies have discretion to provide to the EPA their
recommendations regarding how they intend to meet the modifications to
the PM2.5 monitoring network design criteria to include an
EJ factor that accounts for proximity of populations at increased risk
of adverse health effects from PM2.5 exposures to sources of
concern. Overall, the EPA believes that having sites in the areas of
expected maximum concentrations will best ensure that all communities
are protected. Since there may be multiple choices for sites in EJ
areas near sources of concern, the EPA acknowledges that there may be
many locations that can meet the revised PM2.5 network
design criteria. While, as we explained earlier, we want such sites to
also be in areas of poor air quality, the sites in the area of maximum
concentration will ensure that all communities are protected, there can
be more flexibility afforded in the selection amongst at-risk
communities to meet the revised requirements, since any alternative at-
risk communities would already be protected.
The EPA considered both the concerns and support for the timeline
proposed and clarifies that the component of the proposed requirement
regarding the need to identify potential new sites or an intention to
move sites to be included in the annual monitoring network plan due to
EPA on July 1, 2024, would be satisfied with a statement of intent to
pursue a new site per the revised network design criteria and in
consideration of the minimum monitoring requirements. While monitoring
agencies may provide as much detail as they deem appropriate regarding
the revised PM2.5 network design criteria in their annual
monitoring network plans due on July 1, 2024, there is no expectation
that any details on site-specific information would be included at that
stage. We encourage agencies to provide their initial thinking on the
communities they are most interested in monitoring pursuant to the
revised network design criteria. Therefore, the EPA is finalizing the
timeline as proposed, including the provision that monitoring agencies
report their intention to add or move sites, where required, in their
annual monitoring network plans due to each applicable EPA Regional
office no later than July 1, 2024 (40 CFR 58.10). The
[[Page 16362]]
monitoring agencies will then provide specifics on any new or moved
sites for PM2.5 network design to address at-risk
communities in the annual monitoring network plans due to each
applicable EPA Regional office no later than July 1, 2025 (40 CFR
58.10). And any new or moved sites shall be implemented and fully
operational no later than 24 months from the date of approval of a
Sec. 58.10 plan, or January 1, 2027, whichever comes first.
In summary, the EPA is finalizing modifications to the
PM2.5 network design criteria to include an EJ factor to
address at-risk communities with a focus on exposures from sources of
concern in areas of poor air quality. While this modification to the
PM2.5 network design requires sites to be located in at-risk
communities, particularly those whose air quality is potentially
affected by local sources of concern, such sites must still meet the
requirement for being considered ``area-wide'' air quality. In
finalizing this modification to the PM2.5 network design
requirement, the EPA is making two changes in the final rule response
to the comments received. First, the EPA is broadening our examples of
``sources of concern'' to include localized sources such as point
sources and major transportation facilities or corridors. Second, the
EPA is reinstating ``poor air quality'' in our requirement for the
modified network design criteria, meaning the revised PM2.5
network design requirement now states: ``For areas with additional
required SLAMS, a monitoring station is to be sited in an at-risk
community with poor air quality, particularly where there are
anticipated effects from sources in the area (e.g., a major industrial
area, point source(s), port, rail yard, airport, or other
transportation facility or corridor).'' All other aspects of the
PM2.5 network design requirements are being finalized as
proposed.
5. Revisions to Probe and Monitoring Path Siting Criteria
The EPA proposed changes to monitoring requirements in the Appendix
E--Probe and Monitoring Path Siting Criteria for Ambient Air Quality
Monitoring (88 FR 5676-78, January 27, 2023). Since 2006, the EPA
finalized multiple rule revisions to establish siting requirements for
PM10-2.5 and O3 monitoring sites (71 FR 2748,
January 17, 2006), Near-Road NO2 monitoring sites (75 FR
6535, February 9, 2010), Near-Road CO monitoring sites (76 FR 54342,
August 31, 2011), and Near-Road PM2.5 monitoring sites (78
FR 3285, January 15, 2013). Through these previous revisions to the
regulatory text, some requirements were inadvertently omitted, and,
over time, the clarity of this appendix was reduced through those
omissions that, in a few instances, led to unintended and conflicting
regulatory requirements. The EPA proposed to reinstate portions of
previous Probe and Monitoring Path Siting Criteria Requirements from
previous rulemakings, where appropriate, to restore the original
intent.
The EPA only received a few comments on the proposed rulemaking
pertaining to the proposed changes regarding probe and monitoring path
siting criteria for ambient air quality monitoring, most of which were
supportive of the proposed revisions. One commenter noted that the
image for Figure E-1 in Appendix E to part 58 was distorted and of
extremely poor quality, rendering the text in places almost unreadable
(88 FR 5712, January 27, 2023). The EPA makes several references to
Figure E-1, which provides detailed information needed for assessing a
range of acceptable probe distances from roadways based on a monitor's
spatial scale. The commenter also stated that a higher quality image is
needed for the figure so that agencies can fully interpret the figure
to the extent that EPA requires. The EPA agrees with the commenter that
a higher quality image for Figure E-1 is important and needed. Based on
this comment, the EPA is finalizing the revision to Figure E-1 to
clearly communicate the requirements of appendix E.
The EPA is revising appendix E in its entirety as proposed (88 FR
5709-5717, January 27, 2023) for clarity and as described in detail
below.
a. Separate Section for Open Path Monitoring Requirements
The EPA proposed to relocate all open path monitor siting criteria
requirements to a separate section in appendix E from those
requirements for siting samplers and monitors that utilize probe inlets
(88 FR 5676, January 27, 2023). Separate sections for these distinct
monitoring method types allows the EPA to more clearly articulate
minimum technical siting requirements for each.
The EPA received one supportive comment to adopt this change and
received no adverse comments. Another commenter stated the regulatory
text of the proposal improves the clarity of the appendix but
encouraged the EPA to break the summary tables down further into more
manageable components (perhaps by pollutant). The commenter stated that
summary tables for the proposed appendix continue to be a ``jumbled
mess of regulatory requirements.'' The EPA agrees that the summary
tables E-3 and E-6 in the proposal could be improved further. Also, the
EPA found that footnote 3 of Table E-6 in the proposed rule was
incomplete and corrected this editorial error.
Therefore, the EPA is making editorial changes to both summary
tables E-3 and E-6 and finalizing the remainder of the language as
proposed with the open path monitor siting criteria requirements placed
into a separate section of the appendix.
b. Distance Precision for Spacing Offsets
The EPA proposed to require that when rounding is performed to
assess compliance with these siting requirements, the distance
measurements will be rounded such as to retain at least two significant
figures (88 FR 5676, January 27, 2023). The EPA proposed to communicate
this rounding requirement in the regulatory text using footnotes in the
tables of this appendix.
The EPA received two supportive comments and no adverse comments
regarding this proposed change. While supportive of the proposal, one
of the two supporting comments suggested it would be clearer if EPA
explicitly defined a decimal in the distance values and round to the
nearest tenths place for these assessments. The EPA disagrees with this
recommendation because in some cases it would be more restrictive and
burdensome than the proposed requirement that was intended to provide
both clarity and flexibility. Therefore, the EPA is finalizing the
language as proposed.
c. Summary Table of Probe Siting Criteria
The EPA proposed to provide additional specificity and flexibility
to the summary table for probe siting criteria by changing the ``>''
(greater than) symbols to ``>='' (greater than or equal to) symbols in
the summary table E-4 (88 FR 5676, January 27, 2023). Because one
commenter pointed out to the EPA that in the prior version of the rule
there was no table E-4, as a clerical matter, we have renumbered this
summary table to table E-3 in the final rule. This proposed minor
revision to the summary table more clearly expresses the EPA's intent
that the distance offsets provided in the summary tables in appendix E
are acceptable for NAAQS compliance monitoring.
The EPA received one comment supporting the proposal. The EPA
received no adverse comments. Because
[[Page 16363]]
one commenter pointed out to the EPA that in the prior version of the
rule there was no table E-4, as a clerical matter, we have renumbered
this summary table to table E-3 in the final rule. Therefore, the EPA
is updating the table numbering and otherwise finalizing the tables as
proposed.
d. Spacing From Minor Sources
The EPA proposed to clarify and provide flexibility regarding
siting monitors near minor sources by changing a requirement to a goal
(88 FR 5676-77, January 27, 2023). To accomplish this, the EPA proposed
to replace the ``must'' in the regulation with a ``should.'' While the
EPA proposed to change this requirement to a goal, the EPA reiterated
in the proposal that it recommends that sites with minor sources be
avoided whenever practicable and probe inlets should be spaced as far
from minor sources as possible when alternative monitoring stations are
not suitable.
The EPA received one comment supporting the proposed revision and
received no adverse comments. Therefore, the EPA is finalizing the
language as proposed.
e. Spacing From Obstructions and Trees
The EPA proposed to clarify and redefine that the minimum arc
required to be free of obstructions for a probe inlet or monitoring
path is 270-degrees and that probe inlets must be no closer than 10-
meters to the driplines of any trees (88 FR 5677, January 27, 2023).
These changes were proposed because of inconsistencies introduced into
the rule with the 2006 rulemaking. Both are discussed in more detail in
the following sections.
The majority of comments received were supportive of these proposed
siting amendments and clarifications. Two commenters were not
supportive of this proposal. One adverse comment focused on the
potential that site modifications would be required if the minimum arc
required to be free of obstructions for a probe inlet is 270-degrees.
The second adverse comment pertained to the proposal to clarify
distance requirements from tree driplines. The commenter stated they
would expect significant challenges in meeting the proposed 20-meter
tree dripline distance. This comment is not a substantive negative
comment because the 20-meter distance provided in the proposal is a
goal and not a requirement. As such, monitoring organizations should
not expect additional challenges in meeting the probe siting
requirements. One supportive commenter on the 270-degree minimum arc
proposal also requested that the EPA acknowledge that some cases exist
where monitoring is desired or necessary to protect the public health,
but siting criteria cannot be met.
Based on the only two negative comments received from monitoring
agencies or organizations, one of which was not substantive, the EPA
believes most sites already meet these proposed requirements related to
the arc and distance from dripline. However, the EPA also acknowledges
that there may be limited cases where this proposed revision may
require site modifications, and some sites may not be able to be
achieve the proposed siting requirements, even with modifications to
the site. For cases where long-term trend sites or monitors that
determine the design value for their area cannot reasonably meet these
regulatory siting requirements, the EPA encourages monitoring
organizations to work with their respective EPA Regional offices to
determine if a waiver from this siting criteria would be appropriate
under appendix E, section 10.
These siting requirements are discussed in more detail below in
sections VII.B.5.f and VII.B.5.h.
f. Reinstating Minimum 270-Degree Arc and Clarified 180-Degree Arc
The EPA proposed to correct identified inconsistencies in the 270-
degree requirement for unrestricted airflow to the probe inlet by
reinstating the requirement stated in appendix E, paragraph 4(b), and
to clarify that the continuous 180-degree minimum arc of unrestricted
airflow provision is reserved for monitors sited on the side of a
building or wall to comply with network design criteria requirements
specified in appendix D of part 58 (88 FR 5677, January 27, 2023).
The EPA received two comments regarding this proposal, with one
being supportive and one being negative. The adverse comment focused on
the potential that site modifications would be required if this
revision was made. The commenter supporting the proposal also requested
that the EPA acknowledge that some cases exist where monitoring is
desired or necessary to protect the public health, but siting criteria
cannot be met. The EPA agrees with both commenters and acknowledges
that there does exist limited cases where this proposal would require
site modifications and some sites may not be able to be achieve the
proposed siting requirement even with modifications to the site. For
these cases, and especially when long-term trend sites or monitors that
determine the design value for their area cannot reasonably meet these
regulatory siting requirements, the EPA encourages monitoring
organizations to work with their respective EPA Regional Offices to
determine if a waiver from this siting criteria is appropriate through
the provisions found in Section 10 of this appendix.
Based on the EPA only receiving a single negative comment regard
the 270-degree and 180-degree provisions the EPA thinks most sites
already meet these proposed requirements. Additionally, as stated
above, the EPA is also retaining waiver provisions from these siting
requirements for the remaining cases that can be exercised when
appropriate. Therefore, the EPA is finalizing the language as proposed.
g. Obstacles That Act as Obstructions
The EPA proposed to clarify the definitions of ``obstructions'' and
``obstacles'' in the regulatory text (88 FR 5677, January 27, 2023).
Stating that, ``[o]bstructions to the air flow of the probe inlet are
those obstacles that are horizontally closer than twice the vertical
distance the obstacle protrudes above the probe inlet and can be
reasonably thought to scavenge reactive gases or to restrict the
airflow for any pollutant,'' the EPA proposed to reiterate that the EPA
does not generally consider objects or obstacles such as flag poles or
site towers used for NOy convertors and meteorological
sensors, etc., to be deemed obstructions.
The EPA received one comment supporting the proposal and received
no adverse comments. Therefore, the EPA is finalizing the definitions
as proposed.
h. 10-Meter Tree Dripline Requirement
The EPA proposed to reconcile the conflicting requirements in 5(a)
and the prior table E-4 footnote 3 by clarifying that the probe inlet
must always be no closer than 10 meters to the tree dripline (88 FR
5677, January 27, 2023). The EPA also proposed to reinstate the goal
``that monitor probe inlets should be at least 20-meters from the
driplines of trees,'' a goal that was inadvertently omitted during
previous rule revisions. In addition, the EPA proposed to clarify that
if a tree or group of trees is considered an ``obstruction,'' section
4(a) will apply.
As described above, the majority of comments received were
supportive of the EPA proposed amendments and clarification, with two
commenters focused on the possibility that monitoring agencies may not
be able to meet the revised siting requirements. Specific to the
proposed dripline requirement, the EPA reiterates that the
[[Page 16364]]
20-meter tree dripline offset is not a requirement, but rather a goal.
Monitoring programs should as much as practicable attempt to meet this
20-meter tree dripline offset goal but are only required to be at least
10 meters removed from tree driplines. If these requirements cannot be
met, the EPA encourages monitoring organizations to contact their
respective EPA Regional offices to determine if a waiver from this
siting criteria would be appropriate under appendix E, section 10.
Another commenter recommended that the proposal should also include
an elevation specification. For instance, if a monitor is on the roof
of a shelter, a tree below that roof should not be considered an
obstruction no matter the distance to the dripline. The EPA considers
this scenario to occur in practice only rarely. The EPA agrees that
when the overall tree height is less than the height of the probe
inlet, the tree is not obstructing the airflow to the probe inlet.
However, a tree in such proximity to the probe inlet in many cases is
not likely to remain at a height lower than the probe inlet. The EPA
considers a scenario such as this to be best addressed in the waiver
provisions of this appendix due both to the rarity of this occurring as
well as the need for the EPA to periodically reassess whether tree
growth has adversely impacted the site conditions.
For these reasons, the EPA is finalizing the language as proposed.
i. Spacing Requirement for Microscale Monitoring
The EPA proposed to require that microscale sites for any pollutant
shall have no trees or shrubs blocking the line-of-sight fetch between
the monitor's probe inlet and the source under investigation (88 FR
5677, January 27, 2023). This proposed revision would bring consistency
between near-road monitoring stations and other microscale monitoring.
The EPA received one comment on this proposed requirement
expressing concerns regarding its practicality and legality. The
commenter stated agencies may at times want to site a monitor close to
a source, but the closest location will have trees in the line of sight
on private property. Additionally, in some cases, the trees may have
been planted for the purpose of reducing off-property emissions from a
source such as a Concentrated Animal Feeding Operation (CAFO). The
commenter further stated that the proposal mandates that State agencies
order the removal of trees from private property to collect valid data.
The EPA disagrees that the proposed requirement is impractical or
unlawful. The proposed requirement would not require, mandate, or
otherwise empower monitoring agencies to force the removal of trees on
private property. The EPA agrees with the commenter that trees may at
times be planted as part of control strategies to reduce offsite
emissions and thus protect the public, but the EPA disagrees with the
commenter that the trees must be removed to perform ambient air
monitoring in these locations. Rather, if trees or shrubs block the
line-of-sight fetch between the monitor's probe inlet and the source
under investigation, it is the EPA's position that, for most cases, a
microscale designation does not accurately reflect the monitoring scale
for this location, and instead the EPA would recommend that the
monitoring scale be designated to a more representative monitoring
scale such as middle scale or neighborhood scale.
Moreover, for cases where long-term trend sites or monitors that
determine the design value for an area cannot reasonably meet this
regulatory siting requirement, the EPA encourages monitoring
organizations to work with their respective EPA Regional offices to
determine if a waiver from this siting criteria may be appropriate
under appendix E, section 10.
For these reasons, the EPA is finalizing the language as proposed.
j. Waiver Provisions
The EPA proposed to maintain the appendix E, section 10 waiver
provisions in the current regulation for siting criteria, but to modify
section 10.3 to require that waivers from the probe-siting criteria
must be reevaluated and renewed minimally every 5 years (88 FR 5677-78,
January 27, 2023).
The EPA received one comment supporting the proposal and no adverse
comments. Therefore, the EPA is finalizing the language as proposed.
k. Acceptable Probe Materials
The EPA proposed to expand the list of acceptable probe materials
for sampling reactive gases in appendix E, section 9, from just
borosilicate glass and fluorinated ethylene propylene (FEP)
Teflon[supreg], or their equivalents. The EPA proposed to add
polyvinylidene fluoride (PVDF), also known as Kynar[supreg],
polytetrafluoroethylene (PTFE), and perfluoroalkoxy (PFA) to the list
of approved materials for efficiently transporting gaseous criteria
pollutants, and the use of Nafion\TM\ upstream of ozone analyzers (88
FR 5678, January 27, 2023). Mention of trade names or commercial
products does not constitute endorsement.
The EPA received two comments supporting the proposal and received
no adverse comments. Therefore, the EPA is finalizing the language as
proposed.
D. Incorporating Data From Next Generation Technologies
In the proposal, the EPA requested comment on how to incorporate
data from next generation technologies into Agency efforts (88 FR 5678-
80, January 27, 2023). The near real-time integration of data from
PM2.5 continuous monitors, sensors, and satellites has
allowed the EPA to use data in certain informational applications such
as EPA and USFS's Fire and Smoke Map.\199\ This mapping product uses
Application Program Interfaces (APIs) where data sets are automatically
shared on prespecified computer servers. Given the success of the Fire
and Smoke Map, the EPA indicated interest in exploring the use of next-
generation technologies to develop additional approaches, products, and
applications to help address important non-regulatory air quality data
needs. Therefore, the EPA solicited comment on the most important data
uses and data sets to consider in such future initiatives. Such
approaches and/or products could utilize historical or near real-time
data. The EPA sought this input and prioritization on use of next
generation technologies to help improve the utility of data to better
support air quality management to improve public health and the
environment.
---------------------------------------------------------------------------
\199\ Available at https://fire.airnow.gov/.
---------------------------------------------------------------------------
The EPA received comments from about two dozen entities on its
request for comments on how to incorporate data from next generation
technologies. The entities that provided comment included federal
agencies; representatives of industry and industry groups; public
health, medical, and environmental organizations; State, local and
related multi-state organizations involved in air program management;
Tribes and Tribal organizations involved in air program management; and
other State and local governments.
While there were some differences across commenters, a majority of
the commenters support use of next generation data for non-regulatory
purposes, but not for regulatory decision making due to their inherent
uncertainties and limitations. The EPA also received comments from some
environmental organizations support using alternative data for
regulatory decision making.
[[Page 16365]]
Many commenters pointed out that they are already successfully
using sensor data and networks in supplemental and informational
applications and support further expansion of these capabilities.
Across many commenters, there was support for using next generation
data as ``fit for purpose,'' filling in gaps, finding hot spots,
identifying and addressing EJ concerns, and evaluating and informing
network siting. The EPA acknowledges the successful examples of sensor
data and networks for non-regulatory purposes. A few commenters support
expanding the use of sensor data to provide real-time AQI; the EPA is
interested in this use of next generation data as well. A few
commenters pointed the need for the EPA to work closely with them and
their communities to understand and use next generation data, while
others expressed a desire for help developing best practices around
collecting and using next generation data, developing products with
data analysis/visualization, and developing appropriate QA/QC for
sensor data. The EPA acknowledges each of these requests and expects to
continue to work closely with SLTs and other stakeholders to understand
and develop information on the collection and use of next generation
data.
A few commenters offered more detailed comments. Some recommended
that the EPA repropose implementation provisions related to next
generation technologies with greater clarity to provide for meaningful
comment. For example, the use of low-cost sensor and satellite data
could be used in drawing nonattainment area boundaries or identifying
sources for emissions control, but doing so would be such a significant
change from prior EPA policy that it warrants a more specific proposal,
beyond the scope of this request for comment. In response to this
comment, the EPA notes it did not propose or change the use of non-
regulatory measurement data as part of this proposal, but instead
opened an opportunity to comment about the use of next generation
technologies.
Another commenter stated that while low-cost sensor data can be
invaluable for some purposes, the potentially overwhelming amount of
data produced by sensors may present additional challenges to
communities without the resources or expertise to analyze it. Cost is
another concern associated with some next generation technologies of
which some communities may not be aware, as the initial cost of the
sensor alone is not indicative of the total cost of operation, which
can include costs of internet access and servers. The EPA appreciates
the need to consider all the costs of implementing and maintaining
sensor data.
Another commenter stated that having a dense sensor network
collocated with FRMs and FEMs could help ensure timely maintenance of
the regulatory measurements in the event there appears to be a
divergence of data. The EPA appreciates the comment that emphasizes how
sensors could be used to complement the FRM and FEM data with regard to
ensuring timely maintenance.
Another commentor strongly opposes incorporating sensor data into
any EPA systems unless robust quality assurance (QA) practices are
widely established and managed by qualified personnel. The EPA agrees
that QA is necessary, and notes that the ``fit for purpose'' aspect of
using sensor data will inform the appropriate QA associated with the
intended use of such data.
In summary, the EPA invited comment on how we should consider
incorporating data from next generation technologies into our air
monitoring efforts. In seeking comment on this topic, the EPA did not
propose to add, edit, or delete any regulatory language associated with
the PM NAAQS. The EPA received comments from a variety of entities that
largely support using next generation data for a variety of purposes
that supplement, but cannot replace, the measurement data from
monitoring methods required (i.e., FRMs and FEMs) for regulatory
decision making. Across many commenters, there was support for using
next generation technologies and data as ``fit for purpose,'' filling
in gaps, finding hot spots, identifying, and addressing EJ concerns,
and evaluating and informing network siting. Quality assurance of the
data will be an important component in the use of next generation
technology data. The EPA will consider these comments as it continues
its work with the co-regulated community comprised of SLT agencies and
other stakeholders to understand and use next generation data and joint
efforts to manage the nation's ambient air.
VIII. Clean Air Act Implementation Requirements for the Revised Primary
Annual PM2.5 NAAQS
The EPA's revision to the primary annual PM2.5 NAAQS
discussed in section II above triggers a number of implementation
related activities that were described in the NPRM. The two most
immediate implementation impacts following a final new or revised NAAQS
are related to stationary source permitting and the initial area
designations process. Permitting implications are discussed below in
section VIII.E. With regard to initial area designations, the EPA is
separately issuing a memorandum regarding the Initial Area Designations
for the Revised Primary Annual Fine Particle National Ambient Air
Quality Standard Memorandum (the ``Annual PM2.5 NAAQS
Designations Memorandum'') that will provide information about the
statutory schedule for the designations process. For other
implementation related implications, please refer back to the NPRM
section VIII.
The NPRM also referred to the PM2.5 State Implementation
Plan (SIP) Requirements Rule (81 FR 58010, August 24, 2016), which
specifies planning requirements for areas designated as nonattainment
for purposes of the PM2.5 NAAQS and includes a number of key
recommendations for areas to consider implications of environmental
justice through the attainment planning process, consistent with the
identification of at-risk groups in the 2019 ISA and ISA Supplement and
the statutory requirement to protect the health of at-risk groups. As
stated in the NPRM, State and local air agencies are encouraged to
consider how they might develop implementation plans that encourage
early emission reductions.
A. Designation of Areas
As discussed in section II, with respect to the PM2.5
NAAQS, the EPA is finalizing: (1) Revisions to the level of the primary
annual PM2.5 NAAQS and retaining the current primary 24-hour
PM2.5 NAAQS (section II.B.4); and (2) no change to the
current secondary annual and 24-hour PM2.5 NAAQS at this
time (section V.B.4). Upon promulgation of a new or revised NAAQS,
States and the EPA must initiate the process for initial designations.
The timeline for initial area designations begins with promulgation
of the revised primary annual PM2.5 NAAQS, as stated in the
CAA section 107(d)(1)(B)(i). Through this process, which provides for
input from States and others at various stages, the EPA identifies
areas of the country that either meet or do not meet the revised
primary annual PM2.5 NAAQS, along with the nearby areas
contributing to NAAQS violations. The following includes additional
information regarding the designations process described in the CAA.
Section 107(d)(1) of the CAA states that, ``By such date as the
Administrator may reasonably require, but not later than 1 year after
promulgation of a new or revised national ambient air quality standard
for any pollutant under section
[[Page 16366]]
109, the Governor of each State shall . . . submit to the Administrator
a list of all areas (or portions thereof) in the State'' and make
recommendations for whether the EPA should designate those areas as
nonattainment, attainment, or unclassifiable.\200\ The CAA provides the
EPA with discretion to require States to submit their designations
recommendations within a reasonable amount of time not exceeding one
additional year.\201\ Section 107(d)(1)(A) of the CAA also states that
``the Administrator may not require the Governor to submit the required
list sooner than 120 days after promulgating a new or revised national
ambient air quality standard.'' Section 107(d)(1)(B)(i) further
provides, ``Upon promulgation or revision of a NAAQS, the Administrator
shall promulgate the designations of all areas (or portions thereof) .
. . as expeditiously as practicable, but in no case later than 2 years
from the date of promulgation. Such period may be extended for up to
one year in the event the Administrator has insufficient information to
promulgate the designations.'' With respect to the NAAQS setting
process, courts have interpreted the term ``promulgation'' to be
signature and widespread dissemination of a final rule.\202\
---------------------------------------------------------------------------
\200\ While the CAA says ``designating'' with respect to the
Governor's letter, in the full context of the CAA section it is
clear that the Governor actually makes a recommendation to which the
EPA must respond via a specified process if the EPA does not accept
it.
\201\ In certain circumstances in which the Administrator has
insufficient information to promulgate area designations within two
years from the promulgation of the NAAQS, CAA section
107(d)(1)(B)(i) provides that the EPA may extend the designations
schedule by up to one year.
\202\ API v. Costle, 609 F.2d 20 (D.C. Cir. 1979).
---------------------------------------------------------------------------
If the EPA agrees with the designations recommendation of the
State, then it may proceed to promulgate the designations for such
areas. If, however, the EPA disagrees with the State's recommendation,
then the EPA may elect to make modifications to the recommended
designations. By no later than 120 days prior to promulgating the final
designations, the EPA is required to notify States of any intended
modifications to the State designation recommendations for any areas or
portions thereof, including the boundaries of areas, as the EPA may
deem necessary. States then have an opportunity to comment on the EPA's
intended modification and tentative designation decision. If a State
elects not to provide designation recommendations for any area, then
the EPA must itself promulgate the designation that it deems
appropriate.
While section 107(d) of the CAA specifically addresses the
designations process for States, the EPA intends to follow the same
process for Tribes to the extent practicable, pursuant to section
301(d) of the CAA regarding Tribal authority, and the Tribal Authority
Rule (63 FR 7254, February 12, 1998). To provide clarity and
consistency in doing so, the EPA issued a guidance memorandum to our
Regional Offices on working with Tribes during the designations
process.\203\
---------------------------------------------------------------------------
\203\ ``Guidance to Regions for Working with Tribes during the
National Ambient Air Quality Standards (NAAQS) Designations
Process,'' December 20, 2011, Memorandum from Stephen D. Page to
Regional Air Directors, Regions 1-X available at https://www.epa.gov/sites/default/files/2017-02/documents/12-20-11_guidance_to_regions_for_working_with_tribes_naaqs_designations.pdf
.
---------------------------------------------------------------------------
Consistent with the process used in previous area designations
efforts, the EPA will evaluate each area on a case-by-case basis
considering the specific facts and circumstances unique to the area to
support area boundary decisions for the revised standard. The EPA
intends to issue a designations memorandum which will provide
information regarding the designations process. In broad overview, the
EPA has historically used area-specific analyses to support
nonattainment area boundary recommendations and final boundary
determinations by evaluating factors such as air quality data,
emissions and emissions-related data (e.g., population density and
degree of urbanization, traffic and commuting patterns), meteorology,
geography/topography, and jurisdictional boundaries. We expect to
follow a similar process when establishing area designations for this
revised PM2.5 NAAQS. CAA section 107(d) explicitly requires
that the EPA designate as nonattainment not only the area that is
violating the pertinent standard, but also those nearby areas that
contribute to the violation in the violating area. In the
PM2.5 NAAQS Designations Memorandum, the EPA intends to
include information regarding consideration of federal land boundaries
that may be fully or partially included within the bounds of a county
otherwise identified as nonattainment.
As with past revisions of the PM2.5 NAAQS, the EPA
intends to make the designations decisions for the revised primary
annual PM2.5 NAAQS based on the most recent three years of
quality-assured, certified air quality data in the EPA's Air Quality
System (AQS). Accordingly, the EPA recommends that States base their
initial area designation recommendations on the most current available
three years of complete and certified air quality data at the time of
the recommendations. The EPA will then base the final designations on
the most recent three consecutive years of complete, certified air
quality monitoring data available at the time of final
designations.\204\
---------------------------------------------------------------------------
\204\ In certain circumstances in which the Administrator has
insufficient information to promulgate area designations within two
years from the promulgation of a new or revised NAAQS, CAA section
107(d)(1)(B)(i) provides the EPA may extend the designations
schedule by up to one year.
---------------------------------------------------------------------------
Monitoring data are currently available from numerous existing
PM2.5 Federal Equivalent Methods (FEM) and Federal Reference
Methods (FRM) sites to determine violations of the revised primary
annual PM2.5 NAAQS. As described in section VII.C.3.b, the
EPA took comment on how to deal with cases where an FEM is approved by
the EPA with an update and when it can be implemented in the field. The
EPA took comment on how to approach the data produced during this lag
and received input from over a dozen commenters. The commenters asked
that the EPA be flexible in allowing the use of updated method
correction factors intended to improve the data comparability between
the FRMs and FEMs. The EPA will address any data correction issues
between the FRMs and FEMs through a future Notice of Data Availability
(NOA).
Consistent with past practice and as noted in the NPRM, the EPA
intends to provide additional information concerning the designations
process, including information about the schedule and recommendations
for determining area boundaries in the forthcoming Annual
PM2.5 NAAQS Designations Memorandum. Other topics addressed
in this memorandum include the schedule for preparing and submitting
exceptional events initial notification and exceptional events
demonstrations relevant to the designations process, and information
related to wildfire and prescribed fire on wildlands as it pertains to
initial area designations, as well as addressing back-correction of PM
FEM data when a method has an approved factory calibration as part of a
method update. The Annual PM2.5 NAAQS Designations
Memorandum is intended to assist States and Tribes in formulating their
area recommendations.\205\
---------------------------------------------------------------------------
\205\ See: https://www.epa.gov/particle-pollution-designations.
---------------------------------------------------------------------------
As discussed in the proposal, the ``Treatment of Data Influenced by
Exceptional Events; Final Rule,'' (81 FR 68216, October 3, 2016) and
codified at 40 CFR 50.1, 40 CFR 50.14, and 40 CFR 51.930, contains
instructions and requirements for air agencies that may
[[Page 16367]]
flag air quality data for certain days in the Air Quality System due to
potential impacts from exceptional events (i.e., such as prescribed
fires on wildland, wildfires, or high wind dust storms). Accordingly,
for purposes of initial area designations for a new or revised NAAQS,
an air agency may submit to the EPA an exceptional events demonstration
with supporting information and analyses for each monitoring site and
day the air agency claims the EPA should exclude from design value
calculations for designations purposes.
The EPA has provided tools to assist air agencies in preparing
adequate exceptional events demonstrations.\206\ Further, the EPA will
continue to work with air agencies as they identify exceptional events
that may influence decisions related to the initial area designations
process, and to prepare and submit exceptional events demonstrations if
appropriate. Importantly, air quality monitoring data may be influenced
by emissions from prescribed fires on wildland and wildfires. The EPA's
Exceptional Events Rule provides for both of these types of events to
be considered as exceptional events, provided the affected air agencies
submit exceptional events demonstrations that meet the procedural and
technical requirements of the EPA's Exceptional Events Rule. To that
end, the EPA has issued guidance addressing development of exceptional
events demonstrations for both wildfire and prescribed fires on
wildland.\207\ In light of the growing frequency and severity of
wildfire events, and expected increases in the application of
prescribed fire as a means to achieve long-term reductions in high
severity wildfire risk and associated smoke impacts, the EPA seeks to
ensure that the Agency's exceptional events process provides an
efficient and clear pathway for excluding data that may be affected by
such events in a manner that is consistent with the Clean Air Act and
the public health objectives of the NAAQS. Accordingly, the EPA is
continuing to explore opportunities to develop additional tools that
could assist air agencies in preparing exceptional events
demonstrations for wildfires and prescribed fires on wildland. In
addition, EPA intends to continue engaging with the U.S. Department of
Agriculture, U.S. Department of the Interior, air agencies, and other
stakeholders on these issues. For more information regarding the
exceptional events demonstration submission deadlines for the area
designations process, please see Table 2 to 40 CFR 50.14(c)(2)(vi)--
``Schedule for Initial Notification and Demonstration Submission for
Data Influenced by Exceptional Events for Use in Initial Area
Designations.''
---------------------------------------------------------------------------
\206\ See the EPA's Exceptional Events homepage at https://www.epa.gov/air-quality-analysis/treatment-air-quality-data-influenced-exceptional-events-homepage-exceptional.
\207\ See EPA's ``Final Guidance on the Preparation of
Exceptional Events Demonstrations for Wildfire Events that May
Influence Ozone Concentrations and EPA's Exceptional Events
Guidance: Prescribed Fire on Wildland that May Influence Ozone and
Particulate Matter Concentrations,'' found on EPA's Exceptional
Events homepage at https://www.epa.gov/air-quality-analysis/treatment-air-quality-data-influenced-exceptional-events-homepage-exceptional.
---------------------------------------------------------------------------
B. Section 110(a)(1) and (2) Infrastructure SIP Requirements
As discussed in the NPRM, the CAA directs States to address basic
SIP requirements to implement, maintain, and enforce the NAAQS. Under
CAA sections 110(a)(1) and (2), states are required to have State
implementation plans that provide the necessary air quality management
infrastructure that provides for the implementation, maintenance, and
enforcement of the NAAQS. After the EPA promulgates a new or revised
NAAQS, States are required to make a new SIP submission to establish
that they meet the necessary structural requirements for such new or
revised NAAQS or make changes to do so. The EPA refers to this type of
SIP submission as an ``infrastructure SIP submission.'' Under CAA
section 110(a)(1), all States are required to make these infrastructure
SIP submissions within three years after the effective date of a new or
revised primary standard. While the CAA authorizes the EPA to set a
shorter time for States to make these SIP submissions, the EPA is
requiring submission of infrastructure SIPs within three years of the
effective date of this revised primary annual PM2.5 NAAQS.
The EPA has provided general guidance to States concerning its
interpretation of these requirements of CAA section 110(a)(1) and (2)
in the context of infrastructure SIP submissions for a new or revised
NAAQS.\208\ The EPA encourages States to use this guidance when
developing their infrastructure SIPs for this revised primary annual
PM2.5 NAAQS.
---------------------------------------------------------------------------
\208\ See ``Guidance on Infrastructure State Implementation Plan
(SIP) Elements under Clean Air Act Sections 110(a)(1) and
110(a)(2)'' September 2013, Memorandum from Stephen D. Page to
Regional Air Directors, Regions 1-10.
---------------------------------------------------------------------------
As a reminder, the EPA notes that States are not required to
address nonattainment plan requirements for purposes of the revised
primary annual PM2.5 NAAQS on the same schedule as
infrastructure SIP requirements. The EPA interprets the CAA such that
two elements identified in section 110(a)(2) are not subject to the 3-
year submission deadline of section 110(a)(1) and thus States are not
required to address them in the context of an infrastructure SIP
submission. The elements pertain to part D, in title I of the CAA,
which addresses additional SIP requirements for nonattainment areas.
Therefore, for the reasons explained below, the following section
110(a)(2) elements are considered by the EPA to be outside the scope of
infrastructure SIP actions: (1) The portion of section 110(a)(2)(C),
programs for enforcement of control measures and for construction or
modification of stationary sources that applies to permit programs
applicable in designated nonattainment areas (known as ``nonattainment
new source review'') under part D; and (2) section 110(a)(2)(I), which
requires a SIP submission pursuant to part D, in its entirety.
Accordingly, the EPA does not expect States to address the
requirement for a new or revised NAAQS in the infrastructure SIP
submissions to include regulations or emissions limits developed
specifically for attaining the relevant standard in areas designated
nonattainment for the revised primary annual PM2.5 NAAQS.
States are required to submit infrastructure SIP submissions for the
revised primary annual PM2.5 NAAQS before they will be
required to submit nonattainment plan SIP submissions to demonstrate
attainment with the same NAAQS. States are required to submit
nonattainment plan SIP submissions to provide for attainment and
maintenance of a revised primary annual PM2.5 NAAQS within
18 months from the effective date of nonattainment area designations as
required under CAA section 189(a)(2)(B). The EPA reviews and acts upon
these later SIP submissions through a separate process. For this
reason, the EPA does not expect States to address new nonattainment
area emissions controls per section 110(a)(2)(I) in their
infrastructure SIP submissions.
One of the required infrastructure SIP elements is that each State
SIP must contain adequate provisions to prohibit, consistent with the
provisions of title I of the CAA, emissions from within the State that
will significantly contribute to nonattainment in, or interfere with
maintenance by, any other State of the primary or secondary NAAQS.\209\
This
[[Page 16368]]
element is often referred to as the ``good neighbor'' or ``interstate
transport'' provision.\210\ The provision has two prongs: significant
contribution to nonattainment (prong 1), and interference with
maintenance (prong 2). The EPA and States must give independent
significance to prong 1 and prong 2 when evaluating downwind air
quality problems under CAA section 110(a)(2)(D)(i)(I).\211\ Further,
case law has established that the EPA and States must implement
requirements to meet interstate transport obligations in alignment with
the applicable statutory attainment schedule of the downwind areas
impacted by upwind-state emissions.\212\ Thus, the EPA anticipates that
States will need to address interstate transport obligations associated
with this revised PM NAAQS, in alignment with the provisions of subpart
4 of part D of the CAA, as discussed in more detail in section VIII.C
below. Specifically, States must implement any measures required to
address interstate transport obligations as expeditiously as
practicable and no later than the next statutory attainment date, i.e.,
for this NAAQS revision as expeditiously as practicable, but no later
than the end of the sixth calendar year following nonattainment area
designations. See CAA section 188(c). States may find it efficient to
make SIP submissions to address the interstate transport provisions
separately from other infrastructure SIP elements.
---------------------------------------------------------------------------
\209\ CAA section 110(a)(2)(D)(i)(I).
\210\ CAA section 110(a)(2)(D)(i)(II) also addresses certain
interstate effects that states must address and thus is also
sometimes referred to as relating to ``interstate transport.''
\211\ See North Carolina v. EPA, 531 F.3d 896, 909-11 (D.C. Cir.
2008).
\212\ See id. 911-13. See also Wisconsin v. EPA, 938 F.3d 303,
313-20 (D.C. Cir. 2019); Maryland v. EPA, 958 F.3d 1185, 1203-04
(D.C. Cir. 2020).
---------------------------------------------------------------------------
Each State has the authority and responsibility to review its air
quality management program's existing SIP provisions in light of a new
or revised NAAQS to determine if any revisions are necessary to
implement the new or revised NAAQS. Most States have revised and
updated their SIPs in recent years to address requirements associated
with other revised NAAQS. For certain infrastructure elements, some
States may believe they already have adequate State regulations adopted
and approved into the SIP to address a particular requirement with
respect to the revised primary annual PM2.5 NAAQS.
If a State determines that existing SIP-approved provisions are
adequate in light of this revised primary annual PM2.5 NAAQS
with respect to a given infrastructure SIP element (or sub-element),
then the State may make an infrastructure SIP submission ``certifying''
that the existing State's existing EPA approved SIP already contains
provisions that address one or more specific section 110(a)(2)
infrastructure elements.\213\ In the case of such a submission, the
State does not have to include a copy of the relevant provision (e.g.,
rule or statute) itself. Rather, this certification submission should
provide citations to the SIP-approved State statutes, regulations, or
non-regulatory measures, as appropriate, in or referenced by the
already EPA-approved SIP that meet particular infrastructure SIP
element requirements. The State's infrastructure SIP submission should
also include an explanation as to how the State has determined that
those existing provisions meet the relevant requirements.
---------------------------------------------------------------------------
\213\ A ``certification'' approach would not be appropriate for
the interstate pollution control requirements of CAA section
110(a)(2)(D)(i).
---------------------------------------------------------------------------
Like any other SIP submission, that State can make such an
infrastructure SIP submission certifying that it has already met some
or all of the applicable requirements only after it has provided
reasonable notice and opportunity for public hearing. This ``reasonable
notice and opportunity for public hearing'' requirement for
infrastructure SIP submissions is to meet the requirements of CAA
sections 110(a) and 110(l). Under the EPA's regulations at 40 CFR part
51, if a public hearing is held, an infrastructure SIP submission must
include a certification by the State that the public hearing was held
in accordance with the EPA's procedural requirements for public
hearings. See 40 CFR part 51, appendix V, section 2.1(g), and see 40
CFR 51.102.
In consultation with the EPA's Regional office, a State should
follow all applicable EPA regulations governing infrastructure SIP
submissions in 40 CFR part 51--e.g., subpart I (Review of New Sources
and Modifications), subpart J (Ambient Air Quality Surveillance),
subpart K (Source Surveillance), subpart L (Legal Authority), subpart M
(Intergovernmental Consultation), subpart O (Miscellaneous Plan Content
Requirements), subpart P (Protection of Visibility), and subpart Q
(Reports). For the EPA's general criteria for infrastructure SIP
submissions, refer to 40 CFR part 51, appendix V, Criteria for
Determining the Completeness of Plan Submissions. For additional
information on infrastructure SIP submission requirements, refer to the
EPA's 2013 guidance entitled ``Guidance on Infrastructure State
Implementation Plan (SIP) Elements under Clean Air Act Sections
110(a)(1) and 110(a)(2).'' The EPA recommends that States
electronically submit their infrastructure SIPs to the EPA through the
State Plan Electronic Collaboration System (SPeCS),\214\ an online
system available through the EPA's Central Data Exchange.
---------------------------------------------------------------------------
\214\ https://cdx.epa.gov/.
---------------------------------------------------------------------------
C. Implementing Revised Primary Annual PM2.5 NAAQS in Nonattainment
Areas
As discussed in the NPRM, the EPA issued a SIP Requirements Rule
for implementing the PM2.5 NAAQS (81 FR 58010, August 24,
2016) (PM2.5 SIP Requirements Rule). It provides guidance
and establishes additional regulatory requirements for States regarding
development of attainment plans for nonattainment areas for the 1997,
2006, and 2012 revisions of the PM2.5 NAAQS. The guidance
and regulations in the SIP Requirements Rule also apply to any States
for which the EPA promulgates nonattainment area designations for the
new revised primary annual PM2.5 NAAQS.
The PM2.5 SIP Requirements Rule provides comprehensive
information regarding nonattainment plan requirements including, among
other things: nonattainment area emissions inventories; policies
regarding PM2.5 precursor pollutants (i.e., SO2,
NOX, VOC, and ammonia); control strategies (such as
reasonably available control measures and reasonably available control
technology for direct PM2.5 and relevant precursors); air
quality modeling; attainment demonstrations; reasonable further
progress requirements; quantitative milestones; and contingency
measures. Information provided in the PM2.5 SIP Requirements
Rule is supplemented by other EPA documents, including guidance on
emissions inventory development (80 FR 8787, February 19, 2015; U.S.
EPA, 2017), optional PM2.5 precursor demonstrations (U.S.
EPA, 2019b),\215\ and guidance on air quality modeling for meeting air
quality goals for the ozone and PM2.5 NAAQS and regional
haze program (U.S. EPA, 2018b).
---------------------------------------------------------------------------
\215\ Provides guidance on developing demonstrations under
section 189(e) intended to show that a certain PM2.5
precursor in a particular nonattainment area does not significantly
contribute to PM2.5 concentrations that exceed the
standard.
---------------------------------------------------------------------------
As stated in the NPRM, the PM2.5 SIP Requirements Rule
provides recommendations to States regarding consideration of
environmental justice in the context of PM2.5 attainment
[[Page 16369]]
planning. Some of the considerations for States include: (1)
Identifying areas with overburdened communities where more ambient
monitoring may be warranted; (2) targeting emissions reductions that
may be needed to attain the PM2.5 NAAQS; and (3) increasing
opportunities for meaningful involvement for overburdened populations
(see 88 FR 5558, 5684, January 27, 2023; 80 FR 58010, 58136, August 25,
2016). In light of the identification of at-risk populations for this
reconsideration, the EPA encourages States to consider these and other
factors as part of their attainment plan SIP development process.
The PM2.5 SIP Requirements Rule outlines some examples
of how States can elect to implement these recommendations.\216\ For
instance, States can use modeling and screening tools to better
understand where sources of PM2.5 or PM2.5
precursor emissions are located and identify areas that may be
candidates for additional ambient monitoring. Furthermore, once these
target areas are identified, States can prioritize direct
PM2.5 or PM2.5 precursor control measures and
enforcement strategies in these areas to reduce ambient
PM2.5 and achieve the NAAQS. As articulated in the NPRM and
the PM2.5 SIP Requirements Rule, the EPA recognizes that
States have flexibility under the CAA to concentrate State resources on
controlling sources of PM2.5 emissions in light of
environmental justice considerations (see 88 FR 5558, 5684, January 27,
2023; 81 FR 58010, 58137, August 24, 2016). Moreover, States can
establish opportunities to bolster meaningful involvement in a number
of ways, such as communicating in appropriate languages, ensuring
access to draft SIPs and other information, and developing enhanced
notice-and-comment opportunities, as appropriate (see 88 FR 5558, 5684,
January 27, 2023; 80 FR 58010, 58136, August 25, 2016).
---------------------------------------------------------------------------
\216\ For more information on the EPA's recommendations and
examples, see 81 FR 58010, 58137, August 24, 2016.
---------------------------------------------------------------------------
As previously mentioned, the PM2.5 SIP Requirements Rule
provides guidance and regulatory requirements for remaining
nonattainment areas for the 1997, 2006, and 2012 revisions of the
PM2.5 NAAQS, as well as for nonattainment areas designated
pursuant to any future revisions of the PM2.5 NAAQS,
including the revised annual PM2.5 NAAQS being finalized in
this action. The EPA is not making any changes to the current
PM2.5 SIP Requirements Rule.
D. Implementing the Primary and Secondary PM10 NAAQS
As summarized in sections III.B.4 and V.B.4 above, the EPA is
retaining the current primary and secondary 24-hour PM10
NAAQS to protect against the health effects associated with short-term
exposures to thoracic coarse particles and against the welfare effects
considered in this reconsideration (i.e., visibility, climate, and
materials effects). The EPA is retaining the existing implementation
strategy for meeting the CAA requirements for the PM10
NAAQS. States and emissions sources should continue to follow the
existing regulations and guidance for implementing the current
standards.\217\
---------------------------------------------------------------------------
\217\ CAA Sections 110(a) and 172 contain general nonattainment
planning provisions, regarding the public review, adoption,
submittal, and content of implementation plans. CAA Section 189
specifies additional plan provisions for particulate matter
nonattainment areas. General Preamble for the Implementation of
Title I of the Clean Air Act Amendments of 1990 provides a detailed
discussion of the EPA's interpretation of the Title I requirements
(57 FR 13498, April 16, 1992; 59 FR 41998, August 16, 1994).
---------------------------------------------------------------------------
E. Prevention of Significant Deterioration and Nonattainment New Source
Review Programs for the Revised Primary Annual PM2.5 NAAQS
The CAA, at parts C and D of title I, contains preconstruction
review and permitting programs applicable to new major stationary
sources and major modifications of existing major sources. The
preconstruction review of each new major stationary source and major
modification applies on a pollutant-specific basis, and the
requirements that apply for each pollutant depend on whether the area
in which the source is situated is designated as attainment (or
unclassifiable) or nonattainment for that pollutant. In areas
designated attainment or unclassifiable for a pollutant, the Prevention
of Significant Deterioration (PSD) requirements under part C apply to
construction at major sources. In areas designated nonattainment for a
pollutant, the Nonattainment New Source Review (NNSR) requirements
under part D apply to construction at major sources. Collectively,
those two sets of permit requirements are commonly referred to as the
``major New Source Review'' or ``major NSR'' programs.
Until the EPA designates an area with respect to the revised
primary annual PM2.5 NAAQS, the NSR provisions applicable
under an area's current designation for the 1997, 2006, and 2012
PM2.5 NAAQS would continue to apply. See 40 CFR 51.166(i)(2)
and 52.21(i)(2). That is, for areas designated as attainment/
unclassifiable for the 1997, 2006, and 2012 PM2.5 NAAQS, PSD
will apply to new major stationary sources and major modifications that
trigger major source permitting requirements for PM2.5. For
areas designated nonattainment for the 1997, 2006, or 2012
PM2.5 NAAQS, NNSR requirements will apply for new major
stationary sources and major modifications that trigger major source
permitting requirements for PM2.5. When the initial area
designations for this revised primary annual PM2.5 NAAQS
become effective, those designations will further determine whether PSD
or NNSR applies to PM2.5 in a particular area, depending on
the designation status. New major sources and major modifications will
be subject to the PSD program requirements for PM2.5 if they
are located in an area that does not have a current nonattainment
designation under CAA section 107 for PM2.5.\218\
---------------------------------------------------------------------------
\218\ 40 CFR 51.166(i)(2) and 52.21(i)(2).
---------------------------------------------------------------------------
Under the PSD program, the permit applicant must demonstrate that
the new or modified source emissions increase does not cause or
contribute to a NAAQS violation. In 2017, the EPA revised the Guideline
on Air Quality Models (published as appendix W to 40 CFR part 51) to
address primary and secondary PM2.5 impacts in making this
demonstration. The EPA has since provided associated technical
guidance, models and tools, such as the recent ``Final Guidance for
Ozone and Fine Particulate Matter Permit Modeling'' (July 29,
2022).\219\ Additionally, in light of this NAAQS revision, the EPA is
updating its guidance that provides recommended significant impact
levels (SILs) for PM2.5 and expects that an updated SIL for
the revised primary annual PM2.5 NAAQS will be available
[[Page 16370]]
on or before the effective date of the final NAAQS.
---------------------------------------------------------------------------
\219\ On July 29, 2022, the EPA issued ``Final Guidance for
Ozone and Fine Particulate Matter Permit Modeling,'' available at
https://www.epa.gov/system/files/documents/2022-07/Guidance_for_O3_PM25_Permit_Modeling.pdf. This guidance provides the
EPA's recommendations for how a stationary source seeking a PSD
permit may demonstrate that it will not cause or contribute to a
violation of the National Ambient Air Quality Standards for Ozone
and PM2.5 and PSD increments for PM2.5, as
required under section 165(a)(3) of the Clean Air Act and 40 CFR
51.166(k) and 52.21(k). The EPA has also previously issued two
technical guidance documents for use in conducting these
demonstrations: ``Guidance on the Development of Modeled Emission
Rates for Precursors (MERPs) as a Tier 1 Demonstration Tool for
Ozone and PM2.5 under the PSD Permitting Program,''
available at https://www.epa.gov/sites/default/files/2020-09/documents/epa-454_r-19-003.pdf, and ``Guidance on the Use of Models
for Assessing the Impacts of Emissions from Single Sources on the
Secondarily Formed Pollutants: Ozone and PM2.5,''
available at https://www.epa.gov/sites/default/files/2020-09/documents/epa-454_r-16-005.pdf.
---------------------------------------------------------------------------
The statutory requirements for a PSD permit program set forth under
part C of title I of the CAA (sections 160 through 169) are addressed
by the EPA's PSD regulations found at 40 CFR 51.166 (minimum
requirements for an approvable PSD SIP) and 40 CFR 52.21 (PSD
permitting program for permits issued under the EPA's Federal
permitting authority). These regulations already apply to
PM2.5 in areas that are designated attainment or
unclassifiable for PM2.5 whenever a proposed new major
source or major modification triggers PSD requirements for
PM2.5.
For PSD, a ``major stationary source'' is one with the potential to
emit 250 tons per year (tpy) or more of any regulated NSR pollutant,
unless the new or modified source is classified under a list of 28
source categories contained in the statutory definition of ``major
emitting facility'' in section 169(1) of the CAA. For those 28 source
categories, a ``major stationary source'' is one with the potential to
emit 100 tpy or more of any regulated NSR pollutant. A ``major
modification'' is a physical change or a change in the method of
operation of an existing major stationary source that results, first,
in a significant emissions increase of a regulated NSR pollutant and,
second, in a significant net emissions increase of that pollutant. See
40 CFR 51.166(b)(2)(i), 40 CFR 52.21(b)(2)(i). The EPA PSD regulations
define the term ``regulated NSR pollutant'' to include any pollutant
for which a NAAQS has been promulgated and any pollutant identified by
the EPA as a constituent or precursor to such pollutant. See 40 CFR
51.166(b)(49), 40 CFR 52.21(b)(50). These regulations identify
SO2 and NOX as precursors to PM2.5 in
attainment and unclassifiable areas. See 40 CFR 51.166(b)(49)(i)(b), 40
CFR 52.21(b)(50)(i)(b).\220\ Thus, for PM2.5, the PSD
program currently requires the review and control of emissions of
direct PM2.5 emissions and SO2 and NOX
(as precursors to PM2.5), absent a demonstration otherwise
for NOX. Among other things, for each regulated NSR
pollutant emitted or increased in a significant amount, the PSD program
requires a new major stationary source or a major modification to apply
the ``best available control technology'' (BACT) to limit emissions and
to conduct an air quality impact analysis to demonstrate that the
proposed major stationary source or major modification will not cause
or contribute to a violation of any NAAQS or PSD increment.\221\ See
CAA section 165(a)(3) and (4), 40 CFR 51.166(j) and (k), 40 CFR
52.21(j) and (k). The PSD requirements may also include, in appropriate
cases, an analysis of potential adverse impacts on Class I areas. See
CAA sections 162(a) and 165(d), 40 CFR 51.166(p); 40 CFR
52.21(p)).\222\ The EPA developed the Guideline on Air Quality Models
and other documents to, among other things, provide methods and
guidance for demonstrating that increased emissions from construction
will not cause or contribute to exceedances of the PM2.5
NAAQS and PSD increments for PM2.5.\223\
---------------------------------------------------------------------------
\220\ Sulfur dioxide is a precursor to PM2.5 in all
attainment and unclassifiable areas. NOX is presumed to
be a precursor to PM2.5 in all attainment and
unclassifiable areas, unless a state or the EPA demonstrates that
emissions of NOX from sources in a specific area are not
a significant contributor to that area's ambient PM2.5
concentrations. VOC is presumed not to be a precursor to
PM2.5 in any attainment or unclassifiable area, unless a
state or the EPA demonstrates that emissions of VOC from sources in
a specific area are a significant contributor to that area's ambient
PM2.5 concentrations.
\221\ By establishing the maximum allowable level of ambient
pollutant concentration increase in a particular area, an increment
defines ``significant deterioration'' of air quality in that area.
Increments are defined by the CAA as maximum allowable increases in
ambient air concentrations above a baseline concentration and are
specified in the PSD regulations by pollutant and area
classification (Class I, II and III). 40 CFR 51.166(c), 40 CFR
52.21(c); 75 FR 64864 (October 20, 2010).
\222\ Congress established certain Class I areas in section
162(a) of the CAA, including international parks, national
wilderness areas, and national parks that meet certain criteria.
Such Class I areas, known as mandatory Federal Class I areas, are
afforded special protection under the CAA. In addition, States and
Tribal governments may establish Class I areas within their own
political jurisdictions to provide similar special air quality
protection.
\223\ See 40 CFR part 51, appendix W; 82 FR 5182 (January 17,
2017); See also U.S. EPA, 2021d. The EPA provided an initial version
of the 2021 guidance for public comment on February 10, 2020. Upon
consideration of the comments received, and consistent with
Executive Order 13990, the EPA revised the initial draft guidance
and posted the revised version for additional public comment.
---------------------------------------------------------------------------
Upon the effective date of the revised primary annual
PM2.5 NAAQS, the demonstration required under CAA Section
165(a)(3), and the associated regulations, must include the revised
primary annual PM2.5 NAAQS. In past NAAQS revision rules,
including the 2012 PM2.5 NAAQS (78 FR 3086, January 15,
2013) and 2015 Ozone NAAQS (80 FR 65292, October 26, 2015), the EPA
included limited provision that exempted certain sources with pending
PSD permit applications (those that had reached a particular stage in
the permitting process at the time the revised NAAQS was promulgated or
became effective) from the requirement to demonstrate that the proposed
emissions increases would not cause or contribute to a violation of the
revised NAAQS.\224\ In August 2019, the U.S. Court of Appeals for the
D.C. Circuit vacated the exemption provision in the PSD rules for the
2015 Ozone NAAQS, finding that the provision contradicted ``Congress's
`express policy choice' not to allow construction which will `cause or
contribute to' nonattainment of `any' effective NAAQS, regardless of
when they are adopted or when a permit was completed.'' Murray Energy
Corp. v. EPA, 936 F.3d 597, 627 (D.C. Cir. 2019).\225\ Based on that
court decision, the EPA is not establishing any PSD permitting
exemption provision in this action. Some commenters requested that the
EPA provide the same kind of relief for pending PSD permit applications
by extending the effective date of this new revised NAAQS beyond the 60
days that the EPA has traditionally used for such rules. Such comments
are addressed in the Response to Comments portion of this action. The
EPA is making this revised primary annual PM2.5 NAAQS
effective in 60 days.
---------------------------------------------------------------------------
\224\ This exemption was referred to as ``grandfathering'' in
the 2015 Ozone NAAQS and the D.C. Circuit's Murray Energy Corp.
decision on that exemption. See 80 FR 65292, 65431 (October 26,
2015); Murray Energy Corp. v. EPA, 936 F.3d 597, 627 (D.C. Cir.
2019). The EPA refers to this ``grandfathering'' provision in this
action as an exemption provision.
\225\ While the specifics of this case involved the 2015 ozone
NAAQS, the case was based upon an interpretation of CAA section
165(a) and therefore applies equally to any PSD permitting exemption
provision for a new or revised NAAQS.
---------------------------------------------------------------------------
The EPA anticipates that the existing PM2.5 air quality
in some areas will not be in attainment with the revised primary annual
PM2.5 NAAQS, and the EPA will designate these areas as
nonattainment at a later date, consistent with the designation process
described in the preceding sections. However, until such nonattainment
designation occurs, proposed new major sources and major modifications
located in any area currently designated attainment or unclassifiable
for all preexisting PM2.5 NAAQS will continue to be subject
to the PSD program requirements for PM2.5. Any proposed
major stationary source or major modification triggering PSD
requirements for PM2.5 that does not receive its PSD permit
by the effective date of a new nonattainment designation for the area
where the source would locate would then be required to satisfy
applicable NNSR preconstruction permit requirements for
PM2.5.
In areas where air pollution exceeds the level of the revised
primary annual PM2.5 NAAQS, a PSD permit applicant must
demonstrate that the source or modification will not cause or
[[Page 16371]]
contribute to a violation of the NAAQS. Section 165(a)(3)(B) of the CAA
states that a proposed source may not construct unless it demonstrates
that it will not cause or contribute to a violation of any NAAQS. This
statutory requirement is implemented through a provision contained in
the PSD regulations at 40 CFR 51.166(k) and 52.21(k).\226\ If a source
cannot make this demonstration, or if its initial air quality impact
analysis shows that the source's impact would cause or contribute to a
violation, the reviewing authority may not issue a PSD permit to that
source. However, a PSD permit applicant may be able to make this
demonstration if it compensates for the adverse impact that would
otherwise cause or contribute to a violation of the NAAQS. In contrast
to the NSR requirements for nonattainment areas, the PSD regulations do
not explicitly specify remedial actions that a prospective source must
take to address such a situation, but the EPA has historically
recognized that sources applying for PSD permits may utilize offsetting
reductions in emissions as part of the required PSD demonstration under
CAA section 165(a)(3)(B).\227\
---------------------------------------------------------------------------
\226\ 40 CFR 51.166(k) states that SIPs must require that the
owner or operator of the proposed source or modification demonstrate
that allowable emission increases from the proposed source or
modification, in conjunction with all other applicable emissions
increases or reductions (including secondary emissions), would not
cause or contribute to air pollution in violation of: (i) Any
national ambient air quality standard in any air quality control
region; or (ii) any applicable maximum allowable increase over the
baseline concentration in any area.
\227\ See, e.g., Memorandum from Stephen D. Page, Director,
Office of Air Quality Planning and Standards to Regional Air
Division Directors, Guidance Concerning Implementation of the 1-hour
SO2 NAAQS for the Prevention of Significant Deterioration
Program. August 23, 2010. Office of Air Quality Planning and
Standards U.S. EPA, Research Triangle Park. Available at: https://www.epa.gov/sites/default/files/2015-07/documents/appwso2.pdf; 44 FR
3274, 3278, January 16, 1979; See also In re Interpower of New York,
Inc., 5 E.A.D. 130, 141 (EAB 1994) (describing an EPA Region 2 PSD
permit that relied in part on offsets to demonstrate the source
would not cause or contribute to a violation of the NAAQS). 52 FR
24634, 24684, July 1, 1987; 78 FR 3085, 3261-62, January 15, 2013.
The EPA has recognized the ability of sources to obtain offsets in
the context of PSD though the PSD provisions of the Act do not
expressly reference offsets as the NNSR provisions of the Act do.
See 80 FR 65292, 65441, October 26, 2015.
---------------------------------------------------------------------------
Part D of title I of the CAA includes preconstruction review and
permitting requirements applicable to new major stationary sources and
major modifications located in areas designated nonattainment for a
pollutant for which the EPA has established a NAAQS (i.e., a criteria
pollutant). The relevant part D requirements are typically referred to
as the nonattainment NSR (NNSR) program. The EPA's regulations for the
NNSR program are contained in 40 CFR 51.165 and 52.24 and part 51,
appendix S. Specifically, the EPA has developed minimum program
requirements for a NNSR program that is approvable in a SIP, and those
requirements, which include requirements for PM2.5, are
contained in 40 CFR 51.165. In addition, 40 CFR part 51, appendix S,
contains requirements constituting an interim NNSR program. This
interim program enables NNSR permitting in nonattainment areas by
States that lack a SIP-approved NNSR permitting program during the time
between the date of the relevant designation and the date that the EPA
approves into the SIP a NNSR program. See 40 CFR part 51, appendix S,
section I; 40 CFR 52.24(k).
For NNSR, ``major stationary source'' is generally defined as a
source with the potential to emit at least 100 tpy of the regulated NSR
pollutant for which the area is designated nonattainment. In some
cases, however, the CAA and the NNSR regulations define ``major
stationary source'' for NNSR in terms of a lower rate dependent on the
pollutant and degree of nonattainment in the area. For purposes of the
PM2.5NAAQS, in addition to the general threshold level of
100 tpy in Moderate PM2.5 nonattainment areas, a lower major
source threshold of 70 tpy applies in Serious PM2.5
nonattainment areas pursuant to subpart 4 of part D, title I of the
CAA. See 40 CFR 51.165(a)(1)(iv)(A)(1)(vii) and (viii); 40 CFR part 51,
appendix S, II.A.4(i)(a)(7) and (8).
Under the NNSR program, direct PM2.5 emissions and
emissions of each PM2.5 precursor are considered separately
in accordance with the applicable major source threshold. For example,
the threshold for Serious PM2.5 nonattainment areas is 70
tpy of direct PM2.5, as well as for the PM2.5
precursors SO2, NOX, VOC, and ammonia.\228\ See
40 CFR 51.165(a)(1)(iv)(A)(1)(vii) and (viii); 40 CFR part 51, appendix
S, II.A.4.(i)(a)(7) and (8). A source qualifies as major for
nonattainment NSR in a PM2.5 nonattainment area if it emits
or has the potential to emit direct PM2.5 or any
PM2.5 precursor in an amount equal to or greater than the
applicable threshold.
---------------------------------------------------------------------------
\228\ All of these pollutants are identified as precursors to
PM2.5 in NNSR regulations. See 40 CFR
51.165(a)(1)(xxxvii)(C)(2). No significant emission rate is
established by the EPA for ammonia, and states are required to
define ``significant'' for ammonia for their respective areas unless
the state pursues the optional precursor demonstration to exclude
ammonia from planning requirements. See 40 CFR 51.165(a)(1)(x)(F);
40 CFR 51.165(a)(13).
---------------------------------------------------------------------------
For modifications, NNSR applies to proposed physical changes or
changes in the method of operation of an existing stationary source
where (1) the source is major for the nonattainment pollutant (or a
precursor for that pollutant) and (2) the physical change or change in
the method of operation of a major stationary source results, first, in
a significant emissions increase of a regulated NSR pollutant and,
second, in a significant net emissions increase of that same
nonattainment pollutant (or same precursor for that pollutant). See 40
CFR 51.165(a)(1)(v)(A); 40 CFR part 51, appendix S, II.A.5.(i). For
example, to qualify as a major modification for SO2 (as a
PM2.5 precursor) in a Moderate PM2.5
nonattainment area, the existing source would have to have the
potential to emit 100 tpy or more of SO2, and the project
would have to result in an increase in SO2 emissions of 40
tpy or more. See 40 CFR 51.165(a)(1)(x)(A).
New major stationary sources and major modifications for
PM2.5 subject to NNSR must comply with the ``lowest
achievable emission rate'' (LAER), as defined in the CAA and NNSR
rules. Such sources must also perform other analyses and obtain
emission offsets, as required under section 173 of the CAA and
applicable regulations.
Following the promulgation of this revised primary annual
PM2.5 NAAQS, some new areas may be designated nonattainment
for PM2.5. Where a State does not have an existing NNSR
program or where the current NNSR program does not apply to
PM2.5, that State will be required to submit the necessary
SIP revisions to ensure that new major stationary sources and major
modifications for PM2.5 or a PM2.5 precursor
undergo preconstruction review pursuant to the NNSR program. States
with designated nonattainment areas for the revised primary annual
PM2.5 NAAQS are required to make SIP submissions to meet
nonattainment plan requirements within 18 months from the effective
date of designations, as required under CAA section 189(a)(2)(B).
States that have existing NNSR program requirements that cannot be
interpreted to apply at the time of designation to the revised primary
annual PM2.5 NAAQS may, in the interim, issue permits in
accordance with the applicable nonattainment permitting requirements
contained in 40 CFR part 51, appendix S, which would apply to the
revised primary annual PM2.5 NAAQS upon its effective date.
See 73 FR 28321, 28340, May 16, 2008.
Finally, the EPA has released several documents that discuss air
permitting
[[Page 16372]]
and environmental justice, including, for example, a memorandum \229\
and attached permitting principles.\230\ The EPA recommends that PSD
and NNSR permitting authorities review this memorandum and the
principles and consider applying them in their air permitting actions
as appropriate to help identify, analyze, and address environmental
justice concerns in those air permitting actions to help ensure that
the NAAQS achieve their intended health benefits for at-risk
populations.
---------------------------------------------------------------------------
\229\ Memorandum from Joseph Goffman, Principal Deputy Assistant
Administrator, Office of Air and Radiation, to Air and Radiation
Division Directors, ``Principles for Addressing Environmental
Justice in Air Permitting'' (December 22, 2022), available at
https://www.epa.gov/caa-permitting/ej-air-permitting-principles-addressing-environmental-justice-concerns-air.
\230\ Id., Attachment, ``EJ in Air Permitting: Principles for
Addressing Environmental Justice Concerns in Air Permitting''
(December 2022), available at https://www.epa.gov/caa-permitting/ej-air-permitting-principles-addressing-environmental-justice-concerns-air.
---------------------------------------------------------------------------
F. Transportation Conformity Program
Transportation conformity is required under CAA section 176(c) to
ensure that transportation plans, transportation improvement programs
(TIPs) and federally supported highway and transit projects will not
cause or contribute to any new air quality violation, increase the
frequency or severity of any existing violation, or delay timely
attainment or any required interim emissions reductions or other
milestones. Transportation conformity applies to areas that are
designated as nonattainment or nonattainment areas that have been
redesignated to attainment with an approved CAA section 175A
maintenance plan (i.e., maintenance areas) for transportation-related
criteria pollutants: carbon monoxide, ozone, NO2,
PM2.5, and PM10. Transportation conformity for
the revised primary annual PM2.5 NAAQS does not apply until
one year after the effective date of nonattainment designations for
that NAAQS. See CAA section 176(c)(6) and 40 CFR 93.102(d)). The EPA's
Transportation Conformity Rule \231\ establishes the criteria and
procedures for determining whether transportation activities conform to
the SIP. No changes are being made to the transportation conformity
rule in this final rulemaking. The EPA notes that the transportation
conformity rule already addresses the PM2.5 and
PM10 NAAQS. However, in the future, the EPA intends to
review the need to issue or revise guidance describing how the current
conformity rule applies in nonattainment and maintenance areas for the
revised primary annual PM2.5 NAAQS, as needed.
---------------------------------------------------------------------------
\231\ 40 CFR part 93, subpart A.
---------------------------------------------------------------------------
G. General Conformity Program
The conformity requirement under CAA section 176(c) ensures that
federal activities implemented by federal agencies will not interfere
with a State's ability to attain and maintain the NAAQS. Under CAA
176(c)(1), the requirement prohibits Federal agencies from approving,
permitting, licensing, or funding activities that do not conform to the
purpose of the applicable SIP for the control and prevention of air
pollution. See CAA 176(c)(1)(A). Under CAA 176(c)(1)(B), conformity to
an implementation plan means that federal activities will not cause or
contribute to any new violations of the NAAQS, increase the frequency
or severity of any existing NAAQS violation, or delay timely attainment
or any required interim emissions reductions or other milestones
contained in the applicable SIP.
The general conformity program \232\ implements CAA section
176(c)(4)(A), and the criteria and procedures for determining
conformity of federal activities to the applicable SIP are established
under 40 CFR part 93 subpart B, sections 93.150 through 93.165. General
Conformity applies to federal activities that (1) would cause emissions
of relevant criteria or precursor pollutants to originate within
nonattainment areas or areas that have been redesignated to attainment
with an approved CAA section 175A maintenance plan (i.e., maintenance
areas), as set forth under 40 CFR 93.153, and (2) are not Federal
Highway Administration (FHWA) or Federal Transit Administration (FTA)
transportation projects as defined in 40 CFR 93.101 under the
transportation conformity requirements. See 40 CFR 93.153. General
conformity for the revised primary annual PM2.5 NAAQS does
not apply until one year after the effective date of the nonattainment
designation for that NAAQS. See 40 CFR 93.153(k).
---------------------------------------------------------------------------
\232\ 40 CFR part 93 subpart B.
---------------------------------------------------------------------------
With regard to issues regarding prescribed fires, which were
addressed earlier in this action, here is some additional information
regarding prescribed fires and General Conformity regulations. Under
the General Conformity regulations at 40 CFR 93.153(c)(4), a conformity
evaluation is not required to support a decision by a federal agency to
conduct or carry out prescribed burning when the burn is consistent
with the terms of a land management plan or other plan that includes
the prescribed burn at issue, where the overall plan that includes the
burn was previously evaluated under 40 CFR part 93 subpart B by the
responsible federal agency, and the agency found the plan conforms
under CAA paragraphs 176(c)(1)(A) and (1)(B). This assumes the burn at
issue will be conducted by meeting any conditions specified as
necessary for meeting conformity in the agency's decision to approve
the plan. Alternatively, a presumption of conformity applies also under
40 CFR 93.153(i)(2) for prescribed fires conducted in accordance with a
Smoke Management Program that meets the requirements of the EPA's 1998
Interim Air Quality Policy on Wildland and Prescribed Fires or an
equivalent replacement EPA policy. The preamble to the Exceptional
Events Rule explains that the EPA adapted language associated with the
six basic components of a certifiable Smoke Management Program for
exceptional events purposes from the 1998 Interim Air Quality Policy on
Wildland and Prescribed Fires (see, e.g., 81 FR 68216, 68252 (including
footnote 75), 68256, October 2, 2016). The Exceptional Events Rule at
40 CFR 50.14(a)(3)(ii)(A) also indicates that certain requirements
within the Exceptional Events Rule can be satisfied if a prescribed
fire is conducted under a certified Smoke Management Program or using
appropriate basic smoke management practices such as those identified
in Table 1 to 40 CFR 50.14 (see e.g., 81 FR 68216, 68250-68257, 68277-
68278, October 3, 2016).
No changes are being made to the general conformity regulations in
this final rulemaking and the EPA notes that the courts recognize the
regulations constitute control for the established PM2.5 and
PM10 NAAQS. However, in the future, the EPA intends to
review the need to issue or revise guidance describing how the current
General Conformity regulations apply within nonattainment and
maintenance areas for the revised primary annual PM2.5
NAAQS, as needed.\233\
---------------------------------------------------------------------------
\233\ Further, the EPA's current Unified Agenda and Regulatory
Plan includes its intention to issue a proposed rule to amend the
General Conformity Regulations. The EPA intends to address in that
regulatory action topics regarding prescribed fire, including
consideration of smoke management approaches such as those discussed
in the Exceptional Events Rule, among other topics. See, e.g.,
https://www.reginfo.gov/public/do/eAgendaViewRule?pubId=202310&RIN=2060-AV28.
---------------------------------------------------------------------------
IX. Statutory and Executive Order Reviews
Additional information about these statutes and Executive orders
can be
[[Page 16373]]
found at https://www.epa.gov/laws-regulations/laws-and-executive-orders.
A. Executive Order 12866: Regulatory Planning and Review and Executive
Order 14094: Modernizing Regulatory Review
This action is ``significant regulatory action'' as defined under
section 3(f)(1) of Executive Order 12866, as amended by Executive Order
14094. Accordingly, the EPA submitted this action to the Office of
Management and Budget (OMB) for review. Documentation of any changes
made in response to the Executive Order 12866 review is available in
the docket. The EPA prepared an illustrative analysis of the potential
costs and benefits associated with this action. This analysis,
``Regulatory Impact Analysis for the Reconsideration of the National
Ambient Air Quality Standards for Particulate Matter,'' is available in
the Regulatory Impact Analysis (RIA) docket (EPA-HQ-OAR-2019-0587) and
briefly summarized below. However, the CAA and judicial decisions make
clear that the economic and technical feasibility of attaining ambient
standards are not to be considered in setting or revising NAAQS,
although such factors may be considered in the development of State
plans to implement the standards. Accordingly, although an RIA has been
prepared, the results of the RIA have not been considered in issuing
this final rule.
The RIA estimates the costs and monetized human health benefits in
2032, after implementing existing and expected regulations and
assessing emissions reductions to meet the current primary annual and
24-hour particulate matter NAAQS (12/35 [mu]g/m\3\), associated with
applying national control strategies for the revised annual and 24-hour
standard levels of 9/35 [mu]g/m\3\, as well as the following less and
more stringent alternative standard levels: (1) A less stringent
alternative annual standard level of 10 [mu]g/m\3\ in combination with
the current 24-hour standard (i.e., 10/35 [mu]g/m\3\), (2) a more
stringent alternative annual standard level of 8 [mu]g/m\3\ in
combination with the current 24-hour standard (i.e., 8/35 [mu]g/m\3\),
and (3) a more stringent alternative 24-hour standard level of 30
[mu]g/m\3\ in combination with an annual standard level of 10 [mu]g/
m\3\ (i.e., 10/30 [mu]g/m\3\). Table 3 provides a summary of the
estimated monetized benefits, costs, and net benefits associated with
applying national control strategies toward reaching the revised and
alternative standard levels.
Table 3--Estimated Monetized Benefits, Costs, and Net Benefits of the Illustrative Control Strategies Applied
Toward the Primary Revised and Alternative Annual and Daily Standard Levels of 10/35 [mu]g/m\3\, 10/30 [mu]g/
m\3\, 9/35 [mu]g/m\3\, and 8/35 [mu]g/m\3\ in 2032 for the U.S.
[Millions of 2017$]
----------------------------------------------------------------------------------------------------------------
10/35 10/30 9/35 8/35
----------------------------------------------------------------------------------------------------------------
Benefits \a\.................... $8,500 and $17,000 $10,000 and $22,000 and $48,000 and
$21,000. $46,000. $99,000.
Costs \b\....................... $200.............. $340.............. $590.............. $1,500.
-------------------------------------------------------------------------------
Net Benefits................ $8,300 and $17,000 $9,900 and $21,000 $22,000 and $46,000 and
$46,000. $97,000.
----------------------------------------------------------------------------------------------------------------
Notes: Rows may not appear to add correctly due to rounding. We provide a snapshot of costs and benefits in
2032, using the best available information to approximate social costs and social benefits recognizing
uncertainties and limitations in those estimates. The estimated costs and monetized human health benefits
associated with applying national control strategies do not fully account for all the emissions reductions
needed to reach the final and more stringent alternative standard levels for some standard levels analyzed.
\a\ We assume that there is a cessation lag between the change in PM exposures and the total realization of
changes in mortality effects. Specifically, we assume that some of the incidences of premature mortality
related to PM2.5 exposures occur in a distributed fashion over the 20 years following exposure, which affects
the valuation of mortality benefits at different discount rates. Similarly, we assume there is a cessation lag
between the change in PM exposures and both the development and diagnosis of lung cancer. The benefits are
associated with two point estimates from two different epidemiologic studies, and we present the benefits
calculated at a real discount rate of 3 percent. The monetized benefits exclude additional health and welfare
benefits that could not be quantified.
\b\ The costs are annualized using a 7 percent interest rate.
B. Paperwork Reduction Act (PRA)
This action does not impose any new information collection burden
under the PRA. OMB has previously approved the information collection
activities contained in the existing regulations and has assigned OMB
control number 2060-0084. The data collected through this information
collection consist of ambient air concentration measurements for the
seven air pollutants with national ambient air quality standards (i.e.,
ozone, sulfur dioxide, nitrogen dioxide, lead, carbon monoxide,
PM2.5 and PM10), ozone precursors, air toxics,
meteorological variables at a select number of sites, and other
supporting measurements. Accompanying the pollutant concentration data
are quality assurance/quality control data and air monitoring network
design information. The EPA and others (e.g., State and local air
quality management agencies, tribal entities, environmental
organizations, academic institutions, industrial groups) use the
ambient air quality data for many purposes including informing the
public and other interested parties of an area's air quality, judging
an area's air quality in comparison with the established health or
welfare standards, evaluating an air quality management agency's
progress in achieving or maintaining air pollutant levels below the
national and local standards, developing and revising State
Implementation Plans (SIPs), evaluating air pollutant control
strategies, developing or revising national control policies, providing
data for air quality model development and validation, supporting
enforcement actions, documenting episodes and initiating episode
controls, assessing air quality trends, and conducting air pollution
research.
C. Regulatory Flexibility Act (RFA)
I certify that this action will not have a significant economic
impact on a substantial number of small entities under the RFA. This
action will not impose any requirements on small entities. Rather, this
final rule establishes national standards for allowable concentrations
of PM in ambient air as required by section 109 of the CAA. See also
American Trucking Associations v. EPA, 175 F.3d 1027, 1044-45 (D.C.
Cir. 1999) (NAAQS do not have significant impacts upon small entities
because NAAQS themselves impose no regulations upon small entities),
rev'd in part on other grounds, Whitman v. American Trucking
Associations, 531 U.S. 457 (2001).
[[Page 16374]]
D. Unfunded Mandates Reform Act (UMRA)
This action does not contain an unfunded mandate of $100 million or
more as described in the Unfunded Mandates Reform Act (UMRA), 2 U.S.C.
1531-1538, and does not significantly or uniquely affect small
governments. Furthermore, as indicated previously, in setting a NAAQS
the EPA cannot consider the economic or technological feasibility of
attaining ambient air quality standards, although such factors may be
considered to a degree in the development of State plans to implement
the standards. See also American Trucking Associations v. EPA, 175 F.
3d at 1043 (noting that because the EPA is precluded from considering
costs of implementation in establishing NAAQS, preparation of the RIA
pursuant to the Unfunded Mandates Reform Act would not furnish any
information that the court could consider in reviewing the NAAQS).
E. Executive Order 13132: Federalism
This action will not have substantial direct effects on the States,
on the relationship between the National Government and the States, or
on the distribution of power and responsibilities among the various
levels of government. However, the EPA recognizes that States will have
a substantial interest in this action and any future revisions to
associated requirements.
F. Executive Order 13175: Consultation and Coordination With Indian
Tribal Governments
This action does not have Tribal implications, as specified in
Executive Order 13175. It does not have a substantial direct effect on
one or more Indian Tribes as Tribes are not obligated to adopt or
implement any NAAQS. In addition, Tribes are not obligated to conduct
ambient monitoring for PM or to adopt the ambient monitoring
requirements of 40 CFR part 58. Thus, Executive Order 13175 does not
apply to this action. However, consistent with the EPA Policy on
Consultation and Coordination with Indian Tribes, the EPA offered
consultation to all 574 Federally Recognized Tribes during the
development of this action. Although no Tribes requested consultation,
the EPA provided informational meetings including an informational
meeting with the Pueblo de San Ildefonso and provided information on
the monthly National Tribal Air Association calls.
G. Executive Order 13045: Protection of Children From Environmental
Health Risks and Safety Risks
Executive Order 13045 directs federal agencies to include an
evaluation of the health and safety effects of the planned regulation
on children in federal health and safety standards and explain why the
regulation is preferable to potentially effective and reasonably
feasible alternatives. This action is subject to Executive Order 13045
because it is a significant regulatory action under section 3(f)(1) of
Executive Order 12866, and the EPA believes that the environmental
health or safety risk addressed by this action may have a
disproportionate effect on children. Accordingly, we have evaluated the
environmental health or safety effects of PM exposures on children. The
protection offered by these standards may be especially important for
children because childhood represents a lifestage associated with
increased susceptibility to PM-related health effects. Because children
have been identified as a susceptible population, we have carefully
evaluated the environmental health effects of exposure to PM pollution
among children. Children make up a substantial fraction of the U.S.
population, and often have unique factors that contribute to their
increased risk of experiencing a health effect due to exposures to
ambient air pollutants because of their continuous growth and
development. As described in the 2019 Integrated Science Assessment,
children may be particularly at risk for health effects related to
ambient air PM2.5 exposures compared with adults because
they have (1) a developing respiratory system, (2) increased
ventilation rates relative to body mass compared with adults, and (3)
an increased proportion of oral breathing, particularly in boys,
relative to adults. More detailed information on the evaluation of the
scientific evidence and policy considerations pertaining to children,
including an explanation for why the Administrator judges the revised
standards to be requisite to protect public health, including the
health of children, with an adequate margin of safety, are contained in
section II.A.2. ``Overview of the Health Effects Evidence'', section
II.A.2.b ``Public Health Implications and At-Risk Populations'' and
II.B ``Conclusions on the Primary PM2.5 Standards'' of this
preamble. Copies of all documents have been placed in the public docket
for this action. The Administrator judges that revising the primary
annual PM2.5 standard to a level of 9.0 [micro]g/m\3\ and
retaining the primary 24-hour PM2.5 standard provides
requisite public health protection with an adequate margin of safety,
including for children. Furthermore, the Policy on Children's Health
also applies to this action. Information on how the Policy was applied
is described in section II.A.2 ``Overview of the Health Effects
Evidence'', section II.A.2.b ``Public Health Implications and At-Risk
Populations'' and II.B ``Conclusions on the Primary PM2.5
Standards'' of this preamble.
H. Executive Order 13211: Actions Concerning Regulations That
Significantly Affect Energy Supply, Distribution or Use
This action is not a ``significant energy action'' because it is
not likely to have a significant adverse effect on the supply,
distribution, or use of energy. The purpose of this action is to revise
level of the primary annual PM2.5 NAAQS. The action does not
prescribe specific pollution control strategies by which these ambient
standards and monitoring revisions will be met. Such strategies will be
developed by States on a case-by-case basis, and the EPA cannot predict
whether the control options selected by States will include regulations
on energy suppliers, distributors, or users. Thus, the EPA concludes
that this action does not constitute a significant energy action as
defined in Executive Order 13211.
I. National Technology Transfer and Advancement Act (NTTAA)
This rulemaking involved environmental monitoring or measurement.
The EPA has decided it will continue to use the existing indicators for
fine (PM2.5) and coarse (PM10) particles. The
indicator for fine particles is measured using the Reference Method for
the Determination of Fine Particulate Matter as PM2.5 in the
Atmosphere (appendix L to 40 CFR part 50), which is known as the
PM2.5 FRM, and the indicator for coarse particles is
measured using the Reference Method for the Determination of
Particulate Matter as PM10 in the Atmosphere (appendix J to
40 CFR part 50), which is known as the PM10 FRM.
To the extent feasible, the EPA employs a Performance-Based
Measurement System (PBMS), which does not require the use of specific,
prescribed analytic methods. The PBMS is defined as a set of processes
wherein the data quality needs, mandates or limitations of a program or
project are specified and serve as criteria for selecting appropriate
methods to meet those needs in a cost-effective manner.
[[Page 16375]]
It is intended to be more flexible and cost effective for the regulated
community; it is also intended to encourage innovation in analytical
technology and improved data quality. Though the FRM defines the
particular specifications for ambient monitors, there is some
variability with regard to how monitors measure PM, depending on the
type and size of PM and environmental conditions. Therefore, it is not
practically possible to fully define the FRM in performance terms to
account for this variability. Nevertheless, our approach in the past
has resulted in multiple brands of monitors being approved as FRM for
PM, and we expect this to continue. Also, the FRMs described in 40 CFR
part 50 and the equivalency criteria described in 40 CFR part 53,
constitute a performance-based measurement system for PM, since methods
that meet the field testing and performance criteria can be approved as
FEMs. Since finalized in 2006 (71 FR 61236, October 17, 2006) the new
field and performance criteria for approval of PM2.5
continuous FEMs has resulted in the approval of 13 approved FEMs. In
summary, for measurement of PM2.5 and PM10, the
EPA relies on both FRMs and FEMs, with FEMs relying on a PBMS approach
for their approval. The EPA is not precluding the use of any other
method, whether it constitutes a voluntary consensus standard or not,
as long as it meets the specified performance criteria.
J. Executive Order 12898: Federal Actions To Address Environmental
Justice in Minority Populations and Low-Income Populations and
Executive Order 14096: Revitalizing Our Nation's Commitment to
Environmental Justice for All
The EPA believes that the human health or environmental conditions
associated with the primary PM2.5 NAAQS that exist prior to
this action result in or have the potential to result in
disproportionate and adverse human health or environmental effects on
communities with environmental justice concerns. There is strong
evidence for racial and ethnic disparities in PM2.5
exposures and PM2.5-related health risk, as assessed in the
2019 Integrated Science Assessment and with even more evidence
available since the literature cutoff date for the 2019 Integrated
Science Assessment and evaluated in the Supplement to the 2019
Integrated Science Assessment. There is strong evidence demonstrating
that Black and Hispanic populations, in particular, have higher
PM2.5 exposures than non-Hispanic White populations. Black
populations or individuals that live in predominantly Black
neighborhoods experience higher PM2.5 exposures, in
comparison to non-Hispanic White populations. There is also consistent
evidence across multiple studies that demonstrate increased risk of
PM2.5-related health effects, with the strongest evidence
for health risk disparities for mortality. There is also evidence of
health risk disparities for both Hispanic and non-Hispanic Black
populations compared to non-Hispanic White populations for cause-
specific mortality and incident hypertension.
Socioeconomic status (SES) is a composite measure that includes
metrics such as income, occupation, or education, and can play a role
in access to healthy environments as well as access to healthcare. SES
may be a factor that contributes to differential risk from
PM2.5-related health effects. Studies assessed in the 2019
Integrated Science Assessment and Supplement to the 2019 Integrated
Science Assessment provide evidence that lower SES communities are
exposed to higher concentrations of PM2.5 compared to higher
SES communities. Studies using composite measures of neighborhood SES
consistently demonstrated a disparity in both PM2.5 exposure
and the risk of PM2.5-related health outcomes. There is some
evidence that supports associations larger in magnitude between
mortality and long-term PM2.5 exposures for those with low
income or living in lower income areas compared to those with higher
income or living in higher income neighborhoods. Additionally, evidence
supports conclusions that lower SES is associated with cause-specific
mortality and certain health endpoints (i.e., HI and CHF), but less so
for all-cause or total (non-accidental) mortality.
The EPA believes that this action is likely to reduce existing
disproportionate and adverse effects on communities with environmental
justice concerns.
The EPA additionally identified and addressed environmental justice
concerns by providing opportunities for public input on the proposed
decisions. The EPA held a multi-day virtual public hearing for the
public to provide oral testimony and there was a 60-day public comment
period for the proposed action. As described in section II.A.3 above,
the EPA conducted a risk assessment to support this action that
included an at-risk analysis that evaluates exposure and
PM2.5 mortality risk for older adults (e.g., 65 years and
older), stratified for White, Black, Asian, Native American, Non-
Hispanic, and Hispanic individuals. This at-risk analysis found that
compared to a primary annual PM2.5 standard with a level of
12.0 [micro]g/m\3\, meeting a revised annual standard with a level of
9.0 [micro]g/m\3\ is estimated to reduce PM2.5-associated
health risks in the 30 study areas controlled by the annual standard by
about 22-28% and is expected to reduce disparities in exposure and risk
among these populations.
The information supporting this Executive Order review is is
contained in sections II.A.2, II.B.3.a, II.B.3.c, II.B.2, and II.B.4.
of this preamble and also in the 2019 Integrated Science Assessment,
Supplement to the 2019 Integrated Science Assessment, and 2022 Policy
Assessment. The EPA has carefully evaluated the potential impacts on
minority populations and low SES populations as discussed in sections
II.A.2, II.A.3, II.B.2, and II.B.4 of this preamble. The 2019
Integrated Science Assessment, Supplement to the Integrated Science
Assessment, and 2022 Policy Assessment contain the evaluation of the
scientific evidence, quantitative risk analyses and policy
considerations that pertain to these populations. These documents are
available in the public docket for this action.
K. Congressional Review Act (CRA)
This action is subject to the CRA, and the EPA will submit a rule
report to each House of the Congress and to the Comptroller General of
the United States. This action meets the criteria set forth in 5 U.S.C.
804(2).
References
Abt Associates, Inc. (2001). Assessing public opinions on visibility
impairment due to air pollution: Summary report. U.S. Environmental
Protection Agency. Research Triangle Park, NC. Available at: https://www3.epa.gov/ttn/naaqs/standards/pm/data/vis_rpt_final.pdf.
Abt Associates, Inc. (2005). Particulate matter health risk
assessment for selected urban areas: Draft report. EPA Contract No.
68-D-03-002. U.S. Environmental Protection Agency. Research Triangle
Park, NC. Available at: https://www3.epa.gov/ttn/naaqs/standards/pm/data/PMrisk20051220.pdf.
ATS (2000). What Constitutes an Adverse Health Effect of Air
Pollution? American Journal of Respiratory and Critical Care
Medicine 161(2): 665-673.
BBC Research & Consulting (2003). Phoenix area visibility survey.
Denver, CO. Available at: https://www.regulations.gov/document/EPA-HQ-OAR-2015-0072-0089.
Behbod, B, Urch, B, Speck, M, Scott, JA, Liu, L, Poon, R, Coull, B,
Schwartz, J, Koutrakis, P, Silverman, F and Gold, DR (2013).
Endotoxin in concentrated coarse and fine ambient particles induces
acute systemic inflammation in controlled
[[Page 16376]]
human exposures. Occupational and Environmental Medicine 70(11):
761-767.
Bell, ML, Ebisu, K, Peng, RD, Walker, J, Samet, JM, Zeger, SL and
Dominic, F (2008). Seasonal and regional short-term effects of fine
particles on hospital admissions in 202 U.S. counties, 1999-2005.
American Journal of Epidemiology 168(11): 1301-1310.
Bellavia, A, Urch, B, Speck, M, Brook, RD, Scott, JA, Albetti, B,
Behbod, B, North, M, Valeri, L, Bertazzi, PA, Silverman, F, Gold, D
and Baccarelli, AA (2013). DNA hypomethylation, ambient particulate
matter, and increased blood pressure: Findings from controlled human
exposure experiments. Journal of the American Heart Association
2(3): e000212.
Bennett, JE, Tamura-Wicks, H, Parks, RM, Burnett, RT, Pope, CA,
Bechle, MJ, Marshall, JD, Danaei, G and Ezzati, M (2019).
Particulate matter air pollution and national and county life
expectancy loss in the USA: A spatiotemporal analysis. PLoS Medicine
16(7): e1002856.
Besson, P, Mu[ntilde]oz, C, Ram[iacute]rez-Sagner, G, Salgado, M,
Escobar, R and Platzer, W (2017). Long-Term Soiling Analysis for
Three Photovoltaic Technologies in Santiago Region. IEEE Journal of
Photovoltaics 7(6): 1755-1760.
Bourdrel, T, Annesi-Maesano, I, Alahmad, B, Maesano, CN and Bind, MA
(2021). The impact of outdoor air pollution on COVID-19: a review of
evidence from in vitro, animal, and human studies. 30(159).
Br[auml]uner, EV, M[oslash]ller, P, Barregard, L, Dragsted, LO,
Glasius, M, W[aring]hlin, P, Vinzents, P, Raaschou-Nielsen, O and
Loft, S (2008). Exposure to ambient concentrations of particulate
air pollution does not influence vascular function or inflammatory
pathways in young healthy individuals. Particle and Fibre Toxicology
5: 13.
Brook, RD, Urch, B, Dvonch, JT, Bard, RL, Speck, M, Keeler, G,
Morishita, M, Marsik, FJ, Kamal, AS, Kaciroti, N, Harkema, J, Corey,
P, Silverman, F, Gold, DR, Wellenius, G, Mittleman, MA, Rajagopalan,
S and Brook, JR (2009). Insights into the mechanisms and mediators
of the effects of air pollution exposure on blood pressure and
vascular function in healthy humans. Hypertension 54(3): 659-667.
Burns, J, Boogaard, H, Polus, S, Pfadenhauer, LM, Rohwer, AC, van
Erp, AM, Turley, R and Rehfuess, E (2019). Interventions to reduce
ambient particulate matter air pollution and their effect on health.
Cochrane Database of Systematic Reviews (5).
Cangerana Pereira, FA, Lemos, M, Mauad, T, de Assuncao, JV and
Nascimento Saldiva, PH (2011). Urban, traffic-related particles and
lung tumors in urethane treated mice. Clinics 66(6): 1051-1054.
Chan, EAW, Gantt, B and McDow, S (2018). The reduction of summer
sulfate and switch from summertime to wintertime PM2.5
concentration maxima in the United States. Atmospheric Environment
175: 25-32.
Chen, H, Burnett, RT, Copes, R, Kwong, JC, Villeneuve, PJ, Goldberg,
MS, Brook, RD, van Donkelaar, A, Jerrett, M, Martin, RV, Brook, JR,
Kopp, A and Tu, JV (2016). Ambient fine particulate matter and
mortality among survivors of myocardial infarction: population-based
cohort study. Environmental Health Perspectives 124(9): 1421-1428.
Correia, AW, Pope, CA, III, Dockery, DW, Wang, Y, un, Ezzati, M and
Dominici, F (2013). Effect of air pollution control on life
expectancy in the United States: an analysis of 545 U.S. counties
for the period from 2000 to 2007. Epidemiology 24(1): 23-31.
Corrigan, AE, Becker, MM, Neas, LM, Cascio, WE and Rappold, AG
(2018). Fine particulate matters: The impact of air quality
standards on cardiovascular mortality. Environmental research 161:
364-369.
Cox, LA. (2019a). Letter from Louis Anthony Cox, Jr., Chair, Clean
Air Scientific Advisory Committee, to Administrator Andrew R.
Wheeler. Re: CASAC Review of the EPA's Integrated Science Assessment
for Particulate Matter (External Review Draft--October 2018). April
11, 2019. EPA-CASAC-19-002. Office of the Administrator, Science
Advisory Board U.S. EPA HQ, Washington DC. Available at: https://casac.epa.gov/ords/sab/r/sab_apex/casac/0?report_id=1069&request=APPLICATION_PROCESS%3DREPORT_DOC&session=7184955370570.
Cox, LA. (2019b). Letter from Louis Anthony Cox, Jr., Chair, Clean
Air Scientific Advisory Committee, to Administrator Andrew R.
Wheeler. Re: CASAC Review of the EPA's Policy Assessment for the
Review of the National Ambient Air Quality Standards for Particulate
Matter (External Review Draft--September 2019). December 16, 2019.
EPA-CASAC-20-001. Office of the Administrator, Science Advisory
Board U.S. EPA HQ, Washington DC. Available at: https://casac.epa.gov/ords/sab/r/sab_apex/casac/0?report_id=1073&request=APPLICATION_PROCESS%3DREPORT_DOC&session=6224161457429.
DeFlorio-Barker, S, Crooks, J, Reyes, J and Rappold, AG (2019).
Cardiopulmonary effects of fine particulate matter exposure among
older adults, during wildfire and non-wildfire periods, in the
United States 2008-2010. Environmental health perspectives 127(3):
037006.
DHEW (1969). Air Quality Criteria for Particulate Matter. National
Air Pollution Control Administration. Washington, DC U.S. Department
of Health. January 1969.
Di, Q, Amini, H, Shi, L, Kloog, I, Silvern, R, Kelly, J, Sabath, MB,
Choirat, C, Koutrakis, P and Lyapustin, A (2019). An ensemble-based
model of PM2.5 concentration across the contiguous United
States with high spatiotemporal resolution. Environment
International 130: 104909.
Di, Q, Dai, L, Wang, Y, Zanobetti, A, Choirat, C, Schwartz, JD and
Dominici, F (2017a). Association of short-term exposure to air
pollution with mortality in older adults. JAMA: Journal of the
American Medical Association 318(24): 2446-2456.
Di, Q, Kloog, I, Koutrakis, P, Lyapustin, A, Wang, Y and Schwartz, J
(2016). Assessing PM2.5 exposures with high
spatiotemporal resolution across the Continental United States.
Environmental Science and Technology 50(9): 4712-4721.
Di, Q, Wang, Y, Zanobetti, A, Wang, Y, Koutrakis, P, Choirat, C,
Dominici, F and Schwartz, JD (2017b). Air pollution and mortality in
the Medicare population. New England Journal of Medicine 376(26):
2513-2522.
Dominici, F, Schwartz, J, Di, Q, Braun, D, Choirat, C and Zanobetti,
A (2019). Assessing adverse health effects of long-term exposure to
low levels of ambient air pollution: Phase 1. Health Effects
Institute. Boston, MA. Available at: https://www.healtheffects.org/system/files/dominici-rr-200-report.pdf.
Ely, DW, Leary, JT, Stewart, TR and Ross, DM (1991). The
establishment of the Denver Visibility Standard. Colorado Department
of Health. Denver, Colorado.
Erickson, AC, Christidis, T, Pappin, A, Brook, JR, Crouse, DL,
Hystad, P, Li, C, Martin, RV, Meng, J, Pinault, L, von Donkelaar, A,
Weichenthal, S, Tjepkema, M, Burnett, RT and Brauer, M (2020).
Disease assimilation: The mortality impacts of fine particulate
matter on immigrants to Canada. Health Reports 31(3): 14-26.
Eum, K, Suh, HH, Pun, V and Manjourides, J (2018). Impact of long-
term temporal trends in fine particulate matter (PM2.5)
on association of annual PM2.5 exposure and mortality: an
analysis of over 20 million Medicare beneficiaries. Environmental
Epidemiology 2(2): e009.
Fiore, AM, Naik, V and Leibensperger, EM (2015). Air quality and
climate connections. Journal of the Air and Waste Management
Association 65(6): 645-685.
Frank, N. (2012). Memorandum to PM NAAQS Review Docket (EPA-HQ-OAR-
2007-0492) regarding the Differences between maximum and composite
monitor annual PM2.5 design values by CBSA. Dec 14, 2012.
Docket ID No. EPA-HQ-OAR-2007-0492. Office of Air Quality Planning
and Standards Research Triangle Park, NC. Available at: https://www.regulations.gov/document/EPA-HQ-OAR-2007-0492-10099.
Franklin, M, Zeka, A and Schwartz, J (2007). Association between
PM2.5 and all-cause and specific-cause mortality in 27 US
communities. Journal of Exposure Science and Environmental
Epidemiology 17(3): 279-287.
Gantt, B, Owen, RC and Watkins, N (2021). Characterizing Nitrogen
Oxides and Fine Particulate Matter near Major Highways in the United
States Using the National Near-Road Monitoring Network.
Environmental science & technology 55(5): 2831-2838.
[[Page 16377]]
Ghio, AJ, Hall, A, Bassett, MA, Cascio, WE and Devlin, RB (2003).
Exposure to concentrated ambient air particles alters hematologic
indices in humans. Inhalation Toxicology 15(14): 1465-1478.
Ghio, AJ, Kim, C and Devlin, RB (2000). Concentrated ambient air
particles induce mild pulmonary inflammation in healthy human
volunteers. American Journal of Respiratory and Critical Care
Medicine 162(3): 981-988.
Greven, S, Dominici, F and Zeger, S (2011). An Approach to the
Estimation of Chronic Air Pollution Effects Using Spatio-Temporal
Information. Journal of the American Statistical Association
106(494): 396-406.
Gr[oslash]ntoft, T, Verney-Carron, A and Tidblad, J (2019). Cleaning
Costs for European Sheltered White Painted Steel and Modern Glass
Surfaces Due to Air Pollution Since the Year 2000. Atmosphere 10(4):
167.
Hammer, MS, van Donkelaar, A, Li, C, Lyapustin, A, Sayer, AM, Hsu,
NC, Levy, RC, Garay, MJ, Kalashnikova, OV and Kahn, RA (2020).
Global estimates and long-term trends of fine particulate matter
concentrations (1998-2018). Environmental Science & Technology
54(13): 7879-7890.
Hart, JE, Liao, X, Hong, B, Puett, RC, Yanosky, JD, Suh, H,
Kioumourtzoglou, MA, Spiegelman, D and Laden, F (2015). The
association of long-term exposure to PM2.5 on all-cause
mortality in the Nurses' Health Study and the impact of measurement-
error correction. Environmental Health: A Global Access Science
Source 14: 38.
Hassett-Sipple, B, Schmidt, M and Rajan, P. (2010). Memorandum to PM
NAAQS Review Docket (EPA-HQ-OAR-2007-0492). Analysis of
PM2.5 (Particulate Matter Smaller than 2.5 Micrometers in
Diameter). Mar 30, 2010. Docket ID No. EPA-HQ-OAR-2007-0492. Office
of Air Quality Planning and Standards Research Triangle Park, NC.
Available at: https://www.regulations.gov/document/EPA-HQ-OAR-2007-0492-0077.
Hemmingsen, JG, Jantzen, K, M[oslash]ller, P and Loft, S (2015a). No
oxidative stress or DNA damage in peripheral blood mononuclear cells
after exposure to particles from urban street air in overweight
elderly. Mutagenesis 30(5): 635-642.Hemmingsen, JG, Rissler, J,
Lykkesfeldt, J, Sallsten, G, Kristiansen, J, M[oslash]ller, P and
Loft, S (2015). Controlled exposure to particulate matter from urban
street air is associated with decreased vasodilation and heart rate
variability in overweight and older adults. Particle and Fibre
Toxicology 12(1): 6.
Henneman, LR, Liu, C, Mulholland, JA and Russell, AG (2017).
Evaluating the effectiveness of air quality regulations: A review of
accountability studies and frameworks. Journal of the Air Waste
Management Association 67(2): 144-172.
Henneman, LRF, Choirat, C and Zigler, ACM (2019). Accountability
assessment of health improvements in the United States associated
with reduced coal emissions between 2005 and 2012. Epidemiology
30(4): 477-485.
Hutchinson, JA, Vargo, J, Milet, M, French, NH, Billmire, M,
Johnson, J and Hoshiko, S (2018). The San Diego 2007 wildfires and
Medi-Cal emergency department presentations, inpatient
hospitalizations, and outpatient visits: An observational study of
smoke exposure periods and a bidirectional case-crossover analysis.
PLoS medicine 15(7): e1002601.
IPCC (2013). Climate change 2013: The physical science basis.
Contribution of working group I to the fifth assessment report of
the Intergovernmental Panel on Climate Change. T. F. Stocker, D.
Qin, G. K. Plattner et al. Cambridge University Press. Cambridge,
UK.
Kloog, I, Ridgway, B, Koutrakis, P, Coull, BA and Schwartz, JD
(2013). Long- and short-term exposure to PM2.5 and
mortality: Using novel exposure models. Epidemiology 24(4): 555-561.
Krewski, D, Jerrett, M, Burnett, RT, Ma, R, Hughes, E, Shi, Y,
Turner, MC, Pope, CA, III, Thurston, G, Calle, EE, Thun, MJ,
Beckerman, B, Deluca, P, Finkelstein, N, Ito, K, Moore, DK, Newbold,
KB, Ramsay, T, Ross, Z, Shin, H and Tempalski, B (2009). Extended
follow-up and spatial analysis of the American Cancer Society study
linking particulate air pollution and mortality. ISSN 1041-5505, HEI
Research Report 140. Health Effects Institute. Boston, MA. Available
at: https://www.healtheffects.org/system/files/Krewski140Statement.pdf.
Laden, F, Schwartz, J, Speizer, FE and Dockery, DW (2006). Reduction
in fine particulate air pollution and mortality: extended follow-up
of the Harvard Six Cities study. American Journal of Respiratory and
Critical Care Medicine 173(6): 667-672.
Lavigne, E, Burnett, RT and Weichenthal, S (2018). Association of
short-term exposure to fine particulate air pollution and mortality:
effect modification by oxidant gases. Scientific Reports 8(1):
16097.
Lee, M, Koutrakis, P, Coull, B, Kloog, I and Schwartz, J (2015).
Acute effect of fine particulate matter on mortality in three
Southeastern states from 2007-2011. Journal of Exposure Science and
Environmental Epidemiology 26(2): 173-179.
Lepeule, J, Laden, F, Dockery, D and Schwartz, J (2012). Chronic
exposure to fine particles and mortality: an extended follow-up of
the Harvard Six Cities study from 1974 to 2009. Environmental Health
Perspectives 120(7): 965-970.
Lippmann, M, Chen, LC, Gordon, T, Ito, K and Thurston, GD (2013).
National Particle Component Toxicity (NPACT) Initiative: Integrated
epidemiologic and toxicologic studies of the health effects of
particulate matter components: Investigators' Report. 177. Health
Effects Institute. Boston, MA.
Liu, C, Chen, R, Sera, F, Vicedo-Cabrera, AM, Guo, Y, Tong, S,
Coelho, M, Saldiva, PHN, Lavigne, E, Matus, P, Valdes Ortega, N,
Osorio Garcia, S, Pascal, M, Stafoggia, M, Scortichini, M,
Hashizume, M, Honda, Y, Hurtado-D[iacute]az, M, Cruz, J, Nunes, B,
Teixeira, JP, Kim, H, Tobias, A, [Iacute][ntilde]iguez, C, Forsberg,
B, [Aring]str[ouml]m, C, Ragettli, MS, Guo, YL, Chen, BY, Bell, ML,
Wright, CY, Scovronick, N, Garland, RM, Milojevic, A, Kysel[yacute],
J, Urban, A, Orru, H, Indermitte, E, Jaakkola, JJK, Ryti, NRI,
Katsouyanni, K, Analitis, A, Zanobetti, A, Schwartz, J, Chen, J, Wu,
T, Cohen, A, Gasparrini, A and Kan, H (2019). Ambient Particulate
Air Pollution and Daily Mortality in 652 Cities. New England Journal
of Medicine 381(8): 705-715.
Lowenthal, DH and Kumar, N (2004). Variation of mass scattering
efficiencies in IMPROVE. Journal of the Air and Waste Management
Association (1990-1992) 54(8): 926-934.
Lowenthal, DH and Kumar, N (2016). Evaluation of the IMPROVE
Equation for estimating aerosol light extinction. Journal of the Air
and Waste Management Association 66(7): 726-737.
Lucking, AJ, Lundb[auml]ck, M, Barath, SL, Mills, NL, Sidhu, MK,
Langrish, JP, Boon, NA, Pourazar, J, Badimon, JJ, Gerlofs-Nijland,
ME, Cassee, FR, Boman, C, Donaldson, K, Sandstrom, T, Newby, DE and
Blomberg, A (2011). Particle traps prevent adverse vascular and
prothrombotic effects of diesel engine exhaust inhalation in men.
Circulation 123(16): 1721-1728.
Malm, WC and Hand, JL (2007). An examination of the physical and
optical properties of aerosols collected in the IMPROVE program.
Atmospheric Environment 41(16): 3407-3427.
Malm, WC, Schichtel, B, Molenar, J, Prenni, A and Peters, M (2019).
Which visibility indicators best represent a population's preference
for a level of visual air quality? Journal of the Air & Waste
Management Association 69(2): 145-161.
Malm, WC, Sisler, JF, Huffman, D, Eldred, RA and Cahill, TA (1994).
Spatial and seasonal trends in particle concentration and optical
extinction in the United States. Journal of Geophysical Research
99(D1): 1347-1370.
Mauad, T, Rivero, DH, de Oliveira, RC, Lichtenfels, AJ, Guimaraes,
ET, de Andre, PA, Kasahara, DI, Bueno, HM and Saldiva, PH (2008).
Chronic exposure to ambient levels of urban particles affects mouse
lung development. American Journal of Respiratory and Critical Care
Medicine 178(7): 721-728.
Mie, G (1908). Beitrage zur Optik truber Medien, speziell
kolloidaler Metallosungen [Optics of cloudy media, especially
colloidal metal solutions]. Annalen der Physik 25(3): 377-445.
Miller, KA, Siscovick, DS, Sheppard, L, Shepherd, K, Sullivan, JH,
Anderson, GL and Kaufman, JD (2007). Long-term exposure to air
pollution and incidence of cardiovascular events in women. New
England Journal of Medicine 356(5): 447-458.
Myhre, G, Shindell, D, Br[eacute]on, FM, Collins, W, Fuglestvedt, J,
Huang, J, Koch, D, Lamarque, JF, Lee, D, Mendoza, B, Nakajima, T,
Robock, A, Stephens, G,
[[Page 16378]]
Takemura, T and Zhang, H, Eds. (2013). Anthropogenic and natural
radiative forcing. Cambridge University Press Cambridge, UK.
NHLBI (2017). ``NHLBI fact book, fiscal year 2012: Disease
statistics.'' Retrieved August 23, 2017, from https://web.archive.org/web/20170711012213/https://www.nhlbi.nih.gov/about/documents/factbook/2012/chapter4.
Orr, A, AL Migliaccio, C, Buford, M, Ballou, S and Migliaccio, CT
(2020). Sustained effects on lung function in community members
following exposure to hazardous pm2. 5 levels from wildfire smoke.
Toxics 8(3): 53.
Pitchford, M, Maim, W, Schichtel, B, Kumar, N, Lowenthal, D and
Hand, J (2007). Revised algorithm for estimating light extinction
from IMPROVE particle speciation data. Journal of the Air and Waste
Management Association 57(11): 1326-1336.
Pope, CA, III, Burnett, RT, Thurston, GD, Thun, MJ, Calle, EE,
Krewski, D and Godleski, JJ (2004). Cardiovascular mortality and
long-term exposure to particulate air pollution: epidemiological
evidence of general pathophysiological pathways of disease.
Circulation 109(1): 71-77.
Pope, CA, III, Ezzati, M and Dockery, DW (2009). Fine-particulate
air pollution and life expectancy in the United States. New England
Journal of Medicine 360(4): 376-386.
Pruitt, E. (2018). Memorandum from E. Scott Pruitt, Administrator,
U.S. EPA to Assistant Administrators. Back-to-Basics Process for
Reviewing National Ambient Air Quality Standards. May 9, 2018.
Office of the Administrator U.S. EPA HQ, Washington DC. Available
at: https://www.epa.gov/criteria-air-pollutants/back-basics-process-reviewing-national-ambient-air-quality-standards.
Pryor, SC (1996). Assessing public perception of visibility for
standard setting exercises. Atmospheric Environment 30(15): 2705-
2716.
Puett, RC, Hart, JE, Yanosky, JD, Spiegelman, D, Wang, M, Fisher,
JA, Hong, B and Laden, F (2014). Particulate matter air pollution
exposure, distance to road, and incident lung cancer in the Nurses'
Health Study cohort. Environmental Health Perspectives 122(9): 926-
932.
Pun, VC, Kazemiparkouhi, F, Manjourides, J and Suh, HH (2017). Long-
term PM2.5 exposures and respiratory, cancer and
cardiovascular mortality in American older adults. American Journal
of Epidemiology 186(8): 961-969.
Raaschou-Nielsen, O, Andersen, ZJ, Beelen, R, Samoli, E, Stafoggia,
M, Weinmayr, G, Hoffmann, B, Fischer, P, Nieuwenhuijsen, MJ,
Brunekreef, B, Xun, WW, Katsouyanni, K, Dimakopoulou, K, Sommar, J,
Forsberg, B, Modig, L, Oudin, A, Oftedal, B, Schwarze, PE, Nafstad,
P, De Faire, U, Pedersen, NL, [Ouml]stenson, CG, Fratiglioni, L,
Penell, J, Korek, M, Pershagen, G, Eriksen, KT, S[oslash]rensen, M,
Tj[oslash]nneland, A, Ellermann, T, Eeftens, M, Peeters, PH,
Meliefste, K, Wang, M, Bueno-De-mesquita, B, Key, TJ, De Hoogh, K,
Concin, H, Nagel, G, Vilier, A, Grioni, S, Krogh, V, Tsai, MY,
Ricceri, F, Sacerdote, C, Galassi, C, Migliore, E, Ranzi, A,
Cesaroni, G, Badaloni, C, Forastiere, F, Tamayo, I, Amiano, P,
Dorronsoro, M, Trichopoulou, A, Bamia, C, Vineis, P and Hoek, G
(2013). Air pollution and lung cancer incidence in 17 European
cohorts: Prospective analyses from the European Study of Cohorts for
Air Pollution Effects (ESCAPE). The Lancet Oncology 14(9): 813-822.
Ramanathan, G, Yin, F, Speck, M, Tseng, CH, Brook, JR, Silverman, F,
Urch, B, Brook, RD and Araujo, JA (2016). Effects of urban fine
particulate matter and ozone on HDL functionality. Particle and
Fibre Toxicology 13(1): 26.
Ryan, PA, Lowenthal, D and Kumar, N (2005). Improved light
extinction reconstruction in interagency monitoring of protected
visual environments. Journal of the Air and Waste Management
Association 55(11): 1751-1759.
Sacks, JD, Ito, K, Wilson, WE and Neas, LM (2012). Impact of
covariate models on the assessment of the air pollution-mortality
association in a single- and multipollutant context. American
Journal of Epidemiology 176(7): 622-634.
Sanders, NJ, Barreca, AI and Neidell, MJ (2020a). Estimating causal
effects of particulate matter regulation on mortality. Epidemiology
(Cambridge, Mass.) 31(2): 160.
Sanders, NJ, Barreca, AI and Neidell, MJ (2020b). Estimating Causal
Effects of Particulate Matter Regulation on Mortality. Epidemiology
31(2): 160-167.
Schwartz, J, Austin, E, Bind, MA, Zanobetti, A and Koutrakis, P
(2015). Estimating causal associations of fine particles with daily
deaths in Boston. American Journal of Epidemiology 182(7): 644-650.
Schwartz, J, Bind, MA and Koutrakis, P (2017). Estimating causal
effects of local air pollution on daily deaths: Effect of low
levels. Environmental Health Perspectives 125(1): 23-29.
Schwartz, J, Wei, Y, Di, Q, Dominici, F and Zanobetti, A (2021). A
national difference in differences analysis of the effect of PM2. 5
on annual death rates. Environmental Research 194: 110649.
Schwartz, JD, Wang, Y, Kloog, I, Yitshak-Sade, M, Dominici, F and
Zanobetti, A (2018). Estimating the Effects of PM2.5 on
Life Expectancy Using Causal Modeling Methods. Environmental Health
Perspectives 126(12): 127002.
Sheppard, EA. (2022a). Letter from Elizabeth A. (Lianne) Sheppard,
Chair, Clean Air Scientific Advisory Committee, to Administrator
Michael S. Regan. Re: CASAC Review of the EPA's Policy Assessment
for the Review of the National Ambient Air Quality Standards for
Particulate Matter (External Review Draft--October 2021). March 18,
2022. EPA-CASAC-22-002. Office of the Administrator, Science
Advisory Board U.S. EPA HQ, Washington DC. Available at: https://casac.epa.gov/ords/sab/r/sab_apex/casac/0?report_id=1094&request=APPLICATION_PROCESS%3DREPORT_DOC&session=7184955370570.
Sheppard, EA. (2022b). Letter from Elizabeth A. (Lianne) Sheppard,
Chair, Clean Air Scientific Advisory Committee, to Administrator
Michael S. Regan. Re: CASAC Review of the EPA's Supplement to the
2019 Integrated Science Assessment for Particulate Matter (External
Review Draft--October 2021). EPA-CASAC-22-001. Office of the
Administrator, Science Advisory Board U.S. EPA HQ, Washington DC.
Available at: https://casac.epa.gov/ords/sab/r/sab_apex/casac/0?report_id=1093&request=APPLICATION_PROCESS%3DREPORT_DOC&session=10813926997922.
Shi, L, Zanobetti, A, Kloog, I, Coull, BA, Koutrakis, P, Melly, SJ
and Schwartz, JD (2016). Low-concentration PM2.5 and
mortality: estimating acute and chronic effects in a population-
based study. Environmental Health Perspectives 124(1): 46-52.
Shin, HH, Gogna, P, Maquiling, A, Parajuli, RP, Haque, L and Burr, B
(2021). Comparison of hospitalization and mortality associated with
short-term exposure to ambient ozone and PM2.5 in Canada.
Chemosphere 265: 128683.
Sivagangabalan, G, Spears, D, Masse, S, Urch, B, Brook, RD,
Silverman, F, Gold, DR, Lukic, KZ, Speck, M, Kusha, M, Farid, T,
Poku, K, Shi, E, Floras, J and Nanthakumar, K (2011). The effect of
air pollution on spatial dispersion of myocardial repolarization in
healthy human volunteers. Journal of the American College of
Cardiology 57(2): 198-206.
Smith, AE and Howell, S (2009). An assessment of the robustness of
visual air quality preference study results. CRA International.
Washington, DC. Available at: https://www.regulations.gov/document/EPA-HQ-ORD-2007-0517-0085.
Statistics Canada (2023). Census Profile. 2021 Census of Population.
Statistics Canada Catalogue no. 98-316-X2021001. Ottawa. Released
March 29, 2023. Available at: https://www12.statcan.gc.ca/census-recensement/2021/dp-pd/prof/index.cfm?Lang=E.
Thurston, GD, Kipen, H, Annesi-Maesano, I, Balmes, J, Brook, RD,
Cromar, K, De Matteis, S, Forastiere, F, Forsberg, B, Frampton, MW,
Grigg, J, Heederik, D, Kelly, FJ, Kuenzli, N, Laumbach, R, Peters,
A, Rajagopalan, ST, Rich, D, Ritz, B, Samet, JM, Sandstrom, T,
Sigsgaard, T, Sunyer, J and Brunekreef, B (2017). A joint ERS/ATS
policy statement: what constitutes an adverse health effect of air
pollution? An analytical framework. European Respiratory Journal
49(1): 1600419.
Tong, H, Rappold, AG, Caughey, M, Hinderliter, AL, Bassett, M,
Montilla, T, Case, MW, Berntsen, J, Bromberg, PA, Cascio, WE, Diaz-
Sanchez, D, Devlin, RB and Samet, JM (2015). Dietary
[[Page 16379]]
supplementation with olive oil or fish oil and vascular effects of
concentrated ambient particulate matter exposure in human
volunteers. Environmental Health Perspectives 123(11): 1173-1179.
U.S. EPA (2004a). Air Quality Criteria for Particulate Matter. (Vol
I of II). Office of Research and Development. Research Triangle
Park, NC. U.S. EPA. EPA-600/P-99-002aF. October 2004. Available at:
https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P100LFIQ.txt.
U.S. EPA (2004b). Air Quality Criteria for Particulate Matter. (Vol
II of II). Office of Research and Development. Research Triangle
Park, NC. U.S. EPA. EPA-600/P-99-002bF. October 2004. Available at:
https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P100LG7Q.txt.
U.S. EPA (2005). Review of the National Ambient Air Quality
Standards for Particulate Matter: Policy Assessment of Scientific
and Technical Information, OAQPS Staff Paper. Office of Air Quality
Planning and Standards. Research Triangle Park, NC. U.S. EPA. EPA-
452/R-05-005a. December 2005. Available at: https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P1009MZM.txt.
U.S. EPA (2008). Integrated Review Plan for the National Ambient Air
Quality Standards for Particulate Matter. Office of Research and
Development, National Center for Environmental Assessment; Office of
Air Quality Planning and Standards, Health and Environmental Impacts
Division. Research Triangle Park, NC. U.S. EPA. EPA 452/R-08-004.
March 2008. Available at: https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P1001FB9.txt.
U.S. EPA (2009a). Particulate Matter National Ambient Air Quality
Standards: Scope and Methods Plan for Health Risk and Exposure
Assessment. Office of Air Quality Planning and Standards, Health and
Environmental Impacts Division. Research Triangle Park, NC. U.S.
EPA. EPA-452/P-09-002. February 2009. Available at: https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P100FLWP.txt.
U.S. EPA (2009b). Particulate Matter National Ambient Air Quality
Standards: Scope and Methods Plan for Urban Visibility Impact
Assessment. Office of Air Quality Planning and Standards, Health and
Environmental Impacts Division. Research Triangle Park, NC. U.S.
EPA. EPA-452/P-09-001. February 2009. Available at: https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P100FLUX.txt.
U.S. EPA (2009c). Integrated Science Assessment for Particulate
Matter (Final Report). Office of Research and Development, National
Center for Environmental Assessment. Research Triangle Park, NC.
U.S. EPA. EPA-600/R-08-139F. December 2009. Available at: https://cfpub.epa.gov/ncea/risk/recordisplay.cfm?deid=216546.
U.S. EPA (2010a). Quantitative Health Risk Assessment for
Particulate Matter (Final Report). Office of Air Quality Planning
and Standards, Health and Environmental Impacts Division. Research
Triangle Park, NC. U.S. EPA. EPA-452/R-10-005. June 2010. Available
at: https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P1007RFC.txt.
U.S. EPA (2010b). Particulate Matter Urban-Focused Visibility
Assessment (Final Document). Office of Air Quality Planning and
Standards, Health and Environmental Impacts Division. Research
Triangle Park, NC. U.S. EPA. EPA-452/R-10-004. July 2010. Available
at: https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P100FO5D.txt.
U.S. EPA (2011). Policy Assessment for the Review of the Particulate
Matter National Ambient Air Quality Standards. Office of Air Quality
Planning and Standards, Health and Environmental Impacts Division.
Research Triangle Park, NC. U.S. EPA. EPA-452/R-11-003. April 2011.
Available at: https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P100AUMY.txt.
U.S. EPA (2012). Responses to Significant Comments on the 2012
Proposed Rule on the National Ambient Air Quality Standards for
Particulate Matter (June 29, 2012; 77 FR 38890). Research Triangle
Park, NC. U.S. EPA. Docket ID No. EPA-HQ-OAR-2007-0492. Available
at: https://www3.epa.gov/ttn/naaqs/standards/pm/data/20121214rtc.pdf.
U.S. EPA (2015). Preamble to the integrated science assessments.
U.S. Environmental Protection Agency, Office of Research and
Development, National Center for Environmental Assessment, RTP
Division. Research Triangle Park, NC. U.S. EPA. EPA/600/R-15/067.
November 2015. Available at: https://cfpub.epa.gov/ncea/isa/recordisplay.cfm?deid=310244.
U.S. EPA (2016). Integrated review plan for the national ambient air
quality standards for particulate matter. Office of Air Quality
Planning and Standards. Research Triangle Park, NC. U.S. EPA. EPA-
452/R-16-005. December 2016. Available at: https://www3.epa.gov/ttn/naaqs/standards/pm/data/201612-final-integrated-review-plan.pdf.
U.S. EPA (2017). Emissions Inventory Guidance for Implementation of
Ozone and Particulate Matter National Ambient Air Quality Standards
(NAAQS) and Regional Haze Regulations. Office of Air Quality
Planning and Standards, Office of Air and Radiation. Research
Triangle Park, NC. U.S. EPA. U.S. EPA-454/B-17-002. Available at:
https://www.epa.gov/sites/default/files/2017-07/documents/ei_guidance_may_2017_final_rev.pdf.
U.S. EPA (2018a). Technical Assistance Document (TAD) for the
Reporting of Daily Air Quality--the Air Quality Index (AQI). U.S.
Environmental Protection Agency, Office of Air Quality Planning and
Standards. Research Triangle Park, NC. U.S. EPA. EPA 454/B-18-007.
September 2018. Available at: https://www.airnow.gov/sites/default/files/2020-05/aqi-technical-assistance-document-sept2018.pdf.
U.S. EPA (2018b). Modeling Guidance for Demonstrating Air Quality
Goals for Ozone, PM2.5, and Regional Haze. Office of Air
Quality Planning and Standards, Air Quality Policy Division.
Research Triangle Park, NC. U.S. EPA. EPA 454/R-18-009. November
2018. Available at: https://www.epa.gov/sites/default/files/2020-10/documents/o3-pm-rh-modeling_guidance-2018.pdf.
U.S. EPA (2019a). Integrated Science Assessment (ISA) for
Particulate Matter (Final Report). U.S. Environmental Protection
Agency, Office of Research and Development, National Center for
Environmental Assessment. Washington, DC. U.S. EPA. EPA/600/R-19/
188. December 2019. Available at: https://www.epa.gov/naaqs/particulate-matter-pm-standards-integrated-science-assessments-current-review.
U.S. EPA (2019b). PM2.5 Precursor Demonstration Guidance.
Office of Air Quality Planning and Standards, Air Quality Policy
Division. Research Triangle Park, NC. U.S. EPA. EPA-454/R-19-004.
May 2019. Available at: https://nepis.epa.gov/Exe/ZyPDF.cgi/P100YD1Q.PDF?Dockey=P100YD1Q.PDF.
U.S. EPA (2020a). Responses to significant comments on the 2020
proposed rule on the National Ambient Air Quality Standards for
particulate matter (April 30, 2020; 85 FR 24094). EPA-HQ-OAR-2015-
0072. Available at: https://www.epa.gov/sites/production/files/2020-12/documents/pm_naaqs_response_to_comments_final.pdf.
U.S. EPA (2020b). Policy Assessment for the Review of the National
Ambient Air Quality Standards for Particulate Matter. Office of Air
Quality Planning and Standards, Health and Environmental Impacts
Division. Research Triangle Park, NC. U.S. EPA. EPA-452/R-20-002.
January 2020. Available at: https://www.epa.gov/system/files/documents/2021-10/final-policy-assessment-for-the-review-of-the-pm-naaqs-01-2020.pdf.
U.S. EPA (2021a). Supplement to the 2019 Integrated Science
Assessment for Particulate Matter (External Review Draft). U.S.
Environmental Protection Agency, Office of Research and Development,
Center for Public Health and Environmental Assessment. Research
Triangle Park, NC. U.S. EPA. EPA/600/R-21/198. December 2019.
Available at: https://www.epa.gov/naaqs/particulate-matter-pm-standards-integrated-science-assessments-current-review.
U.S. EPA (2021b). Comparative Assessment of the Impacts of
Prescribed Fire Versus Wildfire (CAIF): A Case Study in the Western
U.S. U.S. Environmental Protection Agency. Washington, DC. U.S. EPA.
EPA/600/R-21/197.
U.S. EPA (2021c). Policy Assessment for the Review of the National
Ambient Air Quality Standards for Particulate Matter (External
Review Draft). Office of Air Quality Planning and Standards, Health
and Environmental Impacts Division. Research Triangle Park, NC. U.S.
EPA. EPA-452/P-21-001. October 2021. Available at: https://
www.epa.gov/system/files/documents/2022-05/
[[Page 16380]]
Final%20Policy%20Assessment%20for%20the%20Reconsideration%20of%20the%
20PM%20NAAQS_May2022_0.pdf.
U.S. EPA (2021d). Guidance for Ozone and Fine Particulate Matter
Permit Modeling. U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards, Air Quality Assessment Division.
Research Triangle Park, NC. U.S. EPA. EPA-454/P-21-001. September
2021.
U.S. EPA (2022a). Policy Assessment for the Reconsideration of the
National Ambient Air Quality Standards for Particulate Matter.
Office of Air Quality Planning and Standards, Health and
Environmental Impacts Division. Research Triangle Park, NC. U.S.
EPA. EPA-452/R-22-004. May 2022. Available at: https://www.epa.gov/system/files/documents/2022-05/Final%20Policy%20Assessment%20for%20the%20Reconsideration%20of%20the%20PM%20NAAQS_May2022_0.pdf.
U.S. EPA (2022b). Supplement to the 2019 Integrated Science
Assessment for Particulate Matter (Final Report). U.S. Environmental
Protection Agency, Office of Research and Development, Center for
Public Health and Environmental Assessment. Research Triangle Park,
NC. U.S. EPA. EPA/600/R 22/028. May 2022. Available at: https://www.epa.gov/naaqs/particulate-matter-pm-standards-integrated-science-assessments-current-review.
Urch, B, Speck, M, Corey, P, Wasserstein, D, Manno, M, Lukic, KZ,
Brook, JR, Liu, L, Coull, B, Schwartz, J, Gold, DR and Silverman, F
(2010). Concentrated ambient fine particles and not ozone induce a
systemic interleukin-6 response in humans. Inhalation Toxicology
22(3): 210-218.
Van de Hulst, H (1981). Light scattering by small particles. Dover
Publications, Inc. New York.
van Donkelaar, A, Martin, RV, Li, C and Burnett, RT (2019). Regional
estimates of chemical composition of fine particulate matter using a
combined geoscience-statistical method with information from
satellites, models, and monitors. Environmental Science & Technology
53(5).
Vieira, JL, Guimaraes, GV, de Andre, PA, Cruz, FD, Nascimento
Saldiva, PH and Bocchi, EA (2016a). Respiratory filter reduces the
cardiovascular effects associated with diesel exhaust exposure a
randomized, prospective, double-blind, controlled study of heart
failure: the FILTER-HF trial. JACC: Heart Failure 4(1): 55-64.
Vieira, JL, Guimaraes, GV, de Andre, PA, Nascimento Saldiva, PH and
Bocchi, EA (2016b). Effects of reducing exposure to air pollution on
submaximal cardiopulmonary test in patients with heart failure:
Analysis of the randomized, double-blind and controlled FILTER-HF
trial. International Journal of Cardiology 215: 92-97.
Wang, Y, Shi, L, Lee, M, Liu, P, Di, Q, Zanobetti, A and Schwartz,
JD (2017). Long-term exposure to PM2.5 and mortality
among older adults in the Southeastern US. Epidemiology 28(2): 207-
214.
Ward-Caviness, CK, Weaver, AM, Buranosky, M, Pfaff, ER, Neas, LM,
Devlin, RB, Schwartz, J, Di, Q, Cascio, WE and Diaz-Sanchez, D
(2020). Associations between long-term fine particulate matter
exposure and mortality in heart failure patients. Journal of the
American Heart Association 9(6): e012517.
Wei, Y, Yazdi, MD, Di, Q, Requia, WJ, Dominici, F, Zanobetti, A and
Schwartz, J (2021). Emulating causal dose-response relations between
air pollutants and mortality in the Medicare population.
Environmental Health: A Global Access Science Source 20(1): 53.
Wu, X, Braun, D, Schwartz, J, Kioumourtzoglou, MA and Dominici, F
(2020). Evaluating the impact of long-term exposure to fine
particulate matter on mortality among the elderly. Science Advances
6(29): eaba5692.
Wyatt, LH, Devlin, RB, Rappold, AG, Case, MW and Diaz-Sanchez, D
(2020). Low levels of fine particulate matter increase vascular
damage and reduce pulmonary function in young healthy adults.
Particle and fibre toxicology 17(1): 1-12.
Yorifuji, T, Kashima, S and Doi, H (2016). Fine-particulate air
pollution from diesel emission control and mortality rates in Tokyo:
a quasi-experimental study. Epidemiology 27(6): 769-778.
Zanobetti, A and Schwartz, J (2009). The effect of fine and coarse
particulate air pollution on mortality: A national analysis.
Environmental Health Perspectives 117(6): 1-40.
Zhang, Z, Wang, J, Kwong, JC, Burnett, RT, van Donkelaar, A, Hystad,
P, Martin, RV, Bai, L, McLaughlin, J and Chen, H (2021). Long-term
exposure to air pollution and mortality in a prospective cohort: The
Ontario Health Study. Environment International 154: 106570.
List of Subjects
40 CFR Part 50
Environmental protection, Air pollution control, Carbon monoxide,
Lead, Nitrogen dioxide, Ozone, Particulate matter, Sulfur oxides.
40 CFR Part 53
Environmental protection, Administrative practice and procedure,
Air pollution control, Reporting and recordkeeping requirements.
40 CFR Part 58
Environmental protection, Administrative practice and procedure,
Air pollution control, Intergovernmental relations, Reporting and
recordkeeping requirements.
Michael S. Regan,
Administrator.
For the reasons set forth in the preamble, chapter I of title 40 of
the Code of Federal Regulations is amended as follows:
PART 50--NATIONAL PRIMARY AND SECONDARY AMBIENT AIR QUALITY
STANDARDS
0
1. The authority citation for part 50 continues to read as follows:
Authority: 42 U.S.C. 7401, et seq.
0
2. Add Sec. 50.20 to read as follows:
Sec. 50.20 National primary ambient air quality standards for PM2.5.
(a) The national primary ambient air quality standards for
PM2.5 are 9.0 micrograms per cubic meter ([micro]g/m\3\)
annual arithmetic mean concentration and 35 [micro]g/m\3\ 24-hour
average concentration measured in the ambient air as PM2.5
(particles with an aerodynamic diameter less than or equal to a nominal
2.5 micrometers) by either:
(1) A reference method based on appendix L to this part and
designated in accordance with part 53 of this chapter; or
(2) An equivalent method designated in accordance with part 53 of
this chapter.
(b) The primary annual PM2.5 standard is met when the
annual arithmetic mean concentration, as determined in accordance with
appendix N to this part, is less than or equal to 9.0 [micro]g/m\3\.
(c) The primary 24-hour PM2.5 standard is met when the
98th percentile 24-hour concentration, as determined in accordance with
appendix N to this part, is less than or equal to 35 [micro]g/m\3\.
0
3. Amend appendix K to part 50 by:
0
a. In section 1.0 revising paragraph (b);
0
b. In section 2.3 adding paragraph (d); and
0
c. In section 3.0 adding paragraphs (a) and (b).
The revision and additions read as follows:
Appendix K to Part 50--Interpretation of the National Ambient Air
Quality Standards for Particulate Matter
1.0 General
* * * * *
(b) The terms used in this appendix are defined as follows:
Average refers to the arithmetic mean of the estimated number of
exceedances per year, as per section 3.1 of this appendix.
Collocated monitors refer to two or more air measurement
instruments for the same parameter (e.g., PM10 mass)
operated at the same site location, and whose placement is
consistent with part 53 of this chapter. For purposes of considering
a combined site record in this appendix, when two or more monitors
are operated at the same site, one
[[Page 16381]]
monitor is designated as the ``primary'' monitor with any additional
monitors designated as ``collocated.'' It is implicit in these
appendix procedures that the primary monitor and collocated
monitor(s) are all reference or equivalent methods; however, it is
not a requirement that the primary and collocated monitors utilize
the same specific sampling and analysis method.
Combined site data record is the data set used for performing
computations in this appendix and represents data for the primary
monitors augmented with data from collocated monitors according to
the procedure specified in section 3.0(a) of this appendix.
Daily value for PM10 refers to the 24-hour average
concentration of PM10 calculated or measured from
midnight to midnight (local time).
Exceedance means a daily value that is above the level of the
24-hour standard after rounding to the nearest 10 [micro]g/m\3\
(i.e., values ending in 5 or greater are to be rounded up).
Expected annual value is the number approached when the annual
values from an increasing number of years are averaged, in the
absence of long-term trends in emissions or meteorological
conditions.
Primary monitors are suitable monitors designated by a State or
local agency in their annual network plan as the default data source
for creating a combined site data record. If there is only one
suitable monitor at a particular site location, then it is presumed
to be a primary monitor.
Year refers to a calendar year.
* * * * *
2.3 Data Requirements
* * * * *
(d) 24-hour average concentrations will be computed from
submitted hourly PM10 concentration data for each
corresponding day of the year and the result will be stored in the
first, or start, hour (i.e., midnight, hour `0') of the 24-hour
period. A 24-hour average concentration shall be considered valid if
at least 75 percent of the hourly averages (i.e., 18 hourly values)
for the 24-hour period are available. In the event that fewer than
all 24 hourly average concentrations are available (i.e., fewer than
24 but at least 18), the 24-hour average concentration shall be
computed on the basis of the hours available using the number of
available hours within the 24-hour period as the divisor (e.g., the
divisor is 19 if 19 hourly values are available). 24-hour periods
with 7 or more missing hours shall also be considered for
computations in this appendix if, after substituting zero for all
missing hourly concentrations, the resulting 24-hour average daily
value exceeds the level of the 24-hour standard specified in Sec.
50.6 after rounding to the nearest 10 [micro]g/m\3\.
* * * * *
3.0 Computational Equations for the 24-Hour Standards
(a) All computations shown in this appendix shall be implemented
on a site-level basis. Site level concentration data shall be
processed as follows:
(1) The default dataset for PM10 mass concentrations
for a site shall consist of the measured concentrations recorded
from the designated primary monitor(s). All daily values produced by
the primary monitor are considered part of the site record.
(2) If a daily value is not produced by the primary monitor for
a particular day, but a value is available from a single collocated
monitor, then that collocated monitor value shall be considered part
of the combined site data record. If daily value data is available
from two or more collocated monitors, the average of those
collocated values shall be used as the daily value. The data record
resulting from this procedure is referred to as the ``combined site
data record.''
(b) In certain circumstances, including but not limited to site
closures or relocations, data from two nearby sites may be combined
into a single site data record for the purpose of calculating a
valid design value. The appropriate Regional Administrator may
approve such combinations if the Regional Administrator determines
that the measured concentrations do not differ substantially between
the two sites, taking into consideration factors such as distance
between sites, spatial and temporal patterns in air quality, local
emissions and meteorology, jurisdictional boundaries, and terrain
features.
* * * * *
0
4. Amend appendix L to part 50 by revising section 7.3.4 and adding
section 7.3.4.5 to read as follows:
Appendix L to Part 50--Reference Method for the Determination of Fine
Particulate Matter as PM[bdi2].[bdi5] in the Atmosphere
* * * * *
7.3.4 Particle size separator. The sampler shall be configured
with one of the three alternative particle size separators described
in this section. One separator is an impactor-type separator (WINS
impactor) described in sections 7.3.4.1, 7.3.4.2, and 7.3.4.3 of
this appendix. One alternative separator is a cyclone-type separator
(VSCC\TM\) described in section 7.3.4.4 of this appendix. The other
alternative separator is also a cyclone-type separator (TE-
PM2.5C) described in section 7.3.4.5 of this appendix.
* * * * *
7.3.4.5 A second cyclone-type separator is identified as a Tisch
TE-PM2.5C Cyclone particle size separator specified as
part of EPA-designated reference method RFPS-1014-219 and as
manufactured by Tisch Environmental Incorporated, 145 S. Miami
Avenue, Village of Cleves, Ohio 45002.
* * * * *
0
5. Amend appendix N to part 50 by:
0
a. In section 1.0 revising paragraph (a);
0
b. In section 3.0 adding paragraph (d)(3);
0
c. In section 4.1 revising paragraph (a); and
0
d. In section 4.2 revising paragraph (a).
The addition and revisions read as follows.
Appendix N to Part 50--Interpretation of the National Ambient Air
Quality Standards for PM[bdi2].[bdi5]
1.0 General
(a) This appendix explains the data handling conventions and
computations necessary for determining when the national ambient air
quality standards (NAAQS) for PM2.5 are met, specifically
the primary and secondary annual and 24-hour PM2.5 NAAQS
specified in Sec. Sec. 50.7, 50.13, 50.18, and 50.20.
PM2.5 is defined, in general terms, as particles with an
aerodynamic diameter less than or equal to a nominal 2.5
micrometers. PM2.5 mass concentrations are measured in
the ambient air by a Federal Reference Method (FRM) based on
appendix L to this part, as applicable, and designated in accordance
with part 53 of this chapter or by a Federal Equivalent Method (FEM)
designated in accordance with part 53 of this chapter. Only those
FRM and FEM measurements that are derived in accordance with part 58
of this chapter (i.e., that are deemed ``suitable'') shall be used
in comparisons with the PM2.5 NAAQS. The data handling
and computation procedures to be used to construct annual and 24-
hour NAAQS metrics from reported PM2.5 mass
concentrations, and the associated instructions for comparing these
calculated metrics to the levels of the PM2.5 NAAQS, are
specified in sections 2.0, 3.0, and 4.0 of this appendix.
* * * * *
3.0 Requirements for Data Use and Data Reporting for Comparisons With
the NAAQS for PM[bdi2].[bdi5]
* * * * *
(d) * * *
(3) In certain circumstances, including but not limited to site
closures or relocations, data from two nearby sites may be combined
into a single site data record for the purpose of calculating a
valid design value. The appropriate Regional Administrator may
approve such site combinations if the Regional Administrator
determines that the measured concentrations do not differ
substantially between the two sites, taking into consideration
factors such as distance between sites, spatial and temporal
patterns in air quality, local emissions and meteorology,
jurisdictional boundaries, and terrain features.
* * * * *
4.1 Annual PM[bdi2].[bdi5] NAAQS
(a) Levels of the primary and secondary annual PM2.5
NAAQS are specified in Sec. Sec. 50.7, 50.13, 50.18, and 50.20 as
applicable.
* * * * *
4.2 Twenty-Four-Hour PM[bdi2].[bdi5] NAAQS
(a) Levels of the primary and secondary 24-hour PM2.5
NAAQS are specified in Sec. Sec. 50.7, 50.13, 50.18, and 50.20 as
applicable.
* * * * *
[[Page 16382]]
PART 53--AMBIENT AIR MONITORING REFERENCE AND EQUIVALENT METHODS
0
6. The authority citation for part 53 continues to read as follows:
Authority: Sec. 301(a) of the Clean Air Act (42 U.S.C.
1857g(a)), as amended by sec. 15(c)(2) of Pub. L. 91-604, 84 Stat.
1713, unless otherwise noted.
Subpart A--General Provisions
0
7. Amend Sec. 53.4 by:
0
a. Revising paragraph (a);
0
b. Adding paragraph (b)(7); and
0
c. Revising paragraph (d).
The revisions and addition read as follows:
Sec. 53.4 Applications for reference or equivalent method
determinations.
(a) Applications for FRM or FEM determinations and modification
requests of existing designated instruments shall be submitted to: U.S.
Environmental Protection Agency, Director, Center for Environmental
Measurement and Modeling, Reference and Equivalent Methods Designation
Program (MD-D205-03), 109 T.W. Alexander Drive, P.O. Box 12055,
Research Triangle Park, North Carolina 27711 (commercial delivery
address: 4930 Old Page Road, Durham, North Carolina 27703).
* * * * *
(b) * * *
(7) All written materials for new FRM and FEM applications and
modification requests must be submitted in English in MS Word format.
For any calibration certificates originally written in a non-English
language, the original non-English version of the certificate must be
submitted to EPA along with a version of the certificate translated to
English. All laboratory and field data associated with new FRM and FEM
applications and modification requests must be submitted in MS Excel
format. All worksheets in MS Excel must be unprotected to enable full
inspection as part of the application review process.
* * * * *
(d) For candidate reference or equivalent methods or for designated
instruments that are the subject of a modification request, the
applicant, if requested by EPA, shall provide to EPA a representative
sampler or analyzer for test purposes. The sampler or analyzer shall be
shipped free on board (FOB) destination to Director, Center for
Environmental Measurements and Modeling, Reference and Equivalent
Methods Designation Program (MD D205-03), U.S. Environmental Protection
Agency, 4930 Old Page Road, Durham, North Carolina 27703, scheduled to
arrive concurrently with or within 30 days of the arrival of the other
application materials. This sampler or analyzer may be subjected to
various tests that EPA determines to be necessary or appropriate under
Sec. 53.5(f), and such tests may include special tests not described
in this part. If the instrument submitted under this paragraph (d)
malfunctions, becomes inoperative, or fails to perform as represented
in the application before the necessary EPA testing is completed, the
applicant shall be afforded the opportunity to repair or replace the
device at no cost to the EPA. Upon completion of EPA testing, the
sampler or analyzer submitted under this paragraph (d) shall be
repacked by EPA for return shipment to the applicant, using the same
packing materials used for shipping the instrument to EPA unless
alternative packing is provided by the applicant. Arrangements for, and
the cost of, return shipment shall be the responsibility of the
applicant. The EPA does not warrant or assume any liability for the
condition of the sampler or analyzer upon return to the applicant.
0
8. Amend Sec. 53.8 by revising paragraph (a) to read as follows:
Sec. 53.8 Designation of reference and equivalent methods.
(a) A candidate method determined by the Administrator to satisfy
the applicable requirements of this part shall be designated as an FRM
or FEM (as applicable) by and upon publication of the designation in
the Federal Register. Applicants shall not publicly announce, market,
or sell the candidate sampler and analyzer as an approved FRM or FEM
(as applicable) until the designation is published in the Federal
Register.
* * * * *
0
9. Amend Sec. 53.14 by revising paragraphs (c)(4), (5), and (6) to
read as follows:
Sec. 53.14 Modification of a reference or equivalent method.
* * * * *
(c) * * *
(4) Send notice to the applicant that additional information must
be submitted before a determination can be made and specify the
additional information that is needed (in such cases, the 90-day period
shall commence upon receipt of the additional information).
(5) Send notice to the applicant that additional tests are
necessary and specify which tests are necessary and how they shall be
interpreted (in such cases, the 90-day period shall commence upon
receipt of the additional test data).
(6) Send notice to the applicant that additional tests will be
conducted by the Administrator and specify the reasons for and the
nature of the additional tests (in such cases, the 90-day period shall
commence 1 calendar day after the additional tests are completed).
* * * * *
0
10. Revise table A-1 to subpart A of part 53 to read as follows:
Table A-1 to Subpart A of Part 53--Summary of Applicable Requirements for Reference and Equivalent Methods for Air Monitoring of Criteria Pollutants
--------------------------------------------------------------------------------------------------------------------------------------------------------
Applicable Applicable subparts of this part
Pollutant Reference or Manual or automated appendix of part 50 -----------------------------------------------------------------------
equivalent of this chapter A B C D E F
--------------------------------------------------------------------------------------------------------------------------------------------------------
SO2.............. Reference.......... Manual............. A-2
Automated.......... A-1 [check] [check]
Equivalent......... Manual............. A-1 [check] .......... [check]
Automated.......... A-1 [check] [check] [check]
CO............... Reference.......... Automated.......... C [check] [check]
Equivalent......... Manual............. C [check] .......... [check]
Automated.......... C [check] [check] [check]
O3............... Reference.......... Automated.......... D [check] [check]
Equivalent......... Manual............. D [check] .......... [check]
Automated.......... D [check] [check] [check]
NO2.............. Reference.......... Automated.......... F [check] [check]
Equivalent......... Manual............. F [check] .......... [check]
Automated.......... F [check] [check] [check]
[[Page 16383]]
Pb............... Reference.......... Manual............. G
Equivalent......... Manual............. G [check] .......... [check]
Automated.......... G [check] .......... [check]
PM10-Pb.......... Reference.......... Manual............. Q
Equivalent......... Manual............. Q [check] .......... [check]
Automated.......... Q [check] .......... [check]
PM10............. Reference.......... Manual............. J [check] .......... .......... [check]
Equivalent......... Manual............. J [check] .......... [check] [check]
Automated.......... J [check] .......... [check] [check]
PM2.5............ Reference.......... Manual............. L [check] .......... .......... .......... [check]
Equivalent Class I. Manual............. L [check] .......... [check] .......... [check]
Equivalent Class II Manual............. L \1\ [check] .......... \2\ .......... [check] 1 2
[check] [check]
Equivalent Class Automated.......... L \1\ [check] .......... [check] .......... [check] \1\
III. [check]
PM10-2.5......... Reference.......... Manual............. L,\2\ O [check] .......... .......... .......... [check]
Equivalent Class I. Manual............. L,\2\ O [check] .......... [check] .......... [check]
Equivalent Class II Manual............. L,\2\ O [check] .......... \2\ .......... [check] \1,2\
[check] [check]
Equivalent Class Automated.......... \1\ L, 1 2 O [check] .......... [check] .......... [check] \1\
III. [check]
--------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ Some requirements may apply, based on the nature of each particular candidate method, as determined by the Administrator.
\2\ Alternative Class III requirements may be substituted.
Subpart B--Procedures for Testing Performance Characteristics of
Automated Methods for SO2, CO, O3, and NO2
0
11. Amend table B-1 to subpart B of part 53 by revising footnote 4 to
read as follows:
Table B-1 to Subpart B of Part 53--Performance Limit Specifications for
Automated Methods
* * * * *
\4\ For nitric oxide interference for the SO2
ultraviolet fluorescence (UVF) method, interference equivalent is
0.003 ppm for the lower range.
* * * * *
0
12. Revise table B-3 to subpart B of part 53 to read as follows:
[[Page 16384]]
Table B-3 to Subpart B of Part 53--Interferent Test Concentration 1
[Parts per million]
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Hydro-
Pollutant Analyzer type chloric Ammonia Hydrogen Sulfur Nitrogen Nitric Carbon Ethylene Ozone M-xylene Water Carbon Methane Ethane Naphthalene
\2\ acid sulfide dioxide dioxide oxide dioxide vapor monoxide
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
SO2........... Ultraviolet ........ ........ \5\ 0.1 \4\ 0.14 0.5 0.5 ......... ......... 0.5 0.2 20,000 ......... ........ ........ \6\ 0.05
fluorescence.
SO2........... Flame ........ ........ 0.01 \4\ 0.14 ......... ........ 750 ......... ........ ......... \3\ 50 ........ ........ ............
photometric. 20,000
SO2........... Gas ........ ........ 0.1 \4\ 0.14 ......... ........ 750 ......... ........ ......... \3\ 50 ........ ........ ............
chromatography. 20,000
SO2........... Spectrophotometr 0.2 0.1 0.1 \4\ 0.14 0.5 ........ 750 ......... 0.5 ......... ........ ......... ........ ........ ............
ic-wet chemical
(pararosanaline
).
SO2........... Electrochemical. 0.2 0.1 0.1 \4\ 0.14 0.5 0.5 ......... 0.2 0.5 ......... \3\ ......... ........ ........ ............
20,000
SO2........... Conductivity.... 0.2 0.1 ......... \4\ 0.14 0.5 ........ 750 ......... ........ ......... ........ ......... ........ ........ ............
SO2........... Spectrophotometr ........ ........ ......... \4\ 0.14 0.5 0.5 ......... ......... 0.5 0.2 ........ ......... ........ ........ ............
ic-gas phase,
including DOAS.
O3............ Ethylene ........ ........ \3\ 0.1 ........ ......... ........ 750 ......... \4\ 0.08 ......... \3\ ......... ........ ........ ............
Chemiluminescen 20,000
ce.
O3............ NO- ........ ........ \3\ 0.1 ........ 0.5 ........ 750 ......... \4\ 0.08 ......... \3\ ......... ........ ........ ............
chemiluminescen 20,000
ce.
O3............ Electrochemical. ........ \3\ 0.1 ......... 0.5 0.5 ........ ......... ......... \4\ 0.08 ......... \3\ ......... ........ ........ ............
20,000
O3............ Spectrophotometr ........ \3\ 0.1 ......... 0.5 0.5 \3\ 0.5 ......... ......... \4\ 0.08 ......... ........ ......... ........ ........ ............
ic-wet chemical
(potassium
iodide).
O3............ Spectrophotometr ........ ........ ......... 0.5 0.5 \3\ 0.5 ......... ......... \4\ 0.08 0.02 20,000 ......... ........ ........ ............
ic-gas phase,
including
ultraviolet
absorption and
DOAS.
CO............ Non-dispersive ........ ........ ......... ........ ......... ........ 750 ......... ........ ......... 20,000 \4\ 10 ........ ........ ............
Infrared.
CO............ Gas ........ ........ ......... ........ ......... ........ ......... ......... ........ ......... 20,000 \4\ 10 ........ 0.5 ............
chromatography
with flame
ionization
detector.
CO............ Electrochemical. ........ ........ ......... ........ ......... 0.5 ......... 0.2 ........ ......... 20,000 \4\ 10 ........ ........ ............
CO............ Catalytic ........ 0.1 ......... ........ ......... ........ 750 0.2 ........ ......... 20,000 \4\ 10 5.0 0.5 ............
combustion-
thermal
detection.
CO............ IR fluorescence. ........ ........ ......... ........ ......... ........ 750 ......... ........ ......... 20,000 \4\ 10 ........ 0.5 ............
CO............ Mercury ........ ........ ......... ........ ......... ........ ......... 0.2 ........ ......... ........ \4\ 10 ........ 0.5 ............
replacement-UV
photometric.
NO2........... Chemiluminescent ........ \3\ 0.1 ......... 0.5 \4\ 0.1 0.5 ......... ......... ........ ......... 20,000 ......... ........ ........ ............
NO2........... Spectrophotometr ........ ........ ......... 0.5 \4\ 0.1 0.5 750 ......... 0.5 ......... ........ ......... ........ ........ ............
ic-wet chemical
(azo-dye
reaction).
NO2........... Electrochemical. 0.2 \3\ 0.1 ......... 0.5 \4\ 0.1 0.5 750 ......... 0.5 ......... 20,000 50 ........ ........ ............
NO2........... Spectrophotometr ........ \3\ 0.1 ......... 0.5 \4\ 0.1 0.5 ......... ......... 0.5 ......... 20,000 50 ........ ........ ............
ic-gas phase.
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ Concentrations of interferent listed must be prepared and controlled to 10 percent of the stated value.
\2\ Analyzer types not listed will be considered by the Administrator as special cases.
\3\ Do not mix interferent with the pollutant.
\4\ Concentration of pollutant used for test. These pollutant concentrations must be prepared to 10 percent of the stated value.
\5\ If candidate method utilizes an elevated-temperature scrubber for removal of aromatic hydrocarbons, perform this interference test.
\6\ If naphthalene test concentration cannot be accurately quantified, remove the scrubber, use a test concentration that causes a full-scale response, reattach the scrubber, and evaluate
response for interference.
[[Page 16385]]
0
13. Amend appendix A to subpart B of part 53 by revising figures B-3
and B-5 to read as follows:
Appendix A to Subpart B of Part 53--Optional Forms for Reporting Test
Results
* * * * *
Figure B-3 to Appendix A to Subpart B of Part 53--Form for Test Data
and Calculations for Lower Detectable Limit (LDL) and Interference
Equivalent (IE) (see Sec. 53.23(c) and (d))
LDL Interference Test Data
Applicant--------------------------------------------------------------
Analyzer---------------------------------------------------------------
Date-------------------------------------------------------------------
Pollutant--------------------------------------------------------------
[GRAPHIC] [TIFF OMITTED] TR06MR24.035
* * * * *
Figure B-5 to Appendix A to Subpart B of Part 53--Form for Calculating
Zero Drift, Span Drift and Precision (see Sec. 53.23(e))
Calculation of Zero Drift, Span Drift, and Precision
Applicant--------------------------------------------------------------
Analyzer---------------------------------------------------------------
Date-------------------------------------------------------------------
Pollutant--------------------------------------------------------------
[GRAPHIC] [TIFF OMITTED] TR06MR24.036
[[Page 16386]]
* * * * *
Subpart C--Procedures for Determining Comparability Between
Candidate Methods and Reference Methods
0
14. Amend Sec. 53.35 by revising paragraph (b)(1)(ii)(D) to read as
follows:
Sec. 53.35 Test procedure for Class II and Class III methods for
PM2.5 and PM10-2.5.
* * * * *
(b) * * *
(1) * * *
(ii) * * *
(D) Site D shall be in a large city east of the Mississippi River,
having characteristically high humidity levels.
* * * * *
0
15. Revise table C-4 to subpart C of part 53 to read as follows:
Table C-4 to Subpart C of Part 53--Test Specifications for PM10, PM2.5, and PM10-2.5 Candidate Equivalent Methods
--------------------------------------------------------------------------------------------------------------------------------------------------------
PM2.5 PM10-2.5
Specification PM10 ---------------------------------------------------------------------------------------------------
Class I Class II Class III Class II Class III
--------------------------------------------------------------------------------------------------------------------------------------------------------
Acceptable concentration range 5-300............. 3-200............. 3-200............. 3-200............. 3-200............. 3-200.
(Rj), [micro]g/m\3\.
Minimum number of test sites.... 2................. 1................. 2................. 4................. 2................. 4.
Minimum number of candidate 3................. 3................. 3\1\.............. 3\1\.............. 3\1\.............. 3.\1\
method samplers or analyzers
per site.
Number of reference method 3................. 3................. 3\1\.............. 3\1\.............. 3\1\.............. 3.\1\
samplers per site.
Minimum number of acceptable
sample sets per site for PM10
methods:
Rj < 20 [micro]g/m\3\....... 3................. .................. .................. .................. .................. ..................
Rj > 20 [micro]g/m\3\....... 3................. .................. .................. .................. .................. ..................
Total................... 10................ .................. .................. .................. .................. ..................
Minimum number of acceptable
sample sets per site for PM2.5
and PM10-2.5 candidate
equivalent methods:
Rj < 15 [micro]g/m\3\ for 24- .................. 3................. 3................. 3................. 3................. 3.
hr or Rj < 8 [micro]g/m\3\
for 48-hr samples..
Rj > 15 [micro]g/m\3\ for 24- .................. 3................. 3................. 3................. 3................. 3.
hr or Rj > 8 [micro]g/m\3\
for 48-hr samples.
Each season................. .................. 10................ 23................ 23................ 23................ 23.
Total, each site........ .................. 10................ 23................ 23 (46 for two- 23................ 23 (46 for two-
season sites). season sites).
Precision of replicate reference 5 [mu]g/m\3\ or 2 [mu]g/m\3\ or 10%\2\............ 10%\2\............ 10%\2\............ 10%.\2\
method measurements, PRj or 7%.. 5%..
RPRj, respectively; RP for
Class II or III PM2.5 or PM10-
2.5, maximum.
Precision of PM2.5 or PM10-2.5 .................. .................. 10%\2\............ 15%\2\............ 15%\2\............ 15%.\2\
candidate method, CP, each site.
Slope of regression relationship 1 0.10 1 0.05 1 0.10 1 0.10 1 0.10 1 0.12.
Intercept of regression 0 5... 0 1... Between: 13.55-- Between: 15.05-- Between: 62.05-- Between: 70.50--
relationship, [micro]g/m\3\. (15.05 x slope), (17.32 x slope), (70.5 x slope), (82.93 x slope),
but not less but not less but not less but not less
than--1.5; and than--2.0; and than--3.5; and than--7.0; and
16.56--(15.05 x 15.05--(13.20 x 78.95--(70.5 x 70.50--(61.16 x
slope), but not slope), but not slope), but not slope), but not
more than +1.5. more than +2.0. more than +3.5. more than +7.0.
-------------------------------------------------------------------------------
Correlation of reference method >= 0.97........... >= 0.97........... >= 0.93--for CCV <= 0.4;
and candidate method >= 0.85 + 0.2 x CCV--for 0.4 <= CCV <= 0.5;
measurements. >= 0.95--for CCV >= 0.5
--------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ Some missing daily measurement values may be permitted; see test procedure.
\2\ Calculated as the root mean square over all measurement sets.
[[Page 16387]]
Subpart D--Procedures for Testing Performance Characteristics of
Methods for PM10
0
16. Amend Sec. 53.43 by revising the formula in paragraph (a)(2)(xvi)
and the formula in paragraph (c)(2)(iv) to read as follows:
Sec. 53.43 Test procedures.
(a) * * *
(2) * * *
(xvi) * * *
[GRAPHIC] [TIFF OMITTED] TR06MR24.037
* * * * *
(c) * * *
(2) * * *
(iv) * * *
[GRAPHIC] [TIFF OMITTED] TR06MR24.038
if Cj is below 80 [micro]g/m\3\, or
[GRAPHIC] [TIFF OMITTED] TR06MR24.039
if Cj is above 80 [micro]g/m\3\.
Subpart E--Procedures for Testing Physical (Design) and Performance
Characteristics of Reference Methods and Class I and Class II
Equivalent Methods for PM2.5 or PM10-2.5
0
17. Amend Sec. 53.51 by revising paragraph (d)(2) to read as follows:
Sec. 53.51 Demonstration of compliance with design specifications and
manufacturing and test requirements.
* * * * *
(d) * * *
(2) VSCC and TE-PM2.5C separators. For samplers and
monitors utilizing the BGI VSCC or Tisch TE-PM2.5C particle
size separators specified in sections 7.3.4.4 and 7.3.4.5 of appendix L
to part 50 of this chapter, respectively, the respective manufacturers
shall identify the critical dimensions and manufacturing tolerances for
the separator, devise appropriate test procedures to verify that the
critical dimensions and tolerances are maintained during the
manufacturing process, and carry out those procedures on each separator
manufactured to verify conformance of the manufactured products. The
manufacturer shall also maintain records of these tests and their test
results and submit evidence that this procedure is incorporated into
the manufacturing procedure, that the test is or will be routinely
implemented, and that an appropriate procedure is in place for the
disposition of units that fail this tolerance tests.
* * * * *
Subpart F--Procedures for Testing Performance Characteristics of
Class II Equivalent Methods for PM2.5
0
18. Amend Sec. 53.61 by revising paragraph (g) introductory text, the
first sentence of paragraph (g)(1), the first sentence of (g)(1)(i),
(g)(2)(i) and adding paragraph (g)(2)(iii) to read as follows:
Sec. 53.61 Test conditions.
* * * * *
(g) Vibrating Orifice Aerosol Generator (VOAG) and Flow-Focusing
Monodisperse Aerosol Generator (FMAG) conventions. This section
prescribes conventions regarding the use of the vibrating orifice
aerosol generator (VOAG) and the flow-focusing monodisperse aerosol
generator (FMAG) for the size-selective performance tests outlined in
Sec. Sec. 53.62, 53.63, 53.64, and 53.65.
(1) Particle aerodynamic diameter. The VOAG and FMAG produce near-
monodisperse droplets through the controlled breakup of a liquid jet. *
* *
(i) The physical diameter of a generated spherical particle can be
calculated from the operational parameters of the VOAG and FMAG as:
* * * * *
(2) * * *
(i) Solid particle tests performed in this subpart shall be
conducted using particles composed of ammonium fluorescein. For use in
the VOAG or FMAG, liquid solutions of known volumetric concentration
can be prepared by diluting fluorescein powder
(C2OH12O5, FW = 332.31, CAS 2321-07-5)
with aqueous ammonia. Guidelines for preparation of fluorescein
solutions of the desired volume concentration (Cvol) are
presented in Vanderpool and Rubow (1988) (Reference 2 in appendix A to
this subpart). For purposes of converting particle physical diameter to
aerodynamic diameter, an ammonium fluorescein particle density of 1.35
g/cm\3\ shall be used.
* * * * *
(iii) Calculation of the physical diameter of the particles
produced by the VOAG and FMAG requires
[[Page 16388]]
knowledge of the liquid solution's volume concentration
(Cvol). Because uranine is essentially insoluble in oleic
acid, the total particle volume is the sum of the oleic acid volume and
the uranine volume. The volume concentration of the liquid solution
shall be calculated as:
[GRAPHIC] [TIFF OMITTED] TR06MR24.040
Where:
Vu = uranine volume, ml;
Voleic = oleic acid volume, ml;
Vsol = total solution volume, ml;
Mu = uranine mass, g;
Pu = uranine density, g/cm\3\;
Moleic = oleic acid mass, g; and
Poleic = oleic acid density, g/cm\3\.
* * * * *
PART 58--AMBIENT AIR QUALITY SURVEILLANCE
0
19. The authority citation for part 58 continues to read as follows:
Authority: 42 U.S.C. 7403, 7405, 7410, 7414, 7601, 7611, 7614,
and 7619.
Subpart A--General Provisions
0
20. Amend Sec. 58.1 by:
0
a. Removing the definition for ``Approved regional method (ARM)''; and
0
b. Revising the definition for ``Traceable.''
The revision reads as follows:
Sec. 58.1 Definitions.
* * * * *
Traceable means a measurement result from a local standard whereby
the result can be related to the International System of Units (SI)
through a documented unbroken chain of calibrations, each contributing
to the measurement uncertainty. Traceable measurement results must be
compared and certified, either directly or via not more than one
intermediate standard, to a National Institute of Standards and
Technology (NIST)-certified reference standard. Examples include but
are not limited to NIST Standard Reference Material (SRM), NIST-
traceable Reference Material (NTRM), or a NIST-certified Research Gas
Mixture (RGM). Traceability to the SI through other National Metrology
Institutes (NMIs) in addition to NIST is allowed if a Declaration of
Equivalence (DoE) exists between NIST and that NMI.
* * * * *
Subpart B--Monitoring Network
0
21. Amend Sec. 58.10 by:
0
a. Revising paragraphs (a)(1) and (b)(10) and (13);
0
b. Adding paragraph (b)(14); and
0
c. Revising paragraph (d).
The revisions and addition read as follows:
Sec. 58.10 Annual monitoring network plan and periodic network
assessment.
(a)(1) Beginning July 1, 2007, the State, or where applicable
local, agency shall submit to the Regional Administrator an annual
monitoring network plan which shall provide for the documentation of
the establishment and maintenance of an air quality surveillance system
that consists of a network of SLAMS monitoring stations that can
include FRM and FEM monitors that are part of SLAMS, NCore, CSN, PAMS,
and SPM stations. The plan shall include a statement of whether the
operation of each monitor meets the requirements of appendices A, B, C,
D, and E to this part, where applicable. The Regional Administrator may
require additional information in support of this statement. The annual
monitoring network plan must be made available for public inspection
and comment for at least 30 days prior to submission to the EPA and the
submitted plan shall include and address, as appropriate, any received
comments.
* * * * *
(b) * * *
(10) Any monitors for which a waiver has been requested or granted
by the EPA Regional Administrator as allowed for under appendix D or
appendix E to this part. For those monitors where a waiver has been
approved, the annual monitoring network plan shall include the date the
waiver was approved.
* * * * *
(13) The identification of any PM2.5 FEMs used in the
monitoring agency's network where the data are not of sufficient
quality such that data are not to be compared to the national ambient
air quality standards (NAAQS). For required SLAMS where the agency
identifies that the PM2.5 Class III FEM does not produce
data of sufficient quality for comparison to the NAAQS, the monitoring
agency must ensure that an operating FRM or filter-based FEM meeting
the sample frequency requirements described in Sec. 58.12 or other
Class III PM2.5 FEM with data of sufficient quality is
operating and reporting data to meet the network design criteria
described in appendix D to this part.
(14) The identification of any site(s) intended to address being
sited in an at-risk community where there are anticipated effects from
sources in the area as required in section 4.7.1(b)(3) of appendix D to
this part. An initial approach to the question of whether any new or
moved sites are needed and to identify the communities in which they
intend to add monitoring for meeting the requirement in this paragraph
(b)(14), if applicable, shall be submitted in accordance with the
requirements of section 4.7.1(b)(3) of appendix D to this part, which
includes submission to the EPA Regional Administrator no later than
July 1, 2024. Specifics on the resulting proposed new or moved sites
for PM2.5 network design to address at-risk communities, if
applicable, would need to be detailed in annual monitoring network
plans due to each applicable EPA Regional office no later than July 1,
2025. The plan shall provide for any required sites to be operational
no later than 24 months from date of approval of a plan or January 1,
2027, whichever comes first.
* * * * *
(d) The State, or where applicable local, agency shall perform and
submit to the EPA Regional Administrator an assessment of the air
quality surveillance system every 5 years to determine, at a minimum,
if the network meets the monitoring objectives defined in appendix D to
this part, whether new sites are needed, whether existing sites are no
longer needed and can be terminated, and whether new technologies are
appropriate for incorporation into the ambient air monitoring network.
The network assessment must consider the ability of existing and
proposed sites to support air quality characterization for areas with
relatively high populations of
[[Page 16389]]
susceptible individuals (e.g., children with asthma) and other at-risk
populations, and, for any sites that are being proposed for
discontinuance, the effect on data users other than the agency itself,
such as nearby States and Tribes or health effects studies. The State,
or where applicable local, agency must submit a copy of this 5-year
assessment, along with a revised annual network plan, to the Regional
Administrator. The assessments are due every 5 years beginning July 1,
2010.
* * * * *
0
22. Amend Sec. 58.11 by revising paragraphs (a)(2) and (e) to read as
follows:
Sec. 58.11 Network technical requirements.
(a) * * *
(2) Beginning January 1, 2009, State and local governments shall
follow the quality assurance criteria contained in appendix A to this
part that apply to SPM sites when operating any SPM site which uses an
FRM or an FEM and meets the requirements of appendix E to this part,
unless the Regional Administrator approves an alternative to the
requirements of appendix A with respect to such SPM sites because
meeting those requirements would be physically and/or financially
impractical due to physical conditions at the monitoring site and the
requirements are not essential to achieving the intended data
objectives of the SPM site. Alternatives to the requirements of
appendix A may be approved for an SPM site as part of the approval of
the annual monitoring plan, or separately.
* * * * *
(e) State and local governments must assess data from Class III
PM2.5 FEM monitors operated within their network using the
performance criteria described in table C-4 to subpart C of part 53 of
this chapter, for cases where the data are identified as not of
sufficient comparability to a collocated FRM, and the monitoring agency
requests that the FEM data should not be used in comparison to the
NAAQS. These assessments are required in the monitoring agency's annual
monitoring network plan described in Sec. 58.10(b) for cases where the
FEM is identified as not of sufficient comparability to a collocated
FRM. For these collocated PM2.5 monitors, the performance
criteria apply with the following additional provisions:
(1) The acceptable concentration range (Rj), [micro]g/m\3\ may
include values down to 0 [micro]g/m\3\.
(2) The minimum number of test sites shall be at least one;
however, the number of test sites will generally include all locations
within an agency's network with collocated FRMs and FEMs.
(3) The minimum number of methods shall include at least one FRM
and at least one FEM.
(4) Since multiple FRMs and FEMs may not be present at each site,
the precision statistic requirement does not apply, even if precision
data are available.
(5) All seasons must be covered with no more than 36 consecutive
months of data in total aggregated together.
(6) The key statistical metric to include in an assessment is the
bias (both additive and multiplicative) of the PM2.5
continuous FEM(s) compared to a collocated FRM(s). Correlation is
required to be reported in the assessment, but failure to meet the
correlation criteria, by itself, is not cause to exclude data from a
continuous FEM monitor.
0
23. Amend Sec. 58.12 by revising paragraph (d)(1):
Sec. 58.12 Operating schedules.
* * * * *
(d) * * *
(1)(i) Manual PM2.5 samplers at required SLAMS stations
without a collocated continuously operating PM2.5 monitor
must operate on at least a 1-in-3 day schedule unless a waiver for an
alternative schedule has been approved per paragraph (d)(1)(ii) of this
section.
(ii) For SLAMS PM2.5 sites with both manual and
continuous PM2.5 monitors operating, the monitoring agency
may request approval for a reduction to 1-in-6 day PM2.5
sampling or for seasonal sampling from the EPA Regional Administrator.
Other requests for a reduction to 1-in-6 day PM2.5 sampling
or for seasonal sampling may be approved on a case-by-case basis. The
EPA Regional Administrator may grant sampling frequency reductions
after consideration of factors (including but not limited to the
historical PM2.5 data quality assessments, the location of
current PM2.5 design value sites, and their regulatory data
needs) if the Regional Administrator determines that the reduction in
sampling frequency will not compromise data needed for implementation
of the NAAQS. Required SLAMS stations whose measurements determine the
design value for their area and that are within plus or minus 10
percent of the annual NAAQS, and all required sites where one or more
24-hour values have exceeded the 24-hour NAAQS each year for a
consecutive period of at least 3 years are required to maintain at
least a 1-in-3 day sampling frequency until the design value no longer
meets the criteria in this paragraph (d)(1)(ii) for 3 consecutive
years. A continuously operating FEM PM2.5 monitor satisfies
the requirement in this paragraph (d)(1)(ii) unless it is identified in
the monitoring agency's annual monitoring network plan as not
appropriate for comparison to the NAAQS and the EPA Regional
Administrator has approved that the data from that monitor may be
excluded from comparison to the NAAQS.
(iii) Required SLAMS stations whose measurements determine the 24-
hour design value for their area and whose data are within plus or
minus 5 percent of the level of the 24-hour PM2.5 NAAQS must
have an FRM or FEM operate on a daily schedule if that area's design
value for the annual NAAQS is less than the level of the annual
PM2.5 standard. A continuously operating FEM or
PM2.5 monitor satisfies the requirement in this paragraph
(d)(1)(iii) unless it is identified in the monitoring agency's annual
monitoring network plan as not appropriate for comparison to the NAAQS
and the EPA Regional Administrator has approved that the data from that
monitor may be excluded from comparison to the NAAQS. The daily
schedule must be maintained until the referenced design values no
longer meets the criteria in this paragraph (d)(1)(iii) for 3
consecutive years.
(iv) Changes in sampling frequency attributable to changes in
design values shall be implemented no later than January 1 of the
calendar year following the certification of such data as described in
Sec. 58.15.
* * * * *
0
24. Revise Sec. 58.15 to read as follows:
Sec. 58.15 Annual air monitoring data certification.
(a) The State, or where appropriate local, agency shall submit to
the EPA Regional Administrator an annual air monitoring data
certification letter to certify data collected by FRM and FEM monitors
at SLAMS and SPM sites that meet criteria in appendix A to this part
from January 1 to December 31 of the previous year. The head official
in each monitoring agency, or his or her designee, shall certify that
the previous year of ambient concentration and quality assurance data
are completely submitted to AQS and that the ambient concentration data
are accurate to the best of her or his knowledge, taking into
consideration the quality assurance findings. The annual data
certification letter is due by May 1 of each year.
(b) Along with each certification letter, the State shall submit to
the Regional Administrator an annual
[[Page 16390]]
summary report of all the ambient air quality data collected by FRM and
FEM monitors at SLAMS and SPM sites. The annual report(s) shall be
submitted for data collected from January 1 to December 31 of the
previous year. The annual summary serves as the record of the specific
data that is the object of the certification letter.
(c) Along with each certification letter, the State shall submit to
the Regional Administrator a summary of the precision and accuracy data
for all ambient air quality data collected by FRM and FEM monitors at
SLAMS and SPM sites. The summary of precision and accuracy shall be
submitted for data collected from January 1 to December 31 of the
previous year.
Subpart C--Special Purpose Monitors
0
25. Amend Sec. 58.20 by revising paragraphs (b) through (e) to read as
follows:
Sec. 58.20 Special purpose monitors (SPM).
* * * * *
(b) Any SPM data collected by an air monitoring agency using a
Federal reference method (FRM) or Federal equivalent method (FEM) must
meet the requirements of Sec. Sec. 58.11 and 58.12 and appendix A to
this part or an approved alternative to appendix A. Compliance with
appendix E to this part is optional but encouraged except when the
monitoring agency's data objectives are inconsistent with the
requirements in appendix E. Data collected at an SPM using a FRM or FEM
meeting the requirements of appendix A must be submitted to AQS
according to the requirements of Sec. 58.16. Data collected by other
SPMs may be submitted. The monitoring agency must also submit to AQS an
indication of whether each SPM reporting data to AQS monitor meets the
requirements of appendices A and E.
(c) All data from an SPM using an FRM or FEM which has operated for
more than 24 months are eligible for comparison to the relevant NAAQS,
subject to the conditions of Sec. Sec. 58.11(e) and 58.30, unless the
air monitoring agency demonstrates that the data came from a particular
period during which the requirements of appendix A, appendix C, or
appendix E to this part were not met, subject to review and EPA
Regional Office approval as part of the annual monitoring network plan
described in Sec. 58.10.
(d) If an SPM using an FRM or FEM is discontinued within 24 months
of start-up, the Administrator will not base a NAAQS violation
determination for the PM2.5 or ozone NAAQS solely on data
from the SPM.
(e) If an SPM using an FRM or FEM is discontinued within 24 months
of start-up, the Administrator will not designate an area as
nonattainment for the CO, SO2, NO2, or 24-hour
PM10 NAAQS solely on the basis of data from the SPM. Such
data are eligible for use in determinations of whether a nonattainment
area has attained one of these NAAQS.
* * * * *
0
26. Amend appendix A to part 58 by:
0
a. Revising section 2.6.1 and adding sections 2.6.1.1 and 2.6.1.2;
0
b. Removing section 3.1.2.2 and redesignating sections 3.1.2.3,
3.1.2.4, 3.1.2.5, and 3.1.2.6 as sections 3.1.2.2, 3.1.2.3, 3.1.2.4,
and 3.1.2.5, respectively;
0
c. Revising sections 3.1.3.3, 3.2.4, 4.2.1, and 4.2.5; and
0
d. In section 6 revising References (1), (4), (6), (7), (9), (10), and
(11) and table A-1.
The revisions and additions read as follows:
Appendix A to Part 58--Quality Assurance Requirements for Monitors used
in Evaluations of National Ambient Air Quality Standards
* * * * *
2.6.1 Gaseous pollutant concentration standards (permeation
devices or cylinders of compressed gas) used to obtain test
concentrations for CO, SO2, NO, and NO2 must
be EPA Protocol Gases certified in accordance with one of the
procedures given in Reference 4 of this appendix.
2.6.1.1 The concentrations of EPA Protocol Gas standards used
for ambient air monitoring must be certified with a 95-percent
confidence interval to have an analytical uncertainty of no more
than 2.0 percent (inclusive) of the certified
concentration (tag value) of the gas mixture. The uncertainty must
be calculated in accordance with the statistical procedures defined
in Reference 4 of this appendix.
2.6.1.2 Specialty gas producers advertising certification with
the procedures provided in Reference 4 of this appendix and
distributing gases as ``EPA Protocol Gas'' for ambient air
monitoring purposes must adhere to the regulatory requirements
specified in 40 CFR 75.21(g) or not use ``EPA'' in any form of
advertising. Monitoring organizations must provide information to
the EPA on the specialty gas producers they use on an annual basis.
PQAOs, when requested by the EPA, must participate in the EPA
Ambient Air Protocol Gas Verification Program at least once every 5
years by sending a new unused standard to a designated verification
laboratory.
* * * * *
3.1.3.3 Using audit gases that are verified against the NIST
standard reference methods or special review procedures and
validated per the certification periods specified in Reference 4 of
this appendix (EPA Traceability Protocol for Assay and Certification
of Gaseous Calibration Standards) for CO, SO2, and
NO2 and using O3 analyzers that are verified
quarterly against a standard reference photometer.
* * * * *
3.2.4 PM2.5 Performance Evaluation Program (PEP)
Procedures. The PEP is an independent assessment used to estimate
total measurement system bias. These evaluations will be performed
under the national performance evaluation program (NPEP) as
described in section 2.4 of this appendix or a comparable program. A
prescribed number of Performance evaluation sampling events will be
performed annually within each PQAO. For PQAOs with less than or
equal to five monitoring sites, five valid performance evaluation
audits must be collected and reported each year. For PQAOs with
greater than five monitoring sites, eight valid performance
evaluation audits must be collected and reported each year. A valid
performance evaluation audit means that both the primary monitor and
PEP audit concentrations are valid and equal to or greater than 2
[micro]g/m3. Siting of the PEP monitor must be consistent with
section 3.2.3.4(c) of this appendix. However, any horizontal
distance greater than 4 meters and any vertical distance greater
than one meter must be reported to the EPA regional PEP coordinator.
Additionally for every monitor designated as a primary monitor, a
primary quality assurance organization must:
* * * * *
4.2.1 Collocated Quality Control Sampler Precision Estimate for
PM10, PM2.5, and Pb. Precision is estimated
via duplicate measurements from collocated samplers. It is
recommended that the precision be aggregated at the PQAO level
quarterly, annually, and at the 3-year level. The data pair would
only be considered valid if both concentrations are greater than or
equal to the minimum values specified in section 4(c) of this
appendix. For each collocated data pair, calculate ti,
using equation 6 to this appendix:
[[Page 16391]]
[GRAPHIC] [TIFF OMITTED] TR06MR24.041
Where Xi is the concentration from the primary
sampler and Yi is the concentration value from the audit
sampler. The coefficient of variation upper bound is calculated
using equation 7 to this appendix:
[GRAPHIC] [TIFF OMITTED] TR06MR24.042
Where k is the number of valid data pairs being aggregated, and
X\2\0.1,k-1 is the 10th percentile of a chi-squared
distribution with k-1 degrees of freedom. The factor of 2 in the
denominator adjusts for the fact that each ti is
calculated from two values with error.
* * * * *
4.2.5 Performance Evaluation Programs Bias Estimate for
PM2.5. The bias estimate is calculated using the PEP
audits described in section 3.2.4. of this appendix. The bias
estimator is based on, si, the absolute difference in
concentrations divided by the square root of the PEP concentration.
[GRAPHIC] [TIFF OMITTED] TR06MR24.043
* * * * *
6. References
(1) American National Standard Institute--Quality Management Systems
For Environmental Information And Technology Programs--Requirements
With Guidance For Use. ASQ/ANSI E4-2014. February 2014. Available
from ANSI Webstore https://webstore.ansi.org/.
* * * * *
(4) EPA Traceability Protocol for Assay and Certification of Gaseous
Calibration Standards. EPA-600/R-12/531. May, 2012. Available from
U.S. Environmental Protection Agency, National Risk Management
Research Laboratory, Research Triangle Park NC 27711. https://www.epa.gov/nscep.
* * * * *
(6) List of Designated Reference and Equivalent Methods. Available
from U.S. Environmental Protection Agency, Center for Environmental
Measurements and Modeling, Air Methods and Characterization
Division, MD-D205-03, Research Triangle Park, NC 27711. https://www.epa.gov/amtic/air-monitoring-methods-criteria-pollutants.
(7) Transfer Standards for the Calibration of Ambient Air Monitoring
Analyzers for Ozone. EPA-454/B-13-004 U.S. Environmental Protection
Agency, Research Triangle Park, NC 27711, October, 2013. https://www.epa.gov/sites/default/files/2020-09/documents/ozonetransferstandardguidance.pdf.
* * * * *
(9) Quality Assurance Handbook for Air Pollution Measurement
Systems, Volume 1--A Field Guide to Environmental Quality Assurance.
EPA-600/R-94/038a. April 1994. Available from U.S. Environmental
Protection Agency, ORD Publications Office, Center for Environmental
Research Information (CERI), 26 W. Martin Luther King Drive,
Cincinnati, OH 45268. https://www.epa.gov/amtic/ambient-air-monitoring-quality-assurance#documents.
(10) Quality Assurance Handbook for Air Pollution Measurement
Systems, Volume II: Ambient Air Quality Monitoring Program Quality
System Development. EPA-454/B-13-003. https://www.epa.gov/amtic/ambient-air-monitoring-quality-assurance#documents.
(11) National Performance Evaluation Program Standard Operating
Procedures. https://www.epa.gov/amtic/ambient-air-monitoring-quality-assurance#npep.
Table A-1 to Section 6 of Appendix A--Minimum Data Assessment Requirements for NAAQS Related Criteria Pollutant Monitors
--------------------------------------------------------------------------------------------------------------------------------------------------------
Method Assessment method Coverage Minimum frequency Parameters reported AQS assessment type
--------------------------------------------------------------------------------------------------------------------------------------------------------
Gaseous Methods (CO, NO2, SO2, O3):
[[Page 16392]]
One-Point QC for SO2, NO2, O3, Response check at Each analyzer......... Once per 2 weeks \5\. Audit concentration One-Point QC.
CO. concentration 0.005- \1\ and measured
0.08 ppm SO2, NO2, concentration.\2\.
O3, and.
0.5 and 5 ppm CO......
Annual performance evaluation for See section 3.1.2 of Each analyzer......... Once per year........ Audit concentration Annual PE.
SO2, NO2, O3, CO. this appendix. \1\ and measured
concentration \2\
for each level.
NPAP for SO2, NO2, O3, CO.......... Independent Audit..... 20% of sites each year Once per year........ Audit NPAP.
concentration\1\ and
measured
concentration \2\
for each level.
Particulate Methods:
Continuous \4\ method-- Collocated samplers... 15%................... 1-in-12 days......... Primary sampler No Transaction
collocated quality control concentration and reported as raw
sampling PM2.5. duplicate sampler data.
concentration.\3\.
Manual method--collocated Collocated samplers... 15%................... 1-in-12 days......... Primary sampler No Transaction
quality control sampling PM10, concentration and reported as raw
PM2.5, Pb-TSP, Pb-PM10. duplicate sampler data.
concentration.\3\.
Flow rate verification PM10 Check of sampler flow Each sampler.......... Once every month \5\. Audit flow rate and Flow Rate
(low Vol) PM2.5, Pb-PM10. rate. measured flow rate Verification.
indicated by the
sampler.
Flow rate verification PM10 Check of sampler flow Each sampler.......... Once every quarter Audit flow rate and Flow Rate
(High-Vol), Pb-TSP. rate. \5\. measured flow rate Verification.
indicated by the
sampler.
Semi-annual flow rate audit Check of sampler flow Each sampler.......... Once every 6 months Audit flow rate and Semi Annual Flow Rate
PM10, TSP, PM10-2.5, PM2.5, Pb- rate using \5\. measured flow rate Audit.
TSP, Pb-PM10. independent standard. indicated by the
sampler.
Pb analysis audits Pb-TSP, Pb- Check of analytical Analytical............ Once each quarter \5\ Measured value and Pb Analysis Audits.
PM10. system with Pb audit audit value (ug Pb/
strips/filters. filter) using AQS
unit code 077.
Performance Evaluation Program Collocated samplers... (1) 5 valid audits for Distributed over all Primary sampler PEP.
PM2.5. primary QA orgs, with 4 quarters \5\. concentration and
<=5 sites. performance
(2) 8 valid audits for evaluation sampler
primary QA orgs, with concentration.
>5 sites.
(3) All samplers in 6
years.
Performance Evaluation Program Collocated samplers... (1) 1 valid audit and Distributed over all Primary sampler PEP.
Pb-TSP, Pb-PM10. 4 collocated samples 4 quarters \5\. concentration and
for primary QA orgs, performance
with <=5 sites. evaluation sampler
(2) 2 valid audits and concentration.
6 collocated samples Primary sampler
for primary QA orgs concentration and
with >5 sites. duplicate sampler
concentration.
--------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ Effective concentration for open path analyzers.
\2\ Corrected concentration, if applicable for open path analyzers.
\3\ Both primary and collocated sampler values are reported as raw data.
\4\ PM2.5 is the only particulate criteria pollutant requiring collocation of continuous and manual primary monitors.
\5\ EPA's recommended maximum number of days that should exist between checks to ensure that the checks are routinely conducted over time and to limit
data impacts resulting from a failed check.
* * * * *
0
27. Amend appendix B to part 58 by:
0
a. Revising section 2.6.1 and adding sections 2.6.1.1 and 2.6.1.2;
0
b. Removing and reserving section 3.1.2.2;
[[Page 16393]]
0
c. Revising sections 3.1.3.3 and 3.2.4;
0
d. Adding sections 3.2.4.1 through 3.2.4.3;
0
e. Revising sections 4.2.1, and 4.2.5; and
0
f. In section 6 revising References (1), (4), (6), (7), (9), (10), and
(11) and table B-1.
The revisions and additions read as follows:
Appendix B to Part 58--Quality Assurance Requirements for Prevention of
Significant Deterioration (PSD) Air Monitoring
* * * * *
2.6.1 Gaseous pollutant concentration standards (permeation devices
or cylinders of compressed gas) used to obtain test concentrations for
CO, SO2, NO, and NO2 must be EPA Protocol Gases
certified in accordance with one of the procedures given in Reference 4
of this appendix.
2.6.1.1 The concentrations of EPA Protocol Gas standards used for
ambient air monitoring must be certified with a 95-percent confidence
interval to have an analytical uncertainty of no more than 2.0 percent (inclusive) of the certified concentration (tag
value) of the gas mixture. The uncertainty must be calculated in
accordance with the statistical procedures defined in Reference 4 of
this appendix.
2.6.1.2 Specialty gas producers advertising certification with the
procedures provided in Reference 4 of this appendix and distributing
gases as ``EPA Protocol Gas'' for ambient air monitoring purposes must
adhere to the regulatory requirements specified in 40 CFR 75.21(g) or
not use ``EPA'' in any form of advertising. The PSD PQAOs must provide
information to the PSD reviewing authority on the specialty gas
producers they use (or will use) for the duration of the PSD monitoring
project. This information can be provided in the QAPP or monitoring
plan but must be updated if there is a change in the specialty gas
producers used.
* * * * *
3.1.3.3 Using audit gases that are verified against the NIST
standard reference methods or special review procedures and validated
per the certification periods specified in Reference 4 of this appendix
(EPA Traceability Protocol for Assay and Certification of Gaseous
Calibration Standards) for CO, SO2, and NO2 and
using O3 analyzers that are verified quarterly against a
standard reference photometer.
* * * * *
3.2.4 PM2.5 Performance Evaluation Program (PEP)
Procedures. The PEP is an independent assessment used to estimate total
measurement system bias. These evaluations will be performed under the
NPEP as described in section 2.4 of this appendix or a comparable
program. Performance evaluations will be performed annually within each
PQAO. For PQAOs with less than or equal to five monitoring sites, five
valid performance evaluation audits must be collected and reported each
year. For PQAOs with greater than five monitoring sites, eight valid
performance evaluation audits must be collected and reported each year.
A valid performance evaluation audit means that both the primary
monitor and PEP audit concentrations are valid and equal to or greater
than 2 [micro]g/m3. Siting of the PEP monitor must be consistent with
section 3.2.3.4(c) of this appendix. However, any horizontal distance
greater than 4 meters and any vertical distance greater than one meter
must be reported to the EPA regional PEP coordinator. Additionally for
every monitor designated as a primary monitor, a primary quality
assurance organization must:
3.2.4.1 Have each method designation evaluated each year; and,
3.2.4.2 Have all FRM and FEM samplers subject to a PEP audit at
least once every 6 years, which equates to approximately 15 percent of
the monitoring sites audited each year.
3.2.4.3 Additional information concerning the PEP is contained in
Reference 10 of this appendix. The calculations for evaluating bias
between the primary monitor and the performance evaluation monitor for
PM2.5 are described in section 4.2.5 of this appendix.
* * * * *
4.2.1 Collocated Quality Control Sampler Precision Estimate for
PM10, PM2.5, and Pb. Precision is estimated via
duplicate measurements from collocated samplers. It is recommended that
the precision be aggregated at the PQAO level quarterly, annually, and
at the 3-year level. The data pair would only be considered valid if
both concentrations are greater than or equal to the minimum values
specified in section 4(c) of this appendix. For each collocated data
pair, calculate ti, using equation 6 to this appendix:
[GRAPHIC] [TIFF OMITTED] TR06MR24.044
Where Xi is the concentration from the primary sampler
and Yi is the concentration value from the audit sampler.
The coefficient of variation upper bound is calculated using equation 7
to this appendix:
[GRAPHIC] [TIFF OMITTED] TR06MR24.045
Where k is the number of valid data pairs being aggregated, and
X\2\0.1,k-1 is the 10th percentile of a chi-squared
distribution with k-1 degrees of freedom. The factor of 2 in the
denominator adjusts for the fact that
[[Page 16394]]
each ti is calculated from two values with error.
* * * * *
4.2.5 Performance Evaluation Programs Bias Estimate for
PM2.5. The bias estimate is calculated using the PEP audits
described in section 3.2.4. of this appendix. The bias estimator is
based on, si, the absolute difference in concentrations
divided by the square root of the PEP concentration.
[GRAPHIC] [TIFF OMITTED] TR06MR24.046
* * * * *
6. References
(1) American National Standard Institute--Quality Management Systems
For Environmental Information And Technology Programs--Requirements
With Guidance For Use. ASQ/ANSI E4-2014. February 2014. Available
from ANSI Webstore https://webstore.ansi.org/.
* * * * *
(4) EPA Traceability Protocol for Assay and Certification of Gaseous
Calibration Standards. EPA-600/R-12/531. May, 2012. Available from
U.S. Environmental Protection Agency, National Risk Management
Research Laboratory, Research Triangle Park NC 27711. https://www.epa.gov/nscep.
* * * * *
(6) List of Designated Reference and Equivalent Methods. Available
from U.S. Environmental Protection Agency, Center for Environmental
Measurements and Modeling, Air Methods and Characterization
Division, MD-D205-03, Research Triangle Park, NC 27711. https://www.epa.gov/amtic/air-monitoring-methods-criteria-pollutants.
(7) Transfer Standards for the Calibration of Ambient Air Monitoring
Analyzers for Ozone. EPA-454/B-13-004 U.S. Environmental Protection
Agency, Research Triangle Park, NC 27711, October, 2013. https://www.epa.gov/sites/default/files/2020-09/documents/ozonetransferstandardguidance.pdf.
* * * * *
(9) Quality Assurance Handbook for Air Pollution Measurement
Systems, Volume 1--A Field Guide to Environmental Quality Assurance.
EPA-600/R-94/038a. April 1994. Available from U.S. Environmental
Protection Agency, ORD Publications Office, Center for Environmental
Research Information (CERI), 26 W. Martin Luther King Drive,
Cincinnati, OH 45268. https://www.epa.gov/amtic/ambient-air-monitoring-quality-assurance#documents.
(10) Quality Assurance Handbook for Air Pollution Measurement
Systems, Volume II: Ambient Air Quality Monitoring Program Quality
System Development. EPA-454/B-13-003. https://www.epa.gov/amtic/ambient-air-monitoring-quality-assurance#documents.
(11) National Performance Evaluation Program Standard Operating
Procedures. https://www.epa.gov/amtic/ambient-air-monitoring-quality-assurance#npep.
Table B-1 to Section 6 of Appendix B- Minimum Data Assessment Requirements for NAAQS Related Criteria Pollutant PSD Monitors
--------------------------------------------------------------------------------------------------------------------------------------------------------
Method Assessment method Coverage Minimum frequency Parameters reported AQS Assessment type
--------------------------------------------------------------------------------------------------------------------------------------------------------
Gaseous Methods (CO, NO2, SO2, O3):
One-Point QC for SO2, NO2, O3, Response check at Each analyzer......... Once per 2 weeks\5\.. Audit One-Point QC.
CO. concentration 0.005- concentration\1\ and
0.08 ppm SO2, NO2, measured
O3, & 0.5 and 5 ppm concentration\2\.
CO.
Quarterly performance See section 3.1.2 of Each analyzer......... Once per quarter\5\.. Audit Annual PE.
evaluation for SO2, NO2, O3, this appendix. concentration\1\ and
CO. measured
concentration\2\ for
each level.
NPAP for SO2, NO2, O3, CO\3\... Independent Audit..... Each primary monitor.. Once per year........ Audit NPAP.
concentration\1\ and
measured
concentration\2\ for
each level.
Particulate Methods:
Collocated sampling PM10, Collocated samplers... 1 per PSD Network per Every 6 days or every Primary sampler No Transaction
PM2.5, Pb. pollutant. 3 days if daily concentration and reported as raw
monitoring required. duplicate sampler data.
concentration\4\.
Flow rate verification PM10, Check of sampler flow Each sampler.......... Once every month\5\.. Audit flow rate and Flow Rate
PM2.5, Pb. rate. measured flow rate Verification.
indicated by the
sampler.
Semi-annual flow rate audit Check of sampler flow Each sampler.......... Once every 6 months Audit flow rate and Semi Annual Flow Rate
PM10, PM2.5, Pb. rate using or beginning, middle measured flow rate Audit.
independent standard. and end of indicated by the
monitoring\5\. sampler.
[[Page 16395]]
Pb analysis audits Pb-TSP, Pb- Check of analytical Analytical............ Each quarter\5\...... Measured value and Pb Analysis Audits.
PM10. system with Pb audit audit value (ug Pb/
strips/filters. filter) using AQS
unit code 077 for
parameters:.
14129--Pb (TSP) LC
FRM/FEM.
85129--Pb (TSP) LC
Non-FRM/FEM..
Performance Evaluation Program Collocated samplers... (1) 5 valid audits for Over all 4 Primary sampler PEP.
PM2.5\3\. PQAOs with <= 5 quarters\5\. concentration and
sites.. performance
(2) 8 valid audits for evaluation sampler
PQAOs with > 5 sites.. concentration.
(3) All samplers in 6
years.
Performance Evaluation Program Collocated samplers... (1) 1 valid audit and Over all 4 Primary sampler PEP.
Pb \3\. 4 collocated samples quarters\5\. concentration and
for PQAOs, with <=5 performance
sites.. evaluation sampler
(2) 2 valid audits and concentration.
6 collocated samples Primary sampler
for PQAOs with >5 concentration and
sites.. duplicate sampler
concentration.
--------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ Effective concentration for open path analyzers.
\2\ Corrected concentration, if applicable for open path analyzers.
\3\ NPAP, PM2.5, PEP, and Pb-PEP must be implemented if data is used for NAAQS decisions otherwise implementation is at PSD reviewing authority
discretion.
\4\ Both primary and collocated sampler values are reported as raw data
\5\ A maximum number of days should be between these checks to ensure the checks are routinely conducted over time and to limit data impacts resulting
from a failed check.
0
28. Amend appendix C to part 58 by:
0
a. Adding sections 2.2 and 2.2.1 through 2.2.19;
0
b. Removing and reserving sections 2.4, 2.4.1;
0
c. Removing sections 2.4.1.1 through 2.4.1.7; and
0
d. Revising section 2.7.1.
The additions and revision reads as follows:
Appendix C to Part 58--Ambient Air Quality Monitoring Methodology
* * * * *
2.2 PM10, PM2.5, or PM10-2.5
continuous FEMs with existing valid designations may be calibrated
using network data from collocated FRM and continuous FEM data under
the following provisions:
2.2.1 Data to demonstrate a calibration may include valid data
from State, local, or Tribal air agencies or data collected by
instrument manufacturers in accordance with 40 CFR 53.35 or other
data approved by the Administrator.
2.2.2 A request to update a designated methods calibration may
be initiated by the instrument manufacturer of record or the EPA
Administrator. State, local, Tribal, and multijusistincional
organizations of these entities may work with an instrument
manufacture to update a designated method calibration.
2.2.3 Requests for approval of an updated PM10,
PM2.5, or PM10-2.5 continuous FEM calibration
must meet the general submittal requirements of section 2.7 of this
appendix.
2.2.4 Data included in the request should represent a subset of
representative locations where the method is operational. For cases
with a small number of collocated FRMs and continuous FEMs sites, an
updated candidate calibration may be limited to the sites where both
methods are in use.
2.2.5 Data included in a candidate method updated calibration
may include a subset of sites where there is a large grouping of
sites in one part of the country such that the updated calibration
would be representative of the country as a whole.
2.2.6 Improvements should be national in scope and ideally
implemented through a firmware change.
2.2.7 The goal of a change to a methods calibration is to
increase the number of sites meeting measurements quality objectives
of the method as identified in section 2.3.1.1 of appendix A to this
part.
2.2.8 For meeting measurement quality objectives (MQOs), the
primary objective is to meet the bias goal as this statistic will
likely have the most influence on improving the resultant data
collected.
2.2.9 Precision data are to be included, but so long as
precision data are at least as good as existing network data or meet
the MQO referenced in section 2.2.8 of this appendix, no further
work is necessary with precision.
2.2.10 Data available to use may include routine primary and
collocated data.
2.2.11 Audit data may be useful to confirm the performance of a
candidate updated calibration but should not be used as the basis of
the calibration to keep the independence of the audit data.
2.2.12 Data utilized as the basis of the updated calibration may
be obtained by accessing EPA's AQS database or future analogous EPA
database.
2.2.13 Years of data to use in a candidate method calibration
should include two recent years where we are past the certification
period for the previous year's data, which is May 1 of each year.
2.2.14 Data from additional years is to be used to test an
updated calibration such that the calibration is independent of the
test years of interest. Data from these additional years need to
minimally demonstrate that a larger number of sites are expected to
meet bias MQO especially at sites near the level of the NAAQS for
the PM indicator of interest.
2.2.15 Outliers may be excluded using routine outlier tests.
2.2.16 The range of data used in a calibration may include all
data available or alternatively use data in the range from the
lowest measured data available up to 125% of the 24-hour NAAQS for
the PM indicator of interest.
[[Page 16396]]
2.2.17 Other improvements to a PM continuous method may be
included as part of a recommended update so long as appropriate
testing is conducted with input from EPA's Office of Research and
Development (ORD) Reference and Equivalent (R&E) Methods Designation
program.
2.2.18 EPA encourages early communication by instrument
manufacturers considering an update to a PM method. Instrument
companies should initiate such dialogue by contacting EPA's ORD R&E
Methods Designation program. The contact information for this can be
found at 40 CFR 53.4.
2.2.19 Manufacturers interested in improving instrument's
performance through an updated factory calibration must submit a
written modification request to EPA with supporting rationale.
Because the testing requirements and acceptance criteria of any
field and/or lab tests can depend upon the nature and extent of the
intended modification, applicants should contact EPA's R&E Methods
Designation program for guidance prior to development of the
modification request.
* * * * *
2.7.1 Requests for approval under sections 2.2, 2.4, 2.6.2, or
2.8 of this appendix must be submitted to: Director, Center for
Environmental Measurement and Modeling, Reference and Equivalent
Methods Designation Program (MD-D205-03), U.S. Environmental
Protection Agency, P.O. Box 12055, Research Triangle Park, North
Carolina 27711.
0
29. Amend appendix D to part 58 by revising sections 1 and 1.1(b), the
introductory text before the table in section 4.7.1(a), and sections
4.7.1(b)(3) and 4.7.2 to read as follows:
Appendix D to Part 58--Network Design Criteria for Ambient Air Quality
Monitoring
* * * * *
1. Monitoring Objectives and Spatial Scales
The purpose of this appendix is to describe monitoring
objectives and general criteria to be applied in establishing the
required SLAMS ambient air quality monitoring stations and for
choosing general locations for additional monitoring sites. This
appendix also describes specific requirements for the number and
location of FRM and FEM sites for specific pollutants, NCore
multipollutant sites, PM10 mass sites, PM2.5
mass sites, chemically-speciated PM2.5 sites, and
O3 precursor measurements sites (PAMS). These criteria
will be used by EPA in evaluating the adequacy of the air pollutant
monitoring networks.
1.1 * * *
(b) Support compliance with ambient air quality standards and
emissions strategy development. Data from FRM and FEM monitors for
NAAQS pollutants will be used for comparing an area's air pollution
levels against the NAAQS. Data from monitors of various types can be
used in the development of attainment and maintenance plans. SLAMS,
and especially NCore station data, will be used to evaluate the
regional air quality models used in developing emission strategies,
and to track trends in air pollution abatement control measures'
impact on improving air quality. In monitoring locations near major
air pollution sources, source-oriented monitoring data can provide
insight into how well industrial sources are controlling their
pollutant emissions.
* * * * *
4.7.1 * * *
(a) State and where applicable, local, agencies must operate the
minimum number of required PM2.5 SLAMS sites listed in
table D-5 to this appendix. The NCore sites are expected to
complement the PM2.5 data collection that takes place at
non-NCore SLAMS sites, and both types of sites can be used to meet
the minimum PM2.5 network requirements. For many State
and local networks, the total number of PM2.5 sites
needed to support the basic monitoring objectives of providing air
pollution data to the general public in a timely manner, support
compliance with ambient air quality standards and emission strategy
development, and support for air pollution research studies will
include more sites than the minimum numbers required in table D-5 to
this appendix. Deviations from these PM2.5 monitoring
requirements must be approved by the EPA Regional Administrator.
* * * * *
(b) * * *
(3) For areas with additional required SLAMS, a monitoring
station is to be sited in an at-risk community with poor air
quality, particularly where there are anticipated effects from
sources in the area (e.g., a major industrial area, point source(s),
port, rail yard, airport, or other transportation facility or
corridor).
* * * * *
4.7.2 Requirement for Continuous PM2.5 Monitoring.
The State, or where appropriate, local agencies must operate
continuous PM2.5 analyzers equal to at least one-half
(round up) the minimum required sites listed in table D-5 to this
appendix. At least one required continuous analyzer in each MSA must
be collocated with one of the required FRM/FEM monitors, unless at
least one of the required FRM/FEM monitors is itself a continuous
FEM monitor in which case no collocation requirement applies. State
and local air monitoring agencies must use methodologies and quality
assurance/quality control (QA/QC) procedures approved by the EPA
Regional Administrator for these required continuous analyzers.
* * * * *
0
30. Revise appendix E to part 58 to read as follows:
Appendix E to Part 58--Probe and Monitoring Path Siting Criteria for
Ambient Air Quality Monitoring
1. Introduction
2. Monitors and Samplers with Probe Inlets
3. Open Path Analyzers
4. Waiver Provisions
5. References
1. Introduction
1.1 Applicability
(a) This appendix contains specific location criteria applicable
to ambient air quality monitoring probes, inlets, and optical paths
of SLAMS, NCore, PAMS, and other monitor types whose data are
intended to be used to determine compliance with the NAAQS. These
specific location criteria are relevant after the general location
has been selected based on the monitoring objectives and spatial
scale of representation discussed in appendix D to this part.
Monitor probe material and sample residence time requirements are
also included in this appendix. Adherence to these siting criteria
is necessary to ensure the uniform collection of compatible and
comparable air quality data.
(b) The probe and monitoring path siting criteria discussed in
this appendix must be followed to the maximum extent possible. It is
recognized that there may be situations where some deviation from
the siting criteria may be necessary. In any such case, the reasons
must be thoroughly documented in a written request for a waiver that
describes whether the resulting monitoring data will be
representative of the monitoring area and how and why the proposed
or existing siting must deviate from the criteria. This
documentation should help to avoid later questions about the
validity of the resulting monitoring data. Conditions under which
the EPA would consider an application for waiver from these siting
criteria are discussed in section 4 of this appendix.
(c) The pollutant-specific probe and monitoring path siting
criteria generally apply to all spatial scales except where noted
otherwise. Specific siting criteria that are phrased with ``shall''
or ``must'' are defined as requirements and exceptions must be
granted through the waiver provisions. However, siting criteria that
are phrased with ``should'' are defined as goals to meet for
consistency but are not requirements.
2. Monitors and Samplers with Probe Inlets
2.1 Horizontal and Vertical Placement
(a) For O3 and SO2 monitoring, and for
neighborhood or larger spatial scale Pb, PM10,
PM10-2.5, PM2.5, NO2, and CO sites,
the probe must be located greater than or equal to 2.0 meters and
less than or equal to 15 meters above ground level.
(b) Middle scale CO and NO2 monitors must have
sampler inlets greater than or equal to 2.0 meters and less than or
equal to 15 meters above ground level.
(c) Middle scale PM10-2.5 sites are required to have
sampler inlets greater than or equal to 2.0 meters and less than or
equal to 7.0 meters above ground level.
(d) Microscale Pb, PM10, PM10-2.5, and
PM2.5 sites are required to have sampler inlets greater
than or equal to 2.0 meters and less than or equal to 7.0 meters
above ground level.
(e) Microscale near-road NO2 monitoring sites are
required to have sampler inlets greater than or equal to 2.0 meters
and less than or equal to 7.0 meters above ground level.
(f) The probe inlets for microscale carbon monoxide monitors
that are being used to
[[Page 16397]]
measure concentrations near roadways must be greater than or equal
to 2.0 meters and less than or equal to 7.0 meters above ground
level. Those probe inlets for microscale carbon monoxide monitors
measuring concentrations near roadways in downtown areas or urban
street canyons must be greater than or equal to 2.5 meters and less
than or equal to 3.5 meters above ground level. The probe must be at
least 1.0 meter vertically or horizontally away from any supporting
structure, walls, parapets, penthouses, etc., and away from dusty or
dirty areas. If the probe is located near the side of a building or
wall, then it should be located on the windward side of the building
relative to the prevailing wind direction during the season of
highest concentration potential for the pollutant being measured.
2.2 Spacing From Minor Sources
(a) It is important to understand the monitoring objective for a
particular site in order to interpret this requirement. Local minor
sources of a primary pollutant, such as SO2, lead, or
particles, can cause high concentrations of that particular
pollutant at a monitoring site. If the objective for that monitoring
site is to investigate these local primary pollutant emissions, then
the site will likely be properly located nearby. This type of
monitoring site would, in all likelihood, be a microscale-type of
monitoring site. If a monitoring site is to be used to determine air
quality over a much larger area, such as a neighborhood or city, a
monitoring agency should avoid placing a monitor probe inlet near
local, minor sources, because a plume from a local minor source
should not be allowed to inappropriately impact the air quality data
collected at a site. Particulate matter sites should not be located
in an unpaved area unless there is vegetative ground cover year-
round, so that the impact of windblown dusts will be kept to a
minimum.
(b) Similarly, local sources of nitric oxide (NO) and ozone-
reactive hydrocarbons can have a scavenging effect causing
unrepresentatively low concentrations of O3 in the
vicinity of probes for O3. To minimize these potential
interferences from nearby minor sources, the probe inlet should be
placed at a distance from furnace or incineration flues or other
minor sources of SO2 or NO. The separation distance
should take into account the heights of the flues, type of waste or
fuel burned, and the sulfur content of the fuel.
2.3 Spacing From Obstructions
(a) Obstacles may scavenge SO2, O3, or
NO2, and can act to restrict airflow for any pollutant.
To avoid this interference, the probe inlet must have unrestricted
airflow pursuant to paragraph (b) of this section and should be
located at a distance from obstacles. The horizontal distance from
the obstacle to the probe inlet must be at least twice the height
that the obstacle protrudes above the probe inlet. An obstacle that
does not meet the minimum distance requirement is considered an
obstruction that restricts airflow to the probe inlet. The EPA does
not generally consider objects or obstacles such as flag poles or
site towers used for NOy convertors and meteorological sensors, etc.
to be deemed obstructions.
(b) A probe inlet located near or along a vertical wall is
undesirable because air moving along the wall may be subject to
removal mechanisms. A probe inlet must have unrestricted airflow
with no obstructions (as defined in paragraph (a) of this section)
in a continuous arc of at least 270 degrees. An unobstructed
continuous arc of 180 degrees is allowable when the applicable
network design criteria specified in appendix D of this part require
monitoring in street canyons and the probe is located on the side of
a building. This arc must include the predominant wind direction for
the season of greatest pollutant concentration potential. For
particle sampling, there must be a minimum of 2.0 meters of
horizontal separation from walls, parapets, and structures for
rooftop site placement.
(c) A sampling station with a probe inlet located closer to an
obstacle than required by the criteria in this section should be
classified as middle scale or microscale, rather than neighborhood
or urban scale, since the measurements from such a station would
more closely represent these smaller scales.
(d) For near-road monitoring stations, the monitor probe shall
have an unobstructed air flow, where no obstacles exist at or above
the height of the monitor probe, between the monitor probe and the
outside nearest edge of the traffic lanes of the target road
segment.
2.4 Spacing From Trees
(a) Trees can provide surfaces for SO2,
O3, or NO2 adsorption or reactions and
surfaces for particle deposition. Trees can also act as obstructions
in locations where the trees are between the air pollutant sources
or source areas and the monitoring site and where the trees are of a
sufficient height and leaf canopy density to interfere with the
normal airflow around the probe inlet. To reduce this possible
interference/obstruction, the probe inlet should be 20 meters or
more from the drip line of trees and must be at least 10 meters from
the drip line of trees. If a tree or group of trees is an obstacle,
the probe inlet must meet the distance requirements of section 2.3
of this appendix.
(b) The scavenging effect of trees is greater for O3
than for other criteria pollutants. Monitoring agencies must take
steps to consider the impact of trees on ozone monitoring sites and
take steps to avoid this problem.
(c) Beginning January 1, 2024, microscale sites of any air
pollutant shall have no trees or shrubs located at or above the
line-of-sight fetch between the probe and the source under
investigation, e.g., a roadway or a stationary source.
2.5 Spacing From Roadways
Table E-1 to Section 2.5 of Appendix E--Minimum Separation Distance
Between Roadways and Probes for Monitoring Neighborhood and Urban Scale
Ozone (O3) and Oxides of Nitrogen (NO, NO2, NOX, NOy)
------------------------------------------------------------------------
Minimum
Roadway average daily traffic, vehicles per Minimum distance\1
day distance\1 2 3\
3\ (meters) (meters)
------------------------------------------------------------------------
<=1,000....................................... 10 10
10,000........................................ 10 20
15,000........................................ 20 30
20,000........................................ 30 40
40,000........................................ 50 60
70,000........................................ 100 100
>=110,000..................................... 250 250
------------------------------------------------------------------------
\1\ Distance from the edge of the nearest traffic lane. The distance for
intermediate traffic counts should be interpolated from the table
values based on the actual traffic count./TNOTE>
\2\ Applicable for ozone monitors whose placement was not approved as of
December 18, 2006.
\3\ All distances listed are expressed as having 2 significant figures.
When rounding is performed to assess compliance with these siting
requirements, the distance measurements will be rounded such as to
retain at least two significant figures.
2.5.1 Spacing for Ozone Probes
In siting an O3 monitor, it is important to minimize
destructive interferences from sources of NO, since NO readily
reacts with O3. Table E-1 of this appendix provides the
required minimum separation distances between a roadway and a probe
inlet for various ranges of daily roadway traffic. A sampling site
with a monitor probe located closer to a roadway than allowed by the
Table E-1 requirements should be classified as middle scale or
microscale, rather than neighborhood or urban scale, since the
measurements from such a site would more closely represent these
smaller scales.
2.5.2 Spacing for Carbon Monoxide Probes
(a) Near-road microscale CO monitoring sites, including those
located in downtown areas, urban street canyons, and other near-road
locations such as those adjacent to highly trafficked roads, are
intended to provide a measurement of the influence of the immediate
source on the pollution exposure on the adjacent area.
(b) Microscale CO monitor probe inlets in downtown areas or
urban street canyon locations shall be located a minimum distance of
2.0 meters and a maximum distance of 10 meters from the edge of the
nearest traffic lane.
(c) Microscale CO monitor probe inlets in downtown areas or
urban street canyon locations shall be located at least 10 meters
from an intersection, preferably at a midblock location. Midblock
locations are preferable to intersection locations because
intersections represent a much smaller portion of downtown space
than do the streets between
[[Page 16398]]
them. Pedestrian exposure is probably also greater in street canyon/
corridors than at intersections.
(d) Neighborhood scale CO monitor probe inlets in downtown areas
or urban street canyon locations shall be located according to the
requirements in Table E-2 of this appendix.
Table E-2 to Section 2.5.2 of Appendix E--Minimum Separation Distance
Between Roadways and Probes for Monitoring Neighborhood Scale Carbon
Monoxide
------------------------------------------------------------------------
Minimum distance 1 2
Roadway average daily traffic, vehicles per day (meters)
------------------------------------------------------------------------
<=10,000....................................... 10
15,000......................................... 25
20,000......................................... 45
30,000......................................... 80
40,000......................................... 115
50,000......................................... 135
>=60,000....................................... 150
------------------------------------------------------------------------
1 Distance from the edge of the nearest traffic lane. The distance for
intermediate traffic counts should be interpolated from the table
values based on the actual traffic count.
2 All distances listed are expressed as having 2 significant figures.
When rounding is performed to assess compliance with these siting
requirements, the distance measurements will be rounded such as to
retain at least two significant figures.
2.5.3 Spacing for Particulate Matter (PM2.5, PM2.5 10, PM10, Pb)
Inlets
(a) Since emissions associated with the operation of motor
vehicles contribute to urban area particulate matter ambient levels,
spacing from roadway criteria are necessary for ensuring national
consistency in PM sampler siting.
(b) The intent is to locate localized hot-spot sites in areas of
highest concentrations, whether it be caused by mobile or multiple
stationary sources. If the area is primarily affected by mobile
sources and the maximum concentration area(s) is judged to be a
traffic corridor or street canyon location, then the monitors should
be located near roadways with the highest traffic volume and at
separation distances most likely to produce the highest
concentrations. For microscale traffic corridor sites, the location
must be greater than or equal 5.0 meters and less than or equal to
15 meters from the major roadway. For the microscale street canyon
site, the location must be greater than or equal 2.0 meters and less
than or equal to 10 meters from the roadway. For the middle scale
site, a range of acceptable distances from the roadway is shown in
Figure E-1 of this appendix. This figure also includes separation
distances between a roadway and neighborhood or larger scale sites
by default. Any PM probe inlet at a site, 2.0 to 15 meters high, and
further back than the middle scale requirements will generally be
neighborhood, urban or regional scale. For example, according to
Figure E-1 of this appendix, if a PM sampler is primarily influenced
by roadway emissions and that sampler is set back 10 meters from a
30,000 ADT (average daily traffic) road, the site should be
classified as microscale, if the sampler's inlet height is between
2.0 and 7.0 meters. If the sampler's inlet height is between 7.0 and
15 meters, the site should be classified as middle scale. If the
sampler is 20 meters from the same road, it will be classified as
middle scale; if 40 meters, neighborhood scale; and if 110 meters,
an urban scale.
[GRAPHIC] [TIFF OMITTED] TR06MR24.047
2.5.4 Spacing for Nitrogen Dioxide (NO2) Probes
(a) In siting near-road NO2 monitors as required in
section 4.3.2 of appendix D of this part, the monitor probe shall be
as near as practicable to the outside nearest edge of the traffic
lanes of the target road segment but shall not be located at a
distance greater than 50 meters, in the horizontal, from the outside
nearest edge of the traffic lanes of the target road segment. Where
possible, the near-road NO2 monitor probe should be
within 20 meters of the target road segment.
(b) In siting NO2 monitors for neighborhood and
larger scale monitoring, it is important to minimize near-road
influences. Table E-1 of this appendix provides the required minimum
separation distances between a roadway and a probe inlet for various
ranges of daily roadway traffic. A site with a monitor probe located
closer to a roadway than allowed by the Table E-1 requirements
should be classified
[[Page 16399]]
as microscale or middle scale rather than neighborhood or urban
scale.
2.6 Probe Material and Pollutant Sampler Residence Time
(a) For the reactive gases (SO2, NO2, and
O3), approved probe materials must be used for monitors.
Studies25 34 have been conducted to determine the
suitability of materials such as polypropylene, polyethylene,
polyvinyl chloride, Tygon[supreg], aluminum, brass, stainless steel,
copper, borosilicate glass, polyvinylidene fluoride (PVDF),
polytetrafluoroethylene (PTFE), perfluoroalkoxy (PFA), and
fluorinated ethylene propylene (FEP) for use as intake sampling
lines. Of the above materials, only borosilicate glass, PVDF, PTFE,
PFA, and FEP have been found to be acceptable for use as intake
sampling lines for all the reactive gaseous pollutants. Furthermore,
the EPA \25\ has specified borosilicate glass, FEP Teflon[supreg],
or their equivalents as the only acceptable probe materials for
delivering test atmospheres in the determination of reference or
equivalent methods. Therefore, borosilicate glass, PVDF, PTFE, PFA,
FEP, or their equivalents must be the only material in the sampling
train (from probe inlet to the back of the monitor) that can be in
contact with the ambient air sample for reactive gas monitors.
Nafion\TM\, which is composed primarily of PTFE, can be considered
equivalent to PTFE; it has been shown in tests to exhibit virtually
no loss of ozone at 20-second residence times.\35\
(b) For volatile organic compound (VOC) monitoring at PAMS, FEP
Teflon[supreg] is unacceptable as the probe material because of VOC
adsorption and desorption reactions on the FEP Teflon[supreg].
Borosilicate glass, stainless steel, or their equivalents are the
acceptable probe materials for VOC and carbonyl sampling. Care must
be taken to ensure that the sample residence time is kept to 20
seconds or less.
(c) No matter how nonreactive the sampling probe material is
initially, after a period of use, reactive particulate matter is
deposited on the probe walls. Therefore, the time it takes the gas
to transfer from the probe inlet to the sampling device is critical.
Ozone in the presence of nitrogen oxide (NO) will show significant
losses, even in the most inert probe material, when the residence
time exceeds 20 seconds.\26\ Other studies 27 28indicate
that a 10-second or less residence time is easily achievable.
Therefore, sampling probes for all reactive gas monitors for
SO2, NO2, and O3 must have a sample
residence time less than 20 seconds.
2.7 Summary
Table E-3 of this appendix presents a summary of the general
requirements for probe siting criteria with respect to distances and
heights. Table E-3 requires different elevation distances above the
ground for the various pollutants. The discussion in this appendix
for each of the pollutants describes reasons for elevating the
monitor or probe inlet. The differences in the specified range of
heights are based on the vertical concentration gradients. For
source oriented and near-road monitors, the gradients in the
vertical direction are very large for the microscale, so a small
range of heights are used. The upper limit of 15 meters is specified
for the consistency between pollutants and to allow the use of a
single manifold for monitoring more than one pollutant.
Table E-3 to Section 2.7 of Appendix E--Summary of Probe Siting Criteria
----------------------------------------------------------------------------------------------------------------
Horizontal or Distance
Height vertical distance from drip
from from supporting line of Distance from
Pollutant Scale \9\ ground to structures \1\ \8\ trees to roadways to probe
probe \8\ to probe inlet probe \8\ \8\ (meters)
(meters) (meters) (meters)
----------------------------------------------------------------------------------------------------------------
SO2\2 3 4 5\................... Middle, 2.0-15 >=1.0 >=10 N/A.
Neighborhood,
Urban, and
Regional.
CO3 4 6....................... Micro [downtown or 2.5-3.5 2.0-10 for
street canyon downtown areas
sites]. or street canyon
microscale.
Micro [Near-Road 2.0-7.0 >=1.0 >=10 <=50 for near-
sites]. road microscale.
Middle and 2.0-15 See Table E-2 of
Neighborhood. this appendix
for middle and
neighborhood
scales.
O32 3 4........................ Middle, 2.0-15 >=1.0 >=10 See Table E-1.
Neighborhood,
Urban, and
Regional.
Micro............. 2.0-7.0 <=50 for near-
road micro-
scale.
NO22 3 4....................... Middle, 2.0-15 >=1.0 >=10 See Table E-1.
Neighborhood,
Urban, and
Regional.
PAMS2 3 4 Ozone precursors..... Neighborhood and 2.0-15 >=1.0 >=10 See Table E-1.
Urban.
PM, Pb 2 3 4 7................. Micro............. 2.0-7.0
Middle, 2.0-15 >=2.0 (horizontal >=10 See Figure E-1.
Neighborhood, distance only)
Urban and
Regional.
----------------------------------------------------------------------------------------------------------------
N/A--Not applicable.
\1\ When a probe is located on a rooftop, this separation distance is in reference to walls, parapets, or
penthouses located on the roof.
\2\ Should be greater than 20 meters from the dripline of tree(s) and must be 10 meters from the dripline.
\3\ Distance from sampler or probe inlet to obstacle, such as a building, must be at least twice the height the
obstacle protrudes above the sampler or probe inlet. Sites not meeting this criterion may be classified as
microscale or middle scale (see paragraphs 2.3(a) and 2.3(c)).
\4\ Must have unrestricted airflow in a continuous arc of at least 270 degrees around the probe or sampler; 180
degrees if the probe is on the side of a building or a wall for street canyon monitoring.
\5\ The probe or sampler should be away from minor sources, such as furnace or incineration flues. The
separation distance is dependent on the height of the minor source emission point(s), the type of fuel or
waste burned, and the quality of the fuel (sulfur, ash, or lead content). This criterion is designed to avoid
undue influences from minor sources.
\6\ For microscale CO monitoring sites, the probe must be >=10 meters from a street intersection and preferably
at a midblock location.
\7\ Collocated monitor inlets must be within 4.0 meters of each other and at least 2.0 meters apart for flow
rates greater than 200 liters/min or at least 1.0 meter apart for samplers having flow rates less than 200
liters/min to preclude airflow interference, unless a waiver has been granted by the Regional Administrator
pursuant to paragraph 3.3.4.2(c) of appendix A of part 58. For PM2.5, collocated monitor inlet heights should
be within 1.0 meter of each other vertically.
\8\ All distances listed are expressed as having 2 significant figures. When rounding is performed to assess
compliance with these siting requirements, the distance measurements will be rounded such as to retain at
least two significant figures.
\9\ See section 1.2 of appendix D for definitions of monitoring scales.
3. Open Path Analyzers
3.1 Horizontal and Vertical Placement
(a) For all O3 and SO2 monitoring sites
and for neighborhood or larger spatial scale NO2, and CO
sites, at least 80 percent of the monitoring path must be located
greater than or equal 2.0 meters and less than or equal to 15 meters
above ground level.
(b) Middle scale CO and NO2 sites must have
monitoring paths greater than or equal 2.0 meters and less than or
equal to 15 meters above ground level.
(c) Microscale near-road monitoring sites are required to have
monitoring paths greater than or equal 2.0 meters and less than or
equal to 7.0 meters above ground level.
(d) For microscale carbon monoxide monitors that are being used
to measure concentrations near roadways, the monitoring path must be
greater than or equal 2.0 meters and less than or equal to 7.0
meters above ground level. If the microscale carbon monoxide
monitors measuring concentrations near roadways are in downtown
areas or urban street canyons, the monitoring path must be greater
than or equal 2.5 meters and less than or equal to 3.5 meters above
ground level and at least 90
[[Page 16400]]
percent of the monitoring path must be at least 1.0 meter vertically
or horizontally away from any supporting structure, walls, parapets,
penthouses, etc., and away from dusty or dirty areas. If a
significant portion of the monitoring path is located near the side
of a building or wall, then it should be located on the windward
side of the building relative to the prevailing wind direction
during the season of highest concentration potential for the
pollutant being measured.
3.2 Spacing From Minor Sources
(a) It is important to understand the monitoring objective for a
particular site in order to interpret this requirement. Local minor
sources of a primary pollutant, such as SO2 can cause
high concentrations of that particular pollutant at a monitoring
site. If the objective for that monitoring site is to investigate
these local primary pollutant emissions, then the site will likely
be properly located nearby. This type of monitoring site would, in
all likelihood, be a microscale type of monitoring site. If a
monitoring site is to be used to determine air quality over a much
larger area, such as a neighborhood or city, a monitoring agency
should avoid placing a monitoring path near local, minor sources,
because a plume from a local minor source should not be allowed to
inappropriately impact the air quality data collected at a site.
(b) Similarly, local sources of nitric oxide (NO) and ozone-
reactive hydrocarbons can have a scavenging effect causing
unrepresentatively low concentrations of O3 in the
vicinity of monitoring paths for O3. To minimize these
potential interferences from nearby minor sources, at least 90
percent of the monitoring path should be at a distance from furnace
or incineration flues or other minor sources of SO2 or
NO. The separation distance should take into account the heights of
the flues, type of waste or fuel burned, and the sulfur content of
the fuel.
3.3 Spacing From Obstructions
(a) Obstacles may scavenge SO2, O3, or
NO2, and can act to restrict airflow for any pollutant.
To avoid this interference, at least 90 percent of the monitoring
path must have unrestricted airflow and should be located at a
distance from obstacles. The horizontal distance from the obstacle
to the monitoring path must be at least twice the height that the
obstacle protrudes above the monitoring path. An obstacle that does
not meet the minimum distance requirement is considered an
obstruction that restricts airflow to the monitoring path. The EPA
does not generally consider objects or obstacles such as flag poles
or site towers used for NOy convertors and meteorological sensors,
etc. to be deemed obstructions.
(b) A monitoring path located near or along a vertical wall is
undesirable because air moving along the wall may be subject to
removal mechanisms. At least 90 percent of the monitoring path for
open path analyzers must have unrestricted airflow with no
obstructions (as defined in paragraph (a) of this section) in a
continuous arc of at least 270 degrees. An unobstructed continuous
arc of 180 degrees is allowable when the applicable network design
criteria specified in appendix D of this part require monitoring in
street canyons and the monitoring path is located on the side of a
building. This arc must include the predominant wind direction for
the season of greatest pollutant concentration potential.
(c) Special consideration must be given to the use of open path
analyzers given their inherent potential sensitivity to certain
types of interferences and optical obstructions. A monitoring path
must be clear of all trees, brush, buildings, plumes, dust, or other
optical obstructions, including potential obstructions that may move
due to wind, human activity, growth of vegetation, etc. Temporary
optical obstructions, such as rain, particles, fog, or snow, should
be considered when siting an open path analyzer. Any of these
temporary obstructions that are of sufficient density to obscure the
light beam will negatively affect the ability of the open path
analyzer to continuously measure pollutant concentrations.
Transient, but significant obscuration of especially longer
measurement paths, could occur as a result of certain meteorological
conditions (e.g., heavy fog, rain, snow) and/or aerosol levels that
are of a sufficient density to prevent the open path analyzer's
light transmission. If certain compensating measures are not
otherwise implemented at the onset of monitoring (e.g., shorter path
lengths, higher light source intensity), data recovery during
periods of greatest primary pollutant potential could be
compromised. For instance, if heavy fog or high particulate levels
are coincident with periods of projected NAAQS-threatening pollutant
potential, the representativeness of the resulting data record in
reflecting maximum pollution concentrations may be substantially
impaired despite the fact that the site may otherwise exhibit an
acceptable, even exceedingly high, overall valid data capture rate.
(d) A sampling station with a monitoring path located closer to
an obstacle than required by the criteria in this section should be
classified as middle scale or microscale, rather than neighborhood
or urban scale, since the measurements from such a station would
more closely represent these smaller scales.
(e) For near-road monitoring stations, the monitoring path shall
have an unobstructed air flow, where no obstacles exist at or above
the height of the monitoring path, between the monitoring path and
the outside nearest edge of the traffic lanes of the target road
segment.
3.4 Spacing From Trees
(a) Trees can provide surfaces for SO2,
O3, or NO2 adsorption or reactions. Trees can
also act as obstructions in locations where the trees are located
between the air pollutant sources or source areas and the monitoring
site, and where the trees are of a sufficient height and leaf canopy
density to interfere with the normal airflow around the monitoring
path. To reduce this possible interference/obstruction, at least 90
percent of the monitoring path should be 20 meters or more from the
drip line of trees and must be at least 10 meters from the drip line
of trees. If a tree or group of trees could be considered an
obstacle, the monitoring path must meet the distance requirements of
section 3.3 of this appendix.
(b) The scavenging effect of trees is greater for O3
than for other criteria pollutants. Monitoring agencies must take
steps to consider the impact of trees on ozone monitoring sites and
take steps to avoid this problem.
(c) Beginning January 1, 2024, microscale sites of any air
pollutant shall have no trees or shrubs located at or above the
line-of-sight fetch between the monitoring path and the source under
investigation, e.g., a roadway or a stationary source.
3.5 Spacing from Roadways
Table E-4 of Section 3.5 of Appendix E--Minimum Separation Distance
Between Roadways and Monitoring Paths for Monitoring Neighborhood and
Urban Scale Ozone (O3) and Oxides of Nitrogen (NO, NO2, NOX, NOy)
------------------------------------------------------------------------
Minimum Minimum
distance distance
Roadway average daily traffic, vehicles per day \1 3\ \1 2 3\
(meters) (meters)
------------------------------------------------------------------------
<=1,000........................................... 10 10
10,000............................................ 10 20
15,000............................................ 20 30
20,000............................................ 30 40
40,000............................................ 50 60
70,000............................................ 100 100
>=110,000......................................... 250 250
------------------------------------------------------------------------
\1\ Distance from the edge of the nearest traffic lane. The distance for
intermediate traffic counts should be interpolated from the table
values based on the actual traffic count.
\2\ Applicable for ozone open path monitors whose placement was not
approved as of December 18, 2006.
\3\ All distances listed are expressed as having 2 significant figures.
When rounding is performed to assess compliance with these siting
requirements, the distance measurements will be rounded such as to
retain at least two significant figures.
3.5.1 Spacing for Ozone Monitoring Paths
In siting an O3 open path analyzer, it is important
to minimize destructive interferences form sources of NO, since NO
readily reacts with O3. Table E-4 of this appendix
provides the required minimum separation distances between a roadway
and at least 90 percent of a monitoring path for various ranges of
daily roadway traffic. A monitoring site with a monitoring path
located closer to a roadway than allowed by the Table E-4
requirements should be classified as microscale or middle scale,
rather than neighborhood or urban scale, since the measurements from
such a site would more closely represent these smaller scales. The
monitoring path(s) must not cross over a roadway with an average
daily traffic count of 10,000 vehicles per day or more. For
locations where a monitoring path crosses a roadway with fewer than
10,000 vehicles per day, monitoring agencies must consider the
entire segment of the monitoring path in the area of potential
atmospheric interference from automobile emissions. Therefore, this
[[Page 16401]]
calculation must include the length of the monitoring path over the
roadway plus any segments of the monitoring path that lie in the
area between the roadway and minimum separation distance, as
determined from Table E-4 of this appendix. The sum of these
distances must not be greater than 10 percent of the total
monitoring path length.
3.5.2 Spacing for Carbon Monoxide Monitoring Paths
(a) Near-road microscale CO monitoring sites, including those
located in downtown areas, urban street canyons, and other near-road
locations such as those adjacent to highly trafficked roads, are
intended to provide a measurement of the influence of the immediate
source on the pollution exposure on the adjacent area.
(b) Microscale CO monitoring paths in downtown areas or urban
street canyon locations shall be located a minimum distance of 2.0
meters and a maximum distance of 10 meters from the edge of the
nearest traffic lane.
(c) Microscale CO monitoring paths in downtown areas or urban
street canyon locations shall be located at least 10 meters from an
intersection, preferably at a midblock location. Midblock locations
are preferable to intersection locations because intersections
represent a much smaller portion of downtown space than do the
streets between them. Pedestrian exposure is probably also greater
in street canyon/corridors than at intersections.
(d) Neighborhood scale CO monitoring paths in downtown areas or
urban street canyon locations shall be located according to the
requirements in Table E-5 of this appendix.
Table E-5 Section 3.5.2 of Appendix E--Minimum Separation Distance
Between Roadways and Monitoring Paths for Monitoring Neighborhood Scale
Carbon Monoxide
------------------------------------------------------------------------
Minimum
Roadway average daily traffic, vehicles per day distance \1 2\
(meters)
------------------------------------------------------------------------
<=10,000................................................ 10
15,000.................................................. 25
20,000.................................................. 45
30,000.................................................. 80
40,000.................................................. 115
50,000.................................................. 135
>=60,000................................................ 150
------------------------------------------------------------------------
\1\ Distance from the edge of the nearest traffic lane. The distance for
intermediate traffic counts should be interpolated from the table
values based on the actual traffic count.
\2\ All distances listed are expressed as having 2 significant figures.
When rounding is performed to assess compliance with these siting
requirements, the distance measurements will be rounded such as to
retain at least two significant figures.
3.5.3 Spacing for Nitrogen Dioxide (NO2) Monitoring Paths
(a) In siting near-road NO2 monitors as required in
section 4.3.2 of appendix D of this part, the monitoring path shall
be as near as practicable to the outside nearest edge of the traffic
lanes of the target road segment but shall not be located at a
distance greater than 50 meters, in the horizontal, from the outside
nearest edge of the traffic lanes of the target road segment.
(b) In siting NO2 open path monitors for neighborhood
and larger scale monitoring, it is important to minimize near-road
influences. Table E-5 of this appendix provides the required minimum
separation distances between a roadway and at least 90 percent of a
monitoring path for various ranges of daily roadway traffic. A site
with a monitoring path located closer to a roadway than allowed by
the Table E-4 requirements should be classified as microscale or
middle scale rather than neighborhood or urban scale. The monitoring
path(s) must not cross over a roadway with an average daily traffic
count of 10,000 vehicles per day or more. For locations where a
monitoring path crosses a roadway with fewer than 10,000 vehicles
per day, monitoring agencies must consider the entire segment of the
monitoring path in the area of potential atmospheric interference
form automobile emissions. Therefore, this calculation must include
the length of the monitoring path over the roadway plus any segments
of the monitoring path that lie in the area between the roadway and
minimum separation distance, as determined from Table E-5 of this
appendix. The sum of these distances must not be greater than 10
percent of the total monitoring path length.
3.6 Cumulative Interferences on a Monitoring Path
The cumulative length or portion of a monitoring path that is
affected by minor sources, trees, or roadways must not exceed 10
percent of the total monitoring path length.
3.7 Maximum Monitoring Path Length
The monitoring path length must not exceed 1.0 kilometer for
open path analyzers in neighborhood, urban, or regional scale. For
middle scale monitoring sites, the monitoring path length must not
exceed 300 meters. In areas subject to frequent periods of dust,
fog, rain, or snow, consideration should be given to a shortened
monitoring path length to minimize loss of monitoring data due to
these temporary optical obstructions. For certain ambient air
monitoring scenarios using open path analyzers, shorter path lengths
may be needed in order to ensure that the monitoring site meets the
objectives and spatial scales defined in appendix D to this part.
The Regional Administrator may require shorter path lengths, as
needed on an individual basis, to ensure that the SLAMS sites meet
the appendix D requirements. Likewise, the Administrator may specify
the maximum path length used at NCore monitoring sites.
3.8 Summary
Table E-6 of this appendix presents a summary of the general
requirements for monitoring path siting criteria with respect to
distances and heights. Table E-6 requires different elevation
distances above the ground for the various pollutants. The
discussion in this appendix for each of the pollutants describes
reasons for elevating the monitoring path. The differences in the
specified range of heights are based on the vertical concentration
gradients. For source oriented and near-road monitors, the gradients
in the vertical direction are very large for the microscale, so a
small range of heights are used. The upper limit of 15 meters is
specified for the consistency between pollutants and to allow the
use of a monitoring path for monitoring more than one pollutant.
Table E-6 Section 3.8 of Appendix E--Summary of Monitoring Path Siting Criteria
----------------------------------------------------------------------------------------------------------------
Horizontal or
vertical
Height from distance from Distance from Distance from
Maximum ground to 80% supporting trees to 90% roadways to
Pollutant monitoring path of monitoring structures 2 of monitoring monitoring path
length 9 10 path 1 8 to 90% of path 1 8 1 8 (meters)
(meters) monitoring (meters)
path 1 8
(meters)
----------------------------------------------------------------------------------------------------------------
SO2 3 4 5 6.................. <= 300 m for 2.0-15 >=1.0 >=10 N/A.
Middle.
<= 1.0 km for
Neighborhood,
Urban, and
Regional.
CO4 5 7...................... <= 300 m for 2.5-3.5 >=1.0 >=10 2.0-10 for
Micro [downtown downtown areas
or street or street
canyon sites]. canyon
microscale.
<= 300 m for 2.0-7.0 <=50 for near-
Micro [Near- road
Road sites]. microscale.
[[Page 16402]]
<= 300 m for 2.0-15 See Table E-5
Middle. of this
appendix for
middle and
neighborhood
scales.
<= 1.0 km for
Neighborhood.
O33 4 5...................... <= 300 m for
Middle.
<= 1.0 km for 2.0-15 >=1.0 >=10 See Table E-4.
Neighborhood,
Urban, and
Regional.
NO23 4 5..................... Between 50 m-300 2.0-7.0 <=50 for near-
m for Micro road micro-
(Near-Road). scale.
<= 300 m for >=1.0 >=10
Middle.
<= 1.0 km for 2.0-15 See Table E-4.
Neighborhood,
Urban, and
Regional.
PAMS3 4 5 Ozone precursors... <= 1.0 km for 2.0-15 >=1.0 >=10 See Table E-4.
Neighborhood
and Urban.
----------------------------------------------------------------------------------------------------------------
N/A--Not applicable.
\1\ Monitoring path for open path analyzers is applicable only to middle or neighborhood scale CO monitoring,
middle, neighborhood, urban, and regional scale NO2 monitoring, and all applicable scales for monitoring SO2,
O3, and O3 precursors.
\2\ When the monitoring path is located on a rooftop, this separation distance is in reference to walls,
parapets, or penthouses located on roof.
\3\ At least 90 percent of the monitoring path should be greater than 20 meters from the dripline of tree(s) and
must be 10-meters from the dripline.
\4\ Distance from 90 percent of monitoring path to obstacle, such as a building, must be at least twice the
height the obstacle protrudes above the monitoring path. Sites not meeting this criterion may be classified as
microscale or middle scale (see text).
\5\ Must have unrestricted airflow 270 degrees around at least 90 percent of the monitoring path; 180 degrees if
the monitoring path is adjacent to the side of a building or a wall for street canyon monitoring.
\6\ The monitoring path should be away from minor sources, such as furnace or incineration flues. The separation
distance is dependent on the height of the minor source's emission point (such as a flue), the type of fuel or
waste burned, and the quality of the fuel (sulfur, ash, or lead content). This criterion is designed to avoid
undue influences from minor sources.
\7\ For microscale CO monitoring sites, the monitoring path must be >=10. meters from a street intersection and
preferably at a midblock location.
\8\ All distances listed are expressed as having 2 significant figures. When rounding is performed to assess
compliance with these siting requirements, the distance measurements will be rounded such as to retain at
least two significant figures.
\9\ See section 1.2 of appendix D for definitions of monitoring scales.
\10\ See section 3.7 of this appendix.
4. Waiver Provisions
Most sampling probes or monitors can be located so that they
meet the requirements of this appendix. New sites, with rare
exceptions, can be located within the limits of this appendix.
However, some existing sites may not meet these requirements and may
still produce useful data for some purposes. The EPA will consider a
written request from the State, or where applicable local, agency to
waive one or more siting criteria for some monitoring sites
providing that the State or their designee can adequately
demonstrate the need (purpose) for monitoring or establishing a
monitoring site at that location.
4.1 For a proposed new site, a waiver may be granted only if
both the following criteria are met:
4.1.1 The proposed new site can be demonstrated to be as
representative of the monitoring area as it would be if the siting
criteria were being met.
4.1.2 The monitor or probe cannot reasonably be located so as to
meet the siting criteria because of physical constraints (e.g.,
inability to locate the required type of site the necessary distance
from roadways or obstructions).
4.2 For an existing site, a waiver may be granted if either the
criterion in section 4.1.1 or the criterion in 4.1.2 of this
appendix is met.
4.3 Cost benefits, historical trends, and other factors may be
used to add support to the criteria in sections 4.1.1 and 4.1.2 of
this appendix; however, by themselves, they will not be acceptable
reasons for the EPA to grant a waiver. Written requests for waivers
must be submitted to the Regional Administrator. Granted waivers
must be renewed minimally every 5 years and ideally as part of the
network assessment as defined in Sec. 58.10(d). The approval date
of the waiver must be documented in the annual monitoring network
plan to support the requirements of Sec. 58.10(a)(1) and
58.10(b)(10).
5. References
1. Bryan, R.J., R.J. Gordon, and H. Menck. Comparison of High
Volume Air Filter Samples at Varying Distances from Los Angeles
Freeway. University of Southern California, School of Medicine, Los
Angeles, CA. (Presented at 66th Annual Meeting of Air Pollution
Control Association. Chicago, IL. June 24-28, 1973. APCA 73-158.)
2. Teer, E.H. Atmospheric Lead Concentration Above an Urban
Street. Master of Science Thesis, Washington University, St. Louis,
MO. January 1971.
3. Bradway, R.M., F.A. Record, and W.E. Belanger. Monitoring and
Modeling of Resuspended Roadway Dust Near Urban Arterials. GCA
Technology Division, Bedford, MA. (Presented at 1978 Annual Meeting
of Transportation Research Board, Washington, DC. January 1978.)
4. Pace, T.G., W.P. Freas, and E.M. Afify. Quantification of
Relationship Between Monitor Height and Measured Particulate Levels
in Seven U.S. Urban Areas. U.S. Environmental Protection Agency,
Research Triangle Park, NC. (Presented at 70th Annual Meeting of Air
Pollution Control Association, Toronto, Canada. June 20-24, 1977.
APCA 77-13.4.)
5. Harrison, P.R. Considerations for Siting Air Quality Monitors
in Urban Areas. City of Chicago, Department of Environmental
Control, Chicago, IL. (Presented at 66th Annual Meeting of Air
Pollution Control Association, Chicago, IL. June 24-28, 1973. APCA
73-161.)
6. Study of Suspended Particulate Measurements at Varying
Heights Above Ground. Texas State Department of Health, Air Control
Section, Austin, TX. 1970. p.7.
7. Rodes, C.E. and G.F. Evans. Summary of LACS Integrated
Pollutant Data. In: Los Angeles Catalyst Study Symposium. U.S.
Environmental Protection Agency, Research Triangle Park, NC. EPA
Publication No. EPA-600/4-77-034. June 1977.
8. Lynn, D.A. et al. National Assessment of the Urban
Particulate Problem: Volume 1, National Assessment. GCA Technology
Division, Bedford, MA. U.S. Environmental Protection Agency,
Research Triangle Park, NC. EPA Publication No. EPA-450/3-75-024.
June 1976.
9. Pace, T.G. Impact of Vehicle-Related Particulates on TSP
Concentrations and Rationale for Siting Hi-Vols in the Vicinity of
Roadways. OAQPS, U.S. Environmental Protection Agency, Research
Triangle Park, NC. April 1978.
10. Ludwig, F.L., J.H. Kealoha, and E. Shelar. Selecting Sites
for Monitoring Total Suspended Particulates. Stanford Research
Institute, Menlo Park, CA. Prepared for U.S. Environmental
Protection Agency, Research Triangle Park, NC. EPA Publication No.
EPA-450/3-77-018. June 1977, revised December 1977.
11. Ball, R.J. and G.E. Anderson. Optimum Site Exposure Criteria
for SO2 Monitoring. The Center for the Environment and
Man,
[[Page 16403]]
Inc., Hartford, CT. Prepared for U.S. Environmental Protection
Agency, Research Triangle Park, NC. EPA Publication No. EPA-450/3-
77-013. April 1977.
12. Ludwig, F.L. and J.H.S. Kealoha. Selecting Sites for Carbon
Monoxide Monitoring. Stanford Research Institute, Menlo Park, CA.
Prepared for U.S. Environmental Protection Agency, Research Triangle
Park, NC. EPA Publication No. EPA-450/3-75-077. September 1975.
13. Ludwig, F.L. and E. Shelar. Site Selection for the
Monitoring of Photochemical Air Pollutants. Stanford Research
Institute, Menlo Park, CA. Prepared for U.S. Environmental
Protection Agency, Research Triangle Park, NC. EPA Publication No.
EPA-450/3-78-013. April 1978.
14. Lead Analysis for Kansas City and Cincinnati, PEDCo
Environmental, Inc., Cincinnati, OH. Prepared for U.S. Environmental
Protection Agency, Research Triangle Park, NC. EPA Contract No. 66-
02-2515, June 1977.
15. Barltrap, D. and C.D. Strelow. Westway Nursery Testing
Project. Report to the Greater London Council. August 1976.
16. Daines, R. H., H. Moto, and D. M. Chilko. Atmospheric Lead:
Its Relationship to Traffic Volume and Proximity to Highways.
Environ. Sci. and Technol., 4:318, 1970.
17. Johnson, D. E., et al. Epidemiologic Study of the Effects of
Automobile Traffic on Blood Lead Levels, Southwest Research
Institute, Houston, TX. Prepared for U.S. Environmental Protection
Agency, Research Triangle Park, NC. EPA-600/1-78-055, August 1978.
18. Air Quality Criteria for Lead. Office of Research and
Development, U.S. Environmental Protection Agency, Washington, DC
EPA-600/8-83-028 aF-dF, 1986, and supplements EPA-600/8-89/049F,
August 1990. (NTIS document numbers PB87-142378 and PB91-138420.)
19. Lyman, D. R. The Atmospheric Diffusion of Carbon Monoxide
and Lead from an Expressway, Ph.D. Dissertation, University of
Cincinnati, Cincinnati, OH. 1972.
20. Wechter, S.G. Preparation of Stable Pollutant Gas Standards
Using Treated Aluminum Cylinders. ASTM STP. 598:40-54, 1976.
21. Wohlers, H.C., H. Newstein and D. Daunis. Carbon Monoxide
and Sulfur Dioxide Adsorption On and Description From Glass, Plastic
and Metal Tubings. J. Air Poll. Con. Assoc. 17:753, 1976.
22. Elfers, L.A. Field Operating Guide for Automated Air
Monitoring Equipment. U.S. NTIS. p. 202, 249, 1971.
23. Hughes, E.E. Development of Standard Reference Material for
Air Quality Measurement. ISA Transactions, 14:281-291, 1975.
24. Altshuller, A.D. and A.G. Wartburg. The Interaction of Ozone
with Plastic and Metallic Materials in a Dynamic Flow System.
Intern. Jour. Air and Water Poll., 4:70-78, 1961.
25. Code of Federal Regulations. 40 CFR 53.22, July 1976.
26. Butcher, S.S. and R.E. Ruff. Effect of Inlet Residence Time
on Analysis of Atmospheric Nitrogen Oxides and Ozone, Anal. Chem.,
43:1890, 1971.
27. Slowik, A.A. and E.B. Sansone. Diffusion Losses of Sulfur
Dioxide in Sampling Manifolds. J. Air. Poll. Con. Assoc., 24:245,
1974.
28. Yamada, V.M. and R.J. Charlson. Proper Sizing of the
Sampling Inlet Line for a Continuous Air Monitoring Station.
Environ. Sci. and Technol., 3:483, 1969.
29. Koch, R.C. and H.E. Rector. Optimum Network Design and Site
Exposure Criteria for Particulate Matter, GEOMET Technologies, Inc.,
Rockville, MD. Prepared for U.S. Environmental Protection Agency,
Research Triangle Park, NC. EPA Contract No. 68-02-3584. EPA 450/4-
87-009. May 1987.
30. Burton, R.M. and J.C. Suggs. Philadelphia Roadway Study.
Environmental Monitoring Systems Laboratory, U.S. Environmental
Protection Agency, Research Triangle Park, N.C. EPA-600/4-84-070
September 1984.
31. Technical Assistance Document for Sampling and Analysis of
Ozone Precursors. Atmospheric Research and Exposure Assessment
Laboratory, U.S. Environmental Protection Agency, Research Triangle
Park, NC 27711. EPA 600/8-91-215. October 1991.
32. Quality Assurance Handbook for Air Pollution Measurement
Systems: Volume IV. Meteorological Measurements. Atmospheric
Research and Exposure Assessment Laboratory, U.S. Environmental
Protection Agency, Research Triangle Park, NC 27711. EPA 600/4-90-
0003. August 1989.
33. On-Site Meteorological Program Guidance for Regulatory
Modeling Applications. Office of Air Quality Planning and Standards,
U.S. Environmental Protection Agency, Research Triangle Park, NC
27711. EPA 450/4-87-013. June 1987F.
34. Johnson, C., A. Whitehill, R. Long, and R. Vanderpool.
Investigation of Gaseous Criteria Pollutant Transport Efficiency as
a Function of Tubing Material. U.S. Environmental Protection Agency,
Research Triangle Park, NC 27711. EPA/600/R-22/212. August 2022.
35. Hannah Halliday, Cortina Johnson, Tad Kleindienst, Russell
Long, Robert Vanderpool, and Andrew Whitehill. Recommendations for
Nationwide Approval of Nafion\TM\ Dryers Upstream of UV-Absorption
Ozone Analyzers. U.S. Environmental Protection Agency, Research
Triangle Park, NC 27711. EPA/600/R-20/390. November 2020.
0
31. Revise appendix G to part 58 to read as follows:
Appendix G to Part 58--Uniform Air Quality Index (AQI) and Daily
Reporting
1. General Information
2. Reporting Requirements
3. Data Handling
1. General Information
1.1 AQI Overview. The AQI is a tool that simplifies reporting
air quality to the public in a nationally uniform and easy to
understand manner. The AQI converts concentrations of pollutants,
for which the EPA has established a national ambient air quality
standard (NAAQS), into a uniform scale from 0-500. These pollutants
are ozone (O3), particulate matter (PM2.5,
PM10), carbon monoxide (CO), sulfur dioxide
(SO2), and nitrogen dioxide (NO2). The scale
of the index is divided into general categories that are associated
with health messages.
2. Reporting Requirements
2.1 Applicability. The AQI must be reported daily for a
metropolitan statistical area (MSA) with a population over 350,000.
When it is useful and possible, it is recommended, but not required
for an area to report a sub-daily AQI as well.
2.2 Contents of AQI Report.
2.2.1 Content of AQI Report Requirements. An AQI report must
contain the following:
a. The reporting area(s) (the MSA or subdivision of the MSA).
b. The reporting period (the day for which the AQI is reported).
c. The main pollutant (the pollutant with the highest index
value).
d. The AQI (the highest index value).
e. The category descriptor and index value associated with the
AQI and, if choosing to report in a color format, the associated
color. Use only the following descriptors and colors for the six AQI
categories:
Table 1 to Section 2 of Appendix G--AQI Categories
------------------------------------------------------------------------
For this AQI Use this descriptor And this color \1\
------------------------------------------------------------------------
0 to 50................ ``Good''............... Green.
51 to 100.............. ``Moderate''........... Yellow.
101 to 150............. ``Unhealthy for Orange.
Sensitive Groups''.
151 to 200............. ``Unhealthy''.......... Red.
201 to 300............. ``Very Unhealthy''..... Purple.
301 and above.......... ``Hazardous''.......... Maroon\1\.
------------------------------------------------------------------------
\1\Specific color definitions can be found in the most recent reporting
guidance (Technical Assistance Document for the Reporting of Daily Air
Quality), which can be found at https://www.airnow.gov/publications/air-quality-index/technical-assistance-document-for-reporting-the-daily-aqi/.
[[Page 16404]]
f. The pollutant specific sensitive groups for any reported
index value greater than 100. The sensitive groups for each
pollutant are identified as part of the periodic review of the air
quality criteria and the NAAQS. For convenience, the EPA lists the
relevant groups for each pollutant in the most recent reporting
guidance (Technical Assistance Document for the Reporting of Daily
Air Quality), which can be found at https://www.airnow.gov/publications/air-quality-index/technical-assistance-document-for-reporting-the-daily-aqi/.
2.2.2 Contents of AQI Report When Applicable. When appropriate,
the AQI report may also contain the following, but such information
is not required:
a. Appropriate health and cautionary statements.
b. The name and index value for other pollutants, particularly
those with an index value greater than 100.
c. The index values for sub-areas of your MSA.
d. Causes for unusually high AQI values.
e. Pollutant concentrations.
f. Generally, the AQI report applies to an area's MSA only.
However, if a significant air quality problem exists (AQI greater
than 100) in areas significantly impacted by the MSA but not in it
(for example, O3 concentrations are often highest
downwind and outside an urban area), the report should identify
these areas and report the AQI for these areas as well.
2.3. Communication, Timing, and Frequency of AQI Report. The
daily AQI must be reported 7 days per week and made available via
website or other means of public access. The daily AQI report
represents the air quality for the previous day. Exceptions to this
requirement are in section 2.4 of this appendix.
a. Reporting the AQI sub-daily is recommended, but not required,
to provide more timely air quality information to the public for
making health-protective decisions.
b. Submitting hourly data in real-time to the EPA's AirNow (or
future analogous) system is recommended, but not required, and
assists the EPA in providing timely air quality information to the
public for making health-protective decisions.
c. Submitting hourly data for appropriate monitors (referenced
in section 3.2 of this appendix) satisfies the daily AQI reporting
requirement because the AirNow system makes daily and sub-daily AQI
reports widely available through its website and other communication
tools.
d. Forecasting the daily AQI provides timely air quality
information to the public and is recommended but not required. Sub-
daily forecasts are also recommended, especially when air quality is
expected to vary substantially throughout the day, like during
wildfires. Long-term (multi-day) forecasts can also be made
available when useful.
2.4. Exceptions to Reporting Requirements.
a. If the index value for a particular pollutant remains below
50 for a season or year, then it may be excluded from the
calculation of the AQI in section 3 of this appendix.
b. If all index values remain below 50 for a year, then the AQI
may be reported at the discretion of the reporting agency. In
subsequent years, if pollutant levels rise to where the AQI would be
above 50, then the AQI must be reported as required in section 2 of
this appendix.
c. As previously mentioned in section 2.3 of this appendix,
submitting hourly data in real-time from appropriate monitors
(referenced in section 3.2 of this appendix) to the EPA's AirNow (or
future analogous) system satisfies the daily AQI reporting
requirement.
3. Data Handling.
3.1 Relationship of AQI and pollutant concentrations. For each
pollutant, the AQI transforms ambient concentrations to a scale from
0 to 500. As appropriate, the AQI is associated with the NAAQS for
each pollutant. In most cases, the index value of 100 is associated
with the numerical level of the short-term standard (i.e., averaging
time of 24-hours or less) for each pollutant. The index value of 50
is associated with the numerical level of the annual standard for a
pollutant, if there is one, at one-half the level of the short-term
standard for the pollutant or at the level at which it is
appropriate to begin to provide guidance on cautionary language.
Higher categories of the index are based on the potential for
increasingly serious health effects to occur following exposure and
increasing proportions of the population that are likely to be
affected. The reported AQI corresponds to the pollutant with the
highest calculated AQI. For the purposes of reporting the AQI, the
sub-indexes for PM10 and PM2.5 are to be
considered separately. The pollutant responsible for the highest
index value (the reported AQI) is called the ``main'' pollutant for
that day.
3.2 Monitors Used for AQI Reporting. Concentration data from
State/Local Air Monitoring Station (SLAMS) or parts of the SLAMS
required by 40 CFR 58.10 must be used for each pollutant except PM.
For PM, calculate and report the AQI on days for which air quality
data has been measured (e.g., from continuous PM2.5
monitors required in appendix D to this part). PM measurements may
be used from monitors that are not reference or equivalent methods
(for example, continuous PM10 or PM2.5
monitors). Detailed guidance for relating non-approved measurements
to approved methods by statistical linear regression is referenced
here:
Reference for relating non-approved PM measurements to approved
methods (Eberly, S., T. Fitz-Simons, T. Hanley, L. Weinstock., T.
Tamanini, G. Denniston, B. Lambeth, E. Michel, S. Bortnick. Data
Quality Objectives (DQOs) For Relating Federal Reference Method
(FRM) and Continuous PM2.5 Measurements to Report an Air
Quality Index (AQI). U.S. Environmental Protection Agency, Research
Triangle Park, NC. EPA-454/B-02-002, November 2002).
3.3 AQI Forecast. The AQI can be forecasted at least 24-hours in
advance using the most accurate and reasonable procedures
considering meteorology, topography, availability of data, and
forecasting expertise. The guidance document, ``Guidelines for
Developing an Air Quality (Ozone and PM2.5) Forecasting
Program,'' can be found at https://www.airnow.gov/publications/weathercasters/guidelines-developing-air-quality-forecasting-program/.
3.4 Calculation and Equations.
a. The AQI is the highest value calculated for each pollutant as
follows:
i. Identify the highest concentration among all of the monitors
within each reporting area and truncate as follows:
(A) Ozone--truncate to 3 decimal places
PM2.5--truncate to 1 decimal place
PM10--truncate to integer
CO--truncate to 1 decimal place
SO2--truncate to integer
NO2--truncate to integer
(B) [Reserved]
ii. Using table 2 to this appendix, find the two breakpoints
that contain the concentration.
iii. Using equation 1 to this appendix, calculate the index.
iv. Round the index to the nearest integer.
Table 2 to Section 3.4 of Appendix G--Breakpoints for the AQI
--------------------------------------------------------------------------------------------------------------------------------------------------------
These breakpoints Equal these AQI's
--------------------------------------------------------------------------------------------------------------------------------------------------------
PM2.5 PM10 ([micro]g/
O3 (ppm) 8-hour O3 (ppm) 1- ([micro]g/ m\3\) 24-hour CO (ppm) 8- SO2 (ppb) 1- NO2 (ppb) 1- AQI Category
hour\1\ m\3\) 24-hour hour hour hour
--------------------------------------------------------------------------------------------------------------------------------------------------------
0.000-0.054......... .............. 0.0-9.0 0-54 0.0-4.4 0-35 0-53 0-50 Good.
0.055-0.070......... .............. 9.1-35.4 55-154 4.5-9.4 36-75 54-100 51-100 Moderate.
0.071-0.085......... 0.125-0.164 35.5-55.4 155-254 9.5-12.4 76-185 101-360 101-150 Unhealthy for
Sensitive Groups.
0.086-0.105......... 0.165-0.204 55.5-125.4 255-354 12.5-15.4 \3\ 186-304 361-649 151-200 Unhealthy.
0.106-0.200......... 0.205-0.404 125.5--225.4 355-424 15.5-30.4 \3\ 305-604 650-1249 201-300 Very Unhealthy.
[[Page 16405]]
0.201-(\2\)......... 0.405+ 225.5+ 425+ 30.5+ \3\ 605+ 1250+ 301+ \4\ Hazardous.
--------------------------------------------------------------------------------------------------------------------------------------------------------
\1\ Areas are generally required to report the AQI based on 8-hour ozone values. However, there are a small number of areas where an AQI based on 1-hour
ozone values would be more precautionary. In these cases, in addition to calculating the 8-hour ozone index value, the 1-hour ozone index value may be
calculated, and the maximum of the two values reported.
\2\ 8-hour O3 concentrations do not define higher AQI values (>301). AQI values > 301 are calculated with 1-hour O3 concentrations.
\3\ 1-hr SO2 concentrations do not define higher AQI values (>=200). AQI values of 200 or greater are calculated with 24-hour SO2 concentration.
\4\ AQI values between breakpoints are calculated using equation 1 to this appendix. For AQI values in the hazardous category, AQI values greater than
500 should be calculated using equation 1 and the concentration specified for the AQI value of 500. The AQI value of 500 are as follows: O3 1-hour--
0.604 ppm; PM2.5 24-hour--325.4 [micro]g/m\3\; PM10 24-hour--604 [micro]g/m\3\; CO ppm--50.4 ppm; SO2 1-hour--1004 ppb; and NO2 1-hour--2049 ppb.
b. If the concentration is equal to a breakpoint, then the index
is equal to the corresponding index value in table 2 to this
appendix. However, equation 1 to this appendix can still be used.
The results will be equal. If the concentration is between two
breakpoints, then calculate the index of that pollutant with
equation 1. It should also be noted that in some areas, the AQI
based on 1-hour O3 will be more precautionary than using
8-hour values (see footnote 1 to table 2). In these cases, the 1-
hour values as well as 8-hour values may be used to calculate index
values and then use the maximum index value as the AQI for
O3.
[GRAPHIC] [TIFF OMITTED] TR06MR24.048
Where:
Ip = the index value for pollutantp.
Cp = the truncated concentration of
pollutantp.
BPHi = the breakpoint that is greater than or equal to
Cp.
BPLo = the breakpoint that is less than or equal to
Cp.
IHi = the AQI value corresponding to BPHi.
Ilo = the AQI value corresponding to BPLo.
c. If the concentration is larger than the highest breakpoint in
table 2 to this appendix then the last two breakpoints in table 2
may be used when equation 1 to this appendix is applied.
Example:
d. Using table 2 and equation 1 to this appendix, calculate the
index value for each of the pollutants measured and select the one
that produces the highest index value for the AQI. For example, if a
PM10 value of 210 [micro]g/m\3\ is observed, a 1-hour
O3 value of 0.156 ppm, and an 8-hour O3 value
of 0.130 ppm, then do this:
i. Find the breakpoints for PM10 at 210 [micro]g/m\3\
as 155 [micro]g/m\3\ and 254 [micro]g/m\3\, corresponding to index
values 101 and 150;
ii. Find the breakpoints for 1-hour O3 at 0.156 ppm
as 0.125 ppm and 0.164 ppm, corresponding to index values 101 and
150;
iii. Find the breakpoints for 8-hour O3 at 0.130 ppm
as 0.116 ppm and 0.374 ppm, corresponding to index values 201 and
300;
iv. Apply equation 21 to this appendix for 210 [micro]g/m\3\,
PM10:
[GRAPHIC] [TIFF OMITTED] TR06MR24.049
v. Apply equation 3 to this appendix for 0.156 ppm, 1-hour
O3:
[GRAPHIC] [TIFF OMITTED] TR06MR24.050
vi. Apply equation 4 to this appendix for 0.130 ppm, 8-hour
O3:
[GRAPHIC] [TIFF OMITTED] TR06MR24.051
[[Page 16406]]
vii. Find the maximum, 206. This is the AQI. A minimal AQI
report could read: ``Today, the AQI for my city is 206, which is
Very Unhealthy, due to ozone.'' It would then reference the
associated sensitive groups.
[FR Doc. 2024-02637 Filed 3-5-24; 8:45 am]
BILLING CODE 6560-50-P