Draft Integrated Review Plan for the
   National Ambient Air Quality
  Standards for Particulate Matter

-------
                                      EPA-452/D-16-001
                                           April 2016
  Draft Integrated Review Plan for
the National Ambient Air Quality
Standards for Particulate Matter
           U. S. Environmental Protection Agency

           National Center for Environmental Assessment
              Office of Research and Development
                       and
            Office of Air Quality Planning and Standards
                 Office of Air and Radiation

           Research Triangle Park, North Carolina 27711

-------
                                    DISCLAIMER

This draft integrated review plan has been prepared by staff in the U.S. Environmental Protection
Agency's (EPA) Office of Air Quality Planning and Standards (OAQPS) and National Center for
Environmental Assessment (NCEA). Any findings and conclusions are those of the authors and
do not necessarily reflect the views of the EPA. This draft document is being circulated to
facilitate discussion with the Clean Air Scientific Advisory Committee and for public comment
to inform the EPA's review of the PM National Ambient Air Quality Standards. This
information is distributed for the purposes of pre-dissemination peer review under applicable
information quality guidelines. It does not represent, and should not be construed to represent,
any Agency determination or policy. Questions or comments related to this draft document
should be addressed to Dr. Scott Jenkins, U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards, C539-06, Research Triangle Park,  North Carolina 27711 (email:
jenkins.scott@epa.gov) and Dr. Jason Sacks, U.S. Environmental Protection Agency, National
Center for Environmental Assessment, MDB243-01, Research Triangle Park, North Carolina
27711 (email: sacks.jason@epa.gov).
         April 2016                        ii             Draft - Do Not Quote or Cite

-------
                              Table of Contents
1      INTRODUCTION	1-1
  1.1   LEGISLATIVE REQUIREMENTS	1-2
  1.2   OVERVIEW OF THE NAAQS REVIEW PROCESS	1-4
  1.3   HISTORY OF REVIEWS OF THE PM NAAQS	1-8
  1.4   GENERAL SCOPE OF THE CURRENT REVIEW	1-15
  1.5   ANTICIPATED SCHEDULE FOR CURRENT REVIEW	1-17
  References	1-19
2      KEY POLICY-RELEVANT ISSUES IN THE CURRENT REVIEW	2-1
  2.1   PRIMARY STANDARDS	2-2
    2.7.7   PM2.s Standards	2-2
    2.1.2   PMw Standard	2-70
    2.7.3   General Approach in the Current Review of the Primary Standards	2-73
  2.2   SECONDARY STANDARDS	2-17
    2.2.7   Non-Visibility Welfare Effects	2-75
    2.2.2   Visibility Effects	2-20
    2.2.3   General Approach in the Current Review of the Secondary Standards.... 2-25
  2.3   PM AMBIENT MONITORING	2-29
    2.3.7   PM Monitoring Networks	2-30
    2.3.2   Consideration Of PM Monitoring Issues In The Current REVIEW	2-3 7
  References	2-38
3      SCIENCE ASSESSMENT	3-1
  3.1   SCOPE OF THE PM ISA	3-1
    3.7.7   General Description	3-7
    3.7.2   Defining Policy-Relevant Literature	3-4
  3.2   ORGANIZATION OF THE PM ISA	3-6
  3.3   ASSESSMENT APPROACH	3-7
    3.3.7   Introduction	3-7
    3.3.2   Literature Search and Selection of Relevant Studies	3-70
    3.3.3   Evaluation of Individual Study Quality	3-77
  3.4   SPECIFIC ISSUES TO BE ADDRESSED IN THE PM ISA	3-12
    3.4.1   Causal Determinations from 2009 PM ISA	3-73
    3.4.2   Uncertainties/Limitations Identified in 2009 PM ISA	3-14
    3.4.3   Integration of Evidence and Causal Determinations	3-7 6
    3.4.4   Atmospheric Chemistry and Ambient Concentrations ofPM	3-75
    3.4.5   Human Exposure	3-79
    3.4.6   Dosimetry	3-20
    3.4.7   Modes of Action	3-27
    3.4.8   Health Effects	3-22
   April 2016                      iii            Draft - Do Not Quote or Cite

-------
    3.4.9  A t-Risk Lifestages and Populations and Public Health Impact	3-25
    3.4.10 Welfare Effects	3-26
  3.5  QUALITY MANAGEMENT	3-29
  3.6  SCIENTIFIC AND PUBLIC REVIEW	3-29
  References	3-30

4     HUMAN HEALTH RISK AND EXPOSURE ASSESSMENT	4-1
  4.1  INTRODUCTION	4-1
  4.2  OVERVIEW OF HEALTH RISK AND EXPOSURE ASSESSMENT FROM
      THE LAST REVIEW	4-2
    4.2.1  Key Observations in the 2010HREA	4-5
    4.2.2  Key Uncertainties in the 2010HREA	4-8
  4.3  CONSIDERATION OF POTENTIAL QUANTITATIVE ASSESSMENTS FOR
      THIS REVIEW	4-10
    4.3.1  Epidemiology-Based Risk Assessment	4-12
    4.3.2  Exposure Assessment	4-22
  4.4  SCIENTIFIC AND PUBLIC REVIEW	4-26
  References	4-27
5     WELFARE-RELATED RISK AND EXPOSURE ASSESSMENTS	5-1
  5.1  INTRODUCTION	5-1
  5.2  SUMMARY OF WELFARE RISK AND EXPOSURE ASSESSMENT FROM
      THE LAST REVIEW	5-2
    5.2.1  Quantitative Assessments of Urban Visibility Impairment in the Last Review
          	5-3
    5.2.2  Consideration of Quantitative Assessments for Additional PM-r elated
          Welfare Effects in the Last Review	5-7
  5.3  CONSIDERATION OF POTENTIAL QUANTITATIVE ASSESSMENTS FOR
      THIS REVIEW	5-8
    5.3.7  Visibility Effects	5-9
    5.3.2  Ecological Effects	5-70
    5.3.3  Climate Effects	5-73
    5.3.4  Materials Effects	5-75
    5.3.5  Characterization of Sources of Uncertainty and Consideration of Information
          Newly Available in this Review	5-16
  5.4  SCIENTIFIC AND PUBLIC REVIEW	5-19
  References	5-20
6     POLICY ASSESSMENT AND RULEMAKING	6-1
  6.1  POLICY ASSESSMENT	6-1
  6.2  RULEMAKING	6-3
  References	6-4
   April 2016                     iv           Draft - Do Not Quote or Cite

-------
                                  List of Tables
Table 1-1. Overview of Documents Developed in NAAQS Reviews	1-5
Table 1-2. Summary of National Ambient Air Quality Standards Promulgated for
         Particulate Matter 1971-2012	1-9
Table 1-3. Anticipated Schedule for the Review of the PM NAAQS	1-18
Table 4-1. Areas of Uncertainty Associated with Epidemiology-Based Risk Estimates and
         Potential Strategies for Addressing these Sources of Uncertainty	4-19
Table 4-2. Uncertainties and Limitations to Input Data used by APEX to Estimate
         Exposures	4-25
Table 5-1 Areas of Uncertainty in the Welfare Assessments and Potential Strategies to
         Address Them	5-17

                                 List of Figures
Figure 1-1. Overview of the EPA's Process for Reviewing NAAQS	1-4
Figure 2-1. Overview of General Approach for Review of Primary PM Standards	2-14
Figure 2-2. Overview of General Approach for Review of Secondary PM Standards	2-26
Figure 3-1. General process for development of Integrated Science Assessments	3-9
         April 2016                      v            Draft - Do Not Quote or Cite

-------
                   List of acronyms/abbreviations

AER         Indoor air exchange rates
APEX       Air Pollutants Exposure model
BenMAP-CE Environmental Benefits Mapping and Analysis Program - Community
             Edition
CAA         Clean Air Act
CASAC      Clean Air Scientific Advisory Committee
CHAD       Consolidated Human Activity Database
CMAQ       Community Multi-scale Air Quality modeling system
CPL         Candidate Protection Levels
C-R         Concentration-response
CSN         Chemical Speciation Network
DEARS      Detroit Exposure Aerosol Research Study
EC          Elemental carbon
U.S. EPA     United States Environmental Protection Agency
FEM         Federal Equivalent Method
FRM         Federal Reference Method
HREA       Health Risk and Exposure Assessment
IHD         Ischemic heart disease
IMPROVE   Interagency Monitoring of Protected Visual Environments
IPCC         Intergovernmental Panel on Climate Change
IRP          Integrated Review Plan
ISA         Integrated Science Assessment
LML         Lowest measured level
METS       Metabolic equivalents of work
NCEA       National  Center for Environmental Assessment
NAAQS      National  Ambient Air Quality Standards
NCh         Nitrogen dioxide
NOx         Oxides of nitrogen
OAR         Office of Air and Radiation
OAQPS      Office of Air Quality Planning and Standards
OC          Organic carbon
ORD         Office of Research and Development
PA          Policy Assessment
PM          Particulate matter
PM2.5         In general terms, particulate matter with an aerodynamic diameter less
             than or equal to a nominal 2.5 um; a measurement of fine particles

             In regulatory terms, particles with an upper 50% cut-point of 2.5 um
             aerodynamic diameter (the 50% cut point diameter is the diameter at
             which the sampler collects 50% of the particles and rejects 50% of the
             particles) and a penetration curve as measured by a reference method
             based on Appendix L of 40 CFR Part 50 and designated in accordance
             with 40 CFR Part 53, by an equivalent method designated in accordance
   April 2016
VI
Draft - Do Not Quote or Cite

-------
PMio
PMlO-2.5
PRB
SES
SHEDS-PM
SO2
SOx
SOPM
TRIM
TSP
UFP
UFVA
VAQ
VOC
WHO
WREA
with 40 CFR Part 53, or by an approved regional method designated in
accordance with Appendix C of 40 CFR Part 58

In general terms, particulate matter with an aerodynamic diameter less
than or equal to a nominal 10 um; a measurement of thoracic particles
(i.e., that subset of inhalable particles thought small enough to penetrate
beyond the larynx into the thoracic region of the respiratory tract)

In regulatory terms, particles with an upper 50% cut-point of 10± 0.5 um
aerodynamic diameter (the 50% cut point diameter is the diameter at
which the sampler collects 50% of the particles and rejects 50% of the
particles) and a penetration curve as  measured by a reference method
based on Appendix J of 40 CFR Part 50 and designated in accordance with
40 CFR Part 53 or by an equivalent method designated in accordance with
40 CFR Part 53

In general terms, particulate matter with an aerodynamic diameter less
than or equal to a nominal 10 um and greater than a nominal 2.5  um; a
measurement of thoracic coarse particulate matter or the coarse fraction of
PMio

In regulatory terms, particles with an upper 50% cut-point of 10 um
aerodynamic diameter and a lower 50% cut-point of 2.5 um aerodynamic
diameter (the 50% cut point diameter is the diameter at which the sampler
collects 50% of the particles and rejects 50% of the particles) as measured
by a reference method based on Appendix O of 40 CFR Part 50 and
designated in accordance with 40 CFR Part 53 or by an equivalent method
designated in accordance with 40 CFR Part 53

Policy relevant background
Socioeconomic status
Stochastic Human Exposure and Dose  Simulation for Particulate Matter
Sulfur dioxide
Sulfur oxides
Secondary Organic Particulate Matter
Total Risk Integrated Methodology
Total Suspended Particles
Ultrafine Particles
Urban-Focused Visibility Assessment
Visual  air quality
Volatile organic compound
World Health Organization
Welfare Risk and Exposure Assessment
   April 2016
                      vn
Draft - Do Not Quote or Cite

-------
 i                                1   INTRODUCTION

 2          The U.S. Environmental Protection Agency (EPA) is conducting a review of the existing
 3    air quality criteria for particulate matter (PM) and of the primary (health-based) and secondary
 4    (welfare-based) national ambient air quality standards (NAAQS) for PM. This review will
 5    provide an integrative assessment of relevant scientific information on PM and will focus on the
 6    basic elements of the PM NAAQS: the indicator, averaging time,  form,1 and level. These
 7    elements, which together serve to define each NAAQS, are considered collectively in evaluating
 8    the protection to public health and public welfare afforded by the  standards. The purpose of this
 9    draft Integrated Review Plan (IRP) is to communicate the plan for reviewing the air quality
10    criteria and the primary and secondary NAAQS for PM.
11          The Clean Air Scientific Advisory Committee (CASAC), an independent scientific
12    advisory committee established under the Clean Air Act, will review this draft IRP at a public
13    meeting and will provide advice to the EPA based on that review.2 The final IRP will reflect the
14    EPA staffs consideration of CASAC advice and of comments received from members of the
15    public as part of the review process. As this review progresses, the plan described in the IRP may
16    be modified to reflect information received during the review process, to address additional
17    advice received from CASAC, and/or to address comments received from the public.
18          This draft IRP is organized into six chapters. Chapter 1 presents the legislative
19    requirements for the review of the NAAQS (section 1.1), background information  on the
20    NAAQS review process (section 1.2), an overview of the decisions made in past reviews of the
21    PM NAAQS (section 1.3), the general scope of the current review (section 1.4), and the
22    anticipated schedule for the current review (section 1.5). Chapter 2 summarizes the supporting
23    rationales for the Administrator's decisions in the last review of the PM NAAQS, including the
24    important uncertainties and limitations in the scientific evidence and quantitative assessments in
25    the last review, and the key policy-relevant issues that will frame the current review. Chapters 3
26    through 6 discuss the key assessment documents for this review (i.e., Integrated Science
27    Assessment (ISA), Risk and Exposure Assessments (REAs), Policy Assessment (PA)), the
      1 The "form" of a standard defines the air quality statistic that is to be compared to the level of the standard in
      determining whether an area attains the standard.
      2 See
      http://voseimte.epa.gov/sab/sabpeople.nsfAVebCominitteesSubcominittees/CASAC%20Particulate%20Matter%20R
      eview%20Panel%20(2015 -2018) for a list of the CASAC PM Panel members.
               April 2016                       1-1             Draft - Do Not Quote or Cite

-------
 1    planned approaches for preparing these documents, and plans for their scientific and public
 2    review. Chapter 6 also includes an overview of the rulemaking process.
 3    1.1   LEGISLATIVE REQUIREMENTS
 4           Two sections of the Clean Air Act (CAA) govern the establishment and revision of the
 5    NAAQS. Section 108 (42 U.S.C. 7408) directs the Administrator to identify and list "air
 6    pollutants" that, in his or her judgment, "cause or contribute to air pollution which may
 7    reasonably be anticipated to endanger public health and welfare" and whose "presence ... in the
 8    ambient air results from numerous or diverse mobile or stationary sources" and to issue air
 9    quality criteria for air pollutants that are listed. 42 U.S.C. § 7408(a) & (b). Air quality criteria are
10    intended to "accurately reflect the latest scientific knowledge useful in indicating the kind and
11    extent of identifiable effects on public health or welfare which may be expected from the
12    presence of [a] pollutant in ambient air  ... ."  42 U.S.C. § 7408(b).
13           Section 109 (42 U.S.C. 7409) directs the Administrator to propose and promulgate
14    "primary" and "secondary" NAAQS for pollutants for which air quality criteria are issued under
15    section 108. 42 U.S.C. § 7409 (a).  Section  109(b)(l) defines a primary standard as one "the
16    attainment and maintenance of which in the judgment of the Administrator, based on such
17    criteria and allowing an adequate margin of safety, are requisite to protect the public health." 42
18    U.S.C. § 7409(b)(l).3 A secondary standard, as defined in section 109(b)(2), "shall specify a
19    level of air quality the attainment and maintenance of which in the judgment of the
20    Administrator, based on such criteria, is requisite to protect the public welfare from any known
21    or anticipated adverse effects associated with the presence of [the] pollutant in the ambient air."4
22    42 U.S.C. §7409(b)(2).
23           The requirement that primary standards provide an adequate margin of safety was
24    intended to address uncertainties associated with inconclusive scientific and technical
25    information available at the time of standard setting. It was also intended to provide a reasonable
      3 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)].
      4 Welfare effects as defined in CAA section 302(h) [42 U.S.C. 7602(h)] include, but are not limited to, "effects on
      soils, water, crops, vegetation, man-made 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."
                April 2016                         1 -2              Draft - Do Not Quote or Cite

-------
 1    degree of protection against hazards that research has not yet identified.5 Both kinds of
 2    uncertainties are components of the risk associated with pollution at levels below those at which
 3    human health effects can be said to occur with reasonable scientific certainty. Thus, in selecting
 4    primary standards that provide an adequate margin of safety, the Administrator is seeking not
 5    only to prevent pollution levels that have been demonstrated to be harmful but also to prevent
 6    lower pollutant levels that may  pose an unacceptable risk of harm, even if the risk is not
 7    precisely identified as to nature or degree.  The CAA does not require the Administrator to
 8    establish a primary NAAQS at a zero-risk level or at background concentration levels, but rather
 9    at a level that reduces risk sufficiently so as to protect public health with an adequate margin of
10    safety.6
11          In addressing the requirement for an adequate margin of safety, the EPA considers such
12    factors as the nature and severity of the health effects involved, the size of the sensitive group(s),
13    and the kind and degree of uncertainties. The selection of any particular approach to providing an
14    adequate margin of safety is a policy choice left specifically to the Administrator's judgment.7
15          In setting standards that are "requisite" to protect public health and welfare, as provided in
16    section 109(b), the EPA's task is to establish standards that are neither more nor less stringent
17    than necessary. In so doing, the EPA may  not consider the costs of implementing the standards.
18    See generally Whitman v. American Trucking Associations, 531 U.S. 457, 465-472, 475-76
19    (2001). Likewise, "[attainability and technological feasibility are not relevant considerations in
20    the promulgation of national ambient air quality standards." American Petroleum Institute v.
21    Costle, 665 F. 2d at 1185.
22          Section 109(d)(l) requires that "not later than December 31, 1980, and at 5-year
23    intervals thereafter, the Administrator shall complete a thorough  review of the criteria
24    published under section 108 and the national ambient air  quality  standards . . . and shall make
25    such revisions in such criteria and standards and promulgate such new standards as may be
26    appropriate . .  . ." 42 U.S.C. § 7409(d)(l). Sections 109(d)(2)(A) and 109(d)(2)(B) require that
27    an independent scientific review committee "shall complete a review of the criteria . . .  and the
28    national primary and secondary ambient air quality standards . . . and shall recommend to the
      5 See Lead Industries Association v. EPA, 647 F.2d 1130, 1154 (D.C. Cir 1980); American Petroleum Institute v.
      Costle, 665 F.2d 1176, 1186 (D.C. Cir. \9K\); American Farm Bureau Federation v. EPA, 559 F. 3d 512, 533 (D.C.
      Cir. 2009); snA Association of Battery Recyclers v. EPA, 604 F. 3d 613, 617-18 (D.C. Cir. 2010).
      6 See Lead Industries v. EPA, 647 F.2d at 1156 n.5\,Mississippi v. EPA, 111, F. 3d 246, 255, 262-63 (D.C. Cir.
      2013).
      7 See Lead Industries Association v. EPA, 647 F.2d at 1161-62; Mississippi v. EPA, 723 F. 3d at 265.
                April 20 J6                        1 -3              Draft - Do Not Quote or Cite

-------
 1    Administrator any new . .  . standards and revisions of existing criteria and standards as may be
 2    appropriate . . . ." 42 U.S.C. § 7409(d)(2). Since the early 1980s, this independent review
 3    function has been performed by CASAC.
 4    1.2   OVERVIEW OF  THE NAAQS REVIEW PROCESS
 5
 6
 7
 8
 9
10
11
12
The current process for reviewing the NAAQS includes five major phases: (1) the planning
phase, (2) the science assessment phase, (3) the risk/exposure assessment phase, (4) the policy
assessment phase, and (5) the rulemaking phase. Figure 1-1 provides an overview of the NAAQS
review process and Table 1-1 provides an overview of each of the documents that are developed
as part of this process.8 The phases of the NAAQS review process, and the documents
developed, are described in more detail below.
         Workshop on
       science-policy issues
                      Integrated Review Plan (IRP): timeline and key
                       policy-relevant issues and scientific questions
                                  Integrated Science Assessment (ISA): evaluation and
                                       synthesis of most policy-relevant studies
                                         Risk/Exposure Assessment (REA):
                                     quantitative assessment, as warranted, focused
                                      on key results, observations, and uncertainties
                                           Policy Assessment (PA): staff analysis of
                                            policy options based on integration and
                                         interpretation of information in the ISA and REA
                  EPA
                proposed
               decisions on
                standards



Public hearings
and comments
on proposal



Agency decision
making and draft
final notice



Interagency
review
Figure 1-1. Overview of the EPA's Process for Reviewing NAAQS
                                                                          Clean Air Scientific
                                                                         Advisory Committee
                                                                           (CASAC) review
                                                                           Public comment
                                                                                EPA final
                                                                               decisions on
                                                                               -standards
      8 The EPA maintains a website on which key documents developed for NAAQS reviews are made available
      (http://www3 .epa.gov/ttn/naaqs/).
                April 2016
                                              1-4
Draft - Do Not Quote or Cite

-------
1   Table 1-1. Overview of Documents Developed in NAAQS Reviews
     Document
Purpose
         CASAC Review
     Integrated Review
     Plan (IRP)
Presents the anticipated schedule and process for the review, and
the key policy-relevant science issues. Not intended to provide
detailed scientific or technical information, plans for quantitative
assessments, or conclusions on existing or alternative standard(s).
         One draft; consideration of CASAC
         advice and public comments on the draft
         IRP is reflected in the final IRP.
     Integrated Science
     Assessment (ISA)
Comprehensive review, synthesis and evaluation of the most
policy-relevant science, including key science judgments. Does not
present quantitative assessments or conclusions on existing or
alternative standard(s).
         Consideration of CASAC advice and
         public comments on drafts of the ISA
         (multiple drafts, as warranted) is
         reflected in the final ISA.
     Risk and Exposure
     Assessment (REA)
     Planning
     Document(s)	
Drawing upon the information in the ISA, presents the planned
approaches and scopes of the quantitative risk and exposure
analyses that are warranted. Does not present conclusions on
existing or alternative standard(s).	
         Consideration of CASAC advice and
         public comments on the REA Planning
         document is reflected in the REA(s).
     Health and
     Welfare REAs
     (HREA, WREA)
Drawing upon the information in the ISA, presents quantitative
assessments of risks and exposures, as warranted, under various air
quality scenarios, including just meeting the existing and potential
alternative standard(s). Does not present conclusions on existing or
alternative standard(s).	
         Consideration of CASAC advice and
         public comments on drafts of the
         REA(s) (multiple drafts, as warranted) is
         reflected in the final REA(s).
     Policy Assessment
     (PA)
Drawing upon the information in the ISA and REA(s), provides a
transparent staff analysis of the scientific basis for policy options
for consideration; facilitates CASAC advice to the Agency and
recommendations to the Administrator on the adequacy of the
existing standards and, as warranted, on the revisions that may be
appropriate to consider.	
         Consideration of CASAC advice and
         public comments on drafts of the PA
         (multiple drafts, as warranted) is
         reflected in the final PA.
     Proposed Rule
Communicates the Administrator's proposed decision(s); informed
by the ISA, REA(s), PA and the advice of CASAC
-  Public comment period, including public hearings as warranted.
     Final Rule
Communicates the Administrator's final decision(s); informed by
the ISA, REA(s), PA, advice of CASAC and public comments on
the proposed rule.	
                              April 2016
                                      1-5
Draft - Do Not Quote or Cite

-------
 1          The planning phase of the NAAQS review process begins with a public workshop
 2    intended to provide the EPA an opportunity to receive input and advice on the key science and
 3    policy issues around which the review will be structured. Workshop participants are asked to
 4    highlight significant new and emerging research related to these key science and policy issues,
 5    and to make recommendations to Agency staff regarding the design and scope of the
 6    review. Drawing from workshop discussions, a draft IRP is prepared jointly by the EPA's
 7    National Center for Environmental Assessment (NCEA), within the Office of Research and
 8    Development (ORD), and the EPA's Office of Air Quality Planning and Standards (OAQPS),
 9    within the Office of Air and Radiation (OAR). The IRP presents the plan for the review,
10    including the anticipated schedule, the key policy-relevant science issues that will guide the
11    review, and the approach to developing the assessment documents that will inform the review. A
12    draft of the IRP is made available for CASAC review and for public comment, and the final IRP
13    reflects the EPA staffs consideration of CASAC advice and public input.
14          Following the IRP, the EPA's NCEA develops the Integrated Science Assessment (ISA).
15    The ISA provides a concise review, synthesis, and evaluation of the most policy-relevant science
16    and includes key science judgments. The ISA provides the scientific foundation for the NAAQS
17    review, and it is intended to provide information useful in forming judgments about  air quality
18    indicator(s), averaging time(s), form(s), and level(s) for the NAAQS. The schedule typically
19    includes production of a first and, if needed,  second draft ISA, both of which undergo CASAC
20    review at public meetings.9 The final ISA reflects staffs consideration of CASAC advice and of
21    the public input provided on drafts of the ISA. Chapter 3 below provides a more detailed
22    description of the planned scope, organization, and assessment approach for the ISA in the
23    current review of the PM NAAQS.
24          Building on the assessment of the evidence in the ISA, staff in the EPA's OAQPS
25    considers the extent to which there is support for the development of Health and/or Welfare Risk
26    and Exposure Assessments (HREA, WREA). As an initial step in these considerations, staff
27    prepares Risk and Exposure Assessment planning document(s) (REA Planning Document(s))
28    that consider the extent to which the  available scientific evidence and tools/methodologies
29    warrant the conduct of quantitative assessments. As discussed in Chapters 4 and 5 below, the
30    REA Planning Document(s) will focus on  the degree to which important uncertainties from
31    previous reviews could be addressed by updated quantitative analyses, and on the degree to
      9 The availability of draft documents (ISA, REA, PA), and a request for public input on those documents, is also
      announced in the Federal Register.


               April 2016                       1 -6             Draft - Do Not Quote or Cite

-------
 1   which updated or additional analyses could improve the understanding of pollutant exposures
 2   and/or risks. To the extent warranted, the REA Planning Document(s) will also outline a general
 3   plan, including scope and methods, for conducting quantitative assessments. REA Planning
 4   Documents are typically prepared in conjunction with the first or second draft ISA and are
 5   reviewed by CASAC at a public meeting.
 6         When developed, REAs provide concise presentations of exposure and/or risk
 7   assessments, including presentations of methods, key results, and uncertainties. One or more
 8   drafts of the REA(s) undergo CASAC review at public meetings. Staff considers the CASAC
 9   advice and public input received in preparing final REAs. Chapters 4 and 5 below provide more
10   detailed descriptions of the approaches in this review of the PM NAAQS to considering the
11   potential support for an HREA and WREA, respectively.
12         Staff in the EPA's OAQPS also prepares  a Policy Assessment (PA), presenting staffs
13   considerations and conclusions regarding the adequacy of the current standards and, if warranted,
14   the range of revised standards that could be supported by the available scientific evidence and
15   exposure/risk information. The PA integrates and interprets the information from the ISA and
16   REA(s) to frame policy options for consideration by the Administrator.  Such an evaluation of
17   policy options is intended to help bridge the gap between the Agency's scientific assessments,
18   presented in the ISA and REA(s) (when available), and the judgments required of the EPA
19   Administrator in determining whether it is appropriate to retain or revise the NAAQS. The PA is
20   also intended to facilitate CAS AC's advice to the Agency and recommendations to the
21   Administrator on the adequacy of the existing standards and on revisions that may be appropriate
22   to consider.  Staffs considerations and conclusions in the PA are based on the available scientific
23   evidence and quantitative exposure and risk information, including the uncertainties and
24   limitations in that evidence and information. The PA focuses on the evidence and information
25   that is most pertinent to evaluating the basic elements of NAAQS: indicator, averaging time,
26   form, and level. The schedule typically includes production of a first and, if needed, second draft
27   PA, both of which undergo CASAC review at public meetings. Staff considers the CASAC
28   advice and public input received in preparing the final PA.
29          Following issuance of the final PA, the EPA develops and publishes a notice of proposed
30   rulemaking that communicates the Administrator's proposed decisions on the NAAQS. These
31   proposed decisions are based on the Administrator's consideration of the scientific evidence
32   assessed in the ISA; quantitative exposure and risk information presented and assessed in the
33   REAs, as available; staffs considerations and conclusions based on the evidence and
34   information, as presented in the PA; CASAC advice received during the development of the ISA,
35   REA(s), and PA; and public input received on drafts of those assessment documents. Prior to
               April 2016                       1 -7             Draft - Do Not Quote or Cite

-------
 1    publication in the Federal Register., a draft proposal notice undergoes interagency review
 2    involving other federal agencies.10 Materials upon which proposed decisions are based, including
 3    the documents described above, are made available to the public in the regulatory docket for
 4    review.11
 5          A public comment period, during which one or more public hearings are generally held,
 6    follows publication of the notice of proposed rulemaking. Taking into account comments
 7    received on the proposed rule,12 the Agency develops a final rule, a draft version of which
 8    undergoes interagency review prior to publication in the Federal Register. Chapter 6 of this IRP
 9    discusses the development of the PA and the rulemaking steps for this review of the PM
10    NAAQS.
11    1.3   HISTORY OF REVIEWS OF THE PM NAAQS
12           Particulate matter is the generic term for a broad class of chemically and physically
13    diverse  substances that exist as discrete  liquid and/or solid particles over a wide range of sizes.
14    Particles originate from a variety of anthropogenic stationary and mobile sources, as well as from
15    natural sources. Particles may be emitted directly, or formed in the atmosphere by
16    transformations of gaseous emissions such as sulfur oxides (SOx), oxides of nitrogen (NOx), and
17    volatile organic compounds (VOC). Examples of secondary particle formation include: (1) the
18    conversion of sulfur dioxide (SCh) to sulfuric acid (IHhSO/O vapor that nucleates new particles or
19    condenses on existing particles and further reacts with gaseous ammonia (NHs) to form various
20    inorganic salts (e.g., ammonium sulfate, (NH4)2SO4, or ammonium bisulfate, NH4HSO4); (2) the
21    conversion of nitrogen dioxide (NCh) to nitric acid (HNCb) vapor that condenses onto existing
22    particles and reacts further with ammonia to form ammonium nitrate (TSEUNOs);  and (3)
23    reactions involving gaseous VOC yielding organic compounds with low vapor pressures that
24    nucleate or condense on existing particles to form secondary organic particulate matter (SOPM)
      10 Where implementation of the proposed decision would have an annual effect on the economy of $100 million of
      more, (e.g., by necessitating the implementation of emissions controls) the EPA also develops and releases a draft
      regulatory impact analysis (RIA) concurrent with the notice of proposed rulemaking. This activity is conducted
      under Executive Order 12866. The RIA is conducted independently of the rulemaking process and, by law, is not
      considered in decisions regarding the NAAQS.
      11 All documents in the docket are listed in the www.regulations.gov index. Publicly available docket materials are
      available either electronically at www.regulations.gov or in hard copy at the Air and Radiation Docket and
      Information Center. The docket ID number for this review of the PM NAAQS is EPA-HQ-OAR-2015-0072.
      12 When issuing the final rulemaking, the Agency responds to all significant comments on the proposed rule.
                April 2016                        1 -8             Draft - Do Not Quote or Cite

-------
1    (U.S. EPA, 2004, Chapter 3).13 The chemical and physical properties of PM vary greatly with
2    time, region, meteorology, and source category, complicating the assessment of health and
3    welfare effects.
4           This section summarizes the PM NAAQS that have been promulgated in past reviews
5    (Table 1-2). Each of these reviews is discussed briefly below.

6    Table 1-2. Summary of National Ambient Air Quality Standards Promulgated for
7    Particulate Matter 1971-201214
Review
Completed
1971
1987
1997
2006
Indicator
Total
Suspended
Particles
(TSP)
PMio
PM25
PMio
PM25
Avg. Time
24-hour
Annual
24-hour
Annual
24-hour
Annual
24-hour
Annual
24-hour
Level
260 ug/m3
(primary)
150 ug/m3
(secondary)
75 ug/m3
(primary)
60 ug/m3
(secondary)
150 ug/m3
50 ug/m3
65 ug/m3
15.0 ug/m3
150 ug/m3
50 ug/m3
35 ug/m3
Form
Not to be exceeded more than once
per year
Annual geometric mean
Not to be exceeded more than once
per year on average over a 3 -year
period
Annual arithmetic mean, averaged
over 3 years
98th percentile, averaged over 3 years
Annual arithmetic mean, averaged
over 3 years15
99th percentile, averaged over 3
years16
Annual arithmetic mean, averaged
over 3 years
98th percentile, averaged over 3 years
     13 As discussed below (section 1.4), the ecological effects associated with deposited paniculate NOx and SOx and
     their transformation products are addressed in the review of the secondary NOx and SOx NAAQS.
     14 When not specified, primary and secondary standards are identical.
     15 The level of the 1997 annual PM2s standard was to be compared to 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"). Spatial averaging is
     discussed further in section 2.1.1, below.
     16 When the 1997 standards were vacated (see below), the form of the 1987 standards remained in place (i.e., not to
     be exceeded more than once per year on average over a 3-year period).
               April 2016
1-9
Draft - Do Not Quote or Cite

-------

2012

PMio
PM25
PMio
Annual
24-hour18
24-hour
Annual
24-hour
15.0 ug/m3
150 ug/m3
35 ug/m3
12.0 ug/m3
(primary)
15.0 ug/m3
(secondary)
150 ug/m3
Annual arithmetic mean, averaged
over 3 years17
Not to be exceed more than once per
year on average over a 3 -year period
98th percentile, averaged over 3 years
Annual mean, averaged over 3
years19
Not to be exceeded more than once
per year on average over 3 years
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
       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).20 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 micrometers (|im) (referred
to as total suspended particulates or TSP). The primary standards were set at 260 |ig/m3, 24-hour
average, not to be  exceeded more than once per year, and 75 |ig/m3, annual geometric mean. The
secondary standards were set at 150 |ig/m3, 24-hour average, not to be exceeded more than once
per year, and 60 |ig/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 PMio, 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.21 The level of the 24-
hour standards (primary and secondary)  was set at 150 |ig/m3, and the form was one expected
17 The EPA tightened the constraints on the spatial averaging criteria by further limiting the conditions under which
some areas may average measurements from multiple community-oriented monitors to determine compliance.
Spatial averaging is discussed further in section 2.1.1, below.
18 The EPA revoked the annual PMio NAAQS in 2006.
19 In the 2012 decision, the EPA eliminated the option for spatial averaging.
20 Prior to the review initiated in 2007 (see below), the AQCD provided the scientific basis for the NAAQS.
21 PMio refers to particles with a nominal mean aerodynamic diameter less than or equal to 10 um. More
specifically, 10 um is the aerodynamic diameter for which the efficiency of particle collection is 50 percent. Larger
particles are not excluded altogether, but are collected with substantially decreasing efficiency while smaller
particles are collected with increasing efficiency.
          April 2016
                                                  1-10
Draft - Do Not Quote or Cite

-------
 1    exceedance per year, on average over three years. The level of the annual standards (primary and
 2    secondary) was set at 50 |ig/m3, and the form was annual arithmetic mean, averaged over three
 3    years.
 4           In April 1994, the EPA announced its plans for the second periodic review of the air
 5    quality criteria and NAAQS for PM, and in 1997 the EPA promulgated revisions to the NAAQS
 6    (62 FR 38652, July 18, 1997). In the 1997 decision, the EPA determined that the fine and coarse
 7    fractions of PMio should be considered separately. This determination was based on evidence
 8    that serious health effects were associated with short- and long-term exposures to fine particles in
 9    areas that met the existing PMio standards. The EPA added new standards, using PM2.5 as the
10    indicator for fine particles (with PM2.5 referring to particles with a nominal mean aerodynamic
11    diameter less than or equal to 2.5  jim). These new standards were as follows: (1) an annual
12    standard with a level of 15.0 |ig/m3, based on the 3-year average of annual arithmetic mean
13    PM2.5 concentrations from  single or multiple community-oriented monitors;22 and (2) a 24-hour
14    standard with a level of 65  |ig/m3, based on the 3-year average of the 98th percentile of 24-hour
15    PM2.5 concentrations at each monitor within an area. Also, the EPA established a new reference
16    method for the measurement of PM2.5 in the ambient air and adopted rules for determining
17    attainment of the new standards. To continue to address the coarse fraction of PMio (referred to
18    as thoracic coarse particles or PMio-2.s; generally including particles with a nominal mean
19    aerodynamic diameter greater than 2.5 jim and less than or equal to 10 jim), the EPA retained the
20    annual PMio standard and revised the form of the  24-hour PMio standard to be based on the 99th
21    percentile of 24-hour PMio concentrations at each monitor in an area. The EPA revised the
22    secondary standards by making them identical in all respects to the primary standards.
23           Following promulgation of the 1997 PM NAAQS, petitions for review were filed by a
24    large number of parties, addressing a broad range of issues. In May 1999, the U.S. Court of
25    Appeals for the District of Columbia Circuit (D.C. Circuit) upheld the EPA's decision to
26    establish fine particle standards, holding that "the growing empirical evidence demonstrating a
27    relationship between fine particle pollution and  adverse health effects amply justifies
28    establishment of new fine particle standards." American Trucking Associations v. EPA, 175 F. 3d
      99
        The level of the 1997 annual PM2s standard was to be compared to 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" as discussed in section
      2.1.1, below). 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 CBS A (78
      FR 3236, January 15,2013).
               April 2016                       1-11             Draft - Do Not Quote or Cite

-------
 1    1027, 1055-56 (D.C. Cir. 1999). The D.C. Circuit also found "ample support" for the EPA's
 2    decision to regulate coarse particle pollution, but vacated the 1997 PMio standards, concluding
 3    that the EPA had not provided a reasonable explanation justifying use of PMio as an indicator for
 4    coarse particles. 175 F. 3d at 1054-55. Pursuant to the D.C. Circuit's decision, the EPA removed
 5    the vacated 1997 PMio standards, and the pre-existing 1987 PMio standards remained in place
 6    (65 FR 80776, December 22, 2000). The D.C. Circuit also upheld the EPA's determination not to
 7    establish more stringent secondary standards for fine particles to address effects on visibility. 175
 8    F. 3d at 1027.
 9           The D.C. Circuit also addressed more general issues related to the NAAQS, including
10    issues related to the consideration of costs in setting NAAQS and the EPA's approach to
11    establishing the levels of NAAQS. Regarding the cost issue, the court reaffirmed prior rulings
12    holding that in setting NAAQS the EPA is "not permitted to consider the cost of implementing
13    those standards." Id. at 1040-41. Regarding the levels of NAAQS, the court held that the EPA's
14    approach to establishing the level of the standards in 1997 (i.e., both for PM and for the ozone
15    NAAQS promulgated on the same day) effected "an unconstitutional delegation of legislative
16    authority." Id. at 1034-40. Although the court stated that "the factors EPA uses in determining
17    the degree of public health concern associated with different levels of ozone and PM are
18    reasonable," it remanded the rule to  the EPA, stating that when the EPA considers these factors
19    for potential non-threshold pollutants "what EPA lacks is  any determinate criterion for drawing
20    lines" to determine where the standards should be set.
21           The D.C. Circuit's holding on the cost and constitutional issues were appealed to the
22    United States Supreme Court. In February 2001, the Supreme Court issued a unanimous
23    decision upholding the EPA's position on both the cost and constitutional issues. Whitman v.
24    American Trucking Associations, 531 U.S. 457, 464, 475-76. On the constitutional issue, the
25    Court held that the statutory requirement that NAAQS be  "requisite" to protect public health
26    with an adequate margin of safety sufficiently guided the EPA's discretion, affirming the EPA's
27    approach of setting standards that are neither more nor less stringent than necessary.23
      23 The Supreme Court remanded the case to the Court of Appeals 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 Court of Appeals rejected
      all remaining challenges to the standards, holding that the EPA's PM2s 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).
               April 2016                        1-12            Draft - Do Not Quote or Cite

-------
 1          In October 1997, the EPA published its plans for the third periodic review of the air
 2    quality criteria and NAAQS for PM (62 FR 55201, October 23, 1997). After CASAC and public
 3    review of several drafts, the EPA's NCEA finalized the AQCD in October 2004 (U. S. EPA,
 4    2004). The EPA's OAQPS finalized a Risk Assessment and Staff Paper in December of 2005
 5    (Abt, 2005; U.S. EPA, 2005).24 On December 20, 2005, the EPA announced its proposed
 6    decision to revise the NAAQS for PM, and solicited comment on a broad range of options (71
 7    FR 2620, January  17, 2006).
 8          On September 21, 2006, the EPA announced its final decisions to revise the primary and
 9    secondary NAAQS for PM to provide increased protection of public health and welfare,
10    respectively (71 FR 61144, October 17, 2006). With regard to the primary and secondary
11    standards for fine particles, the EPA revised the level of the 24-hour PM2.5 standards to 35
12    |ig/m3, retained the level of the annual PM2.5 standards at 15.0 |ig/m3, and revised the form of the
13    annual PM2.5 standards by narrowing the constraints on the optional use of spatial averaging.25
14    With regard to the primary and secondary standards for PMio, the EPA retained the 24-hour
15    standards, with levels at 150 |ig/m3, and revoked the annual standards. The Administrator judged
16    that the available evidence generally did not suggest a link between long-term exposure to
17    existing  ambient levels of coarse particles and health or welfare effects. In addition, a new
18    reference method was added for the measurement of PMio-2.5  in the ambient air, in order to
19    provide a basis for approving federal equivalent methods (FEMs) and to promote the gathering
20    of scientific data to support future reviews of the PM NAAQS.
21          Several parties filed petitions for review following promulgation of the revised PM
22    NAAQS in 2006. These petitions addressed the following issues: (1) selecting the level of the
23    primary  annual PM2.5 standard; (2) retaining PMio as the indicator of a standard for thoracic
24    coarse particles, retaining the level and form of the 24-hour PMio standard, and revoking the
25    PMio annual standard; and (3) setting the secondary PM2.5 standards identical to the primary
26    standards. On February 24, 2009, the U.S.  Court of Appeals for the District of Columbia Circuit
27    issued its opinion in the case American Farm Bureau Federation v. EPA, 559 F. 3d 512 (D.C.
28    Cir. 2009). The court remanded the primary annual PM2.5 NAAQS to EPA because EPA failed to
29    adequately explain why the standards provided the requisite protection from both short- and
      24 Prior to the review initiated in 2007, the Staff Paper, rather than the PA, presented the EPA staffs 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.
      25 Spatial averaging is discussed in more detail in section 2.1.1, below.


               April 2016                       1-13            Draft - Do Not Quote or Cite

-------
 1    long-term exposures to fine particles, including protection for at-risk populations. American
 2    Farm Bureau Federation v. EPA, 559 F. 3d 512, 520-27 (D.C. Cir. 2009). With regard to the
 3    standards for PMio, the court upheld EPA's decisions to retain the 24-hour PMio standard to
 4    provide protection from thoracic coarse particle exposures and to revoke the annual PMio
 5    standard. American Farm Bureau Federation, 559 F. 2d at 533-38. With regard to the secondary
 6    PM2.5 standards, the court remanded the standards to EPA because the Agency failed to
 7    adequately explain why setting the secondary PM standards identical to the primary standards
 8    provided the required protection for public welfare, including protection from visibility
 9    impairment. American Farm Bureau Federation, 559 F. 2d at 528-32. The EPA responded to the
10    court's remands as part of the next review of the PM NAAQS, which was initiated in 2007
11    (discussed below).
12           In June 2007, the EPA initiated the fourth periodic review of the air quality criteria and
13    the PM NAAQS by issuing a call for information in the Federal Register (72 FR 35462, June 28,
14    2007). Based on the NAAQS review process, as revised in 2008 and again in 2009,26 the EPA
15    held science/policy issue workshops on the primary and secondary PM NAAQS (72 FR 34003,
16    June 20, 2007; 72 FR 34005, June 20, 2007), and prepared and released the planning and
17    assessment documents that comprise the review process (i.e., IRP (U.S. EPA, 2008), ISA (U.S.
18    EPA, 2009a), REA planning documents for health and welfare (U. S. EPA, 2009b, c), a
19    quantitative health risk assessment (U.S. EPA, 2010a)27 and an urban-focused visibility
20    assessment (U.S. EPA, 2010b),28 and PA (U.S. EPA, 2011)). In June 2012, the EPA announced
21    its proposed decision to revise the NAAQS for PM (77 FR 38890, June 29, 2012).
      26 The history of the NAAQS review process, including revisions to the process, is discussed at
      http://www3.epa.gov/ttn/naaqs/review2.html.
      27 The quantitative assessment of health risks conducted in the last review was presented in the Quantitative Health
      Risk Assessment for P articulate Matter (U.S. EPA, 2010a). In the current review, quantitative assessments for
      health-related exposures and risks, if warranted, would be presented in the Health Risk and Exposure Assessment
      (HREA). For consistency with the documents developed under the current NAAQS process, the Quantitative Health
      Risk Assessment for P articulate Matter (U.S. EPA, 2010a) from the last review will be referenced in this document
      as the 2010 HREA.
      28 The quantitative assessment of welfare effects conducted in the last review was presented, in part, in the Urban-
      Focused Visibility Assessment (U.S. EPA, 2010b). In the current review, quantitative assessments for welfare
      effects, if warranted, would be presented in the Welfare Risk and Exposure Assessment (WREA). The Urban-
      Focused Visibility Assessment (U.S. EPA, 2010b) from the last review will be referenced in this document as the
      2010UFVA.
                April 2016                        1-14            Draft - Do Not Quote or Cite

-------
 1           In December 2012, the EPA announced its final decisions to revise the primary NAAQS
 2    for PM to provide increased protection of public health (78 FR 3086, January 15, 2013).29 With
 3    regard to primary standards for PM2.5, the EPA revised the level of the annual PM2.5 standard30 to
 4    12.0 |ig/m3 and retained the 24-hour PM2.5 standard, with its level of 35 |ig/m3. For the primary
 5    PMio standard, the EPA retained the 24-hour standard to continue to provide protection against
 6    effects associated with short-term exposure to thoracic coarse particles (i.e., PMio-2.s). With
 7    regard to the secondary PM standards, the EPA generally retained the 24-hour and annual PM2.5
 8    standards31 and the 24-hour PMio standard to address visibility and non-visibility welfare effects.
 9    On judicial review, the revised standards were upheld in all respects. NAM v EPA, 750 F.3d 921
10    (D.C. Cir. 2014).
11    1.4   GENERAL SCOPE OF THE CURRENT REVIEW
12          With regard to  scope, this review is focused on the air quality criteria for PM and on the
13    primary and secondary NAAQS for PM2.5 and PMio. As discussed above, the current primary
14    and secondary PM2.5 standards are meant to protect against the health and welfare effects,
15    respectively, that have been associated with exposures to fine particles. The primary and
16    secondary PMio standards are meant to protect against the effects associated with exposures to
17    thoracic coarse particles (i.e., PMio-2.s). Therefore,  an important aspect of the current review will
18    be the EPA's assessment of the health and welfare  effects that have been associated with size
19    fractionated PM mass, with a particular focus on the PM2.5 and PMio-2.5 size fractions. In
20    addition, as in the last review, the EPA will also assess the available scientific evidence for
21    health or welfare effects associated with additional size fractions (e.g., ultrafine particles) and
22    with particular PM components or groups of components, sources, or environments (e.g., urban
23    and non-urban environments).
24          Based on the available scientific information, the EPA will consider the  extent to which
25    the current PM2.5 and PMio standards are requisite to protect public health and welfare, within
26    the meaning of section 109(b) of the CAA (section 1.1, above). To the extent the available
27    information calls into  question the adequacy of the protection afforded by one  or more of the
28    existing PM standards, the EPA will consider potential alternatives that could be supported by
      31
29 The bases for these decisions are discussed further in sections 2.1 and 2.2, below.
30 The EPA also eliminated the option for spatial averaging (section 2.1.1, below).
31 Consistent with the primary standard, the EPA eliminated the option for spatial averaging with the annual
standard (section 2.1.1, below).
               April 2016                       1-15             Draft - Do Not Quote or Cite

-------
 1    the available scientific evidence and, as available, exposure-/risk-based information, in terms of
 2    the basic elements of the NAAQS (indicator, averaging time, form, level).
 3          With regard to the secondary standards in particular, as in the last review (77 FR 38991,
 4    June 29, 2012), this review of the PM NAAQS will consider the PM-related welfare effects that
 5    are not being addressed in the ongoing review of the secondary NOx and SOx NAAQS.
 6    Specifically, the review of the secondary NOx and SOx NAAQS is addressing the ecological
 7    effects of ecosystem loading of oxides of nitrogen and oxides of sulfur, which includes
 8    particulate nitrogen and sulfur compounds.32 In this review of the PM NAAQS, welfare effects to
 9    be considered include PM-related visibility impairment, climate effects, ecological effects (other
10    than those attributable to deposition of NOx, SOx and their transformation products) and
11    materials damage and soiling (i.e., materials effects). While N and/or S may contribute to these
12    welfare effects, other components of PM play a role as  well. Therefore, it is appropriate to
13    consider these welfare effects in this review of the PM NAAQS to the extent that they are not
14    being covered in the ongoing review of the secondary NOx and SOx NAAQS. In the case of
15    materials effects, the impacts of gaseous and paniculate N and S wet deposition cannot be clearly
16    distinguished, so both will be considered in this review of the PM NAAQS, along with other
17    materials effects of PM-related  deposition.
18          In contrast, as noted above, the ongoing review of the secondary NAAQS for ecological
19    effects of NOx and SOx is considering the ecological effects that result specifically from the
20    deposition of NOx and SOx and their transformation products.33 This approach is consistent with
21    the previous review of the secondary NO2 and SO2 standards,34 in which the EPA conducted a
22    joint review focused on the protection provided against the effects of deposition of NOx and SOx
23    to sensitive aquatic and terrestrial ecosystems. The approach in the last review recognized that
24    NOx and SOx, and their associated transformation products, are linked from the perspectives of
25    both atmospheric chemistry and ecological effects, providing "a strong basis for considering
26    these pollutants together" (77 FR 20222, April 3, 2012).
      32 See http://www3.epa.gov/ttn/naaqs/standards/no2so2sec/index.html for more information on the current review
      of the secondary NOx and SOx NAAQS.
      33 More broadly, the review of the secondary NAAQS for ecological effects of NOx and SOx includes consideration
      of the ecological effects of ecosystem loading of paniculate nitrogen and sulfur compounds.
      34 When discussing specific standards, we refer to those standards by their indicator species (i.e., in this case NO2
      and SO2). As described in the draft NOX/SOXIRP (U.S. EPA, 2015, p. 1-1) the terms NO2 and SO2 are not
      interchangeable with the terms NOx and SOx. NOx and SOx include many more chemical species than just NO2 and
      SO2.
                April 2016                       1-16             Draft - Do Not Quote or Cite

-------
 1   1.5  ANTICIPATED SCHEDULE FOR CURRENT REVIEW
 2         In December 2014, the EPA announced the initiation of the current periodic review of the
 3   air quality criteria for PM and of the PM2.5 and PMio NAAQS and issued a call for information
 4   in the Federal Register (79 FR 71764, December 3, 2014). On February 9 to 11, 2015, the EPA's
 5   NCEA and OAQPS held a public workshop to inform the planning for the current review of the
 6   PM NAAQS (announced in 79 FR 71764, December 3, 2014). This workshop was meant to
 7   provide the EPA an opportunity to receive input and advice on the key science and policy issues
 8   around which the PM NAAQS review will be structured. Workshop participants were asked to
 9   highlight significant new and emerging PM research related to these key science and policy
10   issues,  and to make recommendations to the Agency regarding the design and scope of this
11   review. The input received at this workshop was considered by the EPA staff in developing this
12   draft IRP.  The EPA's anticipated schedule for the remainder of this review is summarized in
13   Table 1-3, below.
              April 2016                     1-17            Draft - Do Not Quote or Cite

-------
1   Table 1-3. Anticipated Schedule for the Review of the PM NAAQS
      Stage of
      Review
Major Milestone
            Actual or Target
            Date
      Planning
Literature Search
Federal Register Call for Information
Workshop on Science/Policy Issues
Release Draft IRP for CASAC/public review
CASAC Review Meeting for Draft IRP
Release Final IRP
            Ongoing
            Decembers, 2014
            February 9-11, 2015
            April 2016
            May 23, 2016
            September 2016
      Science
      Assessment
Release First Draft ISA for CASAC/public review
CASAC Review Meeting for First Draft ISA
Release Second Draft ISA for CASC/public review
CASAC Review Meeting for Second Draft ISA
Release Final ISA
            Spring 2017
            Summer 2017
            Winter 2018
            Spring 2018
            Spring 2019
      Risk/Exposure
      Assessments
Release REA Planning Document(s) for CASAC/public
review
CASAC Review Meeting for REA Planning Document(s)
Release First Draft REA(s) for CASAC/Public Review
CASAC Review Meeting for First Draft REA(s)
Release Second Draft REA(s) for CASAC/Public Review
CASAC Review Meeting for Second Draft REA(s)
Release Final REA(s)
            Spring/Summer 2017

            Summer 2017
            Winter/Spring 2018
            Spring 2018
            Fall 2018
            Fall/Winter 2018
            Winter 2019
      Policy
      Assessment/
      Rulemaking
Release First Draft PA for CASAC/public review
CASAC Review Meeting on First Draft PA
Release Second Draft PA
CASAC Review/Public Comment on Second Draft PA
Release Final PA
Proposed Rulemaking
Final Rulemaking
            Fall 2018
            Fall/Winter 2018
            Fall 2019
            Fall/Winter 2019
            Winter 2020
            2020
            2021
             April 2016
                        1-18
Draft - Do Not Quote or Cite

-------
 1    REFERENCES

 2    DHEW (1969). Air Quality Criteria for Particulate Matter. U. S. Department of Health,
 3          Education, and Welfare. Public Health Service, Environmental Health Service, National
 4          Air Pollution Control Administration, Washington, D.C. January 1969.
 5    U.S. EPA (2004). Air Quality Criteria for Particulate Matter. National Center for Environmental
 6          Assessment, Office of Research and Development, U.S. Environmental Protection
 7          Agency, Research Triangle Park, NC 27711; report no. EPA/600/P-99/002aF and
 8          EPA/600/P-99/002bF. October 2004. Available:
 9          http://www3.epa.gov/ttn/naaqs/standards/pm/s_pm_cr_cd.html.

10    U. S. EPA (2005). Review of the National Ambient Air Quality Standards for Particulate Matter:
11          Policy Assessment of Scientific and Technical Information, OAQPS Staff Paper.
12          Research Triangle Park, NC 27711:  Office of Air Quality Planning and Standards. Report
13          No. EPA-452/R-05-005a. December 2005. Available:
14          http://www3.epa.gov/ttn/naaqs/standards/pm/sjm crsp.html.

15    U. S. EPA (2008). Integrated  Review Plan for the National Ambient Air Quality Standards for
16          Particulate Matter. National Center for Environmental Assessment and Office of Air
17          Quality Planning and Standards, U.S. Environmental Protection Agency, Research
18          Triangle Park, NC. Report No. EPA 452/R-08-004. March 2008. Available:
19          http://www3.epa.gov/ttn/naaqs/standards/pm/sjm 2007jd.html.

20    U.S. EPA (2009a). Integrated Science Assessment for Particulate Matter: Final Report. National
21          Center for Environmental Assessment-RTF Division, Office of Research and
22          Development, Research Triangle Park, NC. EPA/600/R-08/139F. December 2009.
23          Available: http://www3.epa.gov/ttn/naaqs/standards/pm/sjm 2007 isa.html.

24    U.S. EPA (2009b). Particulate Matter National Ambient Air Quality Standards: Scope and
25          Methods Plan for Health Risk and Exposure Assessment. Office of Air Quality Planning
26          and Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC.
27          EPA-452/P-09-002. February 2009.  Available:
28          http://www3.epa.gov/ttn/naaqs/standards/pm/s_pm_2007_pd.html.

29    U. S. EPA (2009c). Particulate Matter National Ambient Air Quality Standards: Scope and
30          Methods Plan for Urban Visibility Impact Assessment. Office of Air Quality Planning
31          and Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC.
32          EPA-452/P-09-001. February 2009.  Available:
33          http://www3.epa.gov/ttn/naaqs/standards/pm/sjm 2007jd.html.

34    U.S. EPA (2010a). Quantitative Health Risk Assessment for Particulate Matter - Final Report.
35          Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency,
36          Research Triangle Park, NC. EPA-452/R-10-005. June 2010. Available:
37          http://www3.epa.gOv/ttn/naaqs/standards/pm/s  pm 2007 risk.html.
               April 2016                       1-19            Draft - Do Not Quote or Cite

-------
 1   U.S. EPA (201 Ob). Paniculate Matter Urban-Focused Visibility Assessment - Final Report.
 2          Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency,
 3          Research Triangle Park, NC. EPA-452/R-10-004. July 2010. Available:
 4          http://www3.epa.gov/ttn/naaqs/standards/pm/sjm 2007  risk.html.

 5   U.S. EPA (2011).  Policy Assessment for the Review of the Particulate Matter National Ambient
 6          Air Quality Standards. Office of Air Quality Planning and Standards, U.S. Environmental
 7          Protection Agency, Research Triangle Park, NC. EPA 452/R-l 1-003. April 2011.
 8          Available: http://www3.epa.gov/ttn/naaqs/standards/pm/s_pm 2007_pa.html.

 9   U.S. EPA (2015).  Draft Integrated Review Plan for the Secondary National Ambient Air Quality
10          Standards for Oxides of Nitrogen and Oxides of Sulfur. Office of Air Quality Planning
11          and Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC.
12          October 2015.
13          https://www3.epa.gov/ttn/naaqs/standards/no2so2sec/data/IRP_NOXSOX_CompleteDraf
14          t  102915  final.pdf

15
               April 2016                       1 -20            Draft - Do Not Quote or Cite

-------
 i    2  KEY POLICY-RELEVANT ISSUES IN THE CURRENT
 2                                          REVIEW

 3          In order to inform the Administrator's decisions on the adequacy of the existing primary
 4    and secondary PM NAAQS, in the current review we35 will address the following overarching
 5    question:
 6    •   Does the currently available scientific evidence and exposure-/risk-based information
 7       support or call into question the adequacy of the protection afforded by the current
 8       primary and/or secondary PM standards?
 9    If warranted, we will also address a second overarching question:
10    •   What alternative standards are supported by the currently available scientific evidence
11       and exposure-/risk-based information, and are appropriate for consideration?
12    To inform our evaluation of these overarching questions, we will identify a number of more
13    specific policy-relevant questions for consideration (see below). These policy-relevant questions
14    will focus on key issues reflecting aspects of the health and welfare  effects evidence, air quality
15    information, and information from quantitative exposure and risk assessments that can inform
16    the Administrator's decisions in the current review. Questions will build upon the conclusions
17    from the last review on the evidence and information, including conclusions on the uncertainties
18    and limitations in that evidence and information.
19          Sections 2.1 and 2.2 below summarize the decisions made in the last review of the PM
20    NAAQS, the interpretations of the underlying scientific evidence and information supporting
21    those decisions, the important uncertainties and limitations in the evidence and information in the
22    last review, and potential key policy-relevant questions for the current review.36 Section 2.1
23    focuses on the primary PM  standards, and section 2.2 focuses on the secondary PM standards.
24    Section 2.3 provides an overview of the PM ambient monitoring networks and the key
25    monitoring-related issues to be considered as  part of the current standards review.
      35 In this document, the terms "we" or "our" refer to staff in the EPA's OAQPS and/or NCEA.
      36 Based on CASAC advice and public input on this draft IRP, we will modify these potential questions as
      appropriate for the final IRP. The key policy-relevant questions identified in the final PM IRP will provide an
      overall framework for the development of the PM ISA, REAs (if warranted), and the PA.


            April 2016                       2-1             Draft - Do Not Quote or Cite

-------
 1    2.1  PRIMARY STANDARDS
 2           The last review of the primary PM NAAQS was completed in 2012 (78 FR 3086, January
 3    15, 2013). As noted above (section 1.3), in the last review the EPA lowered the level of the
 4    primary annual PM2.5 standard from 15.0 to 12.0 |j,g/m3,37 retained the existing 24-hour PM2.5
 5    standard with its level of 35 |J,g/m3, and retained the existing 24-hour PMio standard with its level
 6    of 150  |j,g/m3. Sections 2.1.1 and 2.1.2 below discuss the primary PM2.5 and PMio standards,
 7    respectively. These sections summarize the rationales for the decisions made in the last review,
 8    including the Agency's consideration of important uncertainties and limitations in the scientific
 9    evidence and in the air quality and risk information. Section 2.1.3 provides an overview of the
10    general approach and the potential key policy-relevant questions in the current review of the
11    primary PM2.5 and PMio standards.

12    2.1.1  PM2.5 Standards
13           The 2012 decision to strengthen the suite of primary PM2.5 standards was based on the
14    Administrator's consideration of the extensive body of scientific evidence assessed in the 2009
15    PM ISA (U.S. EPA, 2009); the quantitative risk analyses presented in the 2010 HREA (U.S.
16    EPA, 2010a);38 the advice and recommendations of CAS AC (e.g., Samet, 2009; 2010a, b); and
17    public comments on the proposed rule (78 FR 3086, January 15, 2013;  U.S. EPA, 2012). In
18    particular, the Administrator noted the "strong and generally robust body of evidence of serious
19    health effects associated with both long- and short-term exposures to PM2.s" (78 FR 3120,
20    January 15, 2013). This included epidemiologic studies reporting health effect associations based
21    on long-term average PM2.5 concentrations ranging from about 15.0 |j,g/m3 or above (i.e., at or
22    above the level of the then-existing annual standard) to concentrations "significantly below the
23    level of the annual standard" (78 FR 3120, January 15, 2013).  The Administrator further
24    observed that such studies were part of an overall pattern across a broad range of studies
25    reporting positive associations, which were frequently statistically significant. Based on her
26    "confidence in the association between exposure to PM2.5 and  serious public health effects,
27    combined with evidence of such an association in areas that would meet the current standards"
      37 The Agency also eliminated spatial averaging provisions as part of the form of the annual standard.
      38 In the last review, the EPA generated a quantitative health risk assessment for PM, and did not conduct an
      exposure assessment (U.S. EPA, 2010a). To be consistent with our general process for reviewing the NAAQS
      (section 1.2, above), and with our discussion of potential quantitative analyses in the current review (Chapter 4,
      below), we refer to the 2010 health risk assessment as the 2010 HREA.

            April 2016                        2-2            Draft - Do Not Quote or Cite

-------
 1    (78 FR 3120, January 15, 2013), the Administrator concluded that revision of the suite of
 2    primary PM2.5 standards was necessary in order to provide increased public health protection.
 3    Specifically, she concluded that the then-existing suite of primary PM2.5 standards was not
 4    sufficient, and thus not requisite, to protect public health with an adequate margin of safety. This
 5    decision was consistent with advice received from CAS AC (Samet, 201 Ob).
 6           The Administrator next considered what specific revisions to the existing primary PM2.5
 7    standards were appropriate, given the available evidence and quantitative risk information. She
 8    considered both the annual and 24-hour PM2.5 standards, focusing on the basic  elements of those
 9    standards (i.e., indicator, averaging time, form, and level). These considerations, and the
10    Administrator's conclusions, are summarized below.
11    Indicator
12           In initially setting standards for fine particles in 1997, the EPA concluded it was
13    appropriate to control fine particles as a group, based on PM2.5 mass, rather than singling out any
14    particular component or class of fine particles (62 FR 38667, July 18, 1997). In the review
15    completed in 2006, based on similar considerations, the EPA concluded that the available
16    information supported retaining the PM2.5 indicator and remained too limited to support a distinct
17    standard for any  specific PM2.5 component or group of components associated with particular
18    source categories of fine particles (71 FR 61162 to 61164, October 17, 2006).
19           In the last review, the EPA again considered issues related to the appropriate indicator for
20    fine particles, with a focus on evaluating support for the existing PM2.5  mass-based indicator and
21    for potential alternative indicators based on the ultrafme particle fraction or on  fine particle
22    composition (78  FR 3121, January 15, 2013).39 With regard to PM2.5 mass, as in the 1997 and
23    2006 reviews, the health studies available during the last review continued to link adverse health
24    outcomes (e.g., premature mortality, hospital admissions, emergency department visits) with
25    long- and short-term  exposures to fine particles indexed largely by PM2.5 mass  (78 FR 3121,
26    January 15, 2013).  With regard to the ultrafme fraction of ambient PM, the PA noted the limited
27    body of health evidence assessed in the ISA (summarized in U.S. EPA, 2009, section 2.3.5 and
28    Table 2-6) and the limited monitoring information available to characterize ambient
29    concentrations of ultrafme particles (U.S. EPA, 2011, section 1.3.2 and Appendix B, section
30    B. 1.3). With regard to PM composition, the ISA concluded that "the evidence is not yet
      39 In the last review, the ISA defined ultrafme particles as generally including particles with a mobility diameter less
      than or equal to 0.1 um. Mobility diameter is defined as the diameter of a particle having the same diffusivity or
      electrical mobility in air as the particle of interest, and is often used to characterize particles of 0.5 um or smaller
      (U.S. EPA, 2009a, pp. 3-2 to 3-3).

            April 2016                        2-3             Draft - Do Not Quote or Cite

-------
 1    sufficient to allow differentiation of those constituents or sources that are more closely related to
 2    specific health outcomes" (U.S. EPA, 2009, pp. 2-26 and 6-212; 78 FR 3123, January 15, 2013).
 3    The PA further noted that "many different constituents of the fine particle mixture as well as
 4    groups of components associated with specific source categories of fine particles are linked to
 5    adverse health effects" (U.S. EPA, 2011, p. 2-55; 78 FR 3123, January 15, 2013). Consistent
 6    with the conclusions of the PA, CAS AC advised that it was appropriate to consider retaining
 7    PM2.5 as the indicator for fine particles. CASAC specifically stated that "[fjhere [is] insufficient
 8    peer-reviewed literature to support any other indicator at this time" (Samet, 2010a, p. 12). The
 9    Administrator concurred with the conclusions of the Policy Assessment and with CASAC
10    recommendations, and concluded that it was "appropriate to retain PM2.5 as the indicator for fine
11    particles" (78 FR 3123, January  15, 2013).
12    Averaging time
13           In 1997, the EPA set an annual PM2.5 standard to provide protection from health effects
14    associated with long- and short-term exposures to PM2.5, and a 24-hour standard to supplement
15    the protection afforded by the annual  standard (62 FR 38667 to 38668, July, 18, 1997). In the
16    2006 review, the EPA retained both annual and 24-hour averaging times (71 FR 61164,  October
17    17,2006).
18           In the last review, the EPA again considered issues related to the appropriate averaging
19    times for PM2.5 standards, with a focus on evaluating support for the existing annual and 24-hour
20    averaging times and for potential alternative averaging times based on sub-daily or seasonal
21    metrics. Based on the evidence assessed in the ISA, the PA noted that the overwhelming
22    majority of studies that had been conducted since the 2006 review continued to utilize annual (or
23    multi-year) or 24-hour PM averaging  periods (U.S. EPA, 2011, section 2.3.2). With regard to
24    potential support for an averaging time shorter than 24-hours, the PA noted that studies of
25    cardiovascular effects associated with sub-daily PM concentrations had evaluated a variety  of
26    PM metrics (e.g., PM2.5, PMio, PMio-2.5, ultrafme particles), averaging periods (e.g., 1, 2, and 4
27    hours), and health outcomes (U.S. EPA, 2011,  section 2.3.2). The PA concluded that this
28    information, when viewed as a whole, was too uncertain to  serve as a basis for establishing  a
29    primary PM2.5 standard with an averaging time shorter than 24-hours (U.S. EPA, 2011, p. 2-
30    57).40 With regard to potential support for a seasonal averaging time, few studies were available
      40 For respiratory effects specifically, the Administrator further noted the ISA conclusion that the strongest
      associations were observed with 24-hour average or longer exposures, not with exposures less than 24-hours (U.S.
      EPA, 2009, section 6.3).

            April 2016                       2-4             Draft - Do Not Quote or Cite

-------
 1    to deduce a general pattern in PIVb.s-related risk across seasons, and these studies did not provide
 2    information on health effects associated with season-long exposures to PM2.5 (U. S. EPA, 2011,
 3    p. 2-58; 78 FR 3124, January  15,2013).
 4           The PA reached the overall conclusions that the available information provided strong
 5    support for considering retaining the current annual and 24-hour averaging times and did not
 6    provide support for considering alternative averaging times (U.S. EPA, 2011, p. 2-58). CASAC
 7    agreed that these conclusions were reasonable (Samet, 2010a, p. 13). The Administrator
 8    concurred with the PA conclusions and with CASAC's advice. Specifically, she judged that it
 9    was "appropriate to retain the current annual and 24-hour averaging times for the primary PM2.5
10    standards to protect against health effects associated with long- and short-term exposure periods"
11    (78FR3124, January 15, 2013).
12    Form
13           In 1997, the EPA established the form of the annual PM2.5 standard as an annual
14    arithmetic mean,  averaged over 3 years, from single or multiple community-oriented monitors.41
15    That is, the level  of the annual standard was to be compared to measurements made at each
16    community-oriented monitoring site or, if specific criteria were met, measurements from
17    multiple community-oriented monitoring sites could be averaged together (i.e., spatial
18    averaging)42 (62 FR 38671 to 38672, July 18, 1997). In the 1997 review, the EPA also
19    established the form of the 24-hour PM2.5 standard as the 98th percentile of 24-hour
20    concentrations at each monitor within an area (i.e., no spatial averaging), averaged over three
21    years (62 FR at 38671 to 38674, July 18,  1997). In the 2006 review, the EPA retained these
22    standard forms but tightened the criteria for using spatial  averaging with the annual standard (78
23    FR3124, January 15, 2013).43
      41 As noted above (section 1.3), in the last review 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 CBS A (78 FR
      3236, January 15, 2013). CBSAs are required to have at least one area-wide monitor sited in the area of expected
      maximum PM2 5 concentration.
      42 The original criteria for spatial averaging included: (1) the annual mean concentration at each site shall be within
      20% of the spatially averaged annual mean, and (2) the daily values for each monitoring site pair shall yield a
      correlation coefficient of at least 0.6 for each calendar quarter (62 FR 38671 to 38672, July 18, 1997).
      43 Specifically, the Administrator revised spatial averaging criteria such that "(1) [t]he annual mean concentration at
      each site shall be within 10 percent of the spatially averaged annual mean, and (2) the daily values for each
      monitoring site pair shall yield a correlation coefficient of at least 0.9 for each calendar quarter (71 FR 61167,
      October 17, 2006).
            April 2016                        2-5              Draft - Do Not Quote or Cite

-------
 1           In the last review, the EPA's consideration of the form of the annual PM2.5 standard again
 2    included a focus on the issue of spatial averaging. An analysis of air quality and population
 3    demographic information indicated that the highest PM2.5 concentrations in a given area tended
 4    to be measured at monitors in locations where the surrounding populations were more likely to
 5    live below the poverty line and to include larger percentages of racial and ethnic minorities (U.S.
 6    EPA, 2011, p. 2-60). Based on this analysis, the PA concluded that spatial averaging could result
 7    in disproportionate impacts in at-risk populations, including minority populations and
 8    populations with lower socioeconomic status  (SES). Therefore, the PA concluded that it was
 9    appropriate to consider revising the form of the annual PM2.5 standard such that it did not allow
10    for the use of spatial averaging across monitors (U.S. EPA, 2011, p. 2-60). The CASAC agreed
11    with staff conclusions that it was "reasonable" for the EPA to eliminate the spatial averaging
12    provisions (Samet, 2010b, p.  2), stating the following: "Given mounting evidence showing that
13    persons with lower SES levels are a susceptible group for PM-related health risks, CASAC
14    recommends that the provisions that allow for spatial averaging across monitors be eliminated"
15    (Samet, 201 Oa, p. 13).
16           The Administrator concluded that public health would not be protected with an adequate
17    margin of safety in all locations, as  required by law, if disproportionately higher PM2.5
18    concentrations in low income and minority communities were averaged together with lower
19    concentrations measured at other sites in a large urban area. Therefore, she concluded that the
20    form of the annual PM2.5 standard should be revised to eliminate spatial averaging provisions.
21    Thus, the level of the annual PIVfo.s standard established in the last review is to be compared with
22    measurements from each appropriate monitor in an area, with no allowance for spatial averaging.
23           In the last review, the EPA also considered the form of the 24-hour PM2.5 standard. The
24    Agency recognized that the existing 98th percentile form for the 24-hour standard was originally
25    selected to provide a balance between limiting the occurrence of peak 24-hour PM2.5
26    concentrations and identifying a stable target  for risk management programs.44 Updated air
27    quality analyses in the last review provided additional support for the increased stability of the
28    98th percentile PM2.5 concentration, compared to the 99th percentile (U.S. EPA, 2011, Figure 2-2,
29    p. 2-62). Consistent with the PA conclusions based on this analysis, the Administrator concluded
      44 See ATA III, 283 F.3d at 374-376 which concludes 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
      that is requisite to protect the public health.
            April 2016                       2-6             Draft - Do Not Quote or Cite

-------
 1    that it was appropriate to retain the 98th percentile form for the 24-hour PM2.5 standard (78 FR
 2    3127, January 15, 2013).
 3    Level
 4           The EPA's approach to considering alternative levels of the PM2.5 standards in the last
 5    review was based on evaluating the public health protection afforded by the annual and 24-hour
 6    standards, taken together, against mortality and morbidity effects associated with long-term or
 7    short-term PM2.5 exposures. This approach recognized that there is no bright line clearly
 8    directing the choice of level. Rather, the choice of what is appropriate is a public health policy
 9    judgment entrusted to the Administrator. In the last review, this judgment included consideration
10    of the strengths and limitations of the evidence and the appropriate inferences to be drawn from
11    the evidence and the exposure and risk assessments.
12           In evaluating alternative standards, the Agency considered the extent to which specific
13    alternative PM2.5 standard levels were likely to reduce the magnitudes of both long-term
14    exposure-related mortality risk and short-term exposure-related mortality and morbidity risk.
15    Results of the 2010 Health Risk and Exposure Assessment (HREA) indicated that, compared to
16    revising the 24-hour standard, lowering the level of the annual standard would result in more
17    consistent risk reductions across urban study areas, thereby potentially providing a more
18    consistent degree of public health protection (U.S. EPA, 2010a, pp. 5-15 to 5-17; 78 FR 3128,
19    January 15, 2013). Based on risk results, together with the available evidence, the Administrator
20    concluded that it was appropriate to lower the level of the annual standard in order to increase
21    protection against both long- and short-term PM2.5 exposures. She further concluded that it was
22    appropriate to set the 24-hour standard in order to provide supplemental protection, particularly
23    for areas with high peak-to-mean ratios of 24-hour PM2.5 concentrations (e.g., areas with
24    important local or seasonal sources) and for PM2.s-related effects that may be associated with
25    shorter-than daily exposure periods. The Administrator judged that this approach was the "most
26    effective and efficient way to reduce total PIVb.s-related population risk and to protect public
27    health with an adequate margin of safety" (78 FR 3158, January 15, 2013).
28           In selecting the level of the annual PM2.5 standard, the Administrator recognized the
29    substantial increase in the number and diversity of studies available in the last review, including
30    extended analyses of seminal studies of long-term PM2.5 exposures (i.e., American Cancer
31    Society (ACS) and Harvard Six Cities studies), important new long-term exposure studies, and
32    new U.S. multi-city studies that greatly expanded and reinforced our understanding of mortality
33    and morbidity effects associated with short-term PM2.5 exposures. She placed the greatest
34    emphasis on health endpoints for which  the evidence was strongest, based on the assessment of
35    the evidence in the ISA and on the ISA's causality determinations (U.S. EPA, 2009, section


            April 2016                        2-7             Draft - Do Not Quote or Cite

-------
 1    2.3.1).45 In particular, she noted that the evidence was sufficient to conclude a causal relationship
 2    exists between PM2.5 exposures and mortality and cardiovascular effects (i.e., for both long- and
 3    short-term exposures) and that the evidence was sufficient to conclude a causal relationship is
 4    "likely" to exist between PM2.5 exposures and respiratory effects (i.e., for both long- and short-
 5    term exposures). The Administrator also noted additional, but more limited, evidence for a
 6    broader range of health endpoints including evidence "suggestive  of a causal relationship"
 7    between long-term exposures and developmental and reproductive effects as well as
 8    carcinogenic effects (78 FR 3158, January 15, 2013).
 9           Based on information discussed and presented in the ISA, the Administrator recognized
10    that health effects may occur over the full range of ambient PM2.5  concentrations observed in
11    epidemiologic studies, since no discernible population-level threshold could be identified based
12    on the evidence available in the last review (78 FR 3158, January  15, 2013; U.S. EPA, 2009,
13    section 2.4.3). To inform her decisions on an appropriate level for the annual standard in the
14    absence of a discernible population-level threshold, the Administrator considered the degree to
15    which epidemiologic studies indicate confidence in the magnitude and significance of health
16    effect associations over distributions of ambient PM2.5 concentrations. In doing so, she
17    recognized that epidemiologic studies provide greater confidence in the magnitude and
18    significance of observed  associations for the part of the air quality distribution corresponding to
19    the bulk of the health events evaluated, generally at and around the long-term mean PM2.5
20    concentrations. Accordingly, the Administrator weighed most heavily the long-term mean
21    concentrations reported in key multi-city epidemiologic studies. She also took into account
22    additional population-level information from a subset of studies, beyond the long-term mean
23    concentrations, to identify a broader range of PM2.5 concentrations to consider in judging the
24    need for public health protection.46 In doing so, the Administrator recognized that studies
25    indicate diminished confidence in the magnitude and significance  of observed associations in the
26    lower part of the air quality distribution, corresponding to where a relatively small proportion of
27    the health events are observed.
28           In revising the level of the annual standard to 12.0 |ig/m3, the Administrator noted that
29    such a level was below the long-term mean PM2.5 concentrations reported in key epidemiologic
      45 The ISA framework for reaching causality determinations is discussed in Chapter 3, below.
      46 This information characterized the distribution of health events in the studies, and the corresponding long-term
      meanPM2s concentrations (78 FR 3130 to3134, January 15, 2013). The additional population-level data helped
      inform the Administrator's judgment of how far below the long-term mean concentrations to set the level of the
      annual standard (78 FR 3160).

            April 2016                        2-8             Draft - Do Not Quote or Cite

-------
 1    studies that provided evidence of an array of serious health effects, including premature mortality
 2    and increased hospitalizations for cardiovascular and respiratory effects (78 FR 3161, January
 3    15, 2013). The Administrator further noted that 12.0 |ig/m3 generally corresponded to the lower
 4    portions (i.e., about the 25th percentile) of distributions of health events in the limited number of
 5    epidemiologic studies for which population-level information was available. The Administrator
 6    viewed this population information as  helpful in guiding her determination as to where her
 7    confidence in the magnitude and significance of the PIVb.s  associations were reduced to such a
 8    degree that a standard set at a lower level was not warranted. The Administrator also recognized
 9    that a level of 12.0 jig/m3 reflected placing some weight on studies of reproductive and
10    developmental effects, for which the evidence was less strong (78 FR 3161-3162, January 15,
11    2013).47
12           In conjunction with a revised annual standard with a level of 12.0 |ig/m3, the
13    Administrator concluded that the evidence supported retaining the 35 |ig/m3 level of the 24-hour
14    PM2.5 standard to provide supplemental protection (78 FR 3163, January  15, 2013). Specifically,
15    she judged that by lowering the level of the annual standard, the distribution of 24-hour PM2.5
16    concentrations would be lowered as well, affording additional protection  against effects
17    associated with short-term PM2.5 exposures.48
18           The Administrator recognized that uncertainties remained in the scientific information.
19    She specifically noted uncertainties related to understanding the relative toxicity of the different
20    components in the fine particle mixture, the role of PM2.5 in the complex  ambient mixture,
21    exposure measurement errors in epidemiologic studies, and the nature and magnitude of
22    estimated risks related to relatively low ambient PM2.5 concentrations. Furthermore, the
23    Administrator noted that epidemiologic studies had reported heterogeneity in responses both
24    within and between cities and in geographic regions across the U.S. She recognized that this
25    heterogeneity may be attributed, in part, to differences in fine particle composition in different
26    regions and cities. With regard to evidence for reproductive and developmental effects, the
27    Administrator recognized that there were a number of limitations associated with this body of
      47 With respect to cancer, mutagenic, and genotoxic effects, the Administrator observed that the PM2 5
      concentrations reported in studies evaluating these effects generally included ambient concentrations that are equal
      to or greater than ambient concentrations observed in studies that reported mortality and cardiovascular and
      respiratory effects (U.S. EPA, 2009, section 7.5). Therefore, the Administrator concluded that, in selecting a
      standard level that provides protection from mortality and cardiovascular and respiratory effects, it is reasonable to
      anticipate that protection will also be provided for carcinogenic effects (78 FR 3161-3162, January 15, 2013).

      48 This judgment is supported by risk results presented in the 2010 HREA. For example, see section 4.2.2, and
      Figures 4-4 and 4-6 (U.S. EPA, 2010a).

            April 2016                        2-9             Draft - Do Not Quote or Cite

-------
 1    evidence, including the following: the limited number of studies evaluating such effects;
 2    uncertainties related to identifying the relevant exposure time periods of concern; and limited
 3    toxicological evidence providing little information on the mode of action(s) or biological
 4    plausibility for an association between long-term PM2.5 exposures and adverse birth outcomes.
 5           On balance, the Administrator found that the available evidence, interpreted in light of
 6    the remaining uncertainties (noted above), did not justify an annual standard level set below 12.0
 7    |ig/m3 as being "requisite" (i.e., neither more nor less stringent than necessary) to protect public
 8    health with an adequate margin of safety. Thus, the Administrator concluded that the available
 9    evidence and information supported an annual standard with a level of 12.0 |ig/m3, combined
10    with a 24-hour standard with a level of 35 |ig/m3.  She noted that this combination of standard
11    levels was consistent with CASAC's advice to consider an annual standard level within the range
12    of 13 to 11  ng/m3 and  a 24-hour standard level from 35 to 30 ng/m3 (Samet, 2010b). Taken
13    together, the Administrator concluded that the revised annual PM2.5 standard, with its level of
14    12.0 |ig/m3 and a form that does not allow for spatial averaging, combined with the existing 24-
15    hour standard, would be requisite to protect the public health with an adequate margin of safety
16    from effects associated with long- and short-term PM2.5 exposures.

17    2.1.2  PMio Standard
18           In the last review the EPA retained the existing 24-hour primary PMio standard, with its
19    level of 150 |j,g/m3 and its one-expected-exceedance form, in order to continue to provide public
20    health protection against exposures to PMio-2.5. In support of this decision, the Administrator
21    considered the extent to which a standard with a PMio indicator can provide protection against
22    exposures to PMio-2.5 and the degree of public health protection provided by the existing PMio
23    standard. Her consideration of each of these issues is summarized below.
24           In reaching the conclusion that a standard with a PMio indicator can provide appropriate
25    protection against exposures to PMio-2.5, the Administrator noted that PMio mass includes both
26    coarse PM (PMio-2.s) and  fine PM (PIVb.s). As a result, the concentration of PMio-2.5 allowed by a
27    PMio standard set at a  single level declines as the concentration of PM2.5 increases. Because
28    PM2.5 concentrations tend to be higher in urban areas than rural areas (U.S. EPA, 2005, p. 2-54,
29    and Figures 2-23 and 2-24), the Administrator observed that a PMio standard will generally
30    allow lower PMio-2.5 concentrations in urban areas than in rural areas. She judged it appropriate
31    to maintain such a standard given that the large majority of the evidence for PMio-2.5 toxicity,
32    particularly at relatively low particle concentrations, came from study locations where thoracic
33    coarse particles were of urban origin, and given the possibility that  PMio-2.5 contaminants in
34    urban areas could increase particle toxicity. Thus, in the last review the Administrator concluded

            April 2016                       2-10            Draft - Do Not Quote or Cite

-------
 1    that it remained appropriate to maintain a standard that allows lower ambient concentrations of
 2    PMio-2.5 in urban areas, where the evidence was strongest that thoracic coarse particles are linked
 3    to mortality and morbidity, and higher concentrations in non-urban areas, where the public health
 4    concerns were less certain.
 5           In specifically evaluating the degree of public health protection provided by the primary
 6    PMio standard, with its level of 150 |j,g/m3 and its one-expected-exceedance form, the
 7    Administrator recognized that the available health evidence and air quality information was
 8    much more limited for PMio-2.5 than for PM2.5. In particular, the strongest evidence for health
 9    effects attributable to PMio-2.5 exposures was for cardiovascular effects, respiratory effects,
10    and/or premature mortality following short-term exposures. For each of these categories of
11    effects, the ISA determined that the evidence was "suggestive of a causal relationship" (U.S.
12    EPA, 2009, section 2.3.3). These determinations contrast with those for PM2.5, as described in
13    section 2.1.1 above, which were judged in the ISA to be either "causal" or "likely to be causal"
14    for mortality, cardiovascular effects, and respiratory  effects (U.S. EPA, 2009, Tables 2-1 and 2-
15    2).
16           The Administrator judged that the important uncertainties and limitations associated with
17    the PMio-2.5 evidence and information raised questions as to whether additional public health
18    improvements would be achieved by revising the existing PMio standard. She specifically noted
19    the following:

20    (1)     While PMio-2.5 effect estimates reported for mortality and morbidity were generally
21           positive, most were not statistically significant, even in single-pollutant models. This
22           included effect estimates reported in some study locations with PMio concentrations
23           above those allowed by the current 24-hour PMio standard.
24    (2)     The number of epidemiologic studies that have employed co-pollutant models to address
25           the potential for confounding, particularly by PM2.5, was limited. Therefore, the extent to
26           which PMio-2.5 itself, rather than one or more co-pollutants, contributes to reported health
27           effects remained uncertain.
28    (3)     Only a limited number of experimental studies provided support for the associations
29           reported in epidemiologic studies, resulting in further  uncertainty regarding the
30           plausibility of the associations between PMio-2.5 and mortality and morbidity reported in
31           epidemiologic studies.
32    (4)     Limitations in PMio-2.5 monitoring data and the different approaches used to estimate
33           PMio-2.5 concentrations across epidemiologic studies resulted in uncertainty in the
34           ambient PMio-2.5 concentrations at which the  reported  effects occur, increasing
            April 2016                       2-11             Draft - Do Not Quote or Cite

-------
 1           uncertainty in estimates of the extent to which changes in ambient PMio-2.5 concentrations
 2           would likely impact public health.49
 3    (5)    The composition of PMio-2.5, and the effects associated with the various components,
 4           were uncertain. Without more information on the chemical speciation of PMio-2.5, the
 5           apparent variability in associations across locations was difficult to characterize.
 6    With regard to these uncertainties and limitations, the Administrator noted in particular the
 7    considerable degree of uncertainty in the extent to which health effects reported in epidemiologic
 8    studies are due to PMio-2.5 itself, as opposed to one or more co-occurring pollutants. This
 9    uncertainty reflected the relatively small number of PMio-2.5 studies that had evaluated co-
10    pollutant models, particularly co-pollutant models that included PM2.5, and the very limited body
11    of controlled human exposure evidence supporting the plausibility of PMio-2.5-attributable
12    adverse effects at ambient concentrations.  The Administrator noted that these important
13    limitations in the overall body of health evidence introduce uncertainty into the interpretation of
14    individual  epidemiologic studies, particularly those studies reporting associations with PMio-2.5
15    that are not statistically significant.
16           When  she viewed the evidence as a whole, the Administrator concluded that the degree
17    of public health protection provided against short-term exposures to PMio-2.5 should be
18    maintained but did not need to be increased beyond that provided by the current PMio standard.
19    This conclusion emphasized (1) the important uncertainties and limitations associated with the
20    overall body of health evidence and air quality information for PMio-2.5, as reflected in the ISA
21    causal determinations; (2) information indicating that PMio-2.5 effect estimates for the most
22    serious health effect, mortality, were not statistically significant in U.S. locations that met the
23    current PMio standard and where coarse particle concentrations were either directly measured or
24    estimated based on co-located samplers;50 and (3) that PMio-2.5 effect estimates for morbidity
25    endpoints were both positive and negative in locations that met the current standard, with most
26    not statistically significant. Thus, the Administrator concluded that the existing 24-hour PMio
      49 Such limitations also contributed to the decision not to conduct a quantitative risk assessment for PMio-is. The
      lack of a quantitative PMi 0-2.5 risk assessment further contributed to uncertainty regarding the extent to which any
      revisions to the current PMio standard would be expected to improve the protection of public health, beyond the
      protection provided by the current standard.

      50 The Administrator noted that the study by Zanobetti and Schwartz (2009) was the only study to estimate ambient
      PMi 0-2.5 concentrations as the difference between county-wide PMio mass and county-wide PM2.5 mass (78 FR 3178,
      January 15, 2013). As discussed in the PA, it is not clear how such computed PMi0-2.5 measurements compare with
      the PMio-25 concentrations obtained in other studies either by direct measurement or by calculating the difference
      using co-located samplers (U.S. EPA, 2009, section 6.5.2.3).

            April 2016                       2-12             Draft - Do Not Quote or Cite

-------
 1    standard, with its one-expected exceedance form and a level of 150 ng/m3, is requisite to protect
 2    public health with an adequate margin of safety against effects that have been associated with
 3    PMio-2.5. In light of this conclusion, the EPA retained the existing PMio standard.

 4    2.1.3  General Approach in the Current Review of the Primary Standards
 5          The approach for this review will build on the substantial body of work done during the
 6    course of the last review, taking into account the more recent scientific information and air
 7    quality data now available to inform our understanding of the key policy-relevant issues. The
 8    approach summarized below is most fundamentally based on using the EPA's assessment of the
 9    current scientific evidence, quantitative assessments of exposures and/or risks, and other
10    associated analyses (e.g., air quality analyses) to inform the Administrator's judgments regarding
11    primary standards for PM that are requisite to protect public health with an adequate margin of
12    safety. This approach will involve translating scientific and technical information into the basis
13    for addressing a series of key policy-relevant questions using both evidence- and exposure-/risk-
14    based considerations.51
15          Figure 2-1 summarizes the general approach in the current review to reaching conclusions
16    on the current primary standards and on potential alternative standards, if appropriate. The ISA,
17    HREA (if developed), and PA developed in this review will provide the basis for addressing the
18    key policy-relevant questions and will inform the Administrator's decisions as to whether to
19    retain or revise the primary PM NAAQS. The four basic elements of the NAAQS (i.e., indicator,
20    averaging time, form, and level) will be considered collectively in evaluating the health
21    protection afforded by the current standards, and by any alternatives considered.
22
      51 Evidence-based considerations include those related to the health effects evidence assessed and characterized in
      the ISA. Exposure-/risk-based considerations draw from the results of the quantitative assessments.

            April 2016                        2-13             Draft - Do Not Quote or Cite

-------
                                   Adequacy of Existing Primary PM2 j and PM 0 Standards
                              Evidence-Based Considerations
                 >Does the currently available evidence strengthen support for, or
                 call into question, the nature of PM-attributable health effects? Does
                 it provide an improved understanding of at-risk populations?
                  -   Have uncertainties in the evidence from previous reviews
                      been addressed or have new uncertainties been identified?
                 >ls there support for health effect associations at ambient PM
                 concentrations meeting current standards?
                 >ls there support for the plausibility of PM-attributable adverse
                 effects at ambient PM concentrations meeting current standards?
                                                                        Exposure-./Risk-Based Considerations
                                                                        >Nature, magnitude, and importance of
                                                                        estimated exposures and/or risks
                                                                        associated with current primary PM
                                                                        standards
                                                                        >Uncertainties in the exposure and risk
                                                                        estimates
                                                   Does information call
                                                  into question adequacy
                                                  of current Primary PMZ5
                                                   or PMi0 standards?
                                                                                      Appropriated consider
                                                                                         retaining current
                                                                                            standards
                           Further evaluate the scientific evidence and exposure/risk assessments to
                                        inform Identification of potential alternatives
                 Indicator
       >Support for PM2.5 and/or PM,S?
       >Supportforindicator(s) based
       on other size fraction (e.g.,
       UFPs), PM components, etc,?
                                              Averaging Time
                                    > Support for current 24-hour and/or
                                    annual?
                                    >Supportfor sub-daily, seasonal, or other
                                    averaging time(s)?
                  Form
       > Support for retaining existing
       forms?
       >Supportfor other forms?
                                                          Level
                >Supportfor PM-attributable mortality and/or adverse morbidity effects at PM concentrations corresponding to
                various potential standard levels?
                >Support from quantitative exposure and/or risk assessments for public health improvements with various
                potential standard levels'?
                Uncertainties and limitations in the extent to which revised standard levels could result in public health
                improvements, compared to existing standards
1
2
                           Identify appropriate range of potential alternative standards for consideration
Figure 2-1. Overview of General Approach for Review of Primary PM Standards
              April 2016
                                                  2-14
Draft - Do Not Quote or Cite

-------
 1          The final decisions on the adequacy of the current standards and, if appropriate, potential
 2    alternative standards, are largely public health policy judgments to be made by the
 3    Administrator. The Administrator's final decisions will draw upon the scientific evidence for
 4    PM-related health effects, quantitative analyses of population exposures and/or health risks, and
 5    judgments about how to consider the uncertainties and limitations that are inherent in the
 6    scientific evidence and quantitative analyses. To inform the Administrator's public health policy
 7    judgments and decisions, we will consider the support for, and the potential implications of,
 8    placing more or less emphasis on various aspects of the scientific evidence, the exposure-/risk-
 9    based information, and the associated uncertainties and limitations.
10          The current review of the primary PM2.5 and PMio standards will build upon the
11    conclusions from the last review, taking into account the updated evidence and information that
12    has become available since that review. Our consideration of the evidence and information will
13    inform the answer to the following overarching question:
14    •  Does the currently available scientific evidence and exposure-/risk-based information
15       support or call into question the adequacy of the public health protection afforded by
16       the current primary PMi.s or PMio standards?
17    In order to answer this overarching question, we will consider a series of more specific policy -
18    relevant questions related to the health effects and health risks of short- and long-term PM
19    exposures. These questions will focus on PM exposures indexed by PM2.5 and PMio-2.5 mass, as
20    well as PM exposures indexed by other metrics such as other size fractions (e.g., ultrafine
21    fraction), PM characteristics other than size (e.g., chemical composition), PM from particular
22    sources, or PM present in particular types of environments (e.g.,  urban versus non-urban, various
23    geographic areas). Potential policy-relevant questions52 include the following:
24    •  To what extent has  new evidence strengthened or otherwise altered the scientific support for
25       the occurrence of adverse health effects or premature mortality as a result of exposures to
26       particles in the ambient air? To what  extent have important uncertainties in the evidence from
27       the last review been addressed, and have new uncertainties emerged?
28    •  To what extent has  new evidence improved our understanding of human lifestages and
29       populations that are at increased risk  of experiencing health effects associated with exposures
30       to ambient PM?
      52 As noted above, based on CASAC advice and public input on this draft IRP, we will modify these potential
      questions as appropriate for the final IRP. The key policy-relevant questions identified in the final PM IRP will
      provide an overall framework for the development of the PM ISA, REAs (if warranted), and the PA.

            April 2016                       2-15             Draft - Do Not Quote or Cite

-------
 1    •  What does the evidence indicate with regard to confidence in the occurrence of adverse
 2       health effects attributable to exposures to ambient PM concentrations that would likely have
 3       met the current primary standards?

 4                 -  What do available epidemiologic studies indicate with regard to the shapes of
 5                     PM concentration-response functions and confidence in PM health effect
 6                     associations at ambient particle concentrations that would meet the existing
 7                     standards? Do studies indicate  departures from linearity at such low
 8                     concentrations?

 9                 -  When taken together, what do available epidemiologic and experimental
10                     studies indicate with regard to confidence in PM-attributable adverse effects at
11                     ambient PM concentrations that would meet existing standards?

12    •  To what extent does the evidence indicate that health effects are associated with exposures to
13       PM2.5 or PMio-2.5 themselves (or other indicator of PM), as opposed to one or more co-
14       occurring pollutants, particularly at relatively  low ambient PM concentrations that would
15       meet the existing standards?

16    •  To what extent are PM-attributable health effects larger and/or more serious with co-
17       exposures to other pollutants or other stressors?
18    •  Is new information available to better inform our understanding of factors other than
19       pollutant exposures that might influence the associations between ambient PM concentrations
20       and health effects (e.g., weather-related factors, behavioral factors, heating/air conditioning
21       use, driving patterns, time-activity  patterns)?

22    •  Do studies examining the potential for effect modification, either by co-occurring pollutants
23       or by factors other than pollutant exposures, improve our understanding of the geographic
24       heterogeneity in epidemiologic associations with PM?

25    •  Is new information available to improve our understanding of PM exposures, and how those
26       exposures relate to the ambient concentrations often used as exposure surrogates in
27       epidemiologic studies? Does our understanding of geographic variability in exposure
28       measurement error inform our understanding of geographic heterogeneity in epidemiologic
29       associations?

30    •  Is new information available to improve our understanding of the PM exposure periods (e.g.,
31       exposure durations and/or windows of exposure) that are most relevant for PM-associated
32       mortality and/or morbidity?

33    •  To what extent do quantitative estimates indicate that PM-attributable health risks of public
34       health importance could occur at ambient PM concentrations meeting the current primary
35       standards? To what extent do uncertainties and limitations in the underlying health evidence
36       or in the assessment approaches affect our interpretation of these quantitative estimates?
37           If the available evidence and exposure/risk information call into question the adequacy of
38    the public health protection afforded by the current primary PM2.5 or PMio standards, we will
39    also consider the following overarching question:
            April 2016                       2-16            Draft - Do Not Quote or Cite

-------
 1    •  What alternative standards are supported by the currently available scientific evidence
 2       and exposure-/risk-based information, and are appropriate for consideration?
 3    The answer to this second overarching question will also be informed by our consideration of a
 4    series of more specific questions focused on the basic elements of the NAAQS (indicator,
 5    averaging time, form, level). In addressing these more specific questions, we will draw from our
 6    consideration of the scientific evidence and from the results of any additional air quality,
 7    exposure, and/or risk analyses that focus on potential alternative standards.53 We will consider
 8    the elements of the NAAQS collectively in evaluating the health protection afforded by potential
 9    alternative standards.  Specific policy-relevant questions on potential  alternative standards will
10    include the following:

11    •  Do the available health effects  evidence and air quality  information provide support for
12       consideration of indicators for  fine and thoracic coarse  particles in addition to, or in place of,
13       PM2.5  and PMio, respectively?  Does the evidence support an alternative approach for
14       defining particle pollution, including in terms of other size fractions, specific components,
15       source-related mixtures, or specific environments?

16    •  Do the available health effects  evidence, air quality information, and exposure/risk
17       information provide support for considering averaging times in addition to, or in place of, the
18       current 24-hour and annual averaging times?

19    •  To what extent do the available health effects evidence, air quality information, and
20       exposure/risk information provide support for consideration of alternative standard forms!

21    •  What range of alternative standard levels could be supported based on the scientific evidence,
22       air quality  analyses, and exposure/risk assessments?

23    •  What are the important uncertainties and limitations in the available evidence and in the
24       available quantitative analyses, and how might these uncertainties and limitations be taken
25       into consideration in identifying alternative standard indicators, averaging times, forms
26       and/or levels!

27    2.2   SECONDARY STANDARDS

28           In the last review, the EPA generally retained the existing  suite of secondary PM
29    standards  (78 FR 3228, January 15, 2013).54 As described below, the EPA  retained the
      53 Such additional quantitative analyses can inform conclusions on the magnitude of the public health improvement
      that would be expected upon just meeting various potential alternative standards, and on the exposures and/or risks
      expected to remain.
      54 The previous review of the secondary PM NAAQS focused on the welfare effects associated with particles present
      in the ambient air, including the effects of ambient paniculate species of NOx and SOx. The review did not focus on
      the deposition related effects of paniculate species of NOx and SOx. The EPA reviewed the secondary NOa and SOa
            April 2016                        2-17            Draft - Do Not Quote or Cite

-------
 1    secondary 24-hour PM2.5 standard, with its level of 35 |ig/m3, and the 24-hour PMio standard,
 2    with its level of 150 |ig/m3. The EPA also retained the secondary annual PIVfo.s standard, with its
 3    level of 15.0 |ig/m3, except for a change to the form of the annual standard. Consistent with the
 4    change to the form of the primary annual PM2.5 standard (section 2.1.1 above), the EPA removed
 5    the option for spatial averaging from the form of the secondary annual PIVfo.s standard (78 FR
 6    3228, January 15, 2013). Key aspects of the Administrator's decisions on the secondary PM
 7    standards for non-visibility welfare effects and visibility effects are described below in sections
 8    2.2.1 and 2.2.2, respectively. Key policy-relevant issues for the current review are discussed in
 9    section 2.2.3.

10    2.2.1   Non-Visibility Welfare Effects
11           In the last review of the PM NAAQS, the Administrator concluded that it was important
12    to maintain an appropriate degree of control of both fine and coarse particles to address non-
13    visibility welfare effects. Lacking information that would support revised standards, she further
14    concluded that it was appropriate to retain the existing suite of secondary standards to protect
15    against such effects. Non-visibility welfare effects considered include climate  effects, ecological
16    effects (e.g., effects on plants, soil and nutrient cycling,  wildlife and water),55 and materials
17    effects. The Administrator's consideration of each of these types of effects is discussed below.
18           With regard to the role of PM in climate, the Administrator considered whether it was
19    appropriate to establish any distinct secondary PM standards to address welfare effects
20    associated with climate impacts. In  considering the scientific evidence, she noted the ISA
21    conclusion "that a causal relationship exists between PM and effects on climate" and that
22    aerosols56 alter climate processes directly through radiative forcing  and by indirect effects on
23    cloud brightness, changes in precipitation, and possible  changes in cloud lifetimes (U.S. EPA,
24    2009, section 9.3.10). Additionally, the ISA noted that the major aerosol components
25    contributing to  climate processes (i.e., black carbon (BC), organic carbon (OC), sulfates, nitrates
26    and mineral dusts) vary in their reflectivity, forcing efficiencies, and direction of climate forcing,
      NAAQS separately and evaluated the ecosystem-related welfare effects that result from the deposition of paniculate
      species of NOx, SOx, and their transformation products.
      55 As described above, the on-going NOx/SOx review includes consideration of the ecological effects of ecosystem
      loading of paniculate nitrogen and sulfur compounds. Thus, these endpoints are not discussed in this review.
      56 The term aerosol is used in this document when discussing suspended ambient particles in the context of climate
      impacts.

            April 2016                       2-18             Draft - Do Not Quote or Cite

-------
 1    though there is an overall net climate cooling associated with aerosols in the global atmosphere
 2    (U.S. EPA, 2009,  section 9.3.10).
 3          Noting the strong evidence indicating that aerosols affect climate, the Administrator
 4    further considered whether there was sufficient information to revise the secondary PM
 5    standards. She noted that a number of uncertainties in the scientific information were identified
 6    in the ISA and PA. For example, the ISA and PA noted the spatial and temporal heterogeneity of
 7    PM components that contribute to climate forcing, uncertainties in the measurement of aerosol
 8    components, inadequate  consideration of aerosol impacts in climate modeling, insufficient data
 9    on local and  regional microclimate variations and heterogeneity of cloud formations. In light of
10    these uncertainties and the lack of sufficient data, the PA concluded that it was not feasible in the
11    last review "to conduct a quantitative analysis for the purpose of informing revisions [to the
12    secondary PM NAAQS]  based on climate"  (U.S. EPA, 2011, pp. 5-11 to 5-12) and that there was
13    insufficient information available to base a  national ambient standard on climate impacts
14    associated with ambient concentrations of PM or its constituents (U.S. EPA, 2011, section 5.2.3).
15    The Administrator agreed with this conclusion (78 FR 3225-3226, January 15, 2013).
16          The Administrator also considered ecological effects in the last review, including direct
17    effects on metabolic processes of plants; contribution to total metal loading resulting in alteration
18    of soil biogeochemistry and microbiology, plant and animal growth and reproduction; and
19    contribution  to total organics loading resulting in bioaccumulation and biomagnification across
20    trophic levels. The ISA determined that the evidence was sufficient to conclude that "a causal
21    relationship is likely to exist between deposition of PM and a variety of effects on individual
22    organisms and ecosystems" (U.S. EPA, 2009, p. 2-30; sections 2.5.3 and 9.4.7). However, the
23    ISA also noted that it is generally difficult to characterize the nature and magnitude of effects
24    and to quantify relationships between ambient concentrations of PM and ecosystem responses.
25    Such difficulty is due to  significant data gaps and uncertainties, as well as to the considerable
26    variability that exists in the components of PM and their various ecological effects (U.S. EPA,
27    2009, p. 9-193). Given uncertainties and limitations in the available evidence, the PA concluded
28    that the information available at the time of the last review was insufficient for the purposes of
29    assessing the adequacy of the protection for ecosystems afforded by the existing suite of
30    secondary PM standards  or for establishing a distinct national PM standard based on  ecosystem
31    effects of particulates (U.S. EPA, 2011, p. 5-24). The Administrator agreed with this conclusion
32    (78 FR 3225-3226, January 15, 2013).
33          With regard to  materials effects, the Administrator also considered effects associated with
34    the deposition of PM (i.e., dry and wet deposition), including both physical damage (materials
35    effects) and impaired aesthetic qualities (soiling effects). The deposition of PM can physically


            April  2016                      2-19            Draft - Do Not Quote or Cite

-------
 1    affect materials, adding to the effects of natural weathering processes, by promoting or
 2    accelerating the corrosion of metals; by degrading paints; and by deteriorating building materials
 3    such as stone, concrete, and marble (U.S. EPA, 2009, section 9.5). Additionally, the deposition
 4    of ambient PM can reduce the aesthetic appeal of buildings and objects through soiling. The ISA
 5    concluded that evidence was sufficient to support a causal relationship between PM and effects
 6    on materials (U.S. EPA, 2009, sections 2.5.4 and 9.5.4). However, the PA noted that quantitative
 7    relationships are lacking between particle size, concentrations, and frequency of repainting and
 8    repair of surfaces and that considerable uncertainty exists in the contributions of co-occurring
 9    pollutants to materials damage and soiling processes (U.S. EPA, 2011, p. 5-29). The PA
10    concluded that none of the evidence available in the last review called into question the adequacy
11    of the existing secondary PM standards to protect against material effects and that such effects
12    could play no quantitative role in determining whether revisions to the secondary PM NAAQS
13    were appropriate (U.S. EPA, 2011, p.  5-29). The Administrator agreed with this conclusion (78
14    FR 3225-3226, January 15, 2013).
15          In considering non-visibility welfare effects in the last review, as discussed above, the
16    Administrator concluded that, while it is important to maintain an appropriate degree of control
17    of fine and coarse particles to address non-visibility welfare effects, "[i]n the absence of
18    information that would support any different standards...it is appropriate to retain the existing
19    suite of secondary standards" (FR 78 3225 to 3226, January 15, 2013). Her decision was
20    consistent with CASAC advice related to non-visibility effects. Specifically, CASAC agreed
21    with the PA conclusions that, while these effects are important, "there is not currently a strong
22    technical basis to support revisions of the current standards to protect against these other welfare
23    effects"  (Samet, 2010a, p. 5). Thus, the Administrator concluded that it was appropriate to retain
24    all aspects of the existing 24-hour PM2.5 and PMio secondary standards. With regard to the
25    secondary annual PM2.5 standard, the Administrator concluded that it was appropriate to retain a
26    level of  15.0 |ig/m3 while revising only the form of the  standard to remove the option for spatial
27    averaging, consistent with this change to the primary annual PM2.5 standard (78 FR 3225-3226,
28    January  15, 2013).

29    2.2.2  Visibility Effects
30          The Administrator also considered the level of protection that would be requisite to
31    protect public welfare with regard to visual air quality and whether to adopt a distinct secondary
32    standard to achieve this level of protection. In reaching her final decision that the existing 24-
33    hour PM2.5 standard provides sufficient protection against PM-related visibility impairment (78
34    FR 3228, January 15, 2013), the Administrator considered the evidence assessed in the ISA (U.S.


            April 2016                      2-20            Draft - Do Not Quote or Cite

-------
 1    EPA, 2009) and the analyses included in the Urban Focused Visibility Assessment (UFVA)
 2    (U.S. EPA, 2010b) and the PA (U.S. EPA, 2011). She also considered the degree of protection
 3    for visibility that would be provided by the existing secondary standard, focusing specifically on
 4    the secondary 24-hour PM2.5 standard of 35 |ig/m3. These considerations, and the
 5    Administrator's conclusions regarding visibility, are discussed in more detail below.
 6          In the last review, the ISA concluded that, "collectively, the evidence is sufficient to
 7    conclude that a causal relationship exists between PM and visibility impairment" (U.S. EPA,
 8    2009, p. 2-28). Visibility impairment is caused by light scattering and absorption by suspended
 9    particles and gases, including water content of aerosols.57 The available evidence in the last
10    review indicated that various components of PM have been shown to contribute to visibility
11    impairment. For example, at sufficiently high relative humidity values, sulfate and nitrate are the
12    particulate species that contribute most efficiently to visibility impairment. Elemental carbon
13    (EC) and organic carbon (OC)  also are important contributors, especially in the northwestern
14    U.S. Crustal material can be significant contributors to visibility impairment, particularly for
15    remote areas in the arid southwestern U.S. (U.S. EPA, 2009, section 2.5.1).
16          Visibility impairment can have implications for people's enjoyment of daily activities and
17    for their overall sense of well-being (U.S. EPA, 2009, section 9.2). In consideration of the
18    potential public welfare implications of various degrees of visibility impairment, the
19    Administrator considered the available visibility preference studies which were reviewed by
20    EPA in the 2010 UFVA (U.S. EPA, 2010b, chapter 2).58 These preference studies provided
21    information about the potential public welfare implications of visibility impairment from survey
22    studies in which participants were asked questions about their preferences or the values they
23    placed on various visibility  conditions, as displayed to them in scenic photographs or in images
24    with a range of known light extinction levels.
      57 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
      um (U.S. EPA, 2009, section 2.5.1). Particles with hygroscopic components (e.g., particulate sulfate and nitrate)
      contribute more light extinction at higher relative humidity than at lower relative humidity because they change size
      in the atmosphere in response to ambient relative humidity conditions.
      58 Preference studies were available in four urban areas in the last review. Three western preference studies were
      available, including one in Denver, Colorado (Ely etal., 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 Inc., 2001), and a replicate study with
      26 participants was also conducted for Washington, DC (Smith and Howell, 2009).

            April 2016                       2-21            Draft - Do Not Quote or Cite

-------
 1          In noting the relationship between ambient PM and PM-related light extinction, the
 2    Administrator focused on identifying an adequate level of protection against visibility-related
 3    welfare effects. She first concluded that a standard based on a PM2.5 visibility index would
 4    provide a measure of protection against PM-related light extinction that directly takes into
 5    account the factors (i.e., species composition and relative humidity) that influence the
 6    relationship between PM2.5 in ambient air and PM-related visibility impairment. A PM2.5
 7    visibility index standard would afford a relatively high degree of uniformity of visual air quality
 8    protection in areas across the country by directly incorporating the effects of differences of PM2.5
 9    composition and relative humidity. In defining a target level of protection based on a PM2.5
10    visibility index, as discussed below, the Administrator considered specific elements of the index,
11    including the appropriate indicator, averaging time, level, and form.
12          With regard to the indicator of a visibility index, the Administrator concluded that a
13    calculated PM2.5 light extinction indicator that utilized an adjusted version of the original
14    IMPROVE algorithm,59 in conjunction with monthly average relative humidity data based on
15    long-term climatological means, would be the most appropriate indicator for a PM2.5 visibility
16    index standard (78 FR 3226, January 15, 2013). In reaching her final decision, the Administrator
17    further noted the CAS AC conclusion that it was reasonable to rely on a calculated PM2.5 light
18    extinction indicator based on PM2.5 chemical composition and relative humidity. The
19    Administrator also considered the PM2.5 mass indicator and directly measured PM2.5 light
20    extinction. She concluded that a PM2.5 mass-based indicator would not be appropriate because
21    the available mass monitoring methods did not include measurement of the full water content of
22    ambient PM2.5, nor did they provide information on the composition of PM2.5, both of which
23    contribute to visibility impacts (77 FR 38980, June 29, 2012). In addition, at the time of the
24    proposal, the Administrator provisionally concluded that directly measured PM2.5 light extinction
25    was not an appropriate option because a suitable specification of available equipment or
26    performance-based verification procedures for direct measurement of light extinction could not
27    be developed in the time frame of the review (77 FR 38980-38981, June 29, 2012).
28          With regard to averaging time of the index, the Administrator concluded that a 24-hour
29    averaging time would be appropriate for a visibility index (78 FR 3226, January 15, 2013). She
30    concluded that hourly or sub-daily (4- to 6-hour) averaging times, within daylight hours and
      59 The IMPROVE algorithm uses PM mass concentration measurements and relative humidity estimates to calculate
      light extinction. For more information about the derivation of and input data required for the original and revised
      IMPROVE algorithms, see 78 FR 3186-3177, January 15, 2013.

            April 2016                      2-22             Draft - Do Not Quote or Cite

-------
 1    excluding hours with high relative humidity, are more directly related to the short-term nature of
 2    the perception of PM-related visibility impairment and the relevant exposure periods for
 3    segments of the viewing public than a 24-hour averaging time. However, she also noted the data
 4    quality uncertainties associated with the instruments used to provide the hourly PM2.5 mass
 5    measurements required for an averaging time shorter than 24 hours. The Administrator also
 6    considered comparative analyses of 24-hour and 4-hour averaging times in conjunction with a
 7    calculated PM2.5 indicator in the PA (U.S. EPA, 2011, pp. 4-55 - 4-56, Appendix G, section G.4).
 8    These analyses showed good correlation between 24-hour and 4-hour average PIVfo.s light
 9    extinction, as evidenced by reasonably high city-specific and pooled R-squared values, generally
10    in the range of over 0.6 to over 0.8. The Administrator considered and agreed with the PA
11    conclusion that at a 24-hour averaging time would be a reasonable and appropriate surrogate for
12    a sub-daily averaging time.
13          With regard to form of the index, the Administrator concluded that a multi-year
14    percentile-based form offered greater stability to the air quality management process by reducing
15    the possibility that statistically unusual indicator values would lead to transient violations in the
16    standard. She noted that a three-year average form provided stability  from the occasional effects
17    of inter-annual meteorological variability that can result in unusually high pollution levels for a
18    particular year (U.S. EPA, 2011, p. 4-58). In the UFVA, 90th, 95th, and 98th percentile forms were
19    assessed for alternative PM light extinction standards (U.S. EPA, 2010b, chapter 4). In
20    considering these alternative percentiles, the PA noted that the Regional Haze Program targets
21    the 20 percent most impaired days for improvements in visual air quality in Federal Class I areas.
22    A focus on improving the 20 percent most impaired days suggests that the 90th percentile, which
23    represents the median of the distribution of the 20 percent worst days, would be an appropriate
24    form to consider. Strategies that are implemented so that 90  percent of days would have visual
25    air quality that is at or below the level of the standard would reasonably be expected to lead to
26    improvements in visual air quality for the 20 percent most impaired days. Given that the
27    preference studies did not provide information with regard to the frequency of time that visibility
28    levels should be below these values, the PA found no basis to conclude that it would be
29    appropriate to consider limiting the occurrence of days with peak PM-related light extinction in
30    urban areas to a greater degree (U.S. EPA, 2011, p. 4-59). Based on the above considerations, the
31    Administrator concluded that a 90th percentile form was the most appropriate form (78 FR 3226,
32    January 15, 2013).
33          With regard to level of the index, the Administrator considered the visibility preference
34    studies conducted in four urban areas (U.S. EPA, 2011, p. 4-61) and further quantitative analyses
35    of visibility conditions for 15 urban study areas (U.S. EPA, 2011, Appendix G, Tables G-7 and


            April 2016                       2-23             Draft - Do Not Quote or Cite

-------
 1    G-8). Based on these studies the PA identified a range of levels from 20 to 30 deciviews (dv)60 as
 2    being a reasonable range of "candidate protection levels" (CPLs). In considering this range of
 3    CPLs, the Administrator noted the uncertainties and limitations in public preference studies,
 4    including the small number of stated preference studies available; the relatively small number of
 5    study participants and the extent to which the study participants may not be representative of the
 6    broader study area population in some of the studies; and the variations in the specific materials
 7    and methods used in each study. She concluded that the substantial degrees of variability and
 8    uncertainty in the public preference studies should be reflected in a target protection level at the
 9    upper end of the range of CPLs than if the information were more consistent and certain.
10    Therefore, the Administrator concluded that it was appropriate to set a  target level of protection
11    in terms of a 24-hour PM2.5 visibility index at 30 dv (78 FR 3226-3227, January 15, 2013).
12          Based on her considerations and conclusions summarized above, the Administrator
13    concluded that the protection provided by a secondary standard defined in terms of a PIVfo.s
14    visibility index with a 24-hour averaging time, a 90th percentile form averaged over 3  years, and
15    a level of 30 dv, would be requisite to protect public welfare with regard to visual air quality (78
16    FR 3227, January 15, 2013). Having reached this conclusion, she next  determined whether to
17    adopt such a visibility index as a distinct secondary standard, particularly in the context of the
18    full suite of existing secondary standards.
19           In determining  whether a distinct secondary standard was needed, the Administrator
20    considered the degree of protection from visibility impairment afforded by the existing
21    secondary standards. She considered both whether the existing 24-hour PM2.5 standard of 35
22    |ig/m3 is sufficient (i.e., not under-protective) and whether it is not more stringent than necessary
23    (i.e., not over-protective). In doing so, she noted that the air quality analyses showed that all
24    areas meeting the existing 24-hour PM2.5 standard, with its level of 35  |ig/m3, had visual air
25    quality at least as good as 30  dv  (based on the visibility index  defined above) (Kelly et al.,
26    2012a, 2012b). Thus, the secondary 24-hour PM2.5 standard would likely be controlling relative
27    to a 24-hour visibility index set at  a level of 30 dv. Additionally, areas  would be unlikely to
28    exceed the target level  of protection for visibility of 30 dv without also exceeding the  existing
29    secondary 24-hour standard.61 Thus, the Administrator judged that the  24-hour PM2.5  standard
      60 Deciview 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.
      61 The Administrator also recognized that air quality analyses indicated that "the 24-hour PM2 5 standard of 35 |ag/m3
      also would likely achieve more than the target level of protection of visual air quality (30 dv) in some areas" (78 FR
      3227, January 15, 2013).

            April 2016                       2-24             Draft - Do Not Quote or Cite

-------
 1    "provides sufficient protection in all areas against the effects of visibility impairment—i.e., that
 2    the existing 24-hour PM2.5 standard would provide at least the target level of protection for
 3    visual air quality of 30 dv which the Administrator judges appropriate" (78 FR 3227, January 15,
 4    2013). She further judged that "[s]ince sufficient protection from visibility impairment would be
 5    provided for all areas of the country without adoption of a distinct secondary standard, and
 6    adoption of a distinct secondary standard will not change the degree of over-protection provided
 7    for some areas of the  country... adoption of such a distinct secondary standard is not needed to
 8    provide requisite protection for both visibility and nonvisibility related welfare effects" (78 FR
 9    3228, January 15, 2013).

10    2.2.3  General Approach in the Current Review of the Secondary Standards
11          The approach for this review builds on the substantial body of work completed during the
12    course of the last review, taking into account the more recent scientific information and air
13    quality data now available to inform our understanding of the key-policy relevant issues. The
14    approach described below is most fundamentally based on using the EPA's assessment of the
15    current scientific evidence and associated quantitative analyses to inform the Administrator's
16    judgments regarding secondary standards for PM that are requisite to protect the public welfare
17    from any known or anticipated adverse effects.
18          Figure 2-2 summarizes the general approach, including consideration of the policy -
19    relevant questions which will frame the current review. The ISA, WREA (if warranted), and PA
20    developed in this review will provide the basis for addressing the key policy-relevant questions
21    and will inform the Administrator's judgments as to the adequacy of the current secondary PM2.5
22    and PMio standards, as well as his/her decisions as to whether to retain or revise these standards.
            April 2016                      2-25             Draft - Do Not Quote or Cite

-------
                Adequacy of Public Welfare Protection Provided by Existing Secondary PM Standards
                  Evidence-Based Considerations
  > ISA weight-of-evidence determinations for welfare effects (e.g.,
  visibility impairment, climate, other effects)
  r Preference studies for visibility: Nature of relationship between PM-
  attributable visibility impairment and public perceptions?
  >Sfucfes linking PM to climate impacts: Nature of relationship between
  ambient PM species and climate?
  'r Evidence for other welfare impacts of PM: Nature of relationships
  between ambient PM and other ecological effects?
  > Uncertainties in evidence: how do uncertainties in  the evidence affect
  our understanding of the evidence of effects, particularly for
  concentrations below tie current standard levels?
                                     Quantitative Assessment-Based
                                             Considerations
                                  *-Nature, magnitude, and importance of
                                  estimated welfare impacts associated
                                  with current secondary PM standards
                                  /'Uncertainties in assessment estimates
                                  , including the ability to directly assess
                                  the relationship between changes in PM
                                  concentrations and changes in welfare
                                  impacts
                                                Does information call
                                               into question adequacy
                                              of the welfare protection
                                               provided by the current
                                                  PM standards?
                                                         Consider retaining
                                                        current secondary PM
                                                             standard(s)
                                Consider Potential Alternative Secondary Standards
                                                 Averaging Time
                                     >Support for current 24-hour and/or
                                     annual?
                                              > Support for retaining existing
                                              forms?
> Support for PM25 and/or PM
> Support for mdicator(s) based
on other size fraction, PM
components, light extinction, etc.?
> Support for sub-daily, seasonal, or other
averaging time(s)?
                                                                                   > Support for alternative form based
                                                                                   on daylight hours or other mettle?
                                                                                  	I
                                                       Level
            ^•Support for PM-attributable visibility, climate, other welfare impacts at PM concentrations
            corresponding to various potential standard levels?
            > Support from quantitative exposure and/or risk assessments for public welfare improvements with
            various potential standard levels?
            > Uncertainties and limitations in the extent to which revised standard levels could result in public
            welfare improvements, compared to existing standards	
                           Identify range of potential alternative secondary standards for consideration
      April 2016
               2-26
                                                                          Draft - Do Not Quote or Cite

-------
 1    Figure 2-2. Overview of General Approach for Review of Secondary PM Standards
 2          The current review of the secondary PM2.5 and PMio standards will build upon the
 3    conclusions from the last review, taking into account the updated evidence and information that
 4    has become available since that review. Our consideration of the evidence and information will
 5    inform the answer to the following overarching question:
 6    •  Does the currently available scientific evidence and quantitative information support or
 7       call into question the adequacy of the welfare protection afford by the current
 8       secondary PMi.s or PMio standards?
 9    In order to answer this overarching question, we will consider a series of more specific policy -
10    relevant questions related to the available scientific evidence and information from quantitative
11    assessments. Potential policy-relevant questions62 include the following:
12    •  To what extent has new scientific evidence improved our understanding of the nature and
13       magnitude of visibility, climate, ecosystem responses, and materials effects to ambient PM,
14       including the variability associated with such responses? To what extent have important
15       uncertainties in the evidence from the last review been addressed, and have new uncertainties
16       emerged?

17    •  To what extent is new information available that changes or enhances our understanding of
18       the physics of light extinction and/or its quantification in urban and non-urban areas (e.g.,
19       through light extinction or other monitoring methods or through algorithms such as
20       IMPROVE)?

21    •  To what extent are new studies available on the nature of the relationship between PM-
22       attributable visibility impairment and public perceptions of such impairment? What
23       information is available to inform judgments about the potential adversity to public welfare
24       of PM-attributable visibility impairment?

25                  -  Is evidence available from recent studies in additional urban and non-urban
26                     areas, beyond the studies in the previous  review?

27                  -  To what extent is evidence available that distinguishes visibility preferences
28                     from health preferences?

29                  -  To what extent is evidence available that evaluates the sensitivity of visibility
30                     preferences to survey methods?

31                  -  To what extent is evidence available that examines the potential relationship
32                     between intensity versus frequency of visibility impairment in stated public
33                     preferences?
      62 As noted above (section 2.1), based on CASAC advice and public input on this draft IRP, we will modify these
      potential questions as appropriate for the final IRP. The key policy-relevant questions identified in the final PM IRP
      will provide an overall framework for the development of the PM ISA, REAs (if warranted), and the PA.

            April 2016                      2-27            Draft - Do Not Quote or Cite

-------
 1    •  To what extent is new information available that changes or enhances our understanding of
 2       the climate impacts of PM-related aerosols, particularly regarding the quantification of
 3       anthropogenic aerosol effects on radiative forcing?

 4    •  To what extent is new information available to link PM (excluding nitrates and sulfates) to
 5       ecological effects?

 6    •  To what extent is new information available to link PM to materials effects, including
 7       degradation of surfaces, and deterioration of materials such as metal, stone,  concrete and
 8       marble? Are there studies linking perceptions of reduced aesthetic appeal of buildings and
 9       other objects to PM or wet deposition of N and S species?

10    •  Does the available evidence and/or quantitative analyses suggest that PM-induced visibility
11       impairment or other PM-related welfare effects could occur with ambient concentrations of
12       PM that meet the current standards? If so, could these effects  be of sufficient magnitude
13       and/or frequency such that they might reasonably be judged to be adverse to public welfare?
14       To what degree would updated or additional analyses improve our understanding of the
15       welfare effects that could be allowed by the current standards?

16    •  To what extent have important uncertainties identified in the last review been reduced and/or
17       have new uncertainties emerged?
18          If the available evidence and information from quantitative analyses call into question the
19    adequacy of the welfare protection afforded by the current  secondary PM2.5 or PMio standards,
20    we will also consider the following overarching question:
21    •  What alternative standards  are supported by the currently available scientific evidence
22       and quantitative information and are appropriate for consideration?
23          To answer this second overarching question, we will also  consider a series of more
24    specific questions focused on the  basic elements of the NAAQS (indicator, averaging time, form,
25    level). We will consider these elements collectively in evaluating the public welfare protection
26    afforded by potential alternative standards. With regard to consideration of alternative standards,
27    the specific policy-relevant questions will include the following:
28    •  Do the available welfare effects evidence and air quality information provide support for
29       consideration of indicators in addition to, or in place of, the current mass-based indicators?
30       Do the evidence and information from the quantitative  analyses, if warranted, support
31       alternative indicators based on light extinction, chemical composition, or other factors?

32    •  Do the available welfare effects evidence, air quality information, and information from
33       quantitative assessments, provide support for considering averaging times in addition to, or
34       in place of, the current 24-hour and annual averaging times? Do the evidence and
35       information support a sub-daily or other alternative averaging time?

36    •  To what extent do air quality analyses or information from quantitative assessments provide
37       support for considering alternative standard forms? To what extent do assessments support an
38       alternative form based on daylight hours or other metric?
            April 2016                       2-28            Draft - Do Not Quote or Cite

-------
 1    •  What range of levels should be considered, based on the scientific evidence, air quality
 2       analyses, and quantitative assessments? At what concentrations of ambient PM do adverse
 3       visibility impairment and/or other environmental effects of concern for public welfare occur?
 4    •  What are the important uncertainties and limitations in the available evidence, analyses, and
 5       assessments and how might those uncertainties and limitations be taken into consideration in
 6       identifying alternative  standard indicators, averaging times, forms, and/or levels^

 1    2.3  PM AMBIENT MONITORING
 8           Achieving the degree of public health and welfare protection intended for the NAAQS
 9    depends, in large part, on appropriate ambient monitoring networks.  In the case of PM, existing
10    monitoring networks provide data for a variety of objectives as part of an iterative process in
11    managing air quality. These objectives include: (1) determining compliance with the NAAQS;
12    (2) characterizing air quality status, including providing the public with timely reports and
13    forecasts of the Air Quality Index (AQI); (3) supporting air quality analyses used to conduct
14    assessments of exposure, health risks, and welfare effects; (4) developing and evaluating
15    emissions control strategies; and (5) measuring trends and overall progress for the air pollution
16    control program.
17           Federal rules that regulate ambient monitoring programs are found in 40 CFR parts 50,
18    53 and 58. The EPA amended these regulations in the 2006 and 2012 reviews of the PM
19    NAAQS, in part to support changes necessary for implementation  of the revised PM standards.
20    The EPA expects to follow a similar process for monitoring rule changes during this review, if
21    appropriate. Potential monitoring rule changes include the Federal Reference Methods (FRMs)63
22    that exist as appendices to  part 50, the procedures for approval of Federal Reference and Federal
23    Equivalent Methods (FEMs) contained in part 53, and the rules applicable to ambient monitoring
24    network planning and operations that are the basis for part 58 and Appendices A through E.
25           Section 2.3.1 below provides an overview of the PM monitoring networks and their
26    history. Section 2.3.2 summarizes the key potential monitoring-related issues in the current
27    review.
      63 FRMs provide the methodological basis for comparison to the NAAQS and also serve as the "gold-standard" for
      the comparison of other methods being reviewed for potential approval as equivalent methods. The EPA keeps a
      complete list of designated reference and equivalent methods available on its Ambient Monitoring Technology
      Information Center (AMTIC) website (http://www3.epa.gov/ttn/amtic/files/ambient/criteria/reference-equivalent-
      methods-listpdf).
            April 2016                       2-29             Draft - Do Not Quote or Cite

-------
 1    2.3.1  PM Monitoring Networks
 2          The EPA and its partners at state, local, and tribal monitoring agencies manage and
 3    operate the nations' ambient air monitoring networks. The EPA provides minimum monitoring
 4    requirements for criteria pollutants and related monitoring (e.g., the Chemical Speciation
 5    Network (CSN)) including identification of an FRM for criteria pollutants and guidance
 6    documents to support implementation and operation of the networks. Monitoring agencies carry
 7    out and perform ambient air monitoring in accordance with the EPA's requirements and
 8    guidance as well as often meeting their own state monitoring needs that may go beyond the
 9    minimum federal requirements. This partnership results in a nationally consistent ambient air
10    monitoring program that supports the objectives listed above.  Data from the ambient air
11    monitoring networks are available from two national databases. The AirNow database provides
12    data used in public reporting and forecasting of the AQI, while the Air Quality System (AQS)
13    database is the EPA's long-term repository of ambient air monitoring data.
14          The EPA and states  currently operate robust national networks for both PMio and PM2.5,
15    as these are the two measurement programs directly supporting NAAQS. PMio measurements
16    are based on  gravimetric mass, while PIVb.s measurements include gravimetric mass and
17    chemical speciation. A smaller network of stations is operating and reporting data for PMio-2.5
18    gravimetric mass and a small number of monitors are operated to support special projects,
19    including pilot studies, for continuous speciation and particle count data. Monitoring networks
20    and additional monitoring efforts for each of the various PM size fractions and for PM
21    composition  are discussed below.64 All sampler and monitor counts provided below are based on
22    data submitted to the EPA for calendar year 2014, unless otherwise noted.
23    Total Suspended Particulates (TSP) Sampling
24          The EPA first established NAAQS for PM in 1971, based on the original air quality
25    criteria document (DHEW,  1969). The reference method specified for determining attainment of
26    the original standards was the high-volume sampler, which collects PM up to a nominal size of
27    25 to 45 micrometers (um) (referred to as total suspended particles or TSP). TSP was replaced by
28    PMio as the indicator for particles in a 1987 final notice (52 FR 24854, July 1, 1987). However,
29    TSP sampling remains in operation to provide the aerosol needed for TSP lead sampling as well
30    as for cases where a state may continue to have state standards for TSP. The size of the TSP
31    network peaked in the mid-1970s when over 4,300 TSP samplers were in operation. Today, there
      64 More information on ambient monitoring networks can be found at http://www3.epa.gov/ttn/amtic/.

            April 2016                       2-30            Draft - Do Not Quote or Cite

-------
 1    are about 200 TSP samplers still in operation as part of the lead monitoring program; of these,
 2    about 50 also report TSP mass.
 3    PMio Monitoring
 4          As a result of the 1987 standard for PMio, the EPA and its state and local partners
 5    implemented the first size-selective PM monitoring network in 1990 with the establishment of a
 6    PMio network consisting of mainly high-volume samplers. The PMio monitoring network peaked
 7    in size in 1995 with 1,665 stations reporting data.
 8          Approximately 750 PMio stations with samplers and monitors are currently in operation
 9    to support comparison of the PMio data to the NAAQS, trends, and reporting and forecasting of
10    the AQI. Though the current PMio network is relatively stable, monitoring agencies may
11    continue divesting of some of the PMio monitoring stations where concentration levels are low
12    relative to the NAAQS.
13          While the PMio network is national in scope, there are areas of the west such as
14    California and Arizona with substantially higher PMio station density than the rest of the
15    country. In the PMio mass network, about 320 of the stations operate continuous mass monitors
16    approved as FEMs and 480 operate FRMs. Thus, about 50 of the PMio stations have collocation
17    with both continuous FEMs and FRMs. Almost half of the PMio stations with FRMs operate on a
18    sample frequency of one in every sixth day, with about 140 operating every third day and just
19    over 100 operating every day.
20    PM2.s Monitoring
21          After setting the first PM2.5 NAAQS in 1997, the EPA and states implemented a PM2.5
22    network consisting of ambient air monitoring sites with mass and/or chemical speciation
23    measurements. Network operation began in 1999 with nearly 1000 monitoring stations operating
24    FRMs to provide fine particle mass. The PM2.5 monitoring program remains one of the major
25    ambient air monitoring programs operated across the country.
26          There are three main components of the current PM2.5 monitoring program including
27    FRMs, PM2.5 continuous mass monitors, and Chemical Speciation Network (CSN) samplers. The
28    FRMs are primarily used for comparison to the NAAQS, but also serve other important purposes
29    such as developing trends and evaluating the performance of PM2.5 continuous mass monitors.
30    PM2.5 continuous mass monitors are primarily used to support forecasting and reporting of the
31    AQI, but are also used for comparison to the NAAQS, where approved as FEMs. The CSN and
32    related Interagency Monitoring of Protected Visual Environments (IMPROVE) network are used
33    to provide chemical composition of the aerosol which serve a variety of objectives. This section
34    provides an overview of each of these components of the PM2.5 monitoring program.

           April 2016                      2-31            Draft - Do Not Quote or Cite

-------
 1          As noted above, the PM2.5 monitoring network began operation in 1999 with nearly 1,000
 2   monitoring stations operating FRMs. The PM2.5 FRM network peaked in operation in 2001 with
 3   over 1,150 monitoring stations. In the current PM2.5 network, there are approximately 773 FRM
 4   filter-based samplers that provide 24-hour PM2.5 mass concentration data. Of these operating
 5   FRMs, 111 are providing daily PM2.5 data, 535 every third day, and 125 every sixth day. Close to
 6   800 continuous PM2.5 mass monitors provide hourly data on a near real-time basis. About 363 of
 7   the PM2.5 continuous monitors are FEMs and therefore used for comparison with the NAAQS
 8   and AQI, and another 424 are methods not approved as FEMs and therefore used just for the
 9   AQI.
10          Due to the complex nature of fine particles, in 2000 the EPA and states implemented the
11   CSN to better understand the components of fine particle mass at selected locations across the
12   country.  The CSN was first piloted at 13 sites. After the pilot phase, the program continued with
13   deployment of the Speciation Trends Network (STN) in the fall of 2000. The CSN ultimately
14   grew to 54 trends sites and peaked in operation in 2005 with 252 stations; the 54 trends stations
15   and nearly 200 supplemental stations. The original CSN program had multiple sampler
16   configurations including the Thermo Andersen RAAS, Met One SASS/SuperSASS,  and URG
17   MASS. During the 2000s, the EPA and states worked to align the network to one common
18   sampler for elements and ions, which was the Met One SASS/SuperSASS.  In 2005, CASAC
19   provided recommendations to the EPA for making changes to the CSN. These changes were
20   intended to improve data comparability with the rural IMPROVE carbon concentration data. To
21   accomplish this, the EPA replaced the existing carbon channel sampling and analysis methods
22   with a new modified IMPROVE version III  module C sampler, the URG 3000N. Implementation
23   of the new carbon sampler and analysis was broken into three phases starting in May of 2007
24   through October of 2009.
25          In the current PM2.5 CSN, long-term measurements are made at about 90 largely urban
26   locations comprised of either the STN or the National Core (NCore) network.65 NCore is a
27   multipollutant network measuring particles,  gases, and basic meteorology that has been in formal
28   operation since January 1, 2011.  Particle measurements made at NCore include PM2.5 filter-based
29   mass, which is largely the FRM,  except in some rural locations which utilize the IMPROVE
30   program  PM2.5 mass filter-based measurement; PM2.5 speciation using either the CSN program or
31   IMPROVE program; and PMio-2.5 mass utilizing an FRM, FEM or IMPROVE for some of the
     65 In most cases where a city has an STN station, it is located at the same site as the NCore station. In a few cases, a
     city may have an STN station located at a different location than the NCore station.

           April 2016                      2-32           Draft - Do Not Quote or Cite

-------
 1    rural locations. The NCore network includes a total of 78 stations of which 63 are in urban or
 2    suburban stations designed to provide representative population exposure and another 15 rural
 3    stations designed to provide background and transport information. The NCore network is
 4    deployed in all 50 States, DC, and Puerto Rico with at least one station in each state and two or
 5    more stations in larger population states. Since 2008, the EPA's Office of Research and
 6    Development has approved ten models of PM2.5 continuous monitors as FEMs; about 75% of the
 7    reporting PM2.5 continuous FEMs in the country are the Met One BAM 1020.
 8          Both the STN and NCore networks are intended to remain in operation indefinitely. The
 9    CSN measurements at NCore and STN stations operate every third day. Another approximately
10    70 CSN stations are known as supplemental sites are intended to be potentially less permanent
11    locations used to support State Implementation Plan (SIP) development and other monitoring
12    objectives.66 Supplemental CSN stations typically operate every sixth day. In 2014, about 40
13    supplemental CSN stations that are largely located in the eastern half of the country stopped
14    operations to ensure a sustainable CSN network moving forward. Specific components of fine
15    particles are also measured through the IMPROVE monitoring program67 which supports
16    regional haze characterization and tracks changes in visibility in Class I areas as well as many
17    other rural and some urban areas. The base IMPROVE network includes 110 monitoring
18    locations and. another approximately 20 locations are operated as IMPROVE protocol sites
19    where a monitoring agency has requested participation in the program. These IMPROVE
20    protocol sites are operated the same way as the IMPROVE program, but they may serve a
21    number of monitoring objectives (i.e., the same objectives as the CSN) and are not explicitly tied
22    to the Regional Haze  Program. Samplers at IMPROVE stations operate every third day. In
23    January of 2016, 8 IMPROVE protocol stations stopped operating to ensure a sustainable
24    IMPROVE program moving forward. Details on the process and outcomes of the CSN
25    supplemental and IMPROVE protocol assessments used to identify sites that would no longer be
      66 See http://www.epa.gov/ttn/amtic/speciepg.html for more information on the PM2 5 speciation monitoring
      program.
      67Recognizing the importance of visual air quality, Congress included legislation in the 1977 Clean Air Act to
      prevent future and remedy existing visibility impairment in Class I areas. To aid the implementation of this
      legislation, the IMPROVE program was initiated in 1985 and substantially expanded in 2000-2003. This program
      implemented an extensive long term monitoring program to establish the current visibility conditions, track changes
      in visibility and determine causal mechanism for the visibility impairment in the National Parks and Wilderness
      Areas. For more information see http://www.epa.gov/ttn/amtic/visdata.html.
            April 2016                      2-33             Draft - Do Not Quote or Cite

-------
 1    funded are available on an interactive website68. Together, the CSN and IMPROVE data provide
 2    chemical species information for fine particles that are critical for use in health and
 3    epidemiologic studies to help inform reviews of the PM NAAQS.69
 4           Key changes made to the EPA's monitoring requirements as a result of the 2012 PM
 5    NAAQS review included the addition of PM2.5 monitoring at near road locations in CBSAs over
 6    1 million in population; the clarification of terms used in siting of PM2.5 monitors and their
 7    applicability to the NAAQS; and providing flexibility on data uses to monitoring agencies where
 8    their PM2.5 continuous monitors are not providing  data that meets the performance criteria used
 9    to approve the continuous method as an FEM. The addition of PM2.5 monitoring at near road
10    locations is being phased in 2015 and 2017. By January 1, 2015, 22 CBSAs with a population of
11    2.5 million or more were to have a PM2.5 FRM or FEM operating at a near-road monitoring
12    station. By January 1, 2017, 30 CBSAs with a population between 1  million and 2.5 million are
13    to have a PM2.5 FRM or FEM operating are a near-road monitoring station. The terms clarified in
14    the 2012 rule ensure consistency with other NAAQS and long standing definitions used by the
15    EPA (78 FR 3234, January 15, 2013). The flexibility provided to monitoring agencies such that
16    data from certain PM2.5 continuous FEM monitors are not applicable to the NAAQS, where
17    appropriate, ensures that the incentives of utilizing PM2.5 continuous monitors (e.g., efficiencies
18    in operation and availability of hourly data in near-real time) are realized without having
19    potentially poor performing data be misused in  a NAAQS decision (78 FR 3241, January 15,
20    2013).70
21    PMio-2.5 Monitoring
22           In the 2006 PM NAAQS review, the EPA promulgated a new FRM for the measurement
23    of PMio-2.5 mass in ambient air. Although the standard for thoracic coarse particles uses a PMio
      68 See the Chemical Speciation Network Assessment Interactive Website at:
      https://www.sdas.battelle.org/CSNAssessment/html/Default.html
      69 These data could also be used to better understand visibility through calculation of light extinction using the
      IMPROVE algorithm.
      70 Although not explicitly required under any existing monitoring regulations, the EPA and state and local agencies
      have also been working together to pilot additional PM methods at near-road monitoring stations that may be of
      interest to data users. These methods include such techniques as particle counters, particle size distribution, and
      black carbon by aethalometer. These methods and their rationale for use at near-road monitoring stations are
      described in a Technical Assistance Document (TAD) on NCh near road monitoring (Near-road NCh Monitoring
      Technical Assistance Document, EPA-454/B-12-002, June 2012).
            April 2016                        2-34             Draft - Do Not Quote or Cite

-------
 1    indicator, a new FRM for PMio-2.5 mass was developed to provide a basis for approving FEMs
 2    and to promote the gathering of scientific data to support future reviews of the PM NAAQS. The
 3    PMio-2.5 FRM (or approved FEMs, where available) was implemented at required NCore stations
 4    by January 1, 2011. In addition to NCore, there are other collocated PMio and PM2.5 low-volume
 5    FRMs operating across the country that are essentially providing the PMio-2.5 FRM measurement
 6    by the difference method.
 7          PMio-2.5 measurements are currently performed across the country at NCore stations,
 8    IMPROVE monitoring stations, and at a few additional locations where state or local agencies
 9    choose to operate the method. For urban NCore stations and other State and Local Air
10    Monitoring Stations (SLAMS) the method employed is either a PMio-2.5 FRM, which is
11    performed using a low-volume PMio FRM collocated with a low volume PM2.5 FRM of the same
12    make and model, or FEMs for PMio-2.5, including filter-based dichotomous methods and
13    continuous methods of which several makes and models are approved. Filter-based PMio-2.5
14    measurements at NCore (i.e., the FRM or dichotomous filter-based FEM) operate every third
15    day, while continuous methods have data available every hour of every day. PMio-2.5 filter-based
16    methods at other SLAMS typically operate every third or sixth day. For IMPROVE, which is
17    largely a rural network, PMio-2.5 measurements are made with two sample channels; one each for
18    PMio and the other PM2.5. All IMPROVE program samplers operate every third day. All together
19    there approximately 250 stations  where PMio-2.5 data  are being reported to the EPA.
20          There is no operating chemical speciation network for characterizing the specific
21    components of thoracic coarse particles. Washington University at St. Louis recently reported on
22    a coarse particle speciation pilot study with several objectives aimed at addressing this issue,
23    such as evaluating a coarse particle species analyte list and evaluating sampling and analytical
24    methods.71 The coarse particle speciation pilot study  should provide useful information for any
25    organization wishing to pursue coarse particle speciation.
26    Additional PM Metrics
27          There are additional PM measurement metrics made at a much smaller number of
28    stations. These measurements may be associated with special projects or are complementary
29    measurements to other networks where the monitoring agency has prioritized  having the
      71 Pilot Study on Coarse PM Monitoring, EPA-454/R-15-001, February 2015.
      http://www3.epa.gov/ttn/amtic/files/pmlOpilot/PMc ReportEPA454R-15-001.pdf
            April 2016                       2-3 5            Draft - Do Not Quote or Cite

-------
 1    measurements. None of these measurements are required by regulation. They include PM
 2    measurements such as particle counts, continuous carbon, and continuous sulfate.
 3          Particle count measurements are being implemented at several near-road monitoring
 4    stations as part of the recent deployment of the near-road monitoring program, but may be used
 5    at other locations. Particle counts are one of several measurements identified as being a
 6    secondary priority for multipollutant monitoring at near-road monitoring stations. Details on
 7    priorities for multipollutant monitoring at near-road monitoring stations can be found in Section
 8    16 of the Near-road NCh Monitoring Technical Assistance Document (TAD)72.
 9          Aethalometer data has been measured and submitted to AQS for many years. Data uses
10    include characterizing black carbon and wood smoke. Ambient air monitoring stations that may
11    have aethalometers include some of the near-road monitoring stations and National Air Toxics
12    Trends Stations (NATTS). About 40 monitoring sites across the county are reporting data from
13    aethalometers. While aethalometer data is available at high time resolutions, it is typically
14    reported to the AQS data base in one hour periods.
15          Continuous elemental and organic carbon data are monitored at select locations
16    participating in a pilot of the Sunset EC/OC analyzer as well as a few additional sites that were
17    already operating before the EPA initiated the pilot study. The  Sunset EC/OC analyzer provides
18    high time resolution carbon data, typically every hour, but in some remote locations the
19    instrument is programmed to run every 2 hours to ensure collection of enough aerosol. The data
20    from the Sunset is being compared to filter-based carbon methods from the carbon channel of the
21    CSN program. The six sites participating in the study are Washington DC, Chicago IL, St. Louis
22    MO, Houston TX, Las Vegas NV, and Los Angeles CA. Each of the six sites participating in the
23    pilot study are operating for at least three years.
24          Continuous sulfate continues to be measured at four locations. One station each in North
25    Carolina, Tennessee, Indiana, and Maine are reporting data. The continuous sulfate analyzer
26    provides hourly data and data can be readily compared to 24-hour sulfate data collected from the
27    ion channel in the CSN program.
28          In addition, over the last few years, the EPA has investigated the use of a number of PM
29    sensor technologies as  one of several areas of research intended to address the next generation of
30    air measurements. The investigation into air sensors is envisioned to work towards near real-time
31    or continuous measurement options that are smaller, cheaper, and more portable than traditional
      72 Near-road NO2 Monitoring Technical Assistance Document, Publication No. EPA-454/B-12-002, June 2012.
      Available on EPA's Web site at: http://www3.epa.gov/ttn/amtic/files/nearroad/NearRoadTAD.pdf

            April 2016                       2-36            Draft - Do Not Quote or Cite

-------
 1    FRM or FEM methods. These sensor devices have the potential to be used in a number of
 2    applications such as identifying hot spots, informing network design, providing personal
 3    exposure monitoring, supporting risk assessments, and providing background concentration data
 4    for permitting. The EPA has hosted workshops and published a number of documents and peer-
 5    reviewed articles on this work (See the EPA's web  site at: http://www2.epa.gov/air-
 6    research/next-generation-air-measuring-research).

 7    2.3.2  Consideration Of PM Monitoring Issues In The Current Review
 8           This review of the PM NAAQS will include the consideration of policy-relevant issues
 9    associated with measuring and characterizing PM in ambient air. These issues include the design
10    of the PM network, the performance of existing FRMs and FEMs, the performance of continuous
11    monitors, PM chemical speciation, PM monitoring  near major roadways, the availability and
12    performance of low-cost PM sensors, and data reporting requirements. The EPA will draw upon
13    the information presented in the ISA to inform the evaluation of appropriate ambient monitoring
14    methods and network design for PM. If there is a need for formal  scientific input on important
15    changes to ambient air monitoring, the EPA may request convening the Ambient Air Monitoring
16    Subcommittee of the Clean Air Scientific Advisory Committee. This subcommittee has worked
17    closely with the EPA during past reviews, where appropriate, to provide the scientific review of
18    monitoring options under consideration as part of the PM NAAQS review process.73 Input and
19    development of options to improve ambient air monitoring are also based on input from
20    monitoring agencies and other interested stakeholders.
      73 The EPA will draw upon the information presented in the ISA to inform the evaluation of appropriate ambient
      monitoring methods and network design for PM. Input and development of options to improve ambient air
      monitoring are also based on input from monitoring agencies and other interested stakeholders.

            April 2016                       2-37             Draft - Do Not Quote or Cite

-------
 1   REFERENCES

 2   Abt Associates, Inc. (2001). Assessing Public Opinions on Visibility Impairment due to Air
 3          Pollution: Summary Report. Prepared for EPA Office of Air Quality Planning and
 4          Standards; funded under EPA Contract No. 68-D-98-001. Abt Associates Inc., Bethesda,
 5          MD. January 2001. Available: http://www.epa.gov/ttncaaal/tl/reports/vis_rpt_fmal.pdf
 6   Abt Associates, Inc. (2005). Paniculate Matter Health Risk Assessment For Selected Urban
 1          Areas. Prepared for EPA Office of Air Quality Planning and Standards; funded under
 8          EPA Contract No. 68-D-03-002. Abt Associates Inc., Bethesda, MD. December 2005.
 9          Available: http://www3.epa.gov/ttn/naaqs/standards/pm/data/PMrisk20051220.pdf
10   BBC Research & Consulting (2003). Phoenix Area Visibility Survey. Draft Report. Available:
11          http://www.azdeq.gov/environ/air/download/vis_021903fpdf.
12   Ely, D.W., J.T. Leary, T.R. Stewart, and D.M. Ross (1991). The Establishment of the Denver
13          Visibility Standard. Presented at the 84th Annual Meeting & Exhibition of the Air and
14          Waste Management Association. June 16-21.
15   Pryor,  S.C. (1996). Assessing public perception of visibility for standard setting exercises.
16          Atmospheric Environment 30(15):2705-2716.
17   Samet  J (2009). Letter from Dr. Jonathan M. Samet, Chair, Clean Air Scientific Advisory
18          Committee to the Honorable Lisa P. Jackson, Administrator, U.S. EPA. CASAC Review
19          of EPA's Integrated Science Assessment for Parti culate Matter -Second External Review
20          Draft (July 2009). November 24, 2009. Docket ID number. EPA-HQ-ORD-2007-0517-
21          0121. Available:
22          http://vosemite.epa.gov/sab/sabproduct.nsf/264cbl227d55e02c85257402007446a4/151B
23          lF83B023145585257678006836B9/$File/EP A-CASAC-10-001-unsigned.pdf.

24   Samet  J (2010a). Letter from Dr. Jonathan M. Samet, Chair, Clean Air Scientific Advisory
25          Committee to the Honorable Lisa P. Jackson, Administrator, U.S. EPA. CASAC Review
26          of Policy Assessment for the Review of the PM NAAQS - First External Review Draft
27          (March 2010). May 17, 2010. Docket ID number EPA-HQ-OAR-2007-0492-0113.
28          Available:
29          http://vosemite.epa.gov/sab/sabproduct.nsf/264cbl227d55e02c85257402007446a4/E504
30          EE3276D87A9E8525772700647AFB/$File/EP A-CASAC-10-011-unsigned.pdf.

31   Samet  J (201 Ob). Letter from Dr. Jonathan M. Samet, Chair, Clean Air Scientific Advisory
32          Committee to the Honorable Lisa P. Jackson, Administrator, U.S. EPA. CASAC Review
33          of Policy Assessment for the Review of the PM NAAQS - Second External Review Draft
34          (June 2010). September 10,  2010. Docket ID number EPA-HQ-OAR-2007-0492-0256.
35          Available:
36          http://vosemite.epa.gov/sab/sabproduct.nsf/264cbl227d55e02c85257402007446a4/CCF9
37          F4C0500C500F8525779D0073C593/$File/EP A-CASAC-10-015-unsigned.pdf.

38   U. S. EPA (2005). Review of the National Ambient Air Quality Standards for Parti culate Matter:
39          Policy Assessment of Scientific and Technical Information, OAQPS Staff Paper.
40          Research Triangle Park, NC 27711:  Office of Air Quality Planning  and Standards. Report


            April 2016                      2-3 8            Draft - Do Not Quote or Cite

-------
 1          No. EPA-452/R-05-005a. December 2005. Available:
 2          http://www3.epa.gov/ttn/naaqs/standards/pm/sjm crsp.html.

 3   U.S. EPA (2009). Integrated Science Assessment for Particulate Matter: Final Report. National
 4          Center for Environmental Assessment-RTF Division, Office of Research and
 5          Development, Research Triangle Park, NC. EPA/600/R-08/139F. December 2009.
 6          Available: http://www3.epa.gov/ttn/naaqs/standards/pm/sjm 2007  isa.html.

 7   U. S. EPA (201 Oa). Quantitative Health Risk Assessment for Parti culate Matter - Final Report.
 8          Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency,
 9          Research Triangle Park, NC. EPA-452/R-10-005. June 2010. Available:
10          http://www3.epa.gov/ttn/naaqs/standards/pm/sjm 2007 risk.html.

11   U.S. EPA (201 Ob). Paniculate Matter Urban-Focused Visibility Assessment - Final Report.
12          Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency,
13          Research Triangle Park, NC. EPA-452/R-10-004. July 2010. Available:
14          http://www3.epa.gov/ttn/naaqs/standards/pm/sjm 2007 risk.html.

15   U.S. EPA (2011). Policy Assessment for the Review of the Particulate Matter National Ambient
16          Air Quality Standards. Office of Air Quality Planning and Standards, U.S.  Environmental
17          Protection Agency, Research Triangle Park, NC. EPA 452/R-l 1-003. April 2011.
18          Available: http://www3.epa. gov/ttn/naaqs/standards/pm/sjm 2007 ja. html.

19   U.S. EPA (2012). Responses to Significant Comments  on the 2012 Proposed Rule on the
20          National Ambient Air Quality Standards for Particulate Matter (June 29, 2012;  77 FR
21          38890). Available: http://www3.epa.gov/ttn/naaqs/standards/pm/data/20121214rtc.pdf

22   Zanobetti A; Schwartz J (2009). The effect of fine and  coarse parti culate air  pollution on
23          mortality: A national analysis. Environ Health Perspect,  117: 898-903.
           April 2016                      2-39            Draft - Do Not Quote or Cite

-------
 i                          3   SCIENCE ASSESSMENT

 2          Integrated Science Assessments (ISAs) serve as the scientific foundation of the NAAQS
 3   review process and are developed by the EPA's National Center for Environmental Assessment
 4   in Research Triangle Park (NCEA-RTP). As outlined in sections 1.2 and 1.4 above, the
 5   parti culate matter (PM) ISA will inform the review of both the primary (health-based) and
 6   secondary (welfare-based) PM standards.
 7   3.1   SCOPE OF THE PM ISA

 8   3.1.1  General Description
 9         The ISA provides a comprehensive synthesis and evaluation of the most policy-relevant
10   science, including key science judgments that are important to inform the development of the
11   risk and exposure assessments, as well as other aspects  of the NAAQS review (U.S. EPA, 2016).
12   The shift to the ISA away from the Air Quality Criteria Documents, which originally provided
13   the scientific basis of the NAAQS review process, was initiated in 2006. This change was rooted
14   in the "broad recognition that the Criteria Document is typically 'encyclopedic' in nature, which
15   is seen by many as contributing to an unnecessarily lengthy process for preparing document
16   drafts and for reviews by CAS AC [(Clean Air Scientific Advisory Committee)] and the public,
17   and obscuring a focus on the most policy-relevant scientific information" (U.S. EPA, 2006).
18   CAS AC provided positive feedback on this transition to the ISA, encouraging "the development
19   of a more timely and more concise integrated assessment of the policy-relevant science that
20   would replace the voluminous air quality criteria document"(Peacock, 2008). The purpose of the
21   PM ISA is to provide a critical evaluation and synthesis of the current scientific literature on
22   health and welfare evidence necessary to support the PM NAAQS review process.
23         The PM ISA  is not intended to provide a detailed review of all studies of the
24   aforementioned topics but rather, will draw from the available evidence to synthesize the current
25   state of knowledge of the most policy-relevant issues for the review of the primary and
26   secondary PMNAAQS. The PM ISA will update the scientific assessment upon which the last
27   PM NAAQS review was based. Thus, the PM ISA will  build on the conclusions of the last
28   review of the air quality criteria for PM as presented in  the 2009 PM ISA and focus on peer-
29   reviewed literature published since that document,74 as well as on any new interpretations of
     74 The 2009 PM ISA included studies through May 2009.
         April 2016                             3-1                Draft - Do Quote or Cite

-------
 1    previously available literature. Key findings, conclusions, and uncertainties from the 2009 PM
 2    ISA will be briefly summarized at the beginning of the PM ISA and at the beginning of
 3    individual sections. Important older studies may be discussed in detail to reinforce key concepts
 4    and conclusions and/or if they are open to reinterpretation in light of newer data. Older studies
 5    also may be the primary focus in some subject areas or scientific disciplines (e.g., epidemiology,
 6    controlled human exposure, animal toxicology, atmospheric science, exposure science, visibility
 7    impairment, climate, materials effects, and ecological effects) where research efforts have
 8    subsided, and these older studies remain the definitive works available in the literature.
 9          In order to provide a more focused evaluation of the scientific evidence for health- and
10    welfare-related effects, the PM ISA will discuss the most important topics that address policy -
11    relevant questions. Therefore, the PM ISA will more fully evaluate those health and welfare
12    effects for which the evidence in the 2009 PM ISA was less certain (i.e., effects where the causal
13    determination was "likely to be  causal", "suggestive", or "inadequate" as detailed below in
14    section 3.4.3) and where there is now a larger body of evidence (e.g., diabetes, nervous system
15    effects, etc.). For those health and welfare effects where the 2009 PM  ISA concluded that the
16    evidence was sufficient to infer  a causal relationship (i.e., health: short- and long-term PM2.5
17    exposures and cardiovascular effects; short- and long-term PM2.5 exposures and mortality; and
18    welfare: PM exposures and effects on materials, visibility, and climate), the PM ISA will focus
19    the evaluation on characterizing the extent to which new studies address key uncertainties and
20    limitations identified in the previous review or provide insight on new issues. For the
21    epidemiologic studies  the focus  of the evaluation will include, but not be limited to, addressing
22    the impact of exposure assessment techniques on associations observed; evaluating potential
23    copollutant confounding; assessing the impact of PM components and sources on associations
24    observed; and examining heterogeneity in PM associations. Those epidemiologic studies that do
25    not directly address uncertainties and limitations as identified in the 2009 PM ISA, but still meet
26    our inclusion criteria will be detailed in tables presenting information for consideration in
27    evaluating the adequacy of the current PM NAAQS (e.g., averaging time of exposure,
28    distribution of PM concentrations, etc.). For experimental studies, specifically toxicological
29    studies, the evaluation will focus on those studies that also address key uncertainties and
30    limitations in the evidence identified in the previous review. For example, does the new evidence
31    further our understanding of the biological mechanisms by which PM  elicits a health effect or
32    provide coherence for the effects observed in epidemiologic studies? Similar to the
33    epidemiologic studies, a more detailed description of key pieces of information from the
34    toxicological studies will be detailed in tables.
         April 2016                              3-2                 Draft - Do Quote or Cite

-------
 1          Consistent with the goal of making the PM ISA a focused assessment of the current state
 2    of the science with respect to PM, this ISA incorporates CAS AC advice received during the
 3    review of External Review Drafts of the 2009 PM ISA.75 For example, CAS AC provided
 4    specific advice about the evaluation of PMio. Although the indicator for coarse particles is
 5    PMio,76 studies that examine the health or welfare effects of exposures to only PMio are limited
 6    in their ability to inform the health and welfare effects of PMio-2.5 or PM2.5. As a result, CASAC
 7    suggested revisions to the PM ISA to "remove the impression that PMio is a separate pollutant
 8    from PM2.5 and PMio-2.s", specifically the removal of causal determinations for short- and long-
 9    term exposures to PMio (Samet, 2009) Additionally, CASAC suggested detailing "when
10    possible, the particle size distribution of the PMio mixture" (Samet, 2009). As such, the final
11    2009 PM ISA had minimal discussion of PMio health effect studies, and these studies were used
12    as supporting evidence of PM2.s-related health effects primarily due to the majority of PMio
13    studies being conducted in large urban areas where, in many locations, ambient monitoring
14    typically indicates that PM2.5 comprises a large percent of PMio mass. Therefore, consistent with
15    the previous CASAC panel's recommendation, the current PM ISA will focus on the evaluation
16    of health studies  of PM2.5 and PMio-2.5 as the most pertinent to addressing the key policy-relevant
17    questions of the PM NAAQS review. This is also consistent with the 2012 PM Final Rule, which
18    stated that"... the extent to which PMio effect estimates reflect associations with PMio-2.5 versus
19    PM2.5 can be highly uncertain [and, as a result]... it is preferable to consider PMio-2.5 studies
20    when such studies are available"  (78 FR 3086). Therefore, the evaluation of PMio studies will be
21    limited to those that specifically address remaining uncertainties or limitations in the PM2.5 or
22    PMio-2.5 health or welfare effects evidence or evaluate a new health  or welfare effect not
23    previously examined.
24          The general scope presented above will allow the PM ISA to primarily focus on scientific
25    evaluations that are pertinent to the key policy-relevant questions  described in Chapter 2
26    (sections 2.1 and 2.2) above. Within the discussion of the health and welfare effects information,
27    other scientific information will also be presented and evaluated in order to provide a better
28    understanding of the following issues: (1) the sources of PM in the ambient air; (2) fate and
29    transport of PM in the environment; (3) measurement of PM and recent ambient concentrations
      75 External Review Drafts of the PM ISA are available at:
      http://www3.epa.gOv/ttn/naaqs/standards/pm/s jm_2007_isa.html
      76 As discussed above (sections 1.3 and 2.1.2), the purpose of the PMio standard is to protect against exposures to
         ! 0-2.5.

         April 2016                              3-3                 Draft - Do Quote or Cite

-------
 1    of PM; (4) the validity of inferences that can be drawn about PM health and welfare effects
 2    based on exposure assessment methodology; (5) the potential influence of other factors (e.g.,
 3    other pollutants in the ambient mixture, ambient temperature) shown to be correlated with PM
 4    and health or welfare effects; (6) the shape of the concentration-response (C-R) relationship at
 5    PM concentrations at the low end of the distribution; and (7) populations and lifestages at
 6    greatest risk of PM-related health effects. The process for evaluating and synthesizing scientific
 7    literature and addressing key policy questions is detailed in the Preamble to the Integrated
 8    Science Assessments (U.S. EPA, 2015). Collectively, the approach outlined for the health and
 9    welfare effects will allow for the EPA to provide a focused assessment of the scientific evidence
10    that more directly informs policy-relevant considerations detailed in the Policy Assessment (PA)
11    and Risk and Exposure Assessment(s) (REA) as detailed in Chapter 2.

12    3.1.2  Defining Policy-Relevant Literature
13          PM is unique among the criteria pollutants in that it is composed of multiple components
14    and size fractions. PM is often examined using mass-based metrics for overall mass and the mass
15    of individual components within the following size fractions: PMio (thoracic PM; particulate
16    matter with a nominal mean aerodynamic diameter less than or equal to 10  jim), PM2.5 (fine PM;
17    particulate matter with a nominal mean aerodynamic diameter less than or equal to 2.5 jim), and
18    PMio-2.5 (thoracic coarse or coarse PM; particulate matter with a nominal mean aerodynamic
19    diameter greater than 2.5 jim and less than or equal to 10 jim). Ultrafine particles (UFP,
20    generally considered as particulates with a diameter less than or equal to 0.1 jim  [typically based
21    on physical size, thermal diffusivity or electrical mobility] (U.S. EPA, 2009)) are examined
22    using multiple indices such as mass, number concentration, and surface area. As  discussed in
23    Chapter 1, the current indicators of the PM NAAQS are mass-based for both fine particles,
24    represented by PM2.5, and for thoracic coarse particles, represented by PMio.
25          Given the array of PM size fractions and components examined in studies of atmospheric
26    chemistry, exposure, health effects, and welfare effects, it is important to clearly  define the types
27    of studies that will be evaluated within the PM ISA to ensure its conclusions are pertinent to the
28    key questions of the PM NAAQS review, specifically:
29             .   Does exposure to PM cause a direct health or welfare effect (i.e., an independent
30                 effect)?
31             .   Are the current indicators (i.e., PM2.5 for fine particles and PMio for thoracic
32                 coarse particles), averaging times (e.g., 24-hour average, annual average), and
33                 levels appropriate?
34          When evaluating the broad body of literature across scientific disciplines, the EPA
35    considers whether the studies fall within the scope of the PM ISA (i.e., provide information

         April 2016                              3-4                 Draft - Do Quote or Cite

-------
 1    which can address the above questions). As a result, the focus of the PM ISA with respect to the
 2    health evidence is on studies conducted at concentrations of PM that are relevant to the range of
 3    human exposures across ambient microenvironments (up to 2 mg/m3 PM, which is one to two
 4    orders of magnitude above ambient concentrations) and (1) include a composite measure of PM77
 5    or (2) apply some approach to assess the independent effect of PM when the exposure of interest
 6    is a source-based mixture (e.g., diesel exhaust, gasoline exhaust, wood smoke). However, the
 7    scope of experimental studies included in the PM ISA may be broader when examining modes of
 8    action for PM, and may include in vitro studies, studies examining relative toxicity, and studies
 9    conducted at concentrations > 2 mg/m3. In the case of (1), if a study focuses on a single
10    component, group of components, or source, the study must also examine a composite measure
11    of PM (e.g., mass of PM2.5, PMio-2.5, ultrafine particles [UFP]) to be included in this review.
12    This requirement ensures that the study is relevant to the scope of the PM ISA and evaluated in
13    the proper context; specifically, this approach will facilitate a comparison of effects or
14    associations observed to the current mass-based PM indicator. Case (2) primarily applies to
15    experimental studies that attempt to disentangle the independent effects of PM from a complex
16    air pollution mixture of particles, gases, and semi-volatile components. This may be
17    accomplished by using filtration (e.g., a particle trap) or other approaches to distinguish between
18    effects due to the mixture and effects due to the particles alone.
19          With respect to identifying policy-relevant studies for welfare effects (ecological effects,
20    visibility, materials effects, and climate) PM composition and/or size fraction are important to
21    consider. For ecological effects, studies that focus on deposited components of PM (including
22    metals  and organics, but not nitrogen (N) and sulfur (S)-containing compounds or their
23    transformation products)78 at ambient or near-ambient concentrations of PM will be emphasized,
24    leading to  the evaluation of the direct and indirect effects of PM on vegetation, soils, and biota.
25    Generally, components of PM drive ecosystem responses, although size fraction may play a role
26    in the direct effects of PM such as deposition to vegetation. As a result, the  emphasis will be on
27    studies of ecological effects detailing non-nutrient (N and S) particle chemistry/composition
28    (e.g., cations, trace metals, semi-volatile organics), associated size fraction, and magnitude and
29    rates of wet and dry deposition across the landscape.
      77 Composite measures of PM may include mass, volume, surface area, or number concentration.
      78 The PM ISA will not evaluate nutrient N and S ecosystem effects, which will be addressed in the concurrent
      review of the secondary NOX and SOX NAAQS for ecological effects, see:
      http://www3.epa.gov/ttn/naaqs/standards/no2so2sec/index.html.
         April 2016                              3-5                 Draft - Do Quote or Cite

-------
 1          For materials effects, visibility, and climate effects, this ISA will build on information
 2    available during the last review describing the role of PM on materials damage and soiling,
 3    visibility impairment, and in radiative forcing resulting in global and regional climate change.
 4    For effects on materials, studies will be included that examine the role of PM, or related N and S
 5    species as discussed in Section 1.4, on materials damage and soiling. Specifically, studies will be
 6    considered that examine gaseous and particle contributions from nitrogen and sulfur along with
 7    other PM components due to the difficulty in determining whether an effect on materials is due
 8    to the particle or gaseous component of NOx and SOx and because the NOx/SOx secondary ISA
 9    focuses only on ecological effects, not other welfare effects. For visibility, studies will be
10    included which advance our understanding of visual impairment of airborne PM, including
11    studies of atmospheric chemistry, visibility preference, or other measures of adversity to public
12    welfare, in urban and rural settings. For climate, the ISA will focus on climate as the welfare
13    effect as listed in the 1977 Amendment to the Clean Air Act and will not focus on downstream
14    ecosystem effects, human health effects, or future air quality projections resulting from changes
15    in climate (Clean Air Act Amendments of 1977). The primary literature base for the evaluation
16    of the effects of airborne and deposited PM on climate will come from recent national and
17    international climate assessments such as the National Climate Assessment (Melillo et al., 2014)
18    and International Panel on Climate Change (IPCC, 2014). Focus will be on studies that inform
19    the independent role of PM in  climate forcing as well as effects on U.S. national and regional
20    climate.

21    3.2  ORGANIZATION OF THE PM ISA

22          The broad organization of the PM ISA will be consistent with that used in the recent
23    assessments for other criteria pollutants (e.g., ISA for Ozone and Related Photochemical
24    Oxidants, U.S. EPA, 2013c). The detailed description of the procedures for the assessment of
25    scientific information within the ISA for each of the criteria air pollutants can be found in the
26    Preamble to the Integrated Science Assessments (U.S. EPA, 2015). The PM ISA will begin with
27    a Preface discussing major legal and historical aspects of prior PM NAAQS reviews. An
28    executive summary will succinctly summarize the conclusions of the ISA. An integrative
29    synthesis chapter will provide  a more detailed summary of the key information for each topic
30    area, including the causal  determinations for relationships between exposure to PM and health
31    and welfare effects, information describing the extent to which health and welfare effects can be
32    attributable specifically to PM, evaluation of at-risk lifestages and populations, and other
33    uncertainties related to the interpretation of scientific information. The integrative synthesis
34    chapter also will discuss policy-relevant issues such as the exposure averaging times and the lag

         April 2016                             3-6                 Draft - Do Quote or Cite

-------
 1    structure of associations for health effects; the C-R relationships including the overall shape and
 2    whether or not the evidence supports identification of a discernible threshold below which
 3    effects are not likely to occur; and the public health and welfare impact of effects associated with
 4    exposure to PM. Subsequent chapters will be organized by subject area with the health evidence
 5    presented in separate chapters by PM size fraction (i.e., PM2.5, PMio-2.5, and UFPs) and all of the
 6    welfare (i.e., ecological, visibility, materials, and climate) effects evidence discussed in a
 7    separate chapter(s). Each of the chapters will contain an evaluation of results from recent policy -
 8    relevant studies integrated with previous findings (see section 3.4 below for specific issues to be
 9    addressed). Sections for each broad health effect category (e.g., respiratory effects) will conclude
10    with a causal determination about the relationship with relevant exposures to PM for health
11    outcomes that are specific to the size fraction being evaluated (e.g., PM2.5, PMio-2.5, UFPs).79
12    Sections on welfare will conclude with causal determinations for ecological effects, visibility,
13    materials effects, and climate. The PM ISA will conclude with a chapter that examines studies
14    for evidence of differential exposure and risk for PM-related health effects to draw conclusions
15    about potential at-risk lifestages and populations.
16          The PM ISA may be supplemented with other materials (e.g., tables, figures) if additional
17    documentation is required to support information contained within the PM ISA. These
18    supplementary materials may include more detailed and comprehensive coverage of relevant
19    publications and may accompany the PM  ISA or be available in electronic form as output from
20    the Health and Environmental Research Online (HERO) database developed by the EPA
21    (http://hero.epa.gov).80 Supplementary information that is available in the HERO database will
22    be presented as electronic links in the PM ISA.
23    3.3  ASSESSMENT APPROACH

24    3.3.1  Introduction
25          In each NAAQS review, development of the ISA begins with a "Call for Information"
26    published in the Federal Register. This notice, which for PM was published December 3, 2014
27    (79 FR 71764), announces the EPA's initiation of activities in the preparation of an ISA for the
28    specific NAAQS review and invites the public to assist in this process through the submission of
      79 Within the evaluation of the evidence for specific size fractions, the ISA will also evaluate the strength of
      evidence for PM components and sources.
      80 For more information on the HERO database, see http ://hero .epa. gov/.

         April 2016                              3-7                 Draft - Do Quote or Cite

-------
 1   research studies in the identified subject areas. This and subsequent key components of the
 2   process currently followed for the development of an ISA are presented in Figure 3-1 and are
 3   described in greater detail in the Preamble to the Integrated Science Assessments (U.S. EPA,
 4   2015). Section 1.2 (above) briefly describes how the PM ISA fits into the larger PM NAAQS
 5   review. Important aspects of the development of the PM ISA are described in the sections below,
 6   including the approach for searching the literature, identifying relevant publications, and
 7   informing specific policy-relevant questions that are intended to guide the assessment. These
 8   responsibilities are undertaken by expert authors of the PM ISA chapters, which include the EPA
 9   staff in NCEA-RTP with extensive knowledge in their respective fields and extramural scientists
10   solicited by the EPA for their expertise in specific fields. The process for scientific and public
11   review of drafts of the PM ISA is described in section 3.5 below.
12
         April 2016                             3-8                Draft - Do Quote or Cite

-------
            Integrated Science Assessment Development Process
                                            Literature Search and
                                               Study Selection
                                                 (See Figure III)
                                    Evaluation of Individual Study Quality
         After study selection, the quality of individual studies is evajuated by ITS. EPA or outside experts in the fields of
         atmospheric science, exposure assessment, dosimetry, animal toxicology, controlled human exposure,
         epidemiology,  biogeochemistry, terrestrial and aquatic ecology,  and other welfare effects, considering the design,
         methods, conduct, and documentation of each study. Strengths and limitations of individual studies that may affect
         the interpretation  of the study are considered.
                 Develop Initial Sections
         Review and summarize conclusions from
         previous assessments and new study results
         and findings by discipline and category of
         outcome/effect (e.g., lexicological studies of lung
         function or biogeochemical studies of forests)
                 Peer Input Consultation
          Review of initial draft materials by scientists
          from both outside and within the U.S. EPA in
          public meeting or public teleconference.
                              Evaluation, Synthesis, and Integration of Evidence
         Integrate evidence from scientific disciplines.  Evaluate evidence for related groups of endpoints or outcomes to
         draw conclusions for specific health or welfare effect categories, integrating health or welfare effects evidence with
         information on mode of action and exposure assessment.
                      Development of Scientific Conclusions and Causal  Determinations
         Characterize weight of evidence and devejop judgments regarding causality for health or welfare effect categories.
         Develop conclusions regarding concentration- or dose-response relationships, potentially at-risk populations,
         lifestages, or ecosystems.
            Draft Integrated Science Assessment
         Evaluation and integration of newly published studies
            Final Integrated Science Assessment
      Clean Air Scientific Advisory Committee
      Independent review of draft documents for
      scientific quality and sound implementation of
      causal framework during public meetings.
                                                                        Public Comments
                                                             Comments on draft ISA solicited by the U.S. EPA
2           Source: Modified from Figure II of the Preamble to the Integrated Science Assessments

3                                              (U.S. EPA, 2015).

4    Figure 3-1. General process for development of Integrated Science Assessments.
         April 2016
3-9
Draft - Do Quote or Cite

-------
 1    3.3.2  Literature Search and Selection of Relevant Studies
 2          The EPA uses a structured approach to identify relevant studies for consideration and
 3    inclusion in the IS As. To be deemed relevant, studies must fall within the scope of the PM ISA,
 4    as detailed in Section 3.1. A Federal Register notice is published that announces the initiation of
 5    a review and requests information, including relevant literature, from the public (79 FR 71764).
 6    In addition, the EPA identifies publications by conducting a recursive multi-tiered literature
 7    search process that includes extensive manual and computer-aided citation mining of computer
 8    databases (e.g., PubMed, Web of Science) on specific topics in a variety of disciplines. The
 9    search strategies are designed a priori and iteratively modified to optimize identification of
10    pertinent published papers.
11          For this PM ISA, a broad search string will be developed with keywords including, but
12    not limited to, particulate, PM, PMio, PM2.5, PMio-2.5, and ultrafine particles with the specific
13    syntax tailored for each database (e.g., PubMed, Web of Science). The broad search string will
14    then be used in combination with additional search strings to identify references relevant to
15    health and welfare topic areas. In addition to the use of the broad PM search strings, papers will
16    be identified for inclusion in several additional ways: specialized searches on specific topics;
17    relational searches that identify recent publications that have cited references from previous
18    assessments; identification of relevant literature by expert  scientists; recommendations from the
19    public and CAS AC during the call for information and external review process; and review of
20    citations in previous assessments. These search methods will be used to identify recent research
21    published or accepted for publication since the 2009 PM ISA, i.e., starting in January 2009,
22    through approximately two months before the release of the second external review draft of the
23    PM ISA (see Table 1-3 for target dates). Studies published after the PM ISA cut-off date may
24    also be considered in subsequent phases of the NAAQS review, after assessing whether they
25    provide new information that impacts key scientific issues.
26          Once studies are identified through the multipronged search strategy, PM ISA authors
27    (EPA staff and extramural scientists) will review the studies for relevance.  Relevant to the
28    review of the primary PM NAAQS are epidemiologic, lexicological, and controlled human
29    exposure studies or reports that examine health effects in relation to exposure to PM as well as
30    studies or reports that examine sources, emissions, atmospheric chemistry,  human exposure,
31    dosimetry, and modes of action. For the review of the secondary PM NAAQS relevant studies
32    are those that examine ecological effects, visibility, effects on materials and climate. Specific
33    information detailing the scope of the PM ISA, and subsequently those studies that will be
34    evaluated within it are detailed above in section 3.1.2.
         April 2016                             3-10                Draft - Do Quote or Cite

-------
 1          To be included in the PM ISA, relevant studies and reports must have undergone scientific
 2    peer review and have been published or accepted for publication. Some publications retrieved
 3    from the literature search will be excluded as not being relevant (e.g., do not meet
 4    aforementioned criteria) based on screening of the title or citation (e.g., not about air pollution,
 5    conference abstract, review articles, commentaries) and will not documented further. For other
 6    publications, decisions about relevance require reading beyond the title. These publications will
 7    be labeled as "considered" for inclusion in the PM ISA and are listed in the HERO database
 8    (http://hero.epa.gov).
 9          From the group of "considered" references, studies and reports will be selected for
10    inclusion in the PM ISA based on review of the abstract and full text. The selection process will
11    be based on the extent to which the study is potentially policy-relevant and informative.
12    Potentially policy-relevant and informative studies will include those that provide a basis for or
13    describe the relationship between PM and effects, in particular, those studies that reduce
14    uncertainty or address limitations of critical issues. Also pertinent are studies that offer
15    innovation in method or design or present novel  information on effects or issues previously not
16    identified. Uncertainty can be addressed, for example, by analyses of potential confounding  or
17    effect modification by copollutants or other factors, analyses of C-R or dose-response
18    relationships, or analyses related to time between exposure and response. In keeping with the
19    purpose to accurately reflect the latest scientific knowledge, a majority of the discussion in the
20    PM ISA will describe studies published since the 2009 PM ISA. However, evidence from
21    previous studies will be included and integrated with results from recent studies. In some cases,
22    evidence from  previous studies may  be the key policy-relevant information in a particular subject
23    area or scientific discipline. Analyses conducted by the EPA using publicly available data, for
24    example, air quality and emissions data, will also be considered for inclusion in the PM ISA.
25    Informative studies will not be limited to specific study designs, model  systems, or outcomes.
26    While study quality is important, it is not the sole criteria for study inclusion. The combination of
27    approaches described above are intended to produce a comprehensive collection of pertinent
28    studies needed to address the key scientific issues that form the basis of the PM ISA. References
29    will be cited in the PM ISA by a hyperlink to the HERO database and also are compiled into
30    reference lists.

31    3.3.3  Evaluation of Individual Study Quality

32          After selecting studies for inclusion, individual study quality is evaluated by considering
33    the design, methods, conduct, and documentation of each study, but not by considering whether
34    the study results are positive, negative, or null. In the PM ISA, conclusions about the strength of

         April 2016                             3-11                Draft - Do Quote or Cite

-------
 1    inference from study results will be made by weighing the authors' conclusions and
 2    independently evaluating study quality as detailed in the Preamble to the Integrated Science
 3    Assessments (U.S. EPA, 2015). This uniform approach aims to consider the strengths,
 4    limitations, and possible roles of chance, confounding, and other biases that may affect the
 5    interpretation of the results from individual studies for both health and welfare. In assessing the
 6    scientific quality of studies, the following broad parameters will be considered:
 7       •  How clearly were the study design, study groups, methods, data, and results presented to
 8          allow for study evaluation?
 9       •  To what extent are the air quality data, exposure, or dose metrics of adequate quality to
10          serve as credible exposure indicators?
11       •  Were the study populations, species, subjects, or animal models adequately selected, and
12          are they sufficiently well-defined to allow for meaningful  comparisons between study or
13          exposure groups?
14       •  Are the statistical analyses appropriate, properly performed, and properly interpreted? Do
15          the analytical methods provide adequate sensitivity and precision to support study
16          conclusions?
17       •  Are likely covariates (i.e., potential confounding factors, modifying factors) adequately
18          controlled for or taken into account in the study design or  statistical analyses?
19       •  Are the endpoint measurements meaningful, valid, and reliable?
20          Additional considerations in evaluating individual study quality specific to particular
21    scientific disciplines (e.g. epidemiology, toxicology, ecology) are discussed in detail in the
22    Preamble to the Integrated  Science Assessments (U.S. EPA, 2015).

23    3.4   SPECIFIC ISSUES TO BE ADDRESSED IN THE PM ISA

24          The PM ISA will contain information relevant to considering whether it is appropriate to
25    retain or revise the current primary and secondary PM NAAQS. Decisions on the specific
26    content of the PM ISA will be guided by policy-relevant questions that frame the entire review of
27    the primary and secondary  PM NAAQS as outlined in Chapter 2 above. These policy-relevant
28    questions are related to two overarching issues. The first overarching issue is whether new
29    evidence reinforces or calls into question the evidence presented and evaluated in the last PM
30    NAAQS review: (1) with respect to factors such as the plausibility of health effects caused by
31    exposure to PM, specifically PM2.5 and PMio-2.5, and concentrations of PM associated with health
32    effects; and (2) with respect to welfare-related effects of PM. The second overarching issue is
33    whether uncertainties from the last review have been reduced and/or whether new uncertainties
34    have emerged,  such as whether other measures of PM (e.g., different size fractions or

         April 2016                              3-12                Draft - Do Quote or Cite

-------
 1    components) are better at capturing the health and welfare effects attributed to PM exposures
 2    than those used as the indicators for the current NAAQS. The PM ISA also will address a set of
 3    more specific policy-relevant questions related to the available scientific evidence that stem from
 4    these issues. These questions were derived from the last PM NAAQS review, as well as from
 5    discussions of the scientific evidence that occurred at the February 2015 science policy
 6    workshop which initiated this PM NAAQS review (79 FR 71764, December 3, 2014). The PM
 7    ISA may include supplementary material if additional documentation is required to support
 8    information contained in the PM ISA. As in recently completed ISAs, such supplementary
 9    material will be available in the online HERO database and referenced in the PM  ISA as
10    electronic links to HERO.

11    3.4.1  Causal Determinations from 2009 PM ISA
12          In the 2009 PM ISA, the EPA concluded that the findings of epidemiologic, controlled
13    human exposure, and animal toxicological studies collectively provided evidence of a "causal
14    relationship" for short- and long-term PM2.5 exposures and cardiovascular effects as well as
15    mortality (U.S. EPA, 2009, as summarized in Chapter 2). In evaluating a broader  range of health
16    effects for PM2.5, the 2009 PM ISA concluded there was evidence of a "likely to be causal
17    relationship" for short- and long-term PM2.5 exposures and respiratory effects (U.S. EPA, 2009,
18    as summarized in Chapter 2). Additionally, there was evidence "suggestive of a causal
19    relationship" for long-term PM2.5 exposures and other health effects, including developmental
20    and reproductive effects (e.g., low birth weight, infant mortality) and carcinogenic, mutagenic,
21    and genotoxic effects (e.g., lung cancer mortality) (U.S. EPA, 2009, as summarized in Chapter
22    2). The 2009 PM ISA also formed causal determinations for exposures to PMio-2.5 and ultrafine
23    particles. With respect to PMio-2.5, the 2009 PM ISA concluded that the evidence was
24    "suggestive of a causal relationship" for short-term exposures and cardiovascular effects,
25    respiratory effects, and mortality, with the evidence "inadequate to  infer a causal relationship"
26    for long-term PMio-2.5 exposures and all health effects (U.S. EPA, 2009, as summarized in
27    Chapter 2). For ultrafine particles, the 2009 PM ISA concluded the evidence was  "suggestive of
28    a causal relationship" for short-term exposure and cardiovascular and respiratory  effects, but
29    "inadequate to infer  a causal relationship" for short-term exposures and central nervous  system
30    effects and mortality, and for long-term exposure and all health  effects (U.S. EPA, 2009, as
31    summarized in Chapter 2). For welfare effects, the evidence indicated a "causal relationship" for
32    PM exposures and effects on visibility, climate, and materials; and a "likely to be causal
33    relationship" for PM exposures and ecological effects (U.S. EPA, 2009, as summarized in
34    Chapter 2).

         April 2016                             3-13                Draft - Do Quote or Cite

-------
 1          The causal determinations in the 2009 PM ISA, based on the causal framework and
 2    integration of available evidence from previous and recent studies, were presented with a
 3    summary of the available evidence at the end of the sections for each broad health and welfare
 4    effect category and in the integrative synthesis chapter at the beginning of the PM ISA. In the
 5    current review, specific policy-relevant questions related to the causal determinations that will be
 6    addressed include:
 7       •   Does the evidence base from recent studies contain new information to support or re-
 8           evaluate the causal determinations made for relationships between PM exposure and
 9           various health and welfare effects in the 2009 PM ISA?
10       •   Does new evidence confirm or extend biological plausibility of PM-related health
11           effects?
12       •   What is the strength of inference from epidemiologic studies based on the extent to which
13           they have:
14           o  Examined exposure metrics that capture the spatial and/or temporal pattern of PM in
15              the study area?
16           o  Assessed potential confounding by other pollutants and factors?
17       •   What information is available to  support a rationale for forming causal determinations for
18           PM size fractions other than PM2.5, specific PM components, or PM from specific
19           sources?
20       •  What information is available regarding the health impacts of a decrease in ambient PM
21           concentrations to inform causal determinations?

22    3.4.2  Uncertainties/Limitations Identified in 2009 PM ISA
23          The causal determinations described above for the  relationships between PM2.5, PMio-2.5,
24    and UFPs exposure and health and welfare effects were informed by uncertainties and limitations
25    in the evidence. For example, the 2009 PM ISA noted a number of uncertainties and limitations
26    in the health evidence, such as the heterogeneity often observed in multi-city epidemiologic
27    studies and whether this can be attributed to exposure differences or regional differences in PM
28    components or sources; and uncertainty  related to the use of ambient PM concentrations from
29    central-site monitors and their ability to  represent personal ambient PM exposure (U.S. EPA,
30    2009, Chapter 2). In each of the health and welfare effects sections, and the integrative synthesis
31    chapter, the PM ISA will evaluate uncertainties and limitations in the scientific data. These
32    uncertainties also will inform causal determinations. The PM ISA will evaluate potential
33    confounding by other ambient pollutants. To assess the independent effects of PM, the PM ISA
34    will examine whether epidemiologic associations with PM are observed in copollutant models.
35    Copollutant models are the predominant method used in air pollution epidemiology to estimate

         April 2016                             3-14                Draft - Do Quote or Cite

-------
 1    the effect of one pollutant for a given concentration of a copollutant. The PM ISA also will
 2    evaluate whether PM has interactions with copollutants or joint effects in associations with
 3    health outcomes. The assessment of potential confounding, interactions, or joint effects will draw
 4    upon results from health effects studies, available information on copollutant interactions in the
 5    atmosphere that influence the spatial distributions of PM and copollutants, as well as information
 6    from experimental studies that examine the health effects of PM exposures alone and PM in
 7    combination with other pollutants. In the absence of these aforementioned methods, the PM ISA
 8    will examine whether single-pollutant epidemiologic associations with health effects in a given
 9    study differ between PM and copollutants, and if insights regarding potential copollutant
10    confounding can be gained by examining the magnitude of correlation between pollutants.
11          Drawing from discussion about the strengths and limitations of various exposure
12    assessment methods, the PM ISA will evaluate the strength of inference in epidemiologic studies
13    by considering information such as the exposure duration being examined, the extent of spatial
14    and/or temporal variability in PM in the study area, the distribution of monitoring sites in the
15    study area, the performance of exposure models used, and time-activity patterns of the study
16    population. Additionally, monitoring data will be used to  characterize the spatial distributions in
17    ambient PM concentrations and human exposures,  and in turn, the potential exposure
18    measurement error in particular study areas based on the particular method of exposure
19    assessment used. The adequacy of exposure assessment in epidemiologic studies will be
20    considered in weighing the quality of evidence, and in turn, forming causal determinations.
21          Epidemiologic evidence is unlikely to completely address the uncertainties mentioned
22    above. Any individual study is unlikely to evaluate all potentially correlated copollutants, and the
23    limitations of epidemiologic methods in separating effects of highly correlated pollutants or
24    separating the effects of more than two pollutants in the same model are well recognized. With
25    respect to exposure measurement error, few studies with personal ambient exposure
26    measurements are likely to be available. Thus, coherence  with other lines of evidence may
27    strengthen inferences when there are uncertainties in epidemiologic evidence due to copollutant
28    confounding or exposure measurement error. Controlled human exposure and toxicological
29    studies that demonstrate similar effects at relevant PM exposures may provide coherence with
30    epidemiologic evidence. Further, experimental results that identify key events in the mode of
31    action may provide biological plausibility.
32          In the 2009 PM ISA, a number of uncertainties were identified with respect to quantitative
33    relationships between PM and effects on welfare. Although ecological effects due to PM
34    deposition were described in the previous ISA, there was  a lack of information linking specific
35    ambient concentrations of PM to impairment of ecological receptors. For materials effects, the

         April 2016                             3-15                Draft - Do Quote or Cite

-------
 1    rate of deposition of PM onto surfaces was identified as an uncertainty as was the role of co-
 2    pollutants in soiling effects. The 2009 PM ISA identified uncertainties in assessment of climate
 3    effects including quantification and modeling of aerosols and relating observed effects to
 4    measured concentrations of ambient PM. For visibility in the 2009 PM ISA, there were
 5    uncertainties associated with the visibility preference studies and insufficient urban monitoring
 6    data. The PM ISA will evaluate the status of these uncertainties and limitations in each of the
 7    welfare effects sections and this information will be used in the development of causal
 8    determinations.

 9    3.4.3  Integration of Evidence and Causal Determinations

10          As described in the Preamble to the Integrated Science Assessments (U.S. EPA, 2015,
11    3037426}, the EPA uses a structured framework to provide a consistent and transparent basis for
12    classifying the weight of available evidence for health and welfare effects according to a five-
13    level hierarchy: (1) causal relationship; (2) likely to be a causal relationship; (3) suggestive, but
14    not sufficient, to infer a causal relationship; (4) inadequate to infer a causal relationship; and (5)
15    not likely to be a causal relationship. In the framework, key considerations in drawing
16    conclusions about causality include consistency of findings for an endpoint across studies,
17    coherence of the evidence across  disciplines and across related endpoints, and biological
18    plausibility. As judged by these parameters, evidence that rules out chance, confounding, and
19    other biases with reasonable confidence is sufficient to infer a  causal relationship. Increasing
20    uncertainty due to limited available information, inconsistency, and/or limited coherence and
21    biological plausibility leads to conclusions lower in the hierarchy. Causality determinations are
22    based on the confidence in the integrated body of evidence, considering study design and quality
23    and strengths and weaknesses in the overall collection of previous and recent studies across
24    disciplines. In discussing  each determination of causality, the EPA characterizes the evidence
25    upon which the judgment is based, including the weight of evidence for individual endpoints
26    within the health or welfare effect category or group of related endpoints.
27          For evaluation of human health effects, determinations of causality are made for major
28    health effect categories (e.g., respiratory effects) or more specific groups  of related outcomes and
29    for the range of exposure  concentrations of PM defined to be relevant to ambient concentrations
30    (e.g., up to 2 mg/m3). Findings based on higher exposure concentrations may be considered if
31    they add to the understanding of biological plausibility, provided that they do not reflect different
32    biological mechanisms operating at higher concentrations. The main lines of evidence for use in
33    causal determinations for human health are controlled human exposure, epidemiologic, and
34    toxicological studies. Evidence is integrated from previous and recent studies. Other information

         April 2016                             3-16                 Draft - Do Quote or Cite

-------
 1    including mechanistic evidence, toxicokinetics, and exposure assessment may be drawn upon if
 2    relevant to the evaluation of health effects and if of sufficient importance to affect the overall
 3    evaluation.  The relative importance of different sources of evidence to the conclusions varies by
 4    pollutant or assessment, as does the availability of different sources of evidence when making a
 5    determination regarding causality. In judgments of causality,  scientists will also evaluate
 6    uncertainty in the scientific evidence, considering issues such as generalizing results from a
 7    small number of controlled human exposure subjects to the larger population; extrapolations of
 8    observed pollutant-induced pathophysiological alterations from laboratory animals to humans;
 9    confounding by co-exposure to other ambient pollutants, meteorological factors, or other factors;
10    the potential for effects to be due to exposure to air pollution  mixtures; and the influence of
11    exposure measurement error on epidemiologic study findings. Judgments of causality also are
12    informed by the extent to which uncertainty in one line of evidence (e.g., potential copollutant
13    confounding in  epidemiologic results) is addressed by another line of evidence (e.g., coherence
14    of effects observed in epidemiologic studies with experimental findings, mode of action
15    information). Thus, evidence integration is not a unidirectional process but occurs iteratively
16    within and across scientific disciplines and related outcomes.
17          A similar process is used for the integration of evidence and determination of causality for
18    welfare-related  effects. For ecological effects, several types of experimental approaches (e.g.
19    field studies, mesocosm, controlled laboratory exposures, gradient) may be relevant for assessing
20    PM effects  to ecosystems. Field studies measure biological changes in uncontrolled situations
21    and describe an association between a disturbance and an ecological effect. They are most useful
22    for linking stressors with  effects when stressor and effect levels are measured concurrently.  The
23    presence of confounding factors can make it difficult to attribute observed effects to specific
24    stressors. Mesocosm  studies are performed in the natural environment while controlling for
25    some, but not all, of the environmental conditions and can be considered a hybrid between a field
26    and a laboratory study. Controlled exposure studies provide the strongest evidence for causality,
27    however the scope of inference to conditions in natural systems may be limited. Studies where
28    effects are observed at relevant ambient concentrations are weighted more heavily, however,
29    studies conducted at higher concentrations can inform the understanding of mode of action in
30    aquatic and terrestrial organisms. The EPA evaluates evidence relevant to understanding the
31    quantitative relationships between pollutant exposures and ecological effects. Integration of
32    evidence for visibility and climate draws upon modeling and monitoring data as well as
33    experimental approaches  designed to characterize the role of PM in atmospheric processes. In the
34    case of visibility, new preference studies designed to quantitatively assess the impact of PM on
35    visibility will also be incorporated if available. This includes  evaluating the C-R or deposition-

         April 2016                             3-17                 Draft - Do Quote or Cite

-------
 1    response relationships and, to the extent possible, drawing conclusions on the levels at which
 2    effects are observed. For materials effects, evidence of corrosive and soiling effects on stone,
 3    metals, paint and other surfaces at relevant concentrations of PM will be considered. Generally, a
 4    causal determination is made based on many lines of evidence that reinforce each other and are
 5    based on integrating evidence from both previous and recent studies.

 6    3.4.4  Atmospheric Chemistry and Ambient Concentrations of PM

 7          The PM ISA will present and evaluate data related to ambient concentrations of PM and
 8    its components; sources leading to the presence of PM in the atmosphere; the deposition of PM;
 9    and chemical reactions that determine the formation, transformation, and lifetime of PMin the
10    atmosphere. Specific policy-relevant questions related to atmospheric chemistry and ambient
11    concentrations that will be addressed in the PM ISA include:

12          •   What are the strengths and limitations of existing and new measurement methods and
13              approaches  (including low cost sensors and remote sensing) for both advancing
14              science and providing routine measurements of particulate matter? What are the
15              important biases and uncertainties associated with existing and new monitoring
16              methods? How do new methods and approaches  compare with or complement routine
17              measurement tools (e.g., FRMs)?
18          •   What changes have occurred in primary PM and PM precursor emissions rates over
19              time? Can the impacts of biomass burning (including occurrence of more fires due to
20              climate change and potentially identification of more fires with more sensitive
21              sensors on satellites), new transportation and industrial technologies, improved
22              industrial and motor vehicle emission controls, and strategies for accounting for semi-
23              volatile organic emissions be observed in current atmospheric PM concentrations?
24              What are the uncertainties associated with source apportionment approaches, whereby
25              individual PM sources can be distinguished as a  function of the chemical composition
26              of the emitted PM?
27          •   What changes have occurred in the quantities of sulfate (SO42"),  nitrate (NOs"), and
28              secondary organic PM (SOPM) formed in the atmosphere over time? What new
29              information is available regarding mechanisms of their production, including aqueous
30              phase SOPM formation, the role  of sulfur and nitrogen in SOPM formation, and the
31              role of semi-volatile species? Have the relative contributions of primary and
32              secondary PM components  changed as a result of decreases in precursor emissions?
33          •   What are the current ambient concentrations of different size fractions of PM (i.e.,
34              PM2.5, PMio-2.5, UFP), and how have they changed over time? What are the
35              uncertainties in the spatial and temporal distributions of PM?  How do those patterns
36              vary with particle size fraction?
37          •   What uncertainties exist concerning the sources and atmospheric chemistry of PM
38              components (both primary and secondary particles)? How does PM composition
39              change over various spatial and temporal scales? What new information is available

         April 2016                            3-18                Draft - Do Quote or Cite

-------
 1              regarding the composition of the PM size fractions that could not be well
 2              characterized during the last review (e.g., coarse PM, ultrafme PM)?
 3          »   How does intercontinental or trans-boundary transport of pollutants contribute to
 4              background PM concentrations? What information is available to quantify the
 5              emissions from natural PM sources that contribute to background PM concentrations?
 6              Is there enough information to estimate background concentrations of PM species?
 7          The information provided in the Atmospheric Chemistry and Ambient Concentrations
 8    chapter will provide context for understanding ambient concentration data used as surrogates for
 9    human exposure and for interpretation of the health and welfare effects evidence in  subsequent
10    chapters.

11    3.4.5  Human Exposure
12              The PM ISA will evaluate methods for estimating exposure to ambient PM, the
13    factors that influence exposure to PM, as well as the ability to make inferences about personal
14    exposure to ambient PM when extrapolating from ambient concentration data, particularly in the
15    context of interpreting results from epidemiologic studies, and from controlled human exposure,
16    animal toxicology, and in vitro chamber studies of near-atmospheric conditions. The issues
17    surrounding the ability to make inferences about personal exposure differ by the exposure period
18    of interest. Short-term exposure studies (i.e., exposures ranging from hours to days to weeks)
19    primarily rely on temporal variation in exposure while long-term exposure studies (i.e.,
20    exposures ranging from months to years) rely on spatial variability of exposure.  The PM ISA
21    will consider the available information on differential exposures to UFPs, PM2.5, and PMio-2.5
22    and on particle characteristics such as chemical composition, size, surface area, number, and
23    source. Specific policy-relevant questions related to human exposure that will be addressed in the
24    PM ISA include:
25          •   Are new models or other techniques available to estimate human exposure to ambient
26              PM? What are the strengths and limitations of existing and new techniques? Do any
27              of these techniques characterize PM by size fraction or composition?
28          •   What techniques are available to replicate human exposure to ambient PM in
29              controlled human exposure, animal toxicology, and in vitro experiments? What
30              factors lead to uncertainty in the exposure conditions established within these
31              experiments? What factors affect inference about human exposure from  such
32              experiments?
33          •   What are the limitations in our understanding of exposures to PM components,  size
34              fractions, sources/environments, and relationships between exposure to specific PM
35              size fractions and/or components and corresponding exposure to gaseous co-
36              pollutants? What uncertainties exist regarding the spatial and temporal patterns of PM

         April 2016                             3-19               Draft - Do Quote or Cite

-------
 1              concentrations, size fractions, and components as they relate to patterns of human
 2              exposure?

 3           •   What are the uncertainties when extrapolating from stationary PM monitoring
 4              instruments to personal exposure to PM of ambient origin, especially in populations
 5              at increased risk of PM-related health effects? How do these uncertainties vary for
 6              particles of different size or composition? Issues include measurement error in
 7              outdoor ambient monitors, the use of centralized monitors for estimating community
 8              concentrations, and the use of centralized monitors as a  surrogate for personal
 9              exposure to PM of ambient origin.
10           •   What are the uncertainties associated with using measurements of ambient
11              concentrations of PM to provide an estimate of ambient exposures for health studies,
12              an indicator of personal exposure to PM, and/or an indicator of exposure to other
13              pollutants or pollutant mixtures?
14           •   Higher sampler density is more likely for saturation samplers or low-costs sensors
15              than for other measurement tools with greater sensitivity and precision. How does
16              sampler density affect data quality  objectives needed for individual samplers,
17              particularly for saturation samplers or low-cost sensors?
18           •   What limitations exist for interpretation of sub-daily, daily, and longer-term PM
19              concentration data used to estimate exposures?  How do factors such as air exchange
20              rates, indoor sources, ambient sources,  and methods for measuring personal
21              exposures to ambient PM influence interpretation of PM concentration data over
22              different sampling periods? Does this information differ by PM characteristics (e.g.,
23              size, chemical composition, sources/environments)?
24           •   What new developments have occurred with respect to chemical  transport modeling
25              of short-term and long-term PM concentrations for use in exposure assessment? How
26              might modeling and satellite data supplement monitoring data for understanding
27              human exposures? What are the limitations of using modeling or satellite data in lieu
28              of monitoring data? What advancements have been made with respect to techniques
29              for fusing modeling, monitoring, and/or satellite data for assessing long-term
30              exposures to ambient PM? What are  the uncertainties in data from chemical transport
31              models and satellites at the extremes of the concentration distribution, such as in low
32              concentration areas?
33           •   What new developments have been made in use of source apportionment techniques
34              for assessing human exposure to ambient PM?
35           •   What new developments have been made in assessing and/or correcting the influence
36              of exposure measurement error on health effect estimates for epidemiologic studies of
37              short-term and long-term exposure? How do these methods reduce the uncertainty
38              and/or bias in the health effect estimates for PM exposure?

39    3.4.6  Dosimetry

40           The PM ISA will evaluate literature focusing on dosimetry that may underlie the health
41    outcomes associated with exposure to PM. In Chapter 4 of the 2009 PM ISA, in relation to

         April 2016                             3-20               Draft - Do Quote or Cite

-------
 1    particle deposition in the body, it was concluded that the first line of defense for protecting the
 2    lower respiratory tract from inhaled particles is the nose and mouth; that the lower respiratory
 3    tract of children receive a higher surface dose of ambient PM compared to adults; and that people
 4    with COPD generally have greater total deposition and more heterogeneous deposition patterns
 5    compared to healthy individuals. In relation to particle clearance, it was concluded that particles
 6    depositing on the olfactory mucosa may translocate to the brain; and that a small, but statistically
 7    significant fraction (<1% of deposited material) of poorly soluble particles deposited in the
 8    alveolar region may translocate into circulation. A number of factors such as age and respiratory
 9    disease were recognized as affecting both rates of deposition and clearance. Interspecies
10    differences in deposition and clearance were also discussed. In the current review, specific
11    policy-relevant questions related to dosimetry that will be addressed in the PM ISA include:
12           •   Are there new data that better quantify extrathoracic and thoracic deposition of
13              particles in children, adults, and among species?
14           •   Are new data available evaluating the effects  of respiratory disease on deposition and
15              clearance?
16           •   What new data or models are available that help facilitate interspecies comparisons?
17           "Is there new information related to the translocation of particles  into circulation or the
18              olfactory bulb?
19    3.4.7  Modes  of Action
20          The PM ISA will evaluate literature focusing on modes of action that may underlie the
21    health outcomes associated with exposure to PM. These topic areas will be  developed using both
22    human and animal data. In the current review, specific policy-relevant questions related to modes
23    of action that will be addressed include:
24           •   How do physical-chemical particle characteristics influence biological responses to
25              inhaled PM?
26           •   What new evidence is available to characterize biologic responses of various PM size
27              fractions or PM from different sources?
28           •   What are the time-courses of biological responses to inhaled PM? What are the
29              implications of these time-courses for health effects associated with  short-and long-
30              term exposure to PM?
31           •   What are the interspecies differences in biological responses to inhaled PM? What are
32              the implications of interspecies differences for extrapolation of results to humans?
33           •   Is there a common underlying biological  response (e.g., inflammation) that supports
34              an effect of PM exposure on an array of health outcomes (e.g., cardiovascular,
35              respiratory,  or nervous system)?
         April 2016                              3-21                Draft - Do Quote or Cite

-------
 1           •   What new evidence is available to elucidate mechanisms by which pulmonary
 2              deposition of PM leads to extrapulmonary effects?

 3    3.4.8  Health Effects
 4            In the 2009 PM ISA, the health effects evidence for PM largely focused on PM2.s;
 5    fewer studies were available for PMio-2.5, UFPs, and sources/components of PM. The evidence
 6    indicated that a "causal relationship exists" for short- and long-term exposures to PM2.5 and
 7    cardiovascular effects and mortality, and a "likely to be causal relationship exists" for short- and
 8    long-term PM2.5 exposures and respiratory effects. More limited evidence with a larger degree of
 9    uncertainty formed the basis for the determinations for other health effects and other PM size
10    fractions (i.e., PMio-2.5 and UFPs). Additionally, with respect to sources and PM components, the
11    EPA concluded "that many PM [components] can be linked with differing health effects and the
12    evidence is not yet sufficient to allow differentiation of those [components] or sources that are
13    more closely related to specific health outcomes (U.S. EPA, 2009, p. 2-26)" The EPA will build
14    on the 2009 PM ISA by evaluating the newly available literature related to PM exposures and
15    health effects, including, but not limited to respiratory, cardiovascular, neurologic, reproductive
16    and developmental effects, mortality, and cancer. Depending on data availability, other health
17    effects may be evaluated.
18            The PM ISA will evaluate health effects that occur following both short- and long-term
19    exposures as examined in epidemiologic, controlled human exposure, and animal toxicological
20    studies. Efforts will be directed towards identifying the concentrations at which effects are
21    observed, particularly in potential at-risk lifestages and populations, and assessing the role of PM
22    within the broader mixture of ambient air pollutants. The data will be reviewed for relevance to
23    the current elements of the PM NAAQS (i.e., indicator, averaging time,  form, and level) in order
24    to support decisions regarding the adequacy of the current primary standard. Additionally, the
25    evidence will be evaluated to determine whether specific chemical components and/or
26    source(s)/environment(s) (e.g., near-road)  more fully explain PM health effects. The discussion
27    of health effects will be integrated with relevant information on dosimetry and modes of action,
28    as well as with information from the exposure chapter.
29            In light of recent International Agency for Research on Cancer (IARC) conclusions on
30    the carcinogenicity of outdoor air pollution, and specifically of PM, it is important for the EPA to
31    clearly outline its approach for evaluating  the relationship between long-term PM exposures and
32    cancer in the PM ISA (Loomis et al., 2013). While IARC focuses on whether a chemical is a
33    carcinogen  at any concentration and through any route of exposure, the PM ISA focuses on
34    whether PM can directly cause cancer through inhalation exposures at ambient and near-ambient

         April 2016                             3-22                Draft - Do Quote or Cite

-------
 1    concentrations. When evaluating the epidemiologic evidence for cancer, the EPA will focus on
 2    those studies with composite measures of PM and with exposures consistent with the overall
 3    scope of the PM ISA detailed in Section 3.1, i.e., concentrations of PM that represent the range
 4    of human exposures across ambient microenvironments (up to 2 mg/m3 PM, which is one to two
 5    orders of magnitude above ambient concentrations). For the toxicological evidence, the EPA will
 6    describe in vitro studies and other mechanistic studies related to cancer in the Mode of Action
 7    chapter because these types of studies are pertinent to the biological pathways underlying cancer.
 8    Consistent with previous ISAs, in vivo studies that examine cancer-related endpoints will be
 9    discussed in combination with the epidemiologic evidence in the health chapters. The evaluation
10    of toxicological studies related to mutagenicity, genotoxicity, epigenetic changes, and
11    carcinogenicity will focus on inhalation exposures primarily at ambient relevant concentrations
12    (i.e., up to 2 mg/m3) with the potential inclusion of exposures above this concentration. Lastly,
13    the evaluation of cancer will not focus on studies that use PM filter extracts because they may
14    not mimic what is bioavailable in vivo, or on studies of individual PM components due to the
15    inability to compare effects to the current mass-based PM indicator.
16            In the current review,  specific policy-relevant questions that will be  addressed in
17    consideration of health effects associated with short- and long-term exposure  to PM, include the
18    following:
19    Short-Term Exposure:

20          •   What  new evidence is available to inform policy-relevant considerations of the PM
21              NAAQS (e.g., considerations  related to indicator, averaging time,  form, level) for
22              those health categories where  the 2009 PM ISA concluded that a "causal relationship"
23              exists (i.e., short-term PM2.5 exposures and mortality and cardiovascular effects)? Do
24              new controlled human exposure and toxicological studies continue to provide support
25              for biologically plausible relationships between short-term PM exposures and
26              cardiovascular health effects?  Do new studies report PM-attributable effects at lower
27              PM concentrations than indicated by studies available in the last review?
28          •   How do results of recent studies expand understanding of the relationship between
29              short-term exposure to PM and respiratory effects, such as exacerbation of asthma or
30              chronic obstructive pulmonary disease and respiratory infection? Does recent
31              evidence improve coherence across disciplines for lung function changes, pulmonary
32              inflammation, host defense mechanisms, and outcomes such as symptoms, hospital
33              admissions, or emergency department visits?
34          •   To what extent is short-term exposure to PM related to or associated with the
35              progression of diabetes, other  metabolic diseases, and/or to other endocrine system
36              effects? To what extent are new health outcomes related to or associated with PM
37              exposures?

         April 2016                             3-23                Draft - Do Quote or Cite

-------
 1           •   Across the evaluated health effects, what new evidence is available on effects
 2              occurring from exposures at averaging times different (e.g., 1-hour) than the current
 3              24-hour average PM NAAQS?
 4    Long-Term Exposure:
 5           •   What new evidence is available to inform policy-relevant considerations of the PM
 6              NAAQS (e.g., considerations related to indicator, averaging time, form, level) for
 7              those health categories where the 2009 PM ISA concluded that a "causal relationship"
 8              exists (i.e., long-term PM2.5 exposures and mortality and cardiovascular effects)? Do
 9              new controlled human exposure and toxicological studies continue to provide support
10              for biologically plausible relationships between long-term PM exposures and
11              cardiovascular health effects?  Do new studies report PM-attributable effects at lower
12              PM concentrations than indicated by studies available in the last review?
13           "To what extent do recent studies improve understanding of the relationships  between
14              long-term PM exposure and the development of asthma or to the impairment of lung
15              development? Do recent studies improve coherence across disciplines for respiratory
16              disease incidence, pulmonary inflammation and oxidative stress, and development of
17              allergic responses?
18           "To what extent do recent studies improve understanding of the relationship between
19              long-term PM exposure and reproductive and developmental health outcomes, such
20              as  adverse birth outcomes, fertility and pregnancy outcomes (e.g., infertility, sperm
21              quality, preeclampsia, gestational hypertension), or developmental outcomes (e.g.,
22              neurocognitive ability)?
23           •   To what extent does new literature support a biologically plausible relationship
24              between long-term PM exposures and a number of nervous system effects (e.g.,
25              cognitive decline and autism)?
26           •   How do results of recent studies expand our understanding of the relationship
27              between long-term PM exposure and cancer, mutagenic, genotoxic, and epigenetic
28              effects? To what extent  does the evidence indicate that long-term exposure to PM
29              can increase the incidence of cancer?
30           "To what extent is long-term exposure to PM related to or associated with the
31              development  of diabetes and other metabolic diseases, as well as to health effects in
32              the endocrine system or other organ systems? To what extent are new health
33              outcomes related to or associated with PM exposures?

34    Additional Policy-Relevant Considerations:
35           •   Across the evaluated health effects, to what extent does new evidence inform the
36              understanding of differential effects from exposures to various PM size fractions (i.e.,
37              PM2.5, PMio-2.5, UFPs)?
38           "Do studies of mortality, hospital admissions, or emergency department visits provide
39              new information to improve our understanding of the potential heterogeneity in
40              effects  observed across the U.S. in multi-city studies?
         April 2016                             3-24                Draft - Do Quote or Cite

-------
 1           •   How do recent studies support the attribution of health effects to one or more PM
 2              component(s) or source(s), in addition to PM mass, for health effects for which there
 3              is sufficient evidence of a strong relationship (e.g., cardiovascular effects, mortality)
 4              with PM exposure?
 5           •   How do the results of recent studies inform the shape of the concentration-response
 6              relationship for PM and various health outcomes (e.g., mortality, hospital admissions,
 7              etc.), especially for concentrations near or below the levels of the current PM
 8              NAAQS?
 9           •   What new evidence adds to the understanding of which lifestages and populations are
10              at risk of PM-related health effects?
11           •   What new evidence supports evaluation of inter-individual variability in response to
12              PM exposures?
13           •   Across health effects, what new information is available to delineate the effects of
14              chronic exposure to lower concentrations versus acute, repeated exposures to higher
15              concentrations of PM?
16           •   What is the nature of health effects in persons exposed to multipollutant mixtures that
17              contain PM in comparison to exposure to PM alone?

18    3.4.9  At-Risk Lifestages and Populations and Public Health Impact
19              The PM ISA will evaluate an array of factors that may contribute to increased risk of
20    PM-related health effects for various lifestages or populations. These factors reflect the multiple
21    avenues through which a lifestage or population may be at increased risk of an air pollutant-
22    related health effect, specifically: intrinsic factors (e.g., biological factors such as age or genetic
23    variants), extrinsic factors (nonbiological factors such as diet, low SES), and/or factors affecting
24    dose or exposure (e.g., sex, age, outdoor activity or work, low SES, physical activity). It is also
25    important to recognize the interconnectedness among these factors that may also confer
26    increased risk, an example being pre-existing diseases or conditions and socioeconomic status.
27    The 2009 PM ISA evaluated studies that provided evidence that children, older adults, people
28    with pre-existing cardiopulmonary diseases, and people with lower SES are at increased risk of
29    PM-related health effects (U.S. EPA, 2009, Chapter 8).  Since completion of the 2009 PM ISA,
30    the EPA has developed a more detailed framework to provide a consistent and transparent basis
31    for communicating the overall confidence in the evidence that a particular factor may increase
32    the risk of an air pollutant-related health effect for a lifestage or population according to one of
33    four levels: adequate evidence, suggestive evidence, inadequate evidence, and evidence of no
34    effect (U.S. EPA, 2013c, Table 8-1). Key considerations in characterizing the evidence include
35    consistency of findings for a factor within a discipline and, where available, coherence of the
36    evidence  across disciplines as well as biological plausibility.  Several lines of evidence inform

         April 2016                              3-25                Draft - Do Quote or Cite

-------
 1    conclusions about at-risk lifestages and populations, but primarily include observational or
 2    experimental  studies that compare exposure to PM or relationships with health effects among
 3    groups that vary by some characteristic, such as pre-existing disease or age (i.e., exposure or
 4    effect modification). Also relevant are comparisons of results among observational or
 5    experimental  studies that examine different lifestage or population characteristics or time
 6    windows of exposure and experimental studies that examine health effects of PM in a group with
 7    a particular characteristic (e.g., genetic background, pre-existing disease). Where possible, the
 8    discussion will also include evaluation of the adversity of the health effects potentially associated
 9    with exposure to PM. The assessment of public health impact also may include, as appropriate,
10    an estimation of the sizes of potential at-risk lifestages and populations. Further, to the extent
11    that evidence is available, the at-risk chapter of the PM ISA will discuss what evidence is
12    available regarding interrelationships among risk factors in a particular lifestage or population as
13    described in the preceding section that may add to the understanding of the public health impact
14    of exposure to PM. Specific questions include:
15           "Is  there new information that identifies a combination of factors (i.e., co-occurring)
16              that can lead to one lifestage or population being at greater risk compared to another?
17           •  Have recent studies characterized whether certain lifestages or populations experience
18              differential exposures to PM mass, PM components or PM sources, which may
19              contribute to them being at increased risk?
20           •  What new evidence is available regarding additional lifestages or populations (e.g.,
21              pre-existing diseases such as diabetes, pregnant women) potentially at increased risk
22              of a PM-related health effect?
23           •  Have recent studies been able to delineate whether those lifestages or populations at
24              increased risk of a PM-related health effect differ depending on exposures to different
25              PM size fractions, PM components, or PM sources?
26    3.4.10 Welfare Effects
27           In the 2009 PM ISA, the welfare effects evidence for PM focused on the effects of
28    airborne PM,  including NOx and SOx, on visibility and climate, and on the effects of deposition
29    of PM constituents other than NOx and SOx, primarily metals and carbonaceous compounds on
30    materials and ecology. The  evidence indicated that a "causal relationship exists" for PM
31    exposures and effects on visibility, climate, and materials; and a "likely to be causal relationship
32    exists" for PM exposures and ecological effects. The EPA will build on the 2009 PM ISA by
33    evaluating the newly available literature related to PM exposures and welfare effects, specifically
34    ecological effects, visibility, climate, and materials.
35    Ecological Effects

         April 2016                              3-26                Draft - Do Quote or Cite

-------
 1          The PM ISA will summarize studies of impacts of PM deposition with the secondary
 2    NOx/SOx ISA covering the depositional effects of N and S on ecological receptors. As indicated
 3    in the previous ISA, chemical components of PM determine ecological response to a greater
 4    degree than PM size class. However, size fraction may play a role in direct effects such as
 5    deposition on vegetative surfaces near PM sources. In regards to effects of PM on ecosystem
 6    components (e.g. plants, soils, wildlife, nutrient cycling), both direct and indirect effects of PM
 7    deposition will be considered. PM deposited  directly on vegetative surfaces may impair
 8    photosynthesis and gas exchange between the plant and the atmosphere. Indirect effects of trace
 9    metals and organics associated with PM deposition on ecosystem receptors including aquatic and
10    terrestrial organisms, as well as effects on ecosystem structure and function, will be evaluated. If
11    studies are available that link PM deposition  to ecosystem services these findings will be
12    discussed. Specific questions include:
13           "Is there new information on PM that can be correlated to a specific physiological
14              response in aquatic and terrestrial biota?
15           •   Do new studies show effects on ecosystem structure, function and biodiversity or PM
16              impacts to ecosystem services?

17    Visibility Effects
18           The PM ISA will summarize long-known information needed for placing current
19    information in context with respect to visibility. Previous evaluations have indicated that
20    anthropogenic sulfate and nitrate particles are responsible for most of the regional haze in the
21    eastern U.S. In the West, sulfates may be generally of less importance since less sulfur dioxide is
22    emitted there, but important sub-regional and seasonal differences still exist. Nitrates are a
23    leading factor in haze generation in specific areas such as southern California in winter, for
24    example, while in many of the less-populous areas of the West,  airborne soil particles can be the
25    largest contributor to visibility extinction during the spring.
26           In addition, anthropogenic nitrates and organic species, either emitted directly or formed
27    secondarily from other emissions, can be significant factors in haze formation. Other sources of
28    visibility extinction (e.g., dust, smoke, sea salt) have anthropogenic, biogenic, and geogenic
29    sources that vary in strength and significance sub-regionally as well. Smoke from wildfires and
30    prescribed burnings contribute to visibility extinction both directly and indirectly, though much
31    less is known of these indirect effects through SOPM generation. The PM  ISA will evaluate
32    newly available evidence summarizing the recent important policy-relevant findings and will
33    include sections for aerosol optical characteristics, spatial and temporal trends, and causes of
34    haze. Specific  questions include:
         April 2016                              3-27                Draft - Do Quote or Cite

-------
 1           •   What relationships exist between ambient PM and visibility impairment? How do
 2              these relationships vary between urban and rural settings?
 3           •   What metrics and algorithms are available to estimate visibility impairment as a
 4              function of PM mass, optical characteristics, and composition?
 5           •   What role do spatial patterns (e.g., elevated paniculate nitrate in the midwestern U.S.
 6              and enhancement of sulfate concentrations in the eastern U.S.), seasonal patterns,
 7              multi-year trends, emissions changes, and meteorology play with respect to PM
 8              effects on visibility?
 9           •   What new preference studies or other metrics are available to describe the extent to
10              which PM-related visibility impairment may be considered adverse to public welfare?

11    Climate Effects

12          The PM ISA will present information on radiative forcing and climate change impacts of
13    PM and PM components. Background information on the physics of climate and radiative
14    forcing and observed trends in anthropogenic and natural PM will be presented as well as climate
15    impacts of changes in anthropogenic PM since preindustrial times. Focus will be on information
16    necessary for interpretation of effects described in the chapter and on newly available
17    information since the last ISA. PM size/effect dependencies (e.g., cloud formation and
18    precipitation) and PM composition/effect dependencies (e.g., black carbon and sulfate aerosol)
19    will be addressed as  they pertain to climate effects.

20           •   What are the important and relevant climate metrics used to quantify PM-related
21              climate effects and what are their strengths and limitations in assessing climate effects
22              at different temporal and spatial scales?
23           •   What is the current understanding of the magnitude and direction (warming vs.
24              cooling) of PM climate impacts and what important uncertainties remain?
25           •   To what extent do we understand the independent effect of PM on climate in the
26              broader context of other climate forcers?
27           •   What is the current state of knowledge of PM size and composition in relation to
28              climate forcing and what is the relationship between PM metrics (e.g., size,
29              composition) and important and relevant climate metrics used to quantify PM-related
30              climate effects?
31           •   What recent advancements have been made in understanding PM effects on regional
32              climate in the U.S.?

33    Materials Effects:

34          The PM ISA will summarize soiling and deterioration of materials attributable to PM and
35    related N and S components because of the difficulty associated with isolating the effects of
36    gaseous and paniculate N and S wet deposition and the NOx/SOx secondary ISA only focuses

         April 2016                             3-28                Draft - Do Quote or Cite

-------
 1    on ecological effects, not other welfare effects (See Section 1.4). Previous PM assessments
 2    indicated that stone, paint, metal and other materials can become discolored and/or undergo
 3    corrosion processes from particle pollution. Culturally important items such as buildings, statues
 4    or works of art can be physically damaged and/or lose aesthetic appeal. The focus of the PM ISA
 5    will be on whether there is new information on soiling and corrosion, dose-response relationships
 6    between PM or related N and S species and damage to surfaces. Any new areas of research
 7    focus on PM-related materials effects will also be emphasized in the PM ISA.
 8          "Is there new information on soiling of stone, metal and paint surfaces due to PM or
 9              related N and S species? Corrosive effects?
10          •   Do new studies show effects on additional types of materials?

11    3.5  QUALITY MANAGEMENT
12          Within the EPA, Quality Management Plans (QMP) are developed to ensure that all
13    Agency materials meet a high standard for quality. NCEA-RTP participates in the Agency-wide
14    Quality Management System, which requires the development of a QMP. Implementation of the
15    NCEA QMP ensures that all data generated or used by NCEA scientists are "of the type and
16    quality needed and expected for their intended use" and that all information disseminated by
17    NCEA adheres to a high standard for quality including objectivity, utility, and integrity. Quality
18    assurance (QA) measures detailed in the QMP will be employed for the development of the PM
19    ISA. NCEA QA staff will be responsible for the review and approval of quality-related
20    documentation. NCEA scientists will be responsible for the evaluation of all inputs to the PM
21    ISA, including primary (new) and secondary (existing) data, to ensure their quality is appropriate
22    for their intended purpose. NCEA adheres to Data Quality Objectives, which identify the most
23    appropriate inputs to the science assessment and provide QA instruction for researchers citing
24    secondary information. The approaches utilized to search the literature and criteria applied to
25    select and evaluate studies were detailed in the two  preceding subsections. Generally, NCEA
26    scientists rely on scientific information found in peer-reviewed journal articles, books, and
27    government reports. The PM ISA also can include information that is integrated or summarized
28    from multiple sources to create new figures, tables,  or summation, which is subject to rigorous
29    quality assurance measures to ensure their accuracy.

30    3.6  SCIENTIFIC AND PUBLIC REVIEW
31          Drafts of the PM ISA will be made available for review by  the CAS AC and the public as
32    indicated in Figure 1-1 and Tables 1-1 and 1-3 above. Availability of draft documents will be
33    announced in the Federal Register. The CAS AC will review the draft PM ISA documents and

         April 2016                             3-29                Draft - Do Quote or Cite

-------
1   discuss its comments in public meetings that will be announced in the Federal Register. The
2   EPA will take into account comments, advice, and recommendations received from the CASAC
3   and from the public in revising the draft PM ISA documents. The EPA has established a public
4   docket for the development of the PM ISA.81 After appropriate revision based on comments
5   received from the CASAC and the public, the final document will be made available on the EPA
6   website. A notice announcing the availability of the final PM ISA will be published in the
7   Federal Register.
    81 The ISA docket can be accessed at www.regulations.gov using Docket ID number EPA-HQ-ORD-2014-0859.
        April 2016                            3-30               Draft - Do Quote or Cite

-------
 1   REFERENCES
 2   Clean Air Act Amendments of 1977. U.S. Government Printing Office; 1977.
 3   IPCC (2014). Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II,
 4         and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change.
 5         [Core Writing Team, R.K. Pachauri and L.A. Meyer (eds.)]. IPCC, Geneva, Switzerland.
 6   Loomis D, Groose Y, Lauby-Secretan B, El Ghissassi F, et al. (2013). The carcinogenicity of
 7         outdoor air pollution. Lancet Oncol 14(13): 1262-3.
 8   Melillo JM, Richmond TC, Yohe GW, Eds. (2014). Climate Change Impacts in the United
 9         States: The Third National Climate Assessment. U.S. Global Change  Research Program.
10         Available:
11         http://s3.amazonaws.com/nca2014/low/NCA3 Climate Change Impacts in the  United
12         %20States LowRes.pdf?download=l.
13   Peacock M (2008). Letter from Marcus Peacock, Deputy Administrator, U.S. EPA to Rogene
14         Henderson, Chair, Clean Air Scientific Advisory Committee, September 8, 2008.
15         Available: http://www3.epa.gov/ttn/naaqs/pdfs/PeacocklettertoCASAC090808.pdf
16   Samet J (2009). Letter from Dr. Jonathan M. Samet, Chair, Clean Air Scientific Advisory
17         Committee to the Honorable Lisa P. Jackson, Administrator, U.S. EPA.  CASAC Review
18         of EPA's Integrated Science Assessment for Particulate Matter - First External Review
19         Draft, December 2008. (May 21, 2009). Available:
20         http://ofmpub.epa.gov/eims/eimscomm.getfile?p_download_id=508493.
21   U.S. EPA (2006). Review of the Process for Setting National Ambient Air Quality Standards.
22          NAAQS Process Review Workgroup, U.S. Environmental Protection Agency, March
23          2006. Available:
24          http://www3.epa.gov/ttn/naaqs/pdfs/naaqs_process_report_march2006.pdf.
25   U.S. EPA (2009). Final Report:  Integrated Science Assessment for Paniculate  Matter. U.S.
26         Environmental Protection Agency, Washington, DC, EPA/600/R-08/139F. Available:
27         https://cfpub.epa.gov/ncea/risk/recordisplav.cfm?deid=216546&CFID=50518184&CFTO
28         KEN=64007426.
29   U.S. EPA (2015). Preamble to the Integrated Science Assessments. U.S. Environmental
30         Protection Agency, Washington, DC, EPA/600/R-15/067. Available:
31         https://cfpub.epa.gov/ncea/isa/recordisplay.cfm?deid=310244.
32   U.S. EPA (2016). Process of Reviewing the National Ambient Air Quality Standards. March 4,
33   2016. Available: http://www3.epa.gov/ttn/naaqs/review.html.
        April 2016                            3-31               Draft - Do Quote or Cite

-------
 i             4   HUMAN HEALTH RISK AND EXPOSURE
 2                                     ASSESSMENT

 3   4.1   INTRODUCTION
 4         Within the context of NAAQS reviews, a quantitative health risk and exposure assessment
 5   (HREA) is designed to estimate human exposure and health risks associated with the existing
 6   primary standards and with potential alternative primary standard(s), if any, that might be
 7   appropriate to consider. This assessment can inform conclusions on the adequacy of the public
 8   health protection provided by just meeting these standards. The purpose of this chapter is to
 9   highlight key findings from the quantitative risk assessments conducted in the last review of the
10   PM NAAQS and to identify key issues to be addressed in planning for any additional
11   quantitative assessments that might be warranted for the current review. The scope of any HREA
12   would be informed by the scientific evidence in the upcoming PM ISA; existing and historical
13   air quality patterns and trends; the availability of improved data, methods, tools, and models that
14   may better characterize important uncertainties or provide additional insights beyond those
15   provided by prior HREAs; and available resources.
16         In the upcoming HREA Planning Document (discussed in sections 1.2 and 1.5, above), the
17   EPA will evaluate newly available information within the context of the 2010 HREA from the
18   last review of the PM NAAQS to determine 1) the extent to which important uncertainties may
19   be better characterized by information newly available for the current review and 2) the extent to
20   which this new information may affect the health risks estimated in the 2010 HREA in important
21   ways or suggest new quantitative analyses that can increase our understanding of the health risks
22   associated with ambient PM exposures. The HREA Planning Document will also describe the
23   scope and methods for any new or updated quantitative assessments warranted for this review.
24   CASAC advice and public comments on this draft IRP will be considered in developing the
25   HREA Planning Document, which will also be subject to CASAC review and will be made
26   available for public comment. If warranted, one or more drafts of an HREA would then be
27   prepared and released for CASAC review and public comment prior to completion of a final
28   HREA.
29          Section 4.2 describes the key analyses, findings and uncertainties from the 2010 HREA.
30   Section 4.3 describes the two major components of potential new quantitative health risk
31   assessments (i.e., epidemiology-based risk assessment and exposure assessment) that the EPA
32   will  further evaluate in the HREA Planning Document. Section 4.4 describes the process for
33   obtaining scientific and public review of the HREA Planning Document and the HREA itself.

        April 2016                            4-1               Draft - Do Quote or Cite

-------
 1    4.2  OVERVIEW OF HEALTH RISK AND EXPOSURE ASSESSMENT
 2         FROM THE LAST REVIEW
 3          In the last review of the PM NAAQS, the EPA conducted a quantitative, epidemiology-
 4    based risk assessment for selected health endpoints to provide information and insights that
 5    could help inform decisions on the primary standards, namely, the degree of protection provided
 6    by the then-existing suite of primary standards, as summarized below (in sections 4.2.1 and
 7    4.2.2). The 2010 HREA (U.S. EPA, 2010a) characterized risk associated with ambient fine
 8    particle (PIVfo.s) concentrations because at that time it was decided that any risks estimated using
 9    the limited data available for thoracic coarse (i.e., PMio-2.s) and ultrafme particles would have
10    uncertainties too large to provide reasonable and informative results for the review. The 2010
11    HREA quantified endpoints from health effect categories classified as having a "causal" or
12    "likely to be causal" relationship with exposure to PM2.5 in the 2009 PM ISA (U.S. EPA, 2009a).
13    Endpoints included total, cardiopulmonary, and lung cancer mortality associated with long-term
14    (> 1 year) exposure to PM2.5, and mortality (total non-accidental, cardiovascular, and respiratory)
15    and morbidity (hospital admissions for cardiovascular and respiratory causes and respiratory
16    symptoms not requiring hospitalizations) associated with short-term (24-hour) PM2.5 exposures.
17    The 2010 HREA quantified health risks associated with ambient PM2.5 concentrations (as a
18    surrogate for population exposure) in fifteen urban study areas (U.S. EPA, 2010a, section 3.1).82
19    In addition, the 2010 HREA included a representativeness analysis that contrasted these urban
20    study areas against national-level patterns and trends for key PM related attributes (e.g., PM2.5
21    composition, demographics, and weather) and the total  number of deaths that could be attributed
22    to PM2.5 (U.S. EPA 2010a, sections 4.4.1 and 4.4.2). The 2010 HREA focused on estimating risk
23    remaining with PM2.5 concentrations adjusted to just meet the then-existing primary PM2.5
24    standards (15.0 jig/m3 annual and  35 jig/m3 24-hour, which are abbreviated here as 15/3583), as
25    well as risk reductions from just meeting alternative suites of standard levels (relative to 15/3 5).
26    The EPA did not conduct a quantitative population-based, microenvironmental exposure
27    assessment (such as described below in section 4.3.2) due to concerns related to the utility of
28    such an assessment from both staff and CASAC (Samet, 2009a).
      82 The 2010 HREA only estimated remaining risk and risk reductions for the urban areas where PMis improvements
      were needed to just meet the then-existing standards and alternative suites of standards. As described below, two of
      the 15 urban study areas did not exceed existing standards.
      83 In discussing the then-existing standards and alternative suites of PM2 5 standards, the notation used throughout
      this document is "annual standard/24-hour standard" in units of ug/m3.
         April 2016                              4-2                Draft - Do Quote or Cite

-------
 1           The health endpoints evaluated in the 2010 HREA were selected based on: 1) the overall
 2    weight of evidence and subsequently causal determinations in the 2009 PM ISA for general
 3    health effect categories that is based on the collective body of evidence from epidemiologic,
 4    controlled human exposure, and toxicological studies; 2) the extent to which particular health
 5    endpoints within these broader health effect categories are considered important from a public
 6    health perspective; 3) the availability of epidemiologic studies providing effect estimates for
 7    specific health endpoints; 4) the availability of air quality monitoring data in urban areas that
 8    were evaluated in the epidemiologic studies; 5) the availability of baseline incidence and
 9    prevalence data84 to support population risk (incidence) modeling; and 6) the anticipated
10    usefulness of developing quantitative risk estimates85 for the health endpoint(s) to inform
11    decision-making in the context of the PM NAAQS review. The 2010 HREA also considered
12    information in the 2009 PM ISA on at-risk populations, which identified the life stages  of
13    children and older adults, people with pre-existing cardiovascular and respiratory diseases, and
14    people with lower socioeconomic status as populations at increased risk for PM-related health
15    effects.
16           Criteria used in selecting the urban study areas included 1) the availability of air quality
17    monitoring data; 2) inclusion in an  epidemiologic study providing effect estimates; 3) the
18    availability of study area-specific baseline incidence and prevalence data; 4) the potential for risk
19    reductions by adjusting PM2.5 concentrations to just meet alternative annual and 24-hour standard
20    levels being considered, using the lowest of the alternative suite of standards as the cutoff
      84 The baseline incidence rate is the number of new cases of a health effect at a given time (often per 10,000 or
      100,000 of the general population in a year) in a given location due to all causes, including air pollution. These data
      are an integral part of the C-R functions used to determine the number of new cases that can be attributed to changes
      in PM2 5 concentrations. For some health endpoints (e.g., asthma exacerbation), we also use prevalence rates to
      define the applicable population with a given health condition (e.g., asthmatics). Prevalence refers to the rate of all
      cases in the population (both new and pre-existing).
      85 The term "risk estimate" as used here refers to a quantitative, model-derived estimate of the likelihood for adverse
      health effects within a defined population following exposure to a specific chemical agent or agents. In the context
      of NAAQS reviews, risk estimates typically take the form of incidence (count) estimates for specific morbidity or
      mortality endpoints for a defined population.

         April 2016                                4-3                 Draft - Do Quote or  Cite

-------
 1    (12/25); and 5) regional representation.86 After selecting the 15 urban study areas,87 the 2010
 2    HREA identified the "spatial template" to use in defining the geographical area associated with
 3    each urban study area, including which counties and PM2.5 monitors were associated with a
 4    particular urban  study area. A national-scale assessment evaluated the representativeness of the
 5    risk estimates for the urban study areas compared to national-level results. First, the national-
 6    scale assessment considered key PIVb.s risk-related attributes to determine whether the selected
 7    urban study areas are  nationally representative or more focused on a particular portion of the
 8    population distribution for a given attribute  (e.g., demographics)  (U.S. EPA, 2010a, section
 9    4.4.1). Second, the national-scale assessment analyzed estimates of mortality associated with
10    recent long-term PM2.5 concentrations,88 to assess the extent to which the 15 urban study areas
11    reflected locations within the U.S. likely to experience the highest PIVfo.s-related risk (U.S. EPA,
12    20lOa, section 4.4.2).
13           The 2010 HREA quantified risks using concentration-response (C-R) functions derived
14    from  effect estimates  reported in epidemiologic studies identified in the 2009 PM ISA. These
15    studies generally used ambient air quality data from fixed-site, population-oriented monitors;89
16    thus,  the appropriate application of these estimates  in C-R functions for a PM risk assessment
17    similarly required the use of ambient air quality data at fixed-site, population-oriented monitors.
18    The 2010 HREA adjusted ambient air quality  concentrations using several different methods,
19    including a proportional  "rollback" approach,90 to simulate the distribution of PM2.5 ambient
      86 The goal of the 2010 HREA was to select at least one urban study area from each of the seven geographic regions
      identified in the 1996 PM Criteria Document (U.S. EPA, 1996, section 6.4) (i.e., PM regions). The selected urban
      study areas represented six of these seven regions, with the Upper Midwest not represented.

      87 The 15 urban study areas were Tacoma, Fresno, Los Angeles, Phoenix, Salt Lake City, Dallas, Houston, St. Louis,
      Birmingham, Atlanta, Detroit, Pittsburgh, Baltimore, Philadelphia, and New York. Collectively, these 15 urban
      study areas comprised 31 counties.
      88 These impacts were estimated down to policy relevant background (PRB) for short-term exposure and the lowest
      measured level (LML) observed in the epidemiologic study for long-term exposure. PRB concentrations have
      historically been defined by the  EPA as those concentrations that would occur in the U.S. in the absence of
      anthropogenic emissions in continental North America, defined as the U.S., Canada, and Mexico. The 2010 HREA
      used regional PRB estimates generated using a combination of chemical transport modeling tools, as discussed in
      the 2009 PM ISA (section 3.2.2). LML refers to the lowest measured PM2s concentration within an epidemiologic
      study.
      89 Most studies applied the composite monitor approach, where PM2 5 levels were averaged across monitors within
      an urban study area in order to produce a single, generalizable PM2 5 distribution for that urban area.
      90 "Rollback" is a mathematical  approach to lower recent concentrations at ambient monitors in order to simulate
      just meeting the "design value"  for various standards. "Design values" are the metrics (i.e., statistics) that are
      compared to the NAAQS levels to determine compliance. The proportional rollback method, which had been used in
      previous risk assessments, reflected a uniform percentage of reduction in ambient PM2 5 concentrations across all
      monitors in an urban study area  (U.S EPA, 2010a, section 3.2.3).


          April 2016                                4-4                 Draft - Do Quote  or Cite

-------
 1    concentrations that would "just meet" the then-existing and alternative suites of standards. In
 2    addition, two alternative approaches (locally focused and a hybrid of local and proportional
 3    rollback approaches) were applied to improve our understanding of the uncertainty associated
 4    with the air quality adjustment (U.S. EPA, 2010a, section 3.2.3).91 The general PM health risk
 5    model combined PM2.5 concentrations in specific urban areas, C-R functions derived from
 6    epidemiologic studies, baseline incidence data for specific health  endpoints, and urban area
 7    population estimates to derive estimates of the annual incidence of specified health effects
 8    attributable to ambient PM concentrations under different air quality scenarios.  The human
 9    health risks were estimated using the risk assessment component of the EPA's Total Risk
10    Integrated Methodology model (TRIM.Risk).92 The analyses conducted in the 2010 HREA
11    focused on estimating changes in risks associated with air quality adjusted to just meet the then-
12    existing primary standards, as well as any additional risk reductions estimated to occur upon just
13    meeting alternative suites of standards.  In the short-term exposure analysis, the 2010 HREA
14    modeled risk down to PRB, while the long-term exposure analysis modeled down to the LML
15    observed in the epidemiologic study.

16    4.2.1  Key Observations in the 2010  HREA
17           Section 5.2 of the 2010 HREA discussed the key observations from the quantitative
18    analyses (U.S. EPA, 2010a, pp. 5-9 to 5-10). In summary, the 2010 HREA emphasized
19    cardiovascular-related health effects due to the greater degree of confidence in these endpoints
20    stemming from information provided in the 2009 PM ISA relative to other health effect
21    outcomes, including respiratory effects (U.S. EPA 2010a, p. 5-9). Thus, for long-term exposure-
      91 The locally focused rollback approach reflected a local pattern of reduction in ambient PM2 5 concentrations
      focused exclusively on those monitors within an urban study area that exceeded the 24-hour standard under
      consideration. As such, this approach was only applied to the subset of urban study areas where no adjustment was
      needed to meet the annual standard. The hybrid approach reflected a combination of a more localized pattern of
      rollback focused on source-oriented monitors with relatively elevated ambient PM2 5 levels, followed by a more
      generalized regional pattern of rollback across all monitors in the study area to just meet the standards. For one
      urban study area (Pittsburgh), the 2010 HREA also applied a refined rollback approach that used a dual-zone
      approach to take into account monitor locations and the related topography in that area (U.S. EPA, 2010a, section
      o ^> o \
      3.2.3)
      92 Additional information on the risk characterization module in TRIM can be found at:
      http://www2.epa.gov/fera/total-risk-integrated-methodologv-trim-trimrisk

          April 2016                              4-5                 Draft - Do Quote or Cite

-------
 1    related risk, the focus of the key observations was on ischemic heart disease (IHD)-related
 2    mortality, and for short-term exposures the focus was on cardiovascular-related mortality and
 3    morbidity. The 2010 HREA assessed risks associated with PM2.5 concentrations adjusted to just
 4    meet the then-existing standards and for PM2.5 levels meeting an alternative suite of annual and
 5    24-hour standards. The 2010 HREA also estimated the risks that remained after attaining
 6    alternative annual standard levels of between 10 to 14 jig/m3. This choice accounts both for
 7    epidemiologic evidence in the ISA and advice from CAS AC. The 2010 HREA recognized that
 8    the risks reported at the lower end of the  standard levels were subject to greater uncertainty. The
 9    2010 HREA evaluated five suites of annual and 24-hour standard levels to provide a range of
10    combinations across the urban study areas. These air quality analyses found that the annual
11    standard was controlling in some urban study areas,  and the 24-hour standard was controlling in
12    others.93 The analysis reported separate risk estimates for years 2005, 2006 and 2007. However,
13    the 2010 HREA focused on the 2007-based estimates, as these estimates for 2007 fell in the
14    middle of the risk estimates, in terms of the magnitude of risk (U.S. EPA, 2010a, p. 4-5).
15    Consequently, 2007-based risk estimates are reflected in the key observations below.
16
17    Magnitude of risk associated with air quality adjusted to just meet the then-existing PMi.s
18    standards (15/35)
19           In considering PIVfo.s-related risks associated with air quality just meeting the then-
20    existing PM2.5 standards in the 15 urban study areas, the 2010 HREA focused on the 13 urban
21    study areas that would not meet  15/35 based on air quality from 2005 to 2007.94 These 13 urban
22    study areas had annual and/or 24-hour design values that were above the levels of the then-
23    existing standards. Based  on the risk estimates for these areas, the 2010 HREA made the
24    following key  observations regarding the magnitude of risk remaining95 upon just meeting 15/35
25    (using proportional rollback):
      93 The controlling standard in a particular area is the one that, if met, ensures that the other standard would also be
      met in that area. Because the NAAQS for PM2 5 includes standards using both 24-hour and annual averages, the
      controlling standard in an area was determined by the annual or 24-hour average requiring the largest adjustment or
      "rollback" to meet the existing or potential alternative standard(s), if any. In general, only one of the two standards
      (daily or annual) would be the controlling standard in a particular area and 3-year period.
      94The two urban study areas that did not exceed either the then-existing annual and/or 24-hour design values were
      Dallas and Phoenix. Thus, the 2010 HREA did not estimate risk reductions in this analysis for these urban study
      areas. PM2 5 concentrations in these two areas were not "rolled up" to the then-existing standards.
      95 Risk remaining refers to the estimation of total risk from the then-existing standard down to the lowest bound for
      the analysis (either PRB or LML).
         April 2016                              4-6                 Draft - Do Quote or Cite

-------
 1    •  Long-term exposure-related mortality risk remaining: The analysis estimated that across the
 2       urban study areas, the IHD-related mortality attributable to long-term PIVfo.s exposure ranged
 3       from less than 100 to approximately 2,000 cases per year, with this variability reflecting, to a
 4       great extent, differences in the population size of each urban study area. These estimates
 5       represented from 4 to 17 percent of all IHD-related mortality in a given year across the urban
 6       study areas, which is a measure of risk that takes into account differences in population size
 7       and baseline mortality rates.

 8    •  Short-term exposure-related mortality and morbidity risk remaining: The analysis estimated
 9       that across the urban study areas, the cardiovascular-related mortality attributable to short-
10       term PM2.5 exposure ranged from <10 to 500 cases per year. These estimates represented
11       from approximately  1 to 2 percent of total  cardiovascular-related mortality in a given year
12       across the urban study areas. In terms of morbidity risk, cardiovascular-related hospital
13       admissions ranged from approximately 10 to 800 cases per year across the study areas,
14       representing less than one percent of total cardiovascular-related hospital  admissions in each
15       study area.
16    Magnitude of risk reductions associated with air quality adjusted to just  meet alternative
17    PMi.s standards
18           In characterizing PIVfo.s-related risks associated with air quality adjusted to just meet the
19    alternative annual standards (14/35, 13/35 and 12/35), the 2010 HREA estimated both the
20    magnitude of risk reductions (relative to risk remaining upon just meeting 15/3 5)96 and the
21    magnitude of risk associated with air quality adjusted to just meet the alternative standards.97
22    More uniform risk reductions were estimated to result from just meeting the alternative annual
23    standard levels than just meeting alternative 24-hour standard levels.98 Thus, in discussing these
24    risks, the 2010 HREA focused on the set of urban study areas that would have risk reductions by
25    just meeting each alternative annual standard.

26    •  Reductions in long-term exposure-related mortality risk: Upon just meeting the alternative
27       annual standard  levels considered in conjunction with the existing 24-hour standard (denoted
28       as 14/35, 13/35 and  12/35), the analysis estimated reductions in long-term exposure-related
29       mortality for 1299 of the 15 urban study areas, with the degree of risk reduction increasing
      96 Risk reductions refers to the difference in risk between the then-existing standard (in this case 15/35) and the
      alternative standard being analyzed.
      97 The 2010 HREA also estimated risks at 10 ug/m3 for the annual standard; however, the HREA noted the larger
      degree of uncertainty in these estimates and therefore, these results are not presented here (U.S. EPA, 2010a, section
      5.2.2).
      98 The degree of estimated risk reduction in long-term exposure-related IHD mortality risk provided by the
      alternative 24-hour standards was highly variable across the study areas, in part due to different rollback approaches
      (U.S. EPA, 2010a, section 4.2.2 and section 5.2.3).
      99 For 12/35, three of the urban study areas (i.e., Tacoma, Fresno and Salt Lake City) did not experience any
      decreases in risk when the annual standard was rolled back, as the design values in these urban study areas met the
      annual standard but exceeded the 24-hour standard, and thus would only experience risk reductions under alternative
      24-hour standards.

         April 2016                               4-7                  Draft - Do Quote or Cite

-------
 1       incrementally across the alternative standard levels (both in terms of the number of study
 2       areas experiencing risk reduction and the magnitude of those reductions). For the alternative
 3       annual standard level of 12.0 ug/m3 (in conjunction with the existing 24-hour standard), the
 4       analysis estimated that these urban study areas have reductions in mortality risk (relative to
 5       risk remaining upon just meeting then-existing suite of standards) ranging from about 11 to
 6       35 percent.

 7    •  Reductions in short-term exposure-related mortality and morbidity risk: For the alternative
 8       annual standard level of 12.0 |ig/m3 (in conjunction with the existing 24-hour standard), the
 9       analysis estimated that reductions in both short-term exposure-related cardiovascular
10       mortality and morbidity risk ranged from 5 to 23 percent across the urban study areas. 10°

11    4.2.2   Key Uncertainties in the 2010 HREA

12           The EPA recognized that both the inputs to the risk assessment and the risk model used
13    were each subject to uncertainties that affected the magnitude and distribution of the results. The
14    2010 HREA followed the World Health Organization (WHO, 2008) framework for
15    characterizing uncertainty in the analyses and incorporated elements intended to increase the
16    EPA's  overall confidence in the risks estimated in the fifteen urban study areas (U.S. EPA,
17    2010a,  section 3.5.1 and 4.3), including:

18    •  The use of a rigorous, transparent,  and fully documented process subject to peer and public
19       review in  developing all key elements of the 2010 HREA (U.S. EPA,  2010a, section 5.1.1);

20    •  Integration of key sources of variability into the design and interpretation of results;

21    •  Assessment of the degree to which the urban study areas included in the 2010 HREA are
22       representative of areas in the U.S. experiencing higher PIVfo.s-related risk; and

23    •  Identification and assessment of the impact (i.e., magnitude and direction) of important
24       sources of uncertainty on risk estimates.
25           The 2010 HREA included both qualitative (U.S. EPA, 2010a, section 3.5.3) and
26    quantitative sensitivity analyses (U.S. EPA, 2010a, section 3.5.4  and 4.3) designed to identify
27    sources of uncertainty most influencing the estimated risk. These analyses characterized the
28    sources of uncertainty likely to substantially impact the estimated risk. Since quantitative
29    information was not available to characterize overall levels of confidence in alternative model
      100 The patterns of risk reduction (as a percentage) were similar for both sets of endpoints because the rollback was
      the same for both and the C-R functions were assumed to be linear (U.S. EPA 2010a, section 4.2.2).
         April 2016                              4-8                Draft - Do Quote or Cite

-------
 1    inputs, the uncertainty characterization in the 2010 HREA did not include a full probabilistic
 2    assessment of uncertainty and its impact on risk estimates.
 O

 4    Key observations from analyses of uncertainty
 5           The qualitative analysis of uncertainty identified the following sources of uncertainty as
 6    potentially having a medium to high impact101 on risk estimates (see Table 3-6, p. 3-72 of the
 7    2010 HREA for additional detail, U.S. EPA, 2010a):

 8    •  Intra-urban variability in PM2.5 exposure in epidemiologic studies: Use of composite
 9       monitors as exposure surrogates in epidemiologic studies may not fully capture spatial
10       variability in PM concentrations within an urban area, and introduces exposure measurement
11       error and uncertainty into the effect estimates obtained from the epidemiologic studies.
12    •  Shape and statistical Jit of the C-R functions: There is uncertainty in the shape of the C-R
13       function, particularly at lower ranges of concentrations due to a lack of observations. In
14       addition, we consider the  statistical fit of the of the C-R function. The term "statistical fit" as
15       used here indicates the precision of a statistical model for capturing the observations. It can
16       be influenced by a variety of factors, including exposure measurement error, sample size, and
17       control for confounders.
18    •  Lag structure (short-term exposure studies): Different lags may have varying degrees of
19       association with a particular health endpoint, and it may be difficult to clearly identify the
20       specific lag that produces the majority of a PM-related effect. A lack of information
21       regarding the  specific lag(s) most associated with a particular health endpoint adds
22       uncertainty into risk estimates for that endpoint.
23    •  Transferability of C-R functions from epidemiologic study areas to HREA urban study areas
24       (long-term exposure health endpoints): The use of effects estimates based on data collected
25       in a particular location(s)  as part of the underlying epidemiologic study in different locations
26       (the focus of the risk assessment) introduces uncertainty into the analysis.

27    •  Single-city versus multi-city epidemiologic studies: Often both single-city and multi-city
28       epidemiologic studies are available (for a given health endpoint) for the derivation of C-R
29       functions. Each of these study types has advantages and disadvantages, and both study  types
30       can introduce uncertainty into the analysis.

31    •  Historical PM2.5 exposures: Long-term exposure studies of mortality suggest that different
32       time periods of PM exposure can produce substantially different effects estimates,
33       introducing uncertainty in identifying the exposure window that is most strongly associated
34       with mortality.
      101 The 2010 HREA classified sources as "medium impact" if they have the potential to change the interpretation of
      the risk estimates in the context of the PM NAAQS review, and classified sources as "high impact" if they are likely
      to influence the interpretation of risk.
         April 2016                               4-9                 Draft - Do Quote or Cite

-------
 1    •   PMcompositional differences: The composition of PM can differ across urban study areas. If
 2       these compositional differences contribute to different effect estimates (i.e., heterogeneity),
 3       then substantial uncertainty may be introduced into risk estimates if these compositional
 4       differences are not explicitly addressed.
 5
 6          The 2010 HREA included two types of quantitative sensitivity analyses: single-factor and
 7    multi-factor.  The single-factor sensitivity analysis varied one source of uncertainty at a time and
 8    identified the following sources of uncertainty as having a moderate to large impact102 on risk
 9    estimates (U.S. EPA, 2010a, section 4.3.1):
10    •   Long-term exposure-related mortality: (a) different C-R function model choices (e.g., fixed
11       versus random effects, log-linear versus log-log, single- versus co-pollutant), (b) modeling
12       risk down to PRB rather than LML, (c) use of C-R functions from different epidemiologic
13       studies, and (d) nature of the spatial pattern of ambient PM2.5 reductions (i.e., rollback
14       method).
15    •   Short-term exposure-related mortality and morbidity: (a) use of seasonally differentiated
16       versus annual-based C-R functions and (b) different models, lag  structures and single-versus
17       co-pollutant model forms.
18    The multi-factor sensitivity analysis varied multiple sources of uncertainty simultaneously and
19    covered both long-term exposure-related mortality and short-term exposure-related mortality and
20    morbidity. This multi-factor analysis found that a number of sources of uncertainty, identified
21    above, could work in concert to affect risk estimates, and that these combined effects would be
22    more than additive in certain circumstances (U.S. EPA, 2010a, section 4.3.1.2).
23
24    4.3  CONSIDERATION OF  POTENTIAL QUANTITATIVE
25         ASSESSMENTS FOR THIS REVIEW
26          The goal of an HREA is to provide information relevant to answering questions regarding
27    the adequacy of the existing PM primary standards and, if appropriate, the potential
28    improvements in public health from meeting potential alternative standards. A quantitative
29    HREA for this review, if warranted, would build on the approaches used and lessons learned in
30    the 2010 HREA and would focus on improving the characterization of PM exposure and
      102 The 2010 HREA classified factors as "moderate to large impact" if they had a 20 percent to greater than a 100
      percent difference on risk results.

         April 2016                            4-10                Draft - Do Quote or Cite

-------
 1    associated health risks, including related uncertainties, by incorporating a number of
 2    enhancements, in terms of newly available models, methods, tools, and data.
 3           Characterizing health risks for the current review of the primary PM NAAQS could
 4    include conducting air quality analyses to support quantitative assessments of risk and exposure
 5    in specific urban study areas as well as putting the results into a national public health
 6    perspective. We anticipate that, consistent with the large body of available scientific and air
 7    quality evidence available to support quantitative assessments, the most likely focus of
 8    quantitative analyses in an HREA would be PM2.5. These assessments would be designed to
 9    estimate human exposures and/or to characterize the potential  health risks that are associated
10    with recent ambient levels, with ambient PM2.5 concentrations adjusted to just meet the now-
11    existing standards (12/35) and, if appropriate, with ambient concentrations adjusted to just meet
12    alternative standard(s) that may be considered. To the extent that information becomes available
13    in the PM ISA for the current review that could support quantitative analyses of PM2.5
14    components, PMio-2.5, or ultrafme particles, additional quantitative analyses could be considered
15    which focus on these other size fractions or components. Similar to considerations for PM2.5 risk
16    analyses, we would evaluate whether sufficient information is available for other size fractions
17    and/or  components, including air quality information, endpoints  with causal or likely to be causal
18    determinations in the ISA, availability of concentration-response functions, and baseline
19    incidence data. The considerations for all potential quantitative epidemiology-based risk analyses
20    are described in greater detail in section 4.3.1 below.
21           As described further in this section, an HREA for the current review would consider a
22    variety of health endpoints for which, in staff s judgment, there would be adequate information
23    to develop quantitative risk estimates that can meaningfully inform the review of the primary PM
24    NAAQS. We also intend to evaluate the distribution of risks and patterns of risk reduction (e.g.,
25    among urban  case study areas and across ambient PM2.5 concentrations) and uncertainties in
26    those risk estimates. In presenting the results of any new assessments, we plan to evaluate the
27    influence of various inputs and assumptions on the exposure and risk estimates to more clearly
28    differentiate potential alternative standard(s) that might be considered, including potential
29    impacts on various at-risk populations and lifestages. In addition, we will consider characterizing
30    risk on a national scale. This assessment could also evaluate the  distribution of PM-related health
31    risk across the U.S. (and the factors that influence the distribution of risk) to assess the degree to
32    which risks estimated in each urban  study area are representative of national-level risks.
33           An important issue associated with conducting exposure and human health risk
34    assessments is the treatment of variability and the characterization of uncertainty.  Variability
35    refers to the inherent differences  in a population or variable of interest (e.g., residential air

         April 2016                              4-11                Draft - Do Quote or Cite

-------
 1    exchange rates); it cannot be reduced through further research, but it may be better characterized
 2    with additional measurement. However, models can account for variability. Uncertainty refers to
 3    the lack of knowledge regarding both the actual values of model input variables (i.e., parameter
 4    uncertainty) and the physical systems or relationships (i.e., model uncertainty - e.g., the shapes
 5    of concentration-response relationships). Uncharacterized spatial and temporal  variability in
 6    PM2.5 mass or species concentrations is often an important source of uncertainty in PM risk
 7    assessments. As part of such analyses, variability and uncertainty will be explicitly addressed,
 8    where feasible, in any new air quality, exposure, and health risk assessments.
 9           Consistent with the 2010 HREA, we are considering following the WHO framework
10    (WHO, 2008) to characterize uncertainty and variability, which presents a multi-tiered approach
11    for characterizing uncertainty (and to a lesser extent variability) in the context of a risk
12    assessment. This framework provides a method for systematically linking the characterization of
13    uncertainty to the level of complexity of the underlying risk assessment. Ultimately, the decision
14    as to which tier of uncertainty characterization to include in this risk assessment will depend on
15    the availability of information for characterizing the various sources of uncertainty. The four
16    WHO tiers range from application of default uncertainty factors up to probabilistic assessments.
17    Similar to the 2010 HREA, we anticipate that the uncertainty characterization in any HREA for
18    the current review would likely have sufficient data to conduct uncertainty assessments in the
19    middle tiers (e.g., qualitative assessments of direction/magnitude of the effect on results and
20    sensitivity analyses).
21           Below, we outline the two major components of the quantitative health  risk
22    characterization (i.e., epidemiology-based risk assessments and exposure assessment) that will be
23    described in more detail in the HREA Planning Document. Preparation  of this planning
24    document will draw from the assessments of the scientific evidence in the first  draft PM ISA to
25    facilitate the integration  of policy-relevant science into the planning document.  In particular, the
26    availability of air quality, exposure-response, concentration-response, and baseline incidence and
27    prevalence data as well as consideration of available resources will influence the type of risk and
28    exposure assessments that would be developed.

29    4.3.1  Epidemiology-Based Risk Assessment
30           The goals of an updated epidemiology-based risk assessment for PM would be to: (1)
31    estimate the potential magnitude of mortality and morbidity risks in urban study areas for recent
32    ambient concentrations as well as for PM concentrations adjusted to meet the existing standards
33    and, if appropriate, potential alternative standard(s); (2) develop a better understanding of how
34    various inputs and assumptions influence the risk estimates, including a characterization of the

         April 2016                             4-12                Draft - Do Quote or Cite

-------
 1    confidence in the risk estimates ; and (3) gain insights into the distribution of risks and patterns
 2    of risk reduction across modeled populations (including at-risk populations and lifestages) as
 3    well as the spatial distribution of risk within and across modeled study areas.
 4           The approach for an epidemiology-based risk assessment for the current review would
 5    build upon the methods developed for and insights gained from the 2010 HREA (section 4.2
 6    above). Section 4.3.1.1 below describes the ambient PM concentrations for recent conditions and
 7    adjustments to just meet the existing and potential alternative standard(s). Section 4.3.1.2
 8    discusses factors to be considered in designing any new epidemiology-based risk assessment.
 9    Section 4.3.1.3 identifies the sources of uncertainty that had a substantial impact on risk
10    estimates in the 2010 HREA and discusses potential strategies for reducing and characterizing
11    those sources of uncertainty in the current review. Each of these design factors and strategies for
12    reducing uncertainties, including data availability, would be further evaluated in the HREA
13    Planning Document.

14    4.3.1.1  Ambient PM Concentrations
15    Recent Ambient Conditions
16          For an HREA for the current review, we anticipate relying on data from the national air
17    quality monitoring networks described above (in section 2.3) to characterize ambient PM
18    concentrations for the time period and metric of interest (e.g., size fraction, form). As described
19    above (in section 4.2), the 2010 HREA used data from these monitoring networks to characterize
20    PM2.5 concentrations and associated health risks during 2005 to 2007 for 15 urban study areas. A
21    similar degree of PM2.5  monitoring is now available for more recent years that could be
22    considered in the current review. We would select specific urban study areas to assess in this
23    HREA by considering the available monitoring data in conjunction with the new epidemiologic
24    studies in the upcoming PM ISA, and this new information may suggest we consider the same
25    study areas in the upcoming HREA that were assessed in the 2010 HREA. Criteria for
26    identifying urban study  areas will be identified in the HREA Planning Document.
27          In the 2010 HREA, the air quality analyses focused on PM2.5. Due to the substantial
28    uncertainties and data gaps that were identified, the 2010 HREA did not include quantitative
29    analyses for additional PM size fractions (e.g., PMio-2.5 or ultrafine particles). For PMio-2.5, these
30    uncertainties were related to (1) concerns that the monitoring data for 2005 to 2007 would not be
31    similar to the data used in the epidemiologic studies that provided the effect estimates for the
32    C-R functions;  (2) uncertainty in the prediction of ambient concentrations under the then-existing
33    and alternative standards; (3) concerns that the 15 urban study areas in the 2010 HREA may not
34    be representative of areas experiencing the highest PMio-2.5 concentrations;  and (4) concerns

         April 2016                              4-13                Draft - Do Quote or  Cite

-------
 1    about the much smaller number of relevant epidemiologic studies for PMio-2.5 compared to the
 2    larger number of studies for PM2.5. The 2010 HREA also concluded that the available
 3    information was too limited to support quantitative assessments for ultrafine particles. A lack of
 4    information on the appropriate metric103 for characterizing ultrafine particles in risk assessments,
 5    as well as the lack of a national monitoring network for ultrafine particles, are likely to continue
 6    to challenge our ability to conduct a quantitative assessment for ultrafine particles in the current
 7    review. In the current review, we will consider whether new information on these areas of
 8    uncertainty has become available since the last review of the PM NAAQS that would be
 9    sufficient to support quantitative air quality assessments of recent ambient conditions for PMio-2.5
10    or ultrafine particles. If so, we will consider in the HREA Planning Document the degree to
11    which these ambient concentrations could be adequately characterized using available data from
12    the national monitoring networks (or other datasets in the case of ultrafine particles) to support
13    the HREA.
14          The 2010 HREA did not evaluate PIVb.s composition differences quantitatively because
15    C-R functions for specific PM2.5 components and sources had not been identified at that time.104
16    In the HREA Planning Document, we will consider whether new information on C-R functions
17    for components and/or sources  has become available in the PM ISA for the current review and
18    could necessitate the characterization of ambient PIVfo.s components and/or sources.  If warranted
19    for the HREA for the current review, we would consider the degree to which measurements from
20    the CSN and IMPROVE networks described in Chapter 2 could be used to adequately
21    characterize PM2.5 components or sources.

22    Approach for Adjusting PM Concentrations  to Just Meet the Existing and Potential
23    Alternative Standard(s)
24          In order to estimate changes in health risks from just meeting the existing PM2.5 standards,
25    and potential alternative standard(s) if appropriate, we would first need to adjust recent ambient
26    concentrations to reflect the distribution of PM2.5 concentrations in an urban area that could
27    correspond to these air quality scenarios.  Since there are multiple combinations of emissions
28    profiles that could result in just meeting the current standards or any potential alternative
29    standards, there are many possible scenarios of various spatially and temporally distributed PM2.5
      103 For example, total particle number, solid particle number, active particle surface area, Brunauer-Emmett-Teller
      surface area, or mass of particles in a certain size range.
      104 However, the 2010 HREA indirectly considered PM composition differences by applying region- or city-specific
      effect estimates in modeling health risks for short-term exposure to ambient PM2.5.
         April 2016                             4-14               Draft - Do Quote or Cite

-------
 1    concentrations that would correspond to the target air quality scenarios. Given this fact, the 2010
 2    HREA used multiple air quality adjustment approaches (i.e., proportional, local, and hybrid
 3    rollback) to estimate the effects on the distribution of PM2.5 concentrations at monitors in an
 4    urban study area of just meeting the then-existing and alternative standards. Use of multiple
 5    adjustment approaches enabled exploration of the sensitivity of risk results to different spatial
 6    patterns of PM2.5 reduction, with the goal of largely bounding potential risks.
 7          In the HREA Planning Document, we will consider whether it would be appropriate to
 8    use additional or different approaches to adjusting air quality concentrations including methods
 9    based on results from chemical transport modeling of national emission reduction scenarios.
10    These additional or different adjustment approaches would aim to improve the representation of
11    PM2.5 reduction patterns over an area. If estimates of PRB concentrations are needed for the air
12    quality adjustments in the HREA, we would consider information provided in the PM ISA for
13    the current review. In selecting methods for air quality adjustment, we would consider the ability
14    of the approaches  to represent the relationship between the annual and 24-hour NAAQS.
15           There is relatively limited information on PMio-2.5, specific PM components and sources,
16    or ultrafine particles compared  with PM2.5 to develop approaches for adjusting their
17    concentrations to meet the existing PM2.5 standards and potential alternative standard(s), if
18    appropriate. We will consider in the HREA Planning Document the ability of various adjustment
19    approaches, including those based on statistical and air quality  modeling relationships, to provide
20    reasonable estimates of adjusted concentrations for PMio-2.5, specific PM components and
21    sources, and ultrafine particles, if warranted.

22    4.3.1.2 Factors to  Consider in Designing any New Epidemiology-based Risk Assessment
23    •  PM size fractions, components, and sources: We  anticipate focusing any new quantitative
24       risk assessments primarily on PM2.5 but will consider, to the extent sufficient air quality and
25       epidemiologic data become available in the PM ISA for the current review, characterizing
26       risks associated with PMio-2.5 and ultrafine particles as well as risks associated with specific
27       PM components and/or sources.
28    •  Health endpoints and health input data: For any quantitative risk assessment, we would
29       follow a multi-step process for each health endpoint considered. First, we would identify
30       those health endpoints that the PM ISA for the current review determines have either a
31       "causal" or "likely to be causal" relationship with PM exposure. Next, we would use
32       information from the PM ISA and the epidemiologic studies to specify C-R functions for
33       those health endpoints, including details on model form and associated effect estimates. Last,
34       we would obtain the necessary baseline incidence and/or prevalence rate data for each
35       endpoint. These rates originate from a variety of sources and often reflect varying spatial
36       scales.
         April 2016                             4-15                Draft - Do Quote or Cite

-------
 1    •  Exposure metrics: Similar to previous PM HREAs, we anticipate modeling health endpoints
 2       associated with both short-term (24-hour average) and long-term (annual average) exposure
 3       depending on the epidemiologic study data available in the PM ISA for this review.
 4       Furthermore, for short-term exposure-related endpoints, we could consider whether to
 5       differentiate risk estimates through application of seasonally differentiated effect estimates
 6       paired with seasonal mortality rates, if available.

 7    •  Spatial scale of the analysis: Based on the epidemiologic studies available in the PM ISA for
 8       the current review and the available air quality monitoring data, we would anticipate
 9       analyzing a set of study areas in detail and applying effect estimates specific to each study
10       area. We would  also anticipate adjusting recent PM2.5 concentrations to just meet the existing
11       and, if appropriate, potential alternative standard(s), in each urban area, including
12       consideration of various adjustment approaches (e.g., proportional rollback). In addition, we
13       will consider whether a national-scale analysis of risk for recent ambient concentrations
14       (such as completed in the 2010 HREA) is useful for evaluating the representativeness of the
15       study areas. We will also consider whether to estimate risk reductions at the national-level
16       from just meeting potential alternative standard level(s), if any. However, adjusting air
17       quality can be more challenging at broader regional or national scales than at the urban study
18       area scale. We anticipate that any national-level estimates, if generated, would supplement
19       the higher confidence risks estimated for the study areas.

20    •  At-riskpopulations and lifestages: Based on information from the PM ISA, we would
21       consider whether data would be sufficient to generate more refined risk estimates for these
22       populations groups, including whether effect estimates and baseline incidence rates for these
23       at-risk populations and lifestages are available.

24    •  Risk metrics: The 2010 HREA quantified risks using several risk metrics, including (a)
25       PM2.s-related incidence (for both morbidity and mortality endpoints),  (b) PM2.s-related
26       incidence per 100,000 exposed individuals, representing a standardized risk metric that can
27       be readily compared across urban study areas, (c) percent of baseline incidence attributable to
28       PM2.5 exposure, and (d) percent reduction in PM-related risk from just meeting the then-
29       existing and alternative standard levels. We anticipate including similar risk metrics in any
30       new risk assessment, and we will  consider including some additional risk metrics based on
31       information from the PM ISA to more fully characterize PM-related mortality risk. For
32       example, we  may consider metrics that characterize person-level risk, such as life years
33       gained (from just meeting the existing and potential alternative standard(s), if any), life years
34       lost (from exposure to recent PM  concentrations), and/or estimation of population-level
35       changes in life expectancy.

36    •  Differentiating risk estimates by PM concentration: Based on information from the PM ISA,
37       we would consider whether it would be appropriate and informative to differentiate the
38       magnitude of risks or overall confidence in risk estimates for different concentration ranges.
39       Such an analysis could reflect, among other factors, reduced confidence in specifying C-R
40       functions at lower PM concentrations where less data exist. For such an analysis, we would
41       explore methods such as (a) binning risk estimates into categories of confidence based on
42       data available for C-R functions including differences in the size of standard errors at various


         April 2016                              4-16                Draft - Do Quote or Cite

-------
 1       PM concentrations, and (b) using other types of evidence (clinical and/or lexicological) to
 2       differentiate risk estimates semi-quantitatively within specified PM2.5 concentration ranges.
 3    •  BenMAP-CE: Consistent with the HREA conducted for the last ozone NAAQS review (U.S.
 4       EPA, 2014), we anticipate using the environmental Benefits Mapping and Analysis Program
 5       - Community Edition (BenMAP-CE) (U.S. EPA, 2015), the EPA's GIS-based computer
 6       program for the estimation of health impacts associated with changes in air pollution.105
 7       BenMAP-CE draws upon a database of population, baseline incidence and prevalence, and
 8       effect estimates to automate the calculation of health impacts.106 In addition, BenMAP-CE
 9       can utilize standard errors from the effect estimates to generate 95th percentile confidence
10       intervals around point estimates. These confidence intervals reflect the statistical fit
11       (precision) of the underlying epidemiologic models, assuming that the form of the model is
12       correct.
13    4.3.1.3 Characterization of Sources of Uncertainty and Consideration of Information Newly
14           Available in this Review
15           As described above, the 2010 HREA primarily used sensitivity analyses (including both
16    single and multi-factor) to evaluate the impact of uncertainty in specific input factors on the risk
17    estimates, while a probabilistic analysis (i.e., Monte Carlo) was used to cascade the impact of
18    standard  errors in the effect estimates from epidemiologic studies into 95th percentile confidence
19    intervals  around the risk estimates. Although Monte Carlo analyses could be used to reflect the
20    combined impact of multiple sources of uncertainty affecting the risk estimates, this type of
21    analysis would require information not previously available, including the specification of
22    defensible distributions of values for key inputs and any potential correlations between those
23    inputs. For an HREA in the current review, we anticipate relying on single and multi-factor
24    sensitivity analysis as the primary method for characterizing the combined impact of uncertainty
25    from multiple key inputs on the risk estimates, although we will consider additional tools should
26    they become available. Based on information from the PM ISA for the current review, we would
27    anticipate conducting additional  sensitivity analyses to better characterize uncertainties related to
28    the specification of C-R functions (e.g., model specification, treatment of copollutants, and
29    treatment of lag for short-term exposure endpoints).
      105 The BenMAP software and associated documentation are available for download at
      http://www2.epa.gov/benmap.
      106 As noted in section 4.2, the 2010 HREA used the TPJM model to estimate human health risks (U.S. EPA,
      2010a). For subsequent NAAQS epidemiology-based risk assessments (U.S. EPA, 2014), we utilize BenMAP-CE,
      which provides a more efficient and flexible platform for modeling PM-related health risk. For example, BenMAP-
      CE allows users to readily conduct sensitivity analyses exploring the impact of alternative modeling choices (e.g.,
      effect estimates, thresholds, age ranges for target populations, units of spatial aggregation) on risk estimates.
         April 2016                             4-17                 Draft - Do Quote or Cite

-------
1          In designing an HREA for the current review, our goal will be to reduce and/or better
2    characterize the impact of the sources of uncertainty identified in the 2010 HREA to the extent
3    possible given available information. In addition, we would address any newly identified sources
4    of uncertainty. In Table 4-1 below, we identify several important sources of uncertainty and
5    discuss our considerations for potentially addressing these sources of uncertainty. These
6    strategies would be further described in the HREA Planning Document.
        April 2016                            4-18                Draft - Do Quote or Cite

-------
1    Table 4-1. Areas of Uncertainty Associated with Epidemiology-Based Risk Estimates and
2    Potential Strategies for Addressing these Sources of Uncertainty
         Category of
         Uncertainty/
          Limitation
    Description of Uncertainty/
   Limitation in the Last Review
Consideration of Information
   Newly Available in this
           Review
      Transferability of
      C-R functions from
      epidemiologic study
      areas to HREA
      urban study areas
At times, the EPA applies effect
estimates to urban study areas that
were not included in the original
epidemiologic study providing those
effect estimates. This geographic
extrapolation can introduce
uncertainty into the risk estimates for
those urban study areas.
New epidemiologic studies in
the PM ISA and other available
data may provide additional
perspective on which urban
areas or regions are likely to
have similar population-level
responses to PM pollution and
which are likely to have
different responses. This
information may guide
decision-making regarding the
transferability of effect
estimates to broader
geographic areas.	
      Transferability of
      C-R functions from
      an earlier time
      period (in the
      epidemiologic
      study) to a more
      recent time period in
      the HREA
Depending on the degree to which
conditions associated with PM
exposure and risk (e.g., population
behavior and mobility, baseline
health effects incidence rates,
ambient urban PM profiles) changed
over time, uncertainty may be
associated with the temporal
extrapolation of effect estimates.
New epidemiologic studies in
the PM ISA may provide
additional perspective on the
degree to which effect
estimates vary (for the same
geographic location) over time
and hence the degree to which
uncertainty associated with this
type of temporal extrapolation
of effect estimates may
introduce uncertainty into the
risk results.
      Use of composite
      monitors as
      exposure surrogates
      in the epidemiologic
      studies
The 2010 HREA asserted that the
exposure surrogate used in
epidemiologic studies (i.e.., a single
composite monitor representing an
entire urban study area) could impact
the magnitude of the effect estimates.
New epidemiologic studies in
the PM ISA may characterize
the degree to which different
spatial scales (used in defining
exposure surrogates)  introduce
exposure measurement error,
depending on the endpoint,
location and pollutant being
evaluated.
      Spatial
      heterogeneity in
      effect estimates in
      the epidemiologic
      studies
In the 2010 HREA, effect estimates
for short-term mortality and
morbidity endpoints demonstrated
considerable spatial heterogeneity.
At that time, the EPA noted that this
New epidemiologic studies in
the PM ISA might provide
information regarding which
factors drive spatial
heterogeneity in effect	
        April 2016
                    4-19
   Draft - Do Quote or Cite

-------
     Category of
     Uncertainty/
     Limitation
    Description of Uncertainty/
   Limitation in the Last Review
Consideration of Information
   Newly Available in this
           Review
                      heterogeneity could reflect a number
                      of factors (e.g., differences in
                      behavior across/within cities, PM
                      compositional/source differences,
                      different degrees of exposure
                      measurement error).	
                                     estimates. This information
                                     could help us to differentiate
                                     between effect estimates (and
                                     hence the risk estimates) in
                                     terms of overall confidence.
 Shape and statistical
 fitoftheC-R
 functions107
There is uncertainty in the shape and
statistical fit of the C-R functions,
particularly at lower ranges of
concentrations. Although the 2010
HREA applied linear C-R functions,
these functions could have non-
linearities, thresholds or
concentrations with reduced
confidence in the effect estimate.
New epidemiologic studies in
the PM ISA could provide
additional information
regarding risk across the range
of PM concentrations or
possible non-linearities in the
C-R function, particularly in
lower PM concentrations
where there tends to be less
data. Greater data density at
lower concentrations could also
result in increased precision
(and hence statistical fit) within
that range.	
 Lag structure in
 short-term exposure
 epidemiologic
 studies
Different lags have varying degrees
of association with a particular health
endpoint, and it may be difficult to
clearly identify the specific lag that
produces the majority of a PM-
related effect. A lack of information
regarding the specific lag(s) most
associated with a particular health
endpoint adds uncertainty into risk
estimates for that endpoint. For
short-term mortality, the 2010 HREA
used an average of 0 and 1-day lags,
and other lags, where available, were
considered in sensitivity analyses.
New epidemiologic studies in
the PM ISA could provide
additional information
regarding the lag structure most
strongly associated with
specific health endpoints.
107 The term "statistical fit" as used here indicates the precision of a statistical model for capturing the observations.
It can be influenced by a variety of factors, including exposure measurement error, sample size, and control for
confounders.
   April 2016
                    4-20
    Draft - Do Quote or Cite

-------
    Category of
    Uncertainty/
    Limitation
    Description of Uncertainty/
   Limitation in the Last Review
Consideration of Information
   Newly Available in this
           Review
Single pollutant
models versus
copollutant models
Ozone and other ambient
copollutants may confound effect
estimates associated with PM.
New epidemiologic studies in
the PM ISA could provide
additional information
regarding potential
confounding by copollutants.
Single-city versus
multi-city
epidemiologic
studies
For long-term exposure mortality, the
2010 HREA employed effect
estimates from multi-city
epidemiologic studies. For short-term
exposure mortality, the 2010 HREA
used a combination of single and
multi-city epidemiologic studies,
acknowledging advantages and
disadvantages with both types of
studies.
New epidemiologic studies in
the PM ISA could provide
additional information
regarding whether either type
of study (or a combination of
both) would be preferred in
modeling specific endpoints.
New single-city epidemiologic
studies in the PM ISA could
add confidence to the
individual city risk results
estimated using multi-city
effect estimates.
Historical PM
exposure in long-
term exposure
studies
Long-term epidemiologic studies
suggest that different time periods of
PM exposure can produce
substantially different effects
estimates. This can introduce
uncertainty in identifying the
exposure window is most strongly
associated with an endpoint.
New epidemiologic studies in
the PM ISA could provide
additional information
regarding which exposure
windows are most strongly
associated with long-term
exposure endpoints,
recognizing potential
measurement error linked to
population mobility (within
and between cities) over time.
Differences in PM
composition or
sources
The composition of PM and overall
pollution mixture can differ across
urban study areas. If these
compositional differences contribute
to different effect estimates (i.e.,
heterogeneity), then substantial
uncertainty may be introduced into
the risk estimates if these
compositional differences are not
explicitly addressed.	
New epidemiologic,
toxicological and clinical
studies in the PM ISA could
provide additional information
regarding differences in PM
composition, including whether
there is  sufficient evidence to
model risk by applying
differential effect estimates.
  April 2016
                    4-21
   Draft - Do Quote or Cite

-------
 1    4.3.2  Exposure Assessment
 2          As noted above, neither the 2010 HREA nor any prior review of the PM NAAQS
 3    included a quantitative population-based, microenvironmental exposure assessment, largely
 4    because of the uncertainties surrounding the purpose of such an assessment and the insights it
 5    could provide in better understanding PM-related health risks beyond those provided using the
 6    epidemiology-based approach. In the current review, we will again consider whether to perform
 7    an exposure assessment. In doing so, we will consider the extent to which the results of such an
 8    assessment could be informative when interpreted in light of important uncertainties. Key factors
 9    that we will consider when determining whether to perform an exposure assessment include our
10    technical ability to generate estimates of PM exposure on temporal and spatial scales that would
11    be informative for the review, quantitative and qualitative characterizations of uncertainties in
12    exposure estimates, and the availability in this review of PM human exposure studies that
13    provide insight into the potential public health importance of particular PM exposures.
14          Unlike an epidemiology-based risk assessment that often uses composite ambient  monitor
15    concentrations to represent generalized exposure over an urban area (as described in section
16    4.3.1), an exposure assessment could estimate the magnitude, frequency, duration, and pattern of
17    exposures to PM concentrations, considering activity patterns and microenvironments. Such
18    quantitative assessments could range from a data simplified analysis up to a modeled population-
19    based, microenvironmental exposure assessment. A detailed exposure assessment could provide
20    insight into the populations with the highest PM exposures. In addition, if available human
21    exposure studies provide a basis for identifying ambient-relevant PM exposure concentrations
22    that result in effects of potential public health concern, an exposure assessment could also
23    include a comparison of PM exposure estimates to the exposure concentrations shown to cause
24    such effects (i.e., an  exposure benchmark analysis).

25          If a quantitative population-based, microenvironmental exposure assessment is warranted,
26    exposure calculations would rely largely on measurements of ambient PM concentrations
27    potentially supplemented with information from air quality models and databases of daily
28    time/location/activity patterns and personal attributes, as well as information on physical  factors
29    that influence microenvironmental concentrations, exposures, and internal doses. Important
30    exposure factors include information on atmospheric chemistry and components of PM, temporal
31    and spatial distributions of air quality data, time spent in particular microenvironments and at
32    varying exertion levels, and information on at-risk populations and lifestages. Such an
33    assessment, if warranted, would likely focus primarily on ambient-related exposures to PM2.5 but
34    could consider, to the extent sufficient information becomes available in the PM ISA, exposures
35    associated with other PM size fractions and/or particular PM components. If warranted, such an
         April 2016                              4-22                Draft - Do Quote or Cite

-------
 1    exposure assessment would build upon the information presented in the PM ISA, could employ a
 2    modeling approach recently used for other NAAQS reviews (Nitrogen Dioxide: U.S. EPA, 2008;
 3    Sulfur Dioxide: U.S. EPA, 2009b; Carbon Monoxide: U.S. EPA, 2010c; Ozone: U.S. EPA,
 4    2014), and would depend upon available data and resources.
 5          Section 4.3.2.1 below discusses factors to be considered in designing an exposure
 6    assessment. Section 4.3.2.2 presents challenges in conducting such an assessment and sources of
 7    uncertainty. These design factors and strategies for reducing uncertainties would be further
 8    evaluated in the HREA Planning Document. In the HREA Planning Document, we will consider
 9    whether these strategies, including information from the PM ISA for the current review and
10    available air quality monitoring data, would provide sufficient confidence to move forward with
11    an exposure assessment and the associated air quality analyses in the current review.

12    4.3.2.1 Factors to Consider in Designing an Exposure Assessment

13    •  Characterizing ambient PM concentrations. The air quality data used in exposure
14       assessments generally correspond to the finest spatial and temporal scales in ambient
15       concentrations that substantially influence important exposure metrics and
16       microenvironments of interest.108 Developing highly resolved PM2.5 concentrations from
17       existing monitoring and modeling data to support a population-based, microenvironmental
18       exposure would present a number of challenges, including adequately representing temporal
19       and spatial variability in concentrations at fine scales over urban areas with a variety of
20       microenvironments. The challenges would be greater for PM2.5 components and/or different
21       PM size fractions, because less monitoring is available, gradients may be sharper, and
22       modeling may be less reliable for PM2.5 components and different size fractions compared
23       withPM2.s.
24       •  Exposure model: If a population-based, microenvironmental exposure is warranted for
25       the current review, we would likely use the APEX model109 to estimate exposures. APEX is a
26       probabilistic human exposure model that simulates the movement of individuals through time
27       and space within a study area. Exposures are calculated in APEX by accounting for the
28       complete time series of exposure concentrations and doses as they occur in the simulated
29       individual (minute by minute), and then typically, they are aggregated to a metric of interest
30       at the individual level and summarized for the population simulated in the study area. APEX
31       can use any geographic frame of reference (e.g., census tract centroids,  1x1 km grid blocks)
      IDS por exampie the temporal and spatial scales used in recent APEX simulations for ozone (U.S. EPA, 2014) were
      hourly concentrations at the census tract level, due to observed health effects resulting from 6-8 hour ozone
      exposures in controlled human exposure studies and the availability of comprehensive population demographics.
      109 Additional information on APEX can be found at: http://www2.epa.gov/fera/human-exposure-modeling-air-
      pollutants-exposure-model.

         April 2016                             4-23                Draft - Do Quote or Cite

-------
 1       and time period (e.g., minute, hourly) to represent ambient concentrations for calculating
 2       exposures.

 3    •   Population exposures: An APEX-based exposure assessment could be designed to estimate
 4       population exposures to ambient PM in urban study areas that represent a variety of study
 5       groups of interest (e.g., at-risk populations). Such an assessment could evaluate the influence
 6       of factors that contribute to temporal/spatial patterns of PM concentrations (e.g., climatic
 7       conditions) and variability in activity patterns and microenvironmental  concentrations.
 8       Further, because APEX estimates both short-term and long-term exposures for each
 9       simulated individual, we could determine the importance of short-term  exposure variability
10       with respect to long-term cumulative exposure, which could improve the  understanding of
11       the relationship between short-term and long-term exposure health effects via their linked
12       exposure profiles.
13    •   Ambient concentrations and human activity patterns: An APEX-based exposure assessment
14       can use ambient concentrations with a range of time or spatial scales, and we could design an
15       exposure assessment to focus on the scale of other APEX model input data (e.g., population
16       demographics at the census tract level) and the exposure output of interest. The human
17       activity data used by APEX are drawn from the EPA's Consolidated Human Activity
18       Database (CHAD; McCurdy et al., 2000; U.S. EPA, 2002). The CHAD includes data from
19       several surveys sampling at city, state, and national levels, including diaries of their daily
20       activities, locations visited, activities performed,  time-of-day and their durations. We could
21       consider whether it would be informative to conduct a child-specific analysis that calculates
22       exposure for various lifestages utilizing activity patterns and exposure parameters specific to
23       these lifestages. However, we note that we are currently unaware of exposure parameters for
24       early lifestages (age < 5), and therefore such an analysis may not be feasible for this review.

25    •   Microenvironmental exposure: An APEX-based  exposure assessment estimates exposures
26       that individuals experience while in indoor, outdoor, and inside-vehicle microenvironments
27       using the ambient concentrations and  other input variables specific to microenvironments. In
28       addition to drawing on information from the PM  ISA for the current review to parameterize
29       these and other input variables, we could consider including specific microenvironments and
30       variables in other exposure model analyses (e.g.,  SHEDS-PM; Burke et al., 2001). Because
31       APEX has a flexible, probabilistic approach to estimating exposure and dose, we could
32       consider developing appropriate distributions representing variability and uncertainty for
33       various inputs (e.g., air exchange rates, decay rates, and physiological parameters).

34    •   Spatial and temporal scope of exposure assessment: The overall scope of an exposure
35       assessment, if warranted, could focus  on a subset of urban study areas selected for the
36       epidemiology-based assessment. We could evaluate exposures associated with the recent air
37       quality conditions and, if feasible and informative, evaluate exposures associated with just
38       meeting the existing and potential alternative standard(s), if any. We note that there are likely
39       to be additional challenges and uncertainties associated with characterizing ambient PM
40       concentrations at high spatial and temporal resolution for just meeting the existing and any
41       potential alternative standard(s).
42
         April 2016                             4-24               Draft - Do Quote or Cite

-------
 1   4.3.2.2 Characterization of Sources of Uncertainty and Consideration of Information Newly
 2          Available in this Review
 3          Conducting an exposure assessment for PM2.5 using APEX presents a number of
 4   challenges in obtaining the inputs for the model. Incomplete or uncertain model inputs result in
 5   uncertainties in the estimated exposures. The primary limitations in input data for APEX are
 6   listed in Table 4-2. Identification of available model inputs and data limitations is one of the first
 7   steps in determining whether an APEX-based exposure assessment is warranted for this review.
 8   These  considerations will be discussed, along with possible strategies for addressing the
 9   uncertainties and limitations, in the HREA Planning Document.

10   Table 4-2. Uncertainties and  Limitations to Input Data used by APEX to Estimate
11   Exposures
            Category of
           Uncertainty/
            Limitation
             Description of Uncertainty/ Limitation
       Spatial variability of
       ambient PM2.5
       concentrations
APEX requires spatially variable ambient concentrations in order to
capture the spatial variation of exposure across the study area, and
home/work differential exposures. The spatial resolution of other
APEX inputs would be at a census tract level, so ideally tract-to-tract
spatial variation of ambient PIVb.s would be available for input to
APEX.
      PM aerodynamic
      properties
The aerodynamic properties of particles affect indoor infiltration
rates and deposited PM in different lung regions. The breakdown
into size fractions and density of the ambient PM2.5 (which are input
to APEX) are likely to be fairly uncertain.	
      Near-road and inside-
      vehicle exposures
For some PM size fractions (e.g., PMio-2.5, ultrafme particles), the
near-road and in-vehicle microenvironments may be important
contributors to high ambient-related exposures. How these
concentrations might be characterized could have varying degrees of
uncertainty depending on the data available or the approach taken.
      Activity pattern
      representativeness
The extent to which the human activity database (CHAD) provides a
balanced representation of the population being modeled will vary
across simulated study areas. APEX constructs activity sequences to
account for the effects of population demographics and local climate,
though this adjustment procedure does not fully capture all of the
variability within an urban study area that exists in the people's
activities that actually reside in the study area. Also, the extent to
which a particular at-risk population or lifestage group is represented
by the complete range of CHAD diaries may be uncertain.	
      Approach for
      developing
      longitudinal diary
      profile	
One limitation of the CHAD activity pattern data used in our
exposure model is that, on average, study participants provided less
than two days of diary data. Thus, annual activity sequences of days
for each simulated individual in APEX are a combination of diary
         April 2016
                  4-25
Draft - Do Quote or Cite

-------
           Category of
           Uncertainty/
            Limitation
            Description of Uncertainty/ Limitation
                             days from a single or multiple similarly related subjects (e.g., age,
                             sex), generated using a statistical algorithm and appropriate linking
                             variables (e.g., daily time spent outdoors). Performance evaluations
                             of the APEX algorithms have been limited by the small number of
                             CHAD individuals having multiday diaries.	
      Indoor air exchange
      rates (AER)
Because people generally spend a significant amount of time
indoors, AER are important in estimating exposures. APEX uses
statistical distributions of AER that account for regional variability
and several temperature ranges and considers separately the presence
of mechanical and non-mechanical building ventilation, to the extent
possible. One limitation is that it does not account for spatial
differences that may exist in AER within an urban study area (e.g.,
based on differences in age of residence or building). In addition,
while the set of existing AER distributions were developed from
measurement studies conducted in specific geographic areas in the
U.S., they may not include all specific study areas that could be
selected for a population-based exposure assessment.	
 2   4.4   SCIENTIFIC AND PUBLIC REVIEW

 3         The HREA Planning Document will be distributed to the CASAC for review and
 4   provided to the public for review and comment. The document will be the subject of a review
 5   with the CASAC at a public meeting or teleconference that will be announced in the Federal
 6   Register. The EPA does not produce a final HREA Planning Document, but instead considers
 7   CASAC recommendations and public comments in the design and when conducting the
 8   quantitative assessments either in a new HREA or in updating or expanding the last assessment
 9   as part of the PA. In either case, staff would prepare at least one draft of the assessment for
10   CASAC review and public comment. CASAC would review the document and discuss it at a
11   public meeting that would be announced in the Federal Register. Based on past practice, the
12   EPA expects that CASAC would summarize key advice and recommendations for revision of the
13   assessment in a letter to the EPA Administrator. In revising any draft HREA, the EPA would
14   take into account any such recommendations and also consider comments received from the
15   public, both at the meeting itself and directly in writing. A final assessment would then be made
16   available on an EPA website, with its public availability announced in the Federal Register.
        April 2016
                 4-26
Draft - Do Quote or Cite

-------
 1   REFERENCES

 2   Abt Associates Inc (2008). Environmental Benefits Mapping and Analysis Program (Version
 3          3.0). Bethesda, MD. Prepared for Environmental Protection Agency, Office of Air
 4          Quality Planning and Standards, Air Benefits and Costs Group. Research Triangle Park,
 5          NC.
 6   Byun, D., and K.L. Schere (2006). Review of the Governing Equations, Computational
 7          Algorithms, and Other Components of the Models-3 Community Multiscale Air Quality
 8          (CMAQ) Modeling System. Applied Mechanics Reviews, 59, 51-11.
 9   Burke, J., Zufall, M., Ozkaynak, H. (2001). A population exposure model for particulate matter:
10          Case study results  for PM2.5 in Philadelphia, PA. J Expos Anal Environ Epidemiol. 11:
H          470-489.
12   McCurdy, T.; G. Glen; L.  Smith and Y. Lakkadi (2000). The National Exposure Research
13          Laboratory's consolidated human activity database. Journal Expos Anal Environ
14          Epidemiol. 10:566-578.
15   Samet (2009a). Letter from Dr. Jonathan M. Samet, Chair Clean Air Scientific Advisory
16          Committee, to Lisa P. Jackson, Administrator U.S. Environmental Protection Agency
17          regarding "Consultation on EPA's Particulate Matter National Ambient Air Quality
18          Standards: Scope and Methods Plan for Health Risk and Exposure Assessment." May,
19          21, 2009. (letter in docket #EPA-HQ-OAR-2007-0492).
20   U. S. EPA (2002). Consolidated Human Activities Database (CHAD)  Users Guide: Extracting
21          Human Activity Information from CHAD on the PC. Written for EPA National Exposure
22          Research Laboratory by ManTech Environmental Technologies. Modified in June 2002.
23          Available: https://www.epa.gov/sites/production/files/2015-
24          02/documents/chad_manual.pdf
25   U.S. EPA (2004). Air Quality Criteria for Particulate Matter. U.S. EPA, National Center for
26          Environmental Assessment, Office of Research and Development. Research Triangle
27          Park, NC, EPA/600/P-99/002aF and EPA/600/P-99/002bF. October 2004. Available:
28          http://www.epa.gov/ttn/naaqs/standards/pm/s_pm_cr_cd.html.
29   U.S. EPA (2009a). Integrated Science Assessment (ISA) for Particulate Matter (Final Report).
30          U.S. EPA, National Center for Environmental Assessment, Office of Research and
31          Development. Research Triangle Park, NC. December, 2009. EPA/600/R-08/139F.
32          Available at: http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=216546.
33   U.S. EPA (2011). Policy Assessment for the Review of the Particulate Matter National Ambient
34          Air Quality Standards. U.S. EPA, Office of Air Quality Planning and Standards.
35          Research Triangle Park, NC, EPA 452/R-l 1-003. April, 2011. Available at:
36          http://www.epa.gov/ttn/naaqs/standards/pm/data/20110419pmpafinal. pdf.
37   U. S. EPA (2008). Risk and Exposure  Assessment to Support the Review of the NO2 Primary
38          National Ambient Air Quality Standard. U.S. EPA,  Office of Air Quality Planning and
39          Standards. Research Triangle  Park, NC.  EPA 452/R-08-008a/b. November 2008.
40          Available: http://www.epa.gOv/ttn/naaqs/standards/nox/s nox cr rea.html.
         April 2016                            4-27               Draft - Do Quote or Cite

-------
 1   U. S. EPA (2009b). Risk and Exposure Assessment to Support the Review of the SO 2 Primary
 2          National Ambient Air Quality Standard. U.S. EPA, Office of Air Quality Planning and
 3          Standards. Research Triangle Park, NC. July 2009. EPA-452/R-09-007. Available:
 4          http://www.epa. gov/ttn/naaq s/standards/so2/data/200908 SO2REAFinalReport.pdf.
 5   U.S. EPA (201 Oa). Quantitative Risk Assessment for Paniculate Matter - Final Report. U.S.
 6          EPA, Office of Air Quality Planning and Standards. Research Triangle Park, NC. EPA-
 7          452/R- 10-005. June 2010. Available:
 8          http://www.epa.gov/ttn/naaqs/standards/pm/sjm 2007 risk.html
 9   U.S. EPA (201 Ob). Particulate Matter Urban-Focused Visibility Assessment - Final Report.
10          Office of Air Quality Planning and Standards. Research Triangle Park, NC. EPA-452/R-
11          10-004. June 2010. Available:
12          http://www.epa.gov/ttn/naaqs/standards/pm/sjm 2007 risk.html.
13   U.S. EPA (201 Oc). Quantitative Risk and Exposure Assessment for Carbon Monoxide -
14          Amended. U.S. EPA, Office of Air Quality Planning and Standards. Research Triangle
15          Park, NC EPA-452/R-10-009. July 2010. Available:
16          http://www.epa.gov/ttn/naaqs/standards/co/data/CO-REA-Amended-July2010.pdf.
17   U.S. EPA (2014). Health Risk and Exposure Assessment for Ozone, Final Report. U.S. EPA,
18          Office of Air Quality Planning and Standards. Research Triangle Park, NC. EPA-452/R-
19          14-004a. August 2014. Available:
20          http://www.epa.gov/ttn/naaqs/standards/ozone/data/20140829healthrea.pdf
21   U.S. EPA (2015). BenMAP-CE User's Manual Appendices. U.S. EPA, Office of Air Quality
22          Planning and Standards. Research Triangle Park, NC. Available:
23          http://www2.epa.gov/benmap/manual-and-appendices-benmap-ce
24      World Health Organization (2008). Part 1: Guidance Document on Characterizing and
25          Communicating Uncertainty in Exposure Assessment, Harmonization Project Document
26          No. 6. Published under joint sponsorship of the World Health Organization, the
27          International Labour Organization and the United Nations Environment Programme.
28          WHO Press, World Health Organization, 20 Avenue Appia, 1211 Geneva 27,
29          Switzerland (tel.: +41 22 791 2476).
         April 2016                            4-28               Draft - Do Quote or Cite

-------
 i          5  WELFARE-RELATED RISK AND EXPOSURE
 2                                      ASSESSMENTS

 3    5.1  INTRODUCTION
 4         Within the context of NAAQS reviews, a quantitative welfare risk and exposure
 5    assessment (WREA) is designed to estimate exposure and risks to public welfare110 associated
 6    with the existing secondary standards and with potential alternative secondary standard(s), if any,
 7    that might be appropriate to consider. This assessment can inform conclusions on the adequacy
 8    of the public welfare protection provided by these standards. The purpose of this chapter is to
 9    highlight key findings from the quantitative risk assessments conducted in the last review of the
10    PM NAAQS and to identify key issues to be addressed in planning for any additional
11    quantitative assessments that might be warranted for the current review. The scope of any
12    WREA would be informed by the scientific evidence in the upcoming PM ISA;  existing and
13    historical air quality patterns and trends; the availability of improved data, methods, tools, and
14    models that may better characterize  important uncertainties or provide additional insights beyond
15    those provided by previous assessments; and available resources.
16         In the last review of the PM NAAQS, the quantitative welfare assessments focused
17    primarily on urban visibility and were presented in the Urban-Focused Visibility Assessment
18    (2010 UFVA, U.S. EPA, 201 Ob) and the Policy Assessment (2011  PA,  U.S. EPA, 2011). The
19    EPA did not conduct quantitative assessments for other welfare-related effects from PM (e.g.,
20    ecological effects, climate effects, and materials effects).111
21         In the upcoming WREA Planning Document (discussed in sections 1.2 and 1.5, above),
22    the EPA will evaluate the newly available information within the context of the  2010 UFVA and
23    2011 PA analyses from the last review of the PM NAAQS to determine 1) the extent to which
24    important uncertainties may be better characterized by information newly available for the
25    current review and 2) the extent to which new information may affect the results of the
26    quantitative analyses from the last review in important ways or may suggest additional
      110 Welfare effects as defined in CAA section 302(h) [42 U.S.C. 7602(h)] include, but are not limited to, "effects on
      soils, water, crops, vegetation, man-made 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."
      111 The on-going NOX/SOX review includes consideration of the ecological effects of ecosystem loading of
      paniculate nitrogen and sulfur compounds. Thus, these endpoints are not discussed in this review.

         April 2016                             5-1                Draft - Do Quote or Cite

-------
 1    quantitative analyses that can improve our understanding of the welfare exposures and risks
 2    associated with PM, including consideration of welfare effects beyond visibility impairment. The
 3    WREA Planning Document will also include a preliminary determination of whether a distinct
 4    WREA document would be needed. CASAC advice and public comments on this draft IRP will
 5    be considered in developing the WREA Planning Document, which will also be subject to
 6    CASAC review and will be made available for public comment. If warranted, one or more drafts
 7    of the WREA would then be prepared and released for CASAC review and public comment prior
 8    to completion of a final WREA.
 9           Section 5.2 below describes the key analyses, findings and uncertainties from the last
10    review of the PM NAAQS. Section 5.3 describes the effect categories to be considered for
11    potential welfare assessments (i.e., visibility impairment, ecological effects, climate effects, and
12    materials effects) that the EPA will further evaluate in the WREA Planning Document. Section
13    5.4 describes the process for obtaining scientific and public review of the WREA Planning
14    Document and the WREA itself, if warranted.
15    5.2  SUMMARY OF WELFARE RISK AND EXPOSURE ASSESSMENT
16         FROM THE LAST REVIEW
17           In the last review of the PM NAAQS, as  summarized below, the EPA conducted
18    quantitative assessments of PM-related visibility impairment in urban areas, recognizing that a
19    secondary PM NAAQS developed with an urban focus would complement the EPA's Regional
20    Haze Program, which focuses on Class I areas.112 The purpose of these assessments was to
21    provide information and insights to help  inform decisions on the secondary standards. As noted
22    above, these quantitative analyses were provided in two documents for the last review of the PM
23    NAAQS: the 2010 UFVA and the 2011 PA. The 2010 UFVA combined information from urban-
24    focused public preference  studies, urban air quality data, and urban visibility protection
25    programs. The 2011 PA incorporated some additional air quality analyses and used more recent
26    air quality data. These assessments (described in section 5.2.1) included evaluations of different
27    indicators, levels, averaging times, and forms for their appropriateness in the context of
28    measuring urban visibility impairment and its impacts on public welfare.
      112 In 1977, Congress established as a national goal "the prevention of any future, and the remedying of any
      existing, impairment of visibility in mandatory Federal Class I areas which impairment results from manmade air
      pollution", CAA section 169(a)(l). Currently, 156 national parks and wilderness areas are designated as Class I
      areas. For more information regarding the EPA's Regional Haze Program, please see
      http://www3.epa.gov/visibilitv/program.html.
         April 2016                              5-2                Draft - Do Quote or Cite

-------
 1           As noted above (in section 5.1), in the last review of the PM NAAQS the EPA did not
 2    quantify welfare risks for effects other than visibility impairment. At the time of the last review,
 3    the Agency determined that any risks estimated using the limited data available for other PM-
 4    related welfare effects would have uncertainties too large to provide reasonable and informative
 5    results for the review. We discuss below (in section 5.2.2) the uncertainties identified in the last
 6    review of the PM NAAQS regarding additional PM-related welfare effects.

 7    5.2.1  Quantitative Assessments of Urban Visibility Impairment in the Last Review
 8    5.2.1.1 Reanalyses of Public Preference Survey Studies
 9           The 2010 UFVA included a detailed reanalysis of the available urban visibility
10    preference studies. These studies are surveys designed to elicit  information about the public's
11    preferences regarding visibility impairment by asking participants to rate the acceptability or
12    unacceptability of a haze-obscured scene from each urban area. Each of these studies
13    investigated the same question, "What level of visibility  degradation is acceptable?" The studies
14    available for the last review of the PM NAAQS were composed of three  completed western
15    urban visibility preference survey studies plus a pair of smaller focus studies designed to explore
16    and further develop urban visibility survey methods. The three  western studies included one in
17    Denver, Colorado (Ely et al., 1991),  one in the lower Fraser River valley near Vancouver, British
18    Columbia, Canada (Pryor, 1996), and one in Phoenix, Arizona  (BBC Research & Consulting,
19    2003). A pilot focus group with 9 participants was conducted in Washington, DC (Abt
20    Associates, Inc., 2001), and a replicate study with 26 participants was also conducted for
21    Washington, DC (Smith  and Howell, 2009). When taken together, these  studies included
22    acceptability ratings regarding visual air quality (VAQ)113 from a total of 852 individuals.
23           The approaches in the four studies were all derived from the method first developed for
24    the Denver urban visibility study. Although the approaches used in the four preference studies
25    were similar, the specific materials and methods used in  each study introduced uncertainties to
26    be considered when interpreting the results of the comparison of the studies. Key differences
      113 VAQ is defined as the visibility effect caused solely by air quality conditions and excluding those associated with
      meteorological conditions like fog and precipitation. It is commonly measured as either light extinction (in terms of
      inverse megameters, Mm"1) or the haziness index (in terms of deciview, dv) (Pitchford and Malm, 1993). The
      deciview scale was developed for use in visibility perception studies because it has a more linear relationship to
      perceived changes in haze compared with light extinction. A dv is defined as ten times the natural logarithmic of one
      tenth of the light extinction in Mm"1 (Pitchford and Malm, 1993).

         April 2016                               5-3                 Draft - Do Quote or Cite

-------
 1    between the studies include the following: 1) scene characteristics; 2) image presentation
 2    methods (e.g., projected slides of actual photos, projected images generated using WinHaze114 or
 3    use of a computer monitoring screen); 3) number of participants in each study; 4) participant
 4    representativeness of the general populations of the relevant metropolitan area; and 5) specific
 5    wording used to frame the questions used in the group interview process.
 6           Given the similarities in the approaches used in the preference studies, the EPA used
 7    regression analysis to reanalyze the results of the preferences studies individually and in
 8    combination. Specifically, the 2010 UFVA used a logit model to develop response curves that
 9    depicted the percentage of participants in each study that rated the VAQ (measured in dv) as
10    either "acceptable" or "unacceptable", for each of the visual images presented for a given area.
11    The 2010 UFVA adopted the "50% acceptability" criterion from the Denver preference study
12    (Ely et al., 1991), which identifies the inflection point along the response curve above which the
13    visibility impairment of the scenes was found unacceptable by at least 50% of the participants
14    and below which the level of VAQ in the images was found acceptable by at least 50% of
15    participants, as a useful index for comparing the results across studies.
16           These results were used to identify similarities in public preferences regarding VAQ and
17    to inform the range of candidate protection levels (CPLs)115 in the 2010 UFVA and the 2011 PA.
18    Specifically, the 2010 UFVA used the 50 percent acceptance criteria from each of the four urban
19    preference studies to identify the range of CPLs from 20 dv to 30 dv, with a midpoint of 25
20    dv.116 These values bracketed the daytime urban light extinction conditions that were judged
21    unacceptable by 50 percent of the participants in the preference studies  for those  urban scenes,
22    and provided low,  medium, and high cut-points for use in the remaining assessments (U.S. EPA,
23    201 Ob, pp. 2-28 to 2-29).
24    5.2.1.2 Analyses of Visibility Conditions
25           Building on prior reviews of the PM NAAQS,  in the last review the EPA conducted
26    quantitative analyses to further characterize ambient PM conditions in terms of PM
      114 The WinHaze model (Molenar et al., 1994) uses image processing technology to apply user-specified changes in
      light extinction values to the same base photograph with set scene and lighting characteristics.
      115 The term CPL refers to the target levels of visibility within a range that the EPA staff deemed appropriate for
      consideration that could, in conjunction with other elements of the standard, including indicator, averaging time, and
      form, potentially provide an appropriate degree of visibility protection.
      116 For comparison, 20 dv, 25 dv, and 30 dv are equivalent to 64, 112, and 191 Mm"1, respectively.

         April 2016                               5-4                 Draft - Do Quote or Cite

-------
 1    concentrations and light extinction in 15 urban study areas. Analyses of visibility conditions
 2    were conducted as a part of the 2010 UF VA and the 2011 PA. Analyses in the 2010 UFVA
 3    evaluated calculated PMio light extinction and were expanded in the 2011 PA to evaluate
 4    calculated PM2.5 light extinction.117
 5           The 2010 UFVA focused on calculated PMio light extinction as the indicator for the
 6    quantitative assessments based on the data collected from the IMPROVE network (as described
 7    in section 2.3  above) and the requirements of the EPA's Regional Haze Program (U.S. EPA,
 8    2009a, sections 9.2.3.2 and 9.23.4). These assessments used a combination of several datasets118
 9    as inputs into  the IMPROVE algorithm119 to estimate the daily maximum  1-hour PMio light
10    extinction using data from 2005 to 2007.  The goals of these air quality assessments were to:  1)
11    improve understanding of the extent, patterns, and causes of PM-related impairment of urban
12    visibility during daylight hours; 2) create the basis for projections of PM2.5 mass and PMio light
13    extinction under potential alternative standard scenarios; and 3)  examine the correlation between
14    PMio light extinction and potential alternative indicator(s) based on PM2.5 mass  concentration
15    (U.S. EPA, 2010b, chapters 3, 4, and Appendix D; U.S. EPA,  2011, Appendix F).
16           The 2011 PA contained additional air quality analyses  in order to estimate PM2.5 light
17    extinction. PM2.5 is the size fraction of PM responsible for the visibility impairment in most
18    urban areas, and the methods for estimating PM2.5 light extinction were simpler than the methods
19    used in the 2010 UFVA for estimating PMio light extinction. Benefits of moving to  a simpler
20    approach  included: 1) more transparency in the required calculations; 2) less intensive data
21    processing; 3) an increase in the number of monitoring sites that could meet the data
22    requirements of the approach without adding sampling equipment or additional  laboratory
      117 Analyses in the 2010 UFVA and 2011 PA used the same 15 urban study areas as the 2010 HREA: Tacoma,
      Fresno, Los Angeles, Phoenix, Salt Lake City, Dallas, Houston, St. Louis, Birmingham, Atlanta, Detroit, Pittsburgh,
      Baltimore, Philadelphia, and New York. Collectively, these 15 study areas comprised 31 counties.
      118 Several datasets were combined for the 2010 UFVA analysis: 1) continuous hourly averaged PM25 mass
      concentrations and relative humidity; 2) 24-hour average filter measurements for PM2 5 composition; 3) hourly
      daylight PM2 5 components generated using the Community Multi-scale Air Quality modeling system (CMAQ); and,
      4) estimated/measured hourly PMi0-2.5 concentrations (U.S. EPA, 2010b, section 3). The resulting estimates of
      hourly averaged PM2 5 component concentrations, PMi0-2.5 mass concentration, and the measured relative humidity
      were used in the IMPROVE algorithm to estimate hourly PMio light extinction to be calculated for daylight hours
      with relative humidity no greater than 90 percent (U.S. EPA, 2010b, section 3.2, 3.3).
      119 The IMPROVE algorithm is based on data generated at monitoring sites in the IMPROVE network to estimate
      light extinction as a function of PM concentrations and relative humidity levels. For more detailed information on
      the IMPROVE algorithm, see section 9.2.2.2 of the 2009 PM ISA (U.S. EPA, 2009a). As noted above (in section
      2.2.2), there is both an original and revised version of the IMPROVE algorithm (Pitchford et al. 2007). The revised
      version was developed to address observed biases in rural areas under certain light extinction conditions.
          April 2016                               5-5                 Draft - Do Quote or Cite

-------
 1    analysis; and 4) an increase in the number of days per year for which the calculation of PM light
 2    extinction could be conducted (U.S. EPA, 2011, Appendix F).
 3           The quantitative analyses of visibility conditions can be considered as two distinct
 4    analyses, covering recent conditions and adjustments to just meet the existing standards. The first
 5    analysis presented in both the 2010 UFVA and the 2011 PA characterized the recent visibility
 6    conditions in the 15 urban study areas. However, the 2010 UFVA focused on PMio light
 7    extinction using monitoring data from 2005 to 2007, whereas the 2011 PA focused on PM2.5 light
 8    extinction using monitoring data from 2007 to 2009. The estimates of the percentage of daily
 9    maximum hourly PM2.5 light extinction values exceeding the CPLs from the 2011 PA were
10    somewhat lower than for the PMio light extinction values from the 2010 UFVA. However, the
11    patterns of the estimates across the 15 urban study areas were similar. Specifically, 13 of the  15
12    urban study areas in the 2011 PA (with the exception of two Texas and non-California western
13    urban study areas) were estimated to have exceeded the high CPL (30 dv) from about 10 to 50
14    percent of the days based on the PM2.5 light extinction analysis, while all 15 urban study areas
15    were estimated to exceed the low CPL (20 dv) from 10 to 90 percent of the days (U.S. EPA,
16    2011, pp. 4-29 to 4-30).
17           The second analysis characterized visibility conditions adjusted to just meet the existing
18    secondary standards. This adjustment applied a proportional rollback approach that uniformly
19    reduced all PM2.5 components to just meet the target levels (with a lower bound on potential air
20    quality adjustments at PRB).  For PM2.5 light extinction in the 2011 PA, the analysis was
21    designed to assess the likelihood that PM-related visibility impairment would exceed CPLs upon
22    just meeting the suite of secondary PIVh.sNAAQS (15.0 |ig/m3 annual and 35 |ig/m3 24-hour
23    (98th percentile form)) (U.S. EPA, 2011, p. 4-33).12° The PM2.5 light extinction was estimated in
24    terms of both daily maximum 1-hour average values and multi-hour (i.e., 4-hour) average values
25    for daylight hours (U.S. EPA, 2011, p. 4-33).121 When adjusted to just meet the existing
26    standards, the daily maximum 1-hour average PM2.5 light extinction values in a majority of the
27    urban study areas were estimated to show improvement across the CPLs (U.S. EPA, 2011,
28    Figures 4-4 and 4-7, Tables 4-3 and 4-6). Similar patterns of visibility improvement were also
      120 The 2010 UFVA included additional analyses characterizing visibility conditions on PMio light extinction after
      adjustments to just meet an alternative suite of secondary PM2 5 standards (i.e., 12.0 ug/m3 annual and 25 ug/m3 24-
      hour) (U.S. EPA, 2010b).
      121 As mentioned in section 2.2.2, the EPA focused on sub-daily averaging times as the relevant exposure periods for
      segments of the viewing public due to the short-term nature of the perception of PM-related visibility impairment.
         April 2016                              5-6                 Draft - Do Quote or Cite

-------
 1    shown with daily maximum 4-hour average PM2.5 light extinction values (U.S. EPA, 2011,
 2    Figure 4-8, Appendix G, Figure G-3).

 3    5.2.1.3 Key Uncertainties in the 2010 UFVA and 2011 PA

 4    •  Visibility Preferences: In the last review of the PM NAAQS, a number of uncertainties were
 5       identified with respect to the public preference studies. There was substantial variability
 6       between studies in the range of visibility conditions that were judged by study participants to
 7       be "unacceptable" (U.S. EPA, 2010, Chapter 2). The relative importance of the degree of
 8       visibility impairment (e.g., light extinction level) versus the frequency and duration of
 9       visibility impairment in preference studies was unclear. Additionally, differences in the
10       strength of preference for visibility conditions across the study areas (e.g., why visibility
11       impairment that is acceptable in one area is not acceptable in another area) were not well
12       understood. Moreover,  the value of visibility in the preference studies is not distinctly
13       defined beyond acceptable or unacceptable,  and consideration of other metrics of valuing
14       visibility may be needed. A better understanding of the reasons for differences in preferences
15       across studies, differences in preference in different urban study areas, and alternative
16       measures of acceptable visibility impairments could influence the interpretation of the results
17       in defining additional or alternative range of CPLs for visibility impairment.
18    •  Calculation of Visibility Conditions: In the last review of the PM NAAQS, a number of
19       uncertainties were identified regarding the estimation of recent visibility conditions and
20       adjustments to just meet alternative scenarios.  Calculated light extinction estimates have
21       uncertainties compared to directly measured PM light extinction. A number of uncertainties
22       were identified with the inputs for the IMPROVE algorithm, including the use of measured
23       or estimated PM concentrations and  components and the influence of relative humidity in the
24       IMPROVE calculations. The methods used to temporally apportion 24-hour PM2.5
25       components to calculate hourly-averaged values relied on monthly-averaged PM2.5
26       component variations from chemical transport modeling. The mass balance method used in
27       this modeling to estimate organic carbon concentrations and the loss of nitrate was
28       considered to be a reasonable approach but is not likely to be precise. Moreover, the
29       IMPROVE algorithm was designed for use under the Regional Haze Program, which applies
30       in rural areas, and the algorithm has  not been validated for urban areas. Uncertainties remain
31       with respect to the timeframe (e.g., 1-hour, 4-hour, 24-hour) for evaluating visibility
32       conditions, particularly given that the public preference studies did not provide insight on the
33       frequency or duration of visibility impairment that would be acceptable. In addition, the
34       rollback approach uniformly reduced all non-PRB PM2.5 components, but emission control
35       programs in practice would not likely operate in this manner.

36    5.2.2  Consideration of Quantitative Assessments for Additional PM-related Welfare
37          Effects in the Last Review

38          Based on the determination in the 2009 PM ISA (U.S. EPA, 2009a) that there was a
39    "causal" relationship between PM  and effects on visibility, climate,  and materials, as well as a
40    "likely to be causal" determination with ecological effects, the 2009 Scope and Methods Plan
41    (U.S. EPA, 2009c) considered whether adequate evidence was available to conduct quantitative

         April 2016                             5-7                 Draft - Do  Quote or Cite

-------
 1    assessments for each of these categories of effects. This Plan also noted that the chemical
 2    components of PM (to a greater extent than PM mass or size fraction) largely determine the
 3    nature, degree, and direction of the non-visibility welfare effects. After a careful evaluation of
 4    these categories of effects, the EPA determined in the 2009 PM Scope and Methods Plan that
 5    data were not available to conduct quantitative assessments for non-visibility welfare effects in
 6    the last review, including the following:

 7    •   Ecological effects from deposition of paniculate organics and metals: Data to link PM
 8       toxicity to ecosystem function effects were available for only one tree species, and limited
 9       data on other species hindered additional analyses. Toxic effects of some PM components on
10       plant leaves and soil were not well-characterized and it was difficult to isolate these
11       endpoints from other environmental stressors. A lack of data on the site-specific composition
12       of PM and soil-associated biota prevented quantitative analysis of population shifts from
13       deposited PM. A lack of data on bioavailability of PM components and uncertainties in
14       cumulative exposure effects prevented quantitative analysis of PM metal toxicity to biota.
15       Data on PM components were only available for a few urban areas. Data were unavailable
16       regarding seasonal composition of near roadway PM and trophic transfer to  animals that
17       forage on roadsides.

18    •   Climate effects: Representation of aerosols in climate models was needed to more accurately
19       predict changes in climate.  Complex interactions of aerosols and linkages between clouds
20       and the overall climate system limited the feasibility of conducting a quantitative analysis.
21       Insufficient data on local and regional microclimate variations for many regions of the U.S.
22       and heterogeneity of cloud  formations limited the feasibility of conducting a quantitative
23       analysis.

24    •   Materials effects: New evidence was not sufficient to conduct a quantitative assessment of
25       materials damage or soiling from PM deposition onto surfaces.
26

27    5.3  CONSIDERATION OF POTENTIAL QUANTITATIVE

28         ASSESSMENTS FOR THIS REVIEW

29          The goals of a WREA in the current review would be to provide information relevant to
30    answering questions regarding  the adequacy of the existing PM secondary standards and, if
31    appropriate, the potential improvements in public welfare that could be achieved from meeting
32    potential alternative standard(s). Any quantitative WREA for the current review would build on
33    the approaches used and lessons learned in the last review and would focus on whether newly
34    available models, methods, tools, and data would substantially reduce the previously identified
35    uncertainties.

36          As described further in this section, any WREA for the current review would consider the
37    main categories of PM-related welfare effects (e.g., visibility effects, ecological effects, climate

         April 2016                             5-8                Draft - Do  Quote or Cite

-------
 1    effects, and materials) for which, in staff s judgment, there would be adequate information to
 2    develop quantitative risk estimates that can meaningfully inform the review of the secondary PM
 3    NAAQS. These considerations will be described in more detail in the WREA Planning
 4    Document. Preparation of the WREA Planning Document will draw from the assessments of the
 5    scientific evidence in the first draft PM ISA to facilitate the integration of policy-relevant science
 6    into the planning document. In particular, consideration of the availability of new data regarding
 7    each of these effects as well as consideration of available resources to conduct such assessments
 8    will influence whether we develop additional quantitative assessments in a new WREA. We
 9    anticipate that any quantitative analyses, if warranted, could focus on PM2.5, but we would also
10    consider the extent to which information becomes available in the PM ISA for the current review
11    that could support quantitative analyses of PM components, PMio-2.5, or ultrafme particles.

12    5.3.1  Visibility Effects
13          As described above (in  section 5.2), the EPA conducted extensive quantitative
14    assessments of the impact of visibility impairment on public welfare based on public preferences
15    for VAQ and visibility conditions in 15 urban study areas in the 2010 UFVA and 2011 PA. The
16    goals of any new or additional quantitative assessment for visibility would be to utilize newly
17    available information to address the key uncertainties identified in the last review regarding the
18    two key aspects of those analyses:  visibility preference studies and air quality analyses. We
19    discuss both of these below.

20          In the last review, the EPA heavily relied on surveys of public preferences regarding the
21    acceptability of various levels  of urban VAQ. Quantitative reanalysis of those survey results in
22    the 2010 UFVA informed the development of the range of visibility CPLs (i.e., 20 to 30 dv).
23    Based on information discussed during the 2015 PM NAAQS kick-off workshop (79 FR 71764),
24    limited new evidence related to visibility preferences in urban areas has been identified in the
25    available literature. The WREA Planning Document will evaluate any new visibility preference
26    studies identified in the PM ISA for the current review to determine whether additional
27    quantitative assessments would be warranted. To the extent that additional urban preference
28    studies are identified in the PM ISA, we would consider whether an updated quantitative analysis
29    that considered any new information provided by the new studies would substantially change the
30    range of CPLs identified in the last review. If no additional visibility  preference studies in urban
31    areas become  available in this  review, we anticipate that the upcoming PA would continue to
32    rely upon the quantitative analyses of the preference studies available in the last review. In
33    addition, the PA would consider any additional information beyond preference studies that may


         April 2016                             5-9                Draft - Do Quote or Cite

-------
 1    become available in the PM ISA relevant to assessing adversity to public welfare from urban
 2    visibility impairment.
 3          In addition, we would consider in the WREA Planning Document whether providing
 4    updated air quality assessments to include more recent PM monitoring data would substantially
 5    reduce the uncertainties identified in the last review of the PM NAAQS regarding the
 6    characterization of recent visibility conditions and adjustment approaches to just meet the
 7    existing standards. The determination whether to conduct additional quantitative assessments of
 8    visibility conditions in the current review could depend on (1) whether updated information is
 9    available in the PM ISA to warrant new air quality assessments, such as significantly updated
10    understanding of the variables that affect light extinction including relative humidity, PM species
11    composition, PMio-2.5 data, or the IMPROVE algorithm (Pitchford et al, 2007), and (2) whether
12    we determine that an updated reanalysis of the preference studies would be warranted and would
13    substantially change the range of the CPLs identified in the last review (20 to 30 dv). New
14    quantitative air quality assessments, if warranted, could also use more recent air quality
15    monitoring data, including data for PM concentrations and components to determine the extent to
16    which the existing secondary standards are adequate given recent ambient concentrations. For
17    example, such assessments could determine whether exceedances of the range of CPLs would be
18    estimated to occur under more recent ambient conditions and under the existing secondary
19    standards, if warranted.
20           In the WREA Planning Document, we will consider any new data that have become
21    available in the PM ISA for the current review on preference studies and estimating visibility
22    conditions in order to determine whether conducting additional quantitative analyses on visibility
23    effects would meaningfully inform the current review of the secondary PM NAAQS. The EPA
24    may determine that certain quantitative assessments are more appropriate in a PA rather than in a
25    WREA.

26    5.3.2  Ecological Effects
27           As noted above, the EPA did not include a quantitative ecological risk assessment for any
28    prior reviews of the PM NAAQS. For the current review, we are considering whether
29    information would now be sufficient to conduct a quantitative ecological risk assessment
30    associated with deposition of PM, primarily particulate organics and metals, including whether
31    information and data are now available to address the uncertainties identified in the last review
32    of the PM NAAQS that prohibited a quantitative ecological risk assessment. As noted above (in
33    section 5.2.2), these uncertainties included limited data to link deposition of PM organics and
34    metals to ecosystem functions and specific ecological receptors (e.g., soils, vegetation, and

         April 2016                             5-10                Draft - Do Quote or Cite

-------
 1    fauna) effects, and other site-specific data. In the WREA Planning Document, we would consider
 2    whether any new information in the PM ISA for the current review substantially reduces the
 3    uncertainties identified in the last review of the PM NAAQS. In particular, we would consider
 4    any new information that 1) quantitatively elucidates pathways of exposure, 2) quantitatively
 5    characterizes ecological effects122 from deposition of PM mass or PM components (e.g., specific
 6    organic and metal PM) on ecological receptors, and 3) provides information sufficient to
 7    quantify relationships between ambient PM, deposited PM, specific ecological effects, and, to
 8    the extent feasible, associated ecosystem services.
 9          The first step in potentially assessing ecological effects associated with deposition of
10    organic and metal PM would require the identification and quantification of the pathways of
11    exposure. Pathways  of PM exposure for ecological receptors can include direct deposition to the
12    receptor surface via  wet, dry or occult (e.g., fog) deposition, or transfer from one environmental
13    compartment (e.g., aquatic, terrestrial) or organism to another. The last review of the PM
14    NAAQS noted that the chemical constituents that make up PM (rather than PM mass) largely
15    determine the nature, degree, and direction of the ecological effects (U.S. EPA, 2009c, Appendix
16    A). The components that make up a given mass concentration of PM can vary substantially both
17    temporally and spatially, which would affect the nature and magnitude of resulting ecological
18    effects. Uncertainty  associated with heterogeneity in deposition of PM components, along with
19    significant data gaps, confounded efforts to evaluate PM-related ecological effects on ecosystem
20    function at the ecosystem, regional, watershed, or national scale in the last review of the PM
21    NAAQS (U.S. EPA, 2009c, Appendix A). In the WREA Planning Document, we would evaluate
22    whether sufficient data have become available to estimate the spatial heterogeneity in site-
23    specific PM deposition (across and within ecosystems), particularly of organic and metal
24    components, and then to quantitatively link this deposition to ecological effects in various types
25    of ecosystems.
26           In the last review of the PM NAAQS, vegetation was shown to be sensitive to heavy
27    metal deposition (U.S. EPA,  2009c, Appendix A). For example, PM dry  deposition to leaf
28    surfaces and the inner canopy is well documented (U.S. EPA, 2009a, section 3.3.4.3). Impacts on
29    vegetation from PM metal deposition include growth suppression, toxicity to root colonizing
30    microorganisms, impairment of root development and induction of phytochelatins (U.S. EPA,
      122 As described above, the on-going NOx/SOx review includes consideration of the ecological effects of ecosystem
      loading of paniculate nitrogen and sulfur compounds. Thus, these endpoints are not discussed in this review.

         April 2016                             5-11                Draft - Do Quote or Cite

-------
 1    2009a, section 9.4.5.3).123 However, the last review of the PM NAAQS also noted that toxic
 2    effects of some components of PM on plants are not well characterized, and it was difficult to
 3    isolate these endpoints from other environmental stressors (U.S. EPA, 2009c, Appendix A).
 4    Similarly, based on information discussed during the 2015 PM NAAQS kick-off workshop,
 5    limited new evidence has been identified related to PM effects on vegetation, but most of those
 6    plants were not native to the U.S. In addition, workshop participants also noted that new studies
 7    and models are available to quantify the removal of ambient PM by plants (e.g., air filtration by
 8    dry deposition on leaves), but this is a distinct issue from quantifying how PM deposition affects
 9    those plants (e.g., PM deposition on leaf surfaces that inhibits photosynthesis). In the WREA
10    Planning document, we would evaluate whether sufficient data and tools  are available to apply
11    environmental modeling techniques to PM concentration and deposition data to quantify the
12    responses in U.S. vegetation and the extent to which such information would provide additional
13    insight regarding how PM toxicity affects ecosystem functioning considering the inherent
14    uncertainties in such approaches.
15          Information from the last review of the PM NAAQS indicated a variety of potential
16    adverse effects from PM deposition on soils (U.S. EPA, 2009a, section 9.4.5.2). For example,
17    PM effects  on soil and soil-associated microfauna from heavy metals (such as zinc, copper, and
18    cadmium) have been  shown to be toxic to soil fungi and bacteria, which could have broader
19    implications at the ecosystem level because many plant species are dependent upon fungal  and
20    bacterial associations to obtain nutrients from the rhizosphere (U.S. EPA, 2009a). Shifts in soil
21    microflora populations can also affect nutrient and organic matter cycling and carbon utilization.
22    In addition, vegetation can take up heavy metals from the soil, which could lead to effects on
23    wildlife and humans who consume those plants. These effects were not quantified in the last
24    review of the PM NAAQS because  of a lack of sufficient data. In the WREA Planning
25    Document,  we could evaluate whether sufficient data now exist to  quantify the site-specific
26    composition of PM deposition and the resulting population shifts in soil-associated biota in
27    response to ecological stressors such as PM deposition.
28          The last review of the PM NAAQS indicated that fauna may also  be an ecological
29    receptor for deposition of PM organics and metals through ingestion, absorption, trophic transfer,
30    transfer between aquatic to terrestrial compartments, and bioaccumulation and biomagnification
31    of heavy metals across trophic levels (U.S. EPA, 2009a, section 9.4.5.6).  However the last
      123 Phytochelatins are intracellular metal-binding peptides that act as specific indicators of metal stress (U.S. EPA,
      2009a).

         April 2016                             5-12                Draft - Do Quote or Cite

-------
 1    review of the PM NAAQS noted that quantitative assessment of paniculate metal toxicity to
 2    biota was limited due to the heterogeneous composition of PM, lack of data on the bioavailability
 3    of PM components, and uncertainties in cumulative exposure effects (U.S. EPA, 2009c,
 4    Appendix A). In the WREA Planning Document, we could consider whether new data and
 5    information from the PM ISA in the current review would provide sufficient information on the
 6    effects of PM on aquatic and terrestrial invertebrates, amphibians, birds and mammals to conduct
 7    quantitative analyses of these ecological effects and associated ecosystem services.
 8           In the WREA Planning Document, we could consider whether sufficient data have
 9    become available in the PM ISA for the current review to conduct quantitative analyses on the
10    PM-related ecological risks to vegetation, soils, and fauna associated with PM deposition in light
11    of inherent uncertainties. Although we do not anticipate that  sufficient information would
12    become available to reduce these uncertainties, if it did, we would give consideration to whether
13    conducting such quantitative analyses of PM-related ecological risks and associated ecosystem
14    services would be practicable and would meaningfully inform this review. If uncertainties and
15    data gaps prevent the EPA from conducting a quantitative ecological risk assessment for this
16    review, we anticipate that the upcoming PA would rely on information from the PM ISA
17    regarding PM-related ecological effects.

18    5.3.3  Climate Effects
19           As noted above, the EPA did not include a quantitative climate risk assessment for any
20    prior reviews of the PM NAAQS. The last review of the PM  NAAQS noted that the addition of
21    anthropogenic aerosols124 to the atmosphere perturbs the Earth's energy balance and constitutes
22    an aerosol climate forcing (U.S. EPA, 2009c, Appendix A). The influence of this forcing on
23    various climate metrics across a wide range of scales is an active area of research, and studies
24    have identified both direct and indirect aerosol forcing pathways that can lead to effects on
25    climate (U.S. EPA, 2009a, section 9.3.1). Aerosol direct effects on climate can occur when
26    changes in ambient aerosol properties (e.g., concentration, composition, and size distribution)
27    alter atmospheric radiative budgets by modifying the scattering and absorption of radiation by
28    particles.  Aerosol indirect effects on climate can occur when changes in aerosol properties alter
29    radiative budgets by modifying cloud amount, lifetime, and microphysical and radiative
30    properties. Aerosol indirect effects include the effects on clouds resulting from the role of
      124 As noted above (in section 2.2.1), the term aerosol is used in this document when discussing suspended ambient
      particles in the context of climate impacts.

         April 2016                             5-13                Draft - Do Quote or Cite

-------
 1    particles in the formation of cloud droplets and ice. An aerosol semi-direct effect can occur when
 2    clouds are affected by changes in atmospheric properties (e.g., relative humidity and vertical
 3    temperature structure) resulting from the absorption of radiation by particles.
 4          Recent climate assessments characterize the current scientific understanding of aerosol
 5    effects on climate (Bond et al., 2013; IPCC, 2014; National Climate Assessment, 2014), and we
 6    anticipate that the evaluation of climate evidence in the PM ISA for the current review would be
 7    informed largely by findings of these major assessments. Improvements in measurements and
 8    modeling of atmospheric aerosol properties are continually occurring and will contribute to the
 9    understanding of aerosol effects on climate in this review. For example, remote sensing
10    measurements from satellite systems and surface-based networks (e.g., AERONET125) can
11    facilitate more advanced analysis of variables relevant to aerosol-climate interactions. We
12    anticipate that atmospheric models with improved algorithms will be available in this review to
13    better represent the interactions of atmospheric particles with radiation and clouds and improve
14    the understanding of the role of aerosols in global climate change.
15          However, our current understanding is that considerable uncertainty in the effects of
16    aerosols on climate still exists (Bond et al., 2013; IPCC, 2014). Large spatial and temporal
17    heterogeneities in direct and indirect aerosol climate forcings occur due to the variety of aerosol
18    sources, the intermittency of these sources, the short atmospheric lifetime of aerosols relative to
19    the major greenhouse gases, and the chemical and microphysical processing that occurs in the
20    atmosphere. These features lead to greater uncertainty in quantitative estimates of the effect of
21    aerosols on climate relative to that of the major greenhouse gases. This uncertainty is especially
22    large at the local and regional scales in the U.S. that would likely be most relevant to a
23    quantitative assessment of the potential effects of a national PM standard on climate in this
24    review. In the WREA Planning Document, we will consider the extent to which information in
25    the PM ISA for the current review substantially reduces this uncertainty and whether information
26    and tools would be sufficient to quantify the local and regional effects from aerosols in this
27    review.  Specifically, the WREA Planning Document will consider the extent to which
28    conducting a quantitative climate assessment would provide meaningful information for this
29    review beyond the information available in the PM ISA and the other major scientific
30    assessments and considering the inherent uncertainties in such an assessment. For example, we
31    would consider whether a quantitative assessment in the current review (1) could quantify how
32    the PM NAAQS alone would affect climate and (2) how changes in PM (via changes in the
      125 See http://aeronet. gsfc.nasa. gov/
         April 2016                             5-14                Draft - Do Quote or Cite

-------
 1    standards) would result in changes to climate. If uncertainties suggest that it would not be of
 2    value to the current review for the EPA to conduct a quantitative climate risk assessment for this
 3    review, we anticipate that the upcoming PA would rely on information from the PM ISA
 4    regarding PM-related climate effects.

 5    5.3.4  Materials Effects
 6          As noted above, the EPA did not include a quantitative risk assessment on materials for
 7    any prior reviews of the PM NAAQS due to a lack of data. PM-related materials effects are
 8    generally classified into two categories: materials damage and soiling.
 9          Materials damage associated with deposited PM, particularly sulfates and nitrates,126
10    include the corrosion of metals, degradation of painted surfaces, deterioration of building
11    materials, and weakening of material components. Because of their electrolytic, hygroscopic, and
12    acidic properties and their ability to sorb corrosive gases, particles contribute to materials
13    damage by adding to the effects of natural weathering processes. Deposited pollutants often
14    undergo chemical transformations and are commonly oxidized to acids, leading to materials
15    damage particularly from reactions between materials and NOx and SCh. Wet and dry deposition
16    contribute to PM accumulation and surface damage, although the presence of moisture can
17    accelerate some materials damage (e.g., corrosion of metals). Generally, SCh is more corrosive
18    than NOx, but mixtures of SCh, NOx, and other PM components and species can corrode certain
19    metals at a faster rate than individual pollutants alone. In the last review of the PM NAAQS,
20    sufficient evidence was not available to conduct a quantitative assessment of PM-attributable
21    materials damage (U.S. EPA, 2004, 2009a, b). While ambient particles play a role in the
22    corrosion of metals and in the weathering of materials, no quantitative relationships between
23    ambient particle concentrations and rates of damages had been established (U.S.  EPA, 2009a). If
24    this information becomes available in the PM ISA, we would consider whether this information
25    would be sufficient to support a quantitative assessment for PM-attributable materials damage.
26          Deposition of PM  onto surfaces, such as metal, paint, and stone, can lead to soiling.
27    Soiling is the result of PM accumulation on an object that alters the optical characteristics
28    (appearance). The presence of PM may alter light transmission or change the reflectivity of a
29    surface. These soiling effects can impact the aesthetic value of a structure or result in reversible
      126 In the case of materials effects, it is difficult to isolate the effects of gaseous and paniculate N and S wet
      deposition so both will be considered along with other PM-related deposition effects on materials in this review of
      the PM NAAQS.
         April 2016                              5-15                Draft - Do Quote or Cite

-------
 1    or irreversible damage to the surface. The presence of air pollution can increase the frequency
 2    and duration of cleaning and can enhance biodeterioration processes on the surface of materials.
 3    For example, deposition of carbonaceous components of PM can lead to the formation of black
 4    crusts on surfaces, and the buildup of microbial biofilms127 can discolor surfaces by trapping PM
 5    more efficiently (U.S. EPA, 2009a, p. 9-195). Additionally, the organic or nutrient content of
 6    deposited PM may enhance microbial growth on surfaces. However, in the last review of the PM
 7    NAAQS, sufficient evidence was not available to conduct a quantitative assessment of PM-
 8    attributable soiling effects (U.S. EPA, 2004, 2009a,b). While soiling associated with fine and
 9    coarse particles can result in increased cleaning frequency and repainting of surfaces, no
10    quantitative relationships between particle characteristics and the frequency of cleaning or
11    repainting had been established (U.S. EPA, 2011). Additionally, despite the limited new data in
12    the last review of the PM NAAQS on the role of microbial colonizers in biodeterioration
13    processes and contributions of black crust to soiling, these data were not sufficient for
14    quantitative analyses (U.S. EPA, 2011). Should new information about cleaning frequency and
15    repainting of surfaces and the role of microbial colonizers in soiling effects become available in
16    this review, we would consider whether this information would support a quantitative
17    assessment.
18          In the WREA Planning Document, we will consider whether sufficient data have become
19    available in the PM ISA for the current review to conduct quantitative analyses on either
20    category of PM-related materials effects. If data gaps and uncertainties prevent the EPA from
21    conducting a quantitative materials effects assessment for this review, we anticipate that the
22    upcoming PA would rely on information from the PM ISA regarding PM-related materials
23    effects.

24    5.3.5  Characterization of Sources of Uncertainty and Consideration of Information Newly
25          Available in this Review
26          Table 5-1 summarizes the potentially important uncertainties related to the quantification
27    of PM-related welfare effects where additional information, if available, could improve our
28    understanding of PM-related welfare risks and/or reduce uncertainties identified in the last
29    review of the PM NAAQS. To the extent that we would not be able to reduce these uncertainties
30    in the current review, the utility of conducting additional quantitative welfare assessments would
      127 Microbial biofilms, primarily composed of fungi, can stain exposed rock surfaces with yellow, orange, brown,
      gray, or black colors.
         April 2016                             5-16                Draft - Do Quote or Cite

-------
be limited. We will further discuss these issues, including whether additional information has
come available in the PM ISA for this review, in the WREA Planning Document.

Table 5-1 Areas of Uncertainty in the Welfare Assessments and Potential Strategies to
Address Them
 Major
 Uncertainty or
 Limitation
Uncertainty/Limitation Remaining
from Last Review
Consideration of Information
Newly Available in this Review
 Variability in
 "acceptable"
 VAQ levels
 across
 preference
 studies and
 urban areas
There was substantial variability in
the range of VAQ judged to be
"acceptable" by the preference study
participants across the urban areas
studied. Limited studies and data
were available to better inform our
understanding of the  factors leading
to this variability. In  addition,
differences in the strength of
preferences across urban areas were
not well understood.
If new preference studies are
available in the PM ISA and include
additional explanatory variables
regarding this issue, we would
consider whether incorporating these
studies into another quantitative
assessment of the preference studies
would be warranted for this review.
 Degree of
 visibility
 impairment
 versus
 frequency and
 duration
The relative importance of the
degree of visibility impairment (e.g.,
light extinction level) compared to
the frequency and duration of
visibility impairment in the
preference studies was unclear.
If new preference studies are
available in the PM ISA and include
additional explanatory variables
regarding this issue, we would
consider whether incorporating these
studies into another quantitative
assessment of the preference studies
would be warranted for this review.
 Recent
 visibility
 conditions
Calculated estimates of PM light
extinction have inherent
uncertainties. For example, the
IMPROVE algorithm uses measured
or estimated PM concentrations and
components, is influenced by
relative humidity, and had not been
validated for urban areas at the time
of the last review. Limited hourly
PMio-2.5 monitoring, continuous
PM2.5 speciation monitoring, and
direct measurement of PMio light
extinction contributed to
uncertainties in estimating the light
extinction. In addition, the time
frame for assessing visibility
conditions was uncertain because the
If a new quantitative analysis of
visibility conditions is warranted for
this review (i.e., based on the
availability of new preference
studies), we would evaluate whether
new or different methods are
available to reduce this uncertainty
in characterizing PM2.5 light
extinction for recent conditions.
   April 2016
                         5-17
        Draft - Do Quote or Cite

-------
Major
Uncertainty or
Limitation
Uncertainty/Limitation Remaining
from Last Review
Consideration of Information
Newly Available in this Review
                 preference studies did not provide
                 insights on frequency or duration.
Visibility
conditions
under existing
and alternative
standards
The proportional rollback approach
used to just meet the existing (and
alternative) standards uniformly
reduced all PM2.5 components, but
emission control programs in
practice would not likely operate in
this manner.
If a new quantitative analysis of
visibility conditions under
alternative scenarios is warranted for
this review, we would evaluate
whether new or different methods
are available to reduce this
uncertainty in estimating PM2.5 light
extinction upon just meeting the
existing standards and, if
appropriate, potential alternative
standards under evaluation.
Heterogeneity
and data gaps
in quantifying
ecological
effects
Ecological effects from PM
deposition on soils, flora, and fauna
vary greatly depending on the
specific PM component, the
ecosystem affected, and other site-
specific information. However, data
were limited on PM components
outside of urban areas (other than
near specific point sources) and the
associated ecosystem responses from
deposition of metals and organics.
If studies and data become available
in the PM ISA to sufficiently reduce
the uncertainties associated with
quantifying the ecological effects,
we would consider whether this new
information would be sufficient to
warrant a quantitative ecological
assessment for this review.
Aerosols in
climate models
and
local/regional
variability
Improved representation of aerosols
in climate models is essential to
more accurate predictions of the role
of PM in climate forcing. Most
climate model simulations available
did not consider local variations in
climate forcing due to emissions
sources and local meteorological
patterns.	
If the PM ISA provides an improved
understanding of the local and
regional effects of aerosols on
climate, we would evaluate whether
this new information would be
sufficient to warrant a quantitative
climate assessment for this review.
Materials
damage and
soiling effects
No quantitative relationships were
available between ambient particle
concentrations and rates of corrosion
of metals and in the weathering of
materials. In addition, no
quantitative relationships were
available between particles and the
frequency of cleaning or repainting
of surfaces. Data on the role of
If additional studies become
available in the PM ISA to
substantially reduce the uncertainty
in quantifying materials damage and
soiling effects, we would consider
whether this new information would
be sufficient to warrant a
quantitative materials assessment for
this review.
  April 2016
                         5-18
        Draft - Do Quote or Cite

-------
      Major
      Uncertainty or
      Limitation
Uncertainty/Limitation Remaining
from Last Review
Consideration of Information
Newly Available in this Review
                      microbial colonizers in
                      biodeterioration processes and
                      contributions of black crust to
                      soiling were not sufficient for
                      quantitative analyses.	
 2   5.4  SCIENTIFIC AND PUBLIC REVIEW
 3        The WREA Planning Document will be distributed to the CASAC for review and provided
 4   to the public for review and comment. The document will be the subject of a review with the
 5   CASAC at a public meeting or teleconference that will be announced in the Federal Register.
 6   The EPA does not produce a final WREA Planning Document but instead considers CASAC
 7   recommendations and public comments in the design and when conducting the quantitative
 8   assessments either in a new WREA or in updating or expanding the last assessment as part of the
 9   PA. In either case, staff would prepare at least one draft of the assessment for CASAC review
10   and public comment. CASAC would review the document and discuss it at a public meeting that
11   would be announced in the Federal Register. Based on past practice, the EPA expects that
12   CASAC would summarize key advice and recommendations for revision of the assessment in a
13   letter to the EPA Administrator. In revising any draft WREA, the EPA would take into account
14   any such recommendations and also consider comments received from the public, both at the
15   meeting itself and directly in writing. A final assessment would then be made available on an
16   EPA website, with its public availability announced in the Federal Register.
        April 2016
                       5-19
        Draft - Do Quote or Cite

-------
 1   REFERENCES

 2   Abt Associates, Inc. (2001). Assessing Public Opinions on Visibility Impairment due to Air
 3          Pollution: Summary Report. Prepared for EPA Office of Air Quality Planning and
 4          Standards; funded under EPA Contract No. 68-D-98-001. Abt Associates Inc., Bethesda,
 5          MD. January 2001. Available: http://www.epa.gov/ttncaaal/tl/reports/vis_rpt_fmal.pdf
 6   BBC Research & Consulting (2003). Phoenix Area Visibility Survey. Draft Report. Available:
 7          http://www.azdeq.gov/environ/air/download/vis_021903f.pdf.
 8   Bond, T. C., et al. (2013), Bounding the role of black carbon in the climate system: A scientific
 9          assessment, J. Geophys. Res. Atmos., 118, 5380-5552.
10   Ely, D.W., J.T. Leary, T.R. Stewart, and D.M. Ross (1991). The Establishment of the Denver
11          Visibility Standard. Presented at the 84th Annual Meeting & Exhibition of the Air and
12          Waste Management Association. June 16-21.
13   Intergovernmental Panel on Climate Change (IPCC).  (2013) Climate Change 2013: The Physical
14          Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the
15          Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
16          Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)].
17          Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA,
18          1535 pp, doi:10.1017/CBO9781107415324.
19   Molenar, J.V., W.C. Malm, and C.E. Johnson (1994). Visual air quality simulation techniques.
20          Atmospheric Environment 28(5): 1055-1063.
21   Pitchford, M. and W. Malm (1993). Development and applications of a standard visual index.
22          Atmospheric Environment 28(5): 1049-1054.
23   Pitchford M; Malm W; Schichtel B; Kumar N; Lowenthal D; Hand J (2007). Revised algorithm
24          for estimating light extinction from IMPROVE particle speciation data. J Air Waste
25          ManagAssoc 57: 1326-36.
26   Pryor, S.C. (1996). Assessing public perception of visibility for standard setting exercises.
27          Atmospheric Environment 30(15):2705-2716.
28   Smith, A.E. and S. Howell. (2009). An Assessment of the Robustness of Visual Air Quality
29          Preference Study Results. CRA International, Washington, DC. March 30. Prepared for
30          the Utility Air Regulatory Group. Submitted as supplemental material to presentation by
31          Anne Smith to the public meeting of the EPA Clean Air Science Advisory Council. April
32          2.
33   U.S. EPA (2004). Air Quality Criteria for Paniculate Matter. U.S. EPA, National Center for
34          Environmental Assessment, Office of Research and Development. Research Triangle
35          Park, NC, EPA/600/P-99/002aF and EPA/600/P-99/002bF. October 2004. Available:
36          http://www.epa.gov/ttn/naaqs/standards/pm/s_pm_cr_cd.html.
37   U.S. EPA (2009a). Integrated Science Assessment (ISA) for Paniculate Matter (Final Report).
38          U.S. EPA, National Center for Environmental Assessment, Office of Research and
39          Development. Research Triangle Park, NC. December, 2009. EPA/600/R-08/139F.
40          Available at: http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=216546.

         April 2016                            5-20                Draft - Do Quote or Cite

-------
 1   U.S. EPA (2011). Policy Assessment for the Review of the P articulate Matter National Ambient
 2          Air Quality Standards. U.S. EPA, Office of Air Quality Planning and Standards.
 3          Research Triangle Park, NC, EPA 452/R-l 1-003. April, 2011. Available at:
 4          http://www.epa.gov/ttn/naaqs/standards/pm/data/20110419pmpafmal. pdf.
 5   U. S. EPA (2009c). Paniculate Matter National Ambient Air Quality Standards: Scope and
 6          Methods Plan for Urban Visibility Impact Assessment. U.S. EPA, Office of Air Quality
 7          Planning  and Standards, Research Triangle Park, NC. EPA-452/P-09-001. February
 8          2009. Available:
 9          http://www3.epa.gov/ttn/naaqs/standards/pm/data/20090227PMNAAQSWelfare.pdf.
10   U.S. EPA (201 Ob). Paniculate Matter Urban-Focused Visibility Assessment - Final Report.
11          Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency,
12          Research Triangle Park, NC. EPA-452/R-10-004. June 2010. Available:
13          http://www.epa.gov/ttn/naaqs/standards/pm/s_pm_2007_risk.html.
14   U.S. Global  Change Research Program (2014). Climate Change Impacts in the United States:
15          The Third National Climate Assessment. May. Available:
16          http://nca2014.globalchange.gov/.
17
18
        April 2016                            5-21               Draft - Do Quote or Cite

-------
 i           6  POLICY ASSESSMENT AND RULEMAKING

 2          As outlined in section 1.2 above, the PA and rulemaking comprise the final phases of the
 3    review. These phases are described briefly in sections 6.1 (PA) and 6.2 (rulemaking) below.
 4    6.1  POLICY ASSESSMENT
 5          The PA will provide a transparent analysis and presentation of staff conclusions
 6    regarding the adequacy of the existing PM standards and the potential alternatives, if any, that
 7    are appropriate to consider in the current review. The PA will integrate and interpret the
 8    information from the ISA and, if available, REA(s) to frame policy options for consideration by
 9    the Administrator. The PA is also intended to facilitate CASAC's advice to the Agency, and
10    recommendations to the Administrator, on the adequacy of the existing standards and on
11    revisions that may be appropriate to consider, as provided for in the CAA. Staff conclusions will
12    be based on the assessment of the scientific evidence in the PM ISA; the results of exposure and
13    risk assessments in the REA(s), as available; and any additional staff evaluations and analyses
14    that are included in the PA.
15          The discussion in the PA will be framed by consideration of a series of policy-relevant
16    questions drawn from those presented  in sections 2.1.3 and 2.2.3 above. These will include
17    questions on the adequacy of the current PM standards and, as appropriate, on the elements of
18    potential alternative  standards (i.e., indicator, averaging time, form, level). The PA will identify
19    evidence-based and exposure-/risk-based approaches for addressing these policy-relevant
20    questions and for reaching public health and welfare policy judgments.
21          The PA will identify the range  of policy options that the staff concludes could be
22    supported by the available scientific evidence and the information from available quantitative
23    assessments. In so doing, the PA will describe the interpretations of this evidence and
24    information that could support various policy options, as appropriate, and that could be
25    considered by the Administrator in making decisions on the PM NAAQS. This will include the
26    identification of key uncertainties and  limitations in the underlying scientific evidence and in the
27    information available from quantitative assessments.128
28          In identifying ranges of primary and secondary standard options for consideration, the PA
29    will recognize that the Administrator's final decisions will reflect public health and public
      128 In addition to presenting staff conclusions on the NAAQS, the PA will also highlight areas for future health- and
      welfare-related research, model development, and data collection.
               April 2016                       6-1            Draft - Do Not Quote or Cite

-------
 1    welfare policy judgments. It will further recognize that the Administrator's final decisions will
 2    draw upon scientific information and analyses of health or welfare effects and risks, as well as
 3    judgments about how to deal with the range of uncertainties that are inherent in the scientific
 4    evidence and analyses.
 5          Our approach in the PA to informing the Administrator's judgments on the primary PM
 6    standards will recognize that the available health effects evidence reflects a continuum consisting
 7    of PM exposure concentrations at which scientists generally agree that health effects are likely to
 8    occur, through lower exposure concentrations at which the likelihood and magnitude of the
 9    response become increasingly uncertain. This approach is consistent with the requirements of the
10    NAAQS provisions of the CAA and with how the EPA and the courts have historically
11    interpreted the CAA. These provisions require the Administrator to establish primary standards
12    that are requisite to protect public health and that are neither more nor less stringent than
13    necessary for this purpose. As discussed in section 1.1 above, the  provisions do not require that
14    primary  standards be set at a zero-risk level, but rather at a level that avoids unacceptable risks to
15    public health,  including the health of at-risk populations.
16          Similarly, our approach in the PA to informing the Administrator's judgments on the
17    secondary PM standards will recognize that a final decision must  draw upon scientific  evidence
18    and analyses about effects on public welfare, as well as judgments about how to deal with the
19    range of uncertainties that are inherent in the relevant information. As is the case for the primary
20    standards discussed above, this approach is consistent with the requirements of the NAAQS
21    provisions of the CAA and with how the EPA and the courts have historically interpreted the
22    CAA. These provisions require the Administrator to establish secondary standards that are
23    requisite to protect public welfare from any known or anticipated  adverse effects associated with
24    the presence of the pollutant in the ambient air. In  so doing, the Administrator seeks to establish
25    standards that are neither more nor less stringent than necessary for this purpose. The provisions
26    do not require that secondary standards be set to eliminate all welfare effects, but rather at a level
27    that protects public welfare from those effects that are judged to be adverse.
28          Staff will prepare at least one draft of the PA for CAS AC review and public comment.
29    CASAC will review the draft PA at a public meeting that will be announced in the Federal
30    Register. Based on past practice by CASAC, the EPA expects that CASAC will  summarize key
31    advice and recommendations for revision of the document in a letter to the EPA Administrator.
32    In revising the draft PA, we will take into account such recommendations, comments received
33    from CASAC and from the public at the meeting itself, and any written comments received.  The
34    final document will be made available on the EPA website, with its public availability
3 5    announced in the Federal Register.

               April 2016                       6-2             Draft - Do Not Quote  or Cite

-------
 1    6.2  RULEMAKING
 2          Following the issuance of the final PA and the EPA management consideration of staff
 3    analyses and conclusions presented therein, and taking into consideration CASAC advice and
 4    recommendations, the Agency will develop a notice of proposed rulemaking. The notice of
 5    proposed rulemaking will convey the Administrator's proposed conclusions regarding the
 6    adequacy of the current standard(s) and any revision that may be appropriate. A draft notice of
 7    proposed rulemaking will be submitted to the Office of Management and Budget (OMB) for
 8    interagency review, in which OMB and other federal agencies are provided the opportunity for
 9    review and comment. After the completion of interagency review, the EPA will publish the
10    notice in the Federal Register seeking comment on the proposed agency  action - namely
11    whether or not to revise the current standard, and if so, how. Monitoring  rule changes associated
12    with review of the PM standards will be developed and proposed, as appropriate, in conjunction
13    with the NAAQS rulemaking.
14          At the time of publication of the notice of proposed rulemaking, all materials on which the
15    proposal is based will be made available in the public docket of the rulemaking.129 Publication of
16    the proposal notice is followed by a public comment period, generally lasting 60 to 90 days,
17    during which the public is invited to submit comments on the proposal to the rulemaking docket.
18          Taking into account comments received on the proposed rule, the  Agency will then
19    develop a notice of final rulemaking, which again undergoes OMB-coordinated interagency
20    review prior to issuance by the EPA of the final rule. At  the time of the final rulemaking, the
21    Agency responds to all significant comments on the proposed rule.130 Publication of the final rule
22    in the Federal Register completes the rulemaking process.
23
      129 The rulemaking docket for the current PM NAAQS review is identified as the EPA-HQ-OAR-2015-0072. This
      docket has incorporated the ISA docket (EPA-HQ-OPJ)-2014-0859) by reference. Both dockets are publicly
      accessible at www.regulations.gov.
      130 For example, Agency responses to all significant comments on the 2009 notice of proposed rulemaking in the last
      review were provided in the preamble to the final rule and in a document titled Responses to Significant Comments
      on the 2012 Proposed Rule on the National Ambient Air Quality Standards for Particulate Matter (June 29,  2012;
      77FR 28890) (U.S. EPA, 2012).
               April 2016                        6-3             Draft - Do Not Quote or Cite

-------
REFERENCES

U.S. EPA (2012). Responses to Significant Comments on the 2012 Proposed Rule on the
      National Ambient Air Quality Standards for Particulate Matter. U.S. Environmental
      Protection Agency, Research Triangle Park, NC. Available at:
      http://www.epa.gov/ttn/naaq s/standards/pm/data/20121214rtc. pdf
         April 2016                      6-4           Draft - Do Not Quote or Cite

-------
United States                             Office of Air Quality Planning and Standards             Publication No. EP A-452/D-16-001
Environmental Protection                  Health and Environmental Impacts Division                                    April 2016
Agency                                          Research Triangle Park, NC

-------