UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                                    WASHINGTON, D C  20460
                                        MAY  25  2010
                                                                     OFFICE OF THE ADMINISTRATOR
                                                                       SCIENCE ADVISORY BOARD
SUBJECT:  Transmitted of Science Advisory Board Report
FROM:     Vanessa T. Vu
            Director, Science Advisory Board Staff Office (HOOF)

TO:         Karen Sheffer
            EPA Headquarters Library Repository (3404T)
       This is to advise you that the Science Advisory Board, Clean Air Scientific Advisory
Committee, Particulate Matter Review Panel, issued a report numbered EPA-CASAC-10-011,
CAS AC Review of Policy Assessment for the Review of the PMNAAQS - First External
Review Draft, (March 2010), dated May 17,2010.

       Two copies of the report are attached and a third copy has been sent electronically to
the attention of Ms. Jeannie Turner at tumer.ieannie(S).epa.gov.  The report is available in
electronic format on the Science Advisory Board's Web site at http://www.epa.gov/sab.

       If you have any questions regarding this report, please contact the Designated Federal
Officer, Dr. Holly Stallworth directly at (202) 343-9867.
Attachments (2)
                                   Internet Address (URL) • http //www epa gov
            Recycled/Recyclable • Printed with Vegetable Oil Based Inks on 100% Postconsumer, Process Chlorine Free Recycled Paper

-------
             UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                              WASHINGTON D.C. 20460
                                                              OFFICE OF THE ADMINISTRATOR
                                                                SCIENCE ADVISORY BOARD
                                     May 17,2010

EPA-CASAC-10-011

The Honorable Lisa P. Jackson
Administrator
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, N.W.
Washington, D.C. 20460

       Subject:  CASAC Review of Policy Assessment for the Review of the PMNAAQS - First
               External Review Draft (March 2010)

Dear Administrator Jackson,

       The Clean Air Scientific Advisory Committee (CASAC) Particulate Matter (PM) Review
Panel teleconferenced on April 8, 2010 and April 9, 2010 to review the Policy Assessment for the
Review of the Particulate Matter National Ambient Air Quality Standards (NAAQS) - First
External Review Draft (March 2010).  The Policy Assessment is intended to help "bridge the gap"
between the relevant scientific information and assessments and the judgment required of the EPA
Administrator in determining whether and how to revise the National Ambient Air Quality
Standards (NAAQS) for Particulate Matter (PM).  This letter highlights CASAC's main comments
on the Policy Assessment, followed by our consensus responses to the charge questions and
comments from individual Panel members. Before addressing specific issues associated with the
Policy Assessment, we offer some general thoughts below for your consideration.

General Comments

       In 2006, EPA revised the primary and secondary NAAQS for PM to provide increased
protection of public health and welfare, setting the 24-h PM2 5 standard at 35  ug/m3 and retaining
the annual standard at 15 ug/m3. In the Policy Assessment released earlier this year, EPA is
considering revisions to these primary standards, using PM2S as the indicator for fine particles, and
either PMio-2 s (as measured), or the combination (difference) of PM2sand PMioas the indicator
for thoracic coarse particles.  EPA has also, for the first time, proposed consideration of a
secondary PM standard that would be based on a different indicator- PM light extinction - than
the mass-based indicators which have  long been used for the primary PM standards.  The CASAC
applauds this initiative and encourages the Agency to move aggressively forward toward
developing the next generation of indicators for primary PM standards as well.

-------
       EPA has a history of continuously improving indicators for primary PM NAAQS as new
evidence becomes available. Since the first standard for airborne PM. Total Suspended Particles
(TSP), was promulgated in 1971, the EPA has implemented PM NAAQS with progressively more
specific indicators from both the pollution source and biological perspectives. In 1987, EPA
changed the indicator from TSP to PMm (thoracic fraction penetrating into the thorax) and in 1997,
fine PM (PM2 5), which differs substantially in chemical composition from the coarse particle
fraction within PMio and which penetrates to the gas exchange region of the lung, was adopted to
supplement the protection provided by the PMio standard.

       PM2 5 has been a useful surrogate index since it was adopted in the 1997 PM NAAQS
promulgation, but may become an increasingly inadequate index of health risk as the mass
concentration limits are reduced to the levels being contemplated in the current Policy Assessment.
PMio is a non-specific indicator of the respiratory system risks associated with PMio-2s, as widely
acknowledged in the Panel's discussions of the Integrated Science Assessment OSA), Quantitative
Health Risk Assessment, and Policy Assessment documents. Neither index is directly related to
health risks associated with ultra-fine particles (UFP). The Panel also notes the challenge posed by
the promulgation of a PM NAAQS directed at coarse particles, which vary widely in mass and
composition across the country and between urban and rural areas.

       While research evidence on PM and health has evolved, we urge the Agency to undertake
additional efforts to leverage the gains made thus far. Now is the time to look ahead to future
review cycles and reinvigorate support for the development of evidence that might lead to newer
indicators that may correlate better with the health effects associated with ambient air
concentrations of paniculate matter (PM) and for more comprehensive PM monitoring data,
including expanding the range of sizes, and adding measurements of numbers, chemistry, species,
and related emissions characteristics of particles. There is an inherent feedback in the cycle from
research to policy formulation whereby researchers use the monitoring data that are gathered
primarily for regulatory purposes with available indicators, and, in turn, expand the scientific basis
for regulation. If EPA initiates efforts with air pollution research and monitoring communities
now to create a more robust monitoring platform for research, the Agency will be better positioned
to make an evidence-based transition to the "next generation" of indicators of PM-related health
risks.

Draft Policy Assessment (March 2010)

       Overall, the first draft Policy Assessment provides a solid start as EPA completes further
analyses and deliberations related to thoracic coarse particles and visibility-related effects.  As the
first instance of moving through the sequence from an Integrated Science Assessment (ISA)
through a Policy Assessment, the documents and the underlying flow of evidence and analytic
findings provide an ordered and transparent basis for decision-making.  CASAC was unanimous
in recommending that the document be condensed and carefully edited to remove redundancy.
While the document needs to be self-contained, the text should quote only the most critical
evidence and rely more extensively on citations to the ISA and Risk Assessment. CASAC also
suggests that the underlying decision-making algorithm that flows through the major questions
posed in the text be made more explicit. (Text, and perhaps  a figure, could be provided in Chapter
1, or at the start of Chapters 2-5.)

-------
       Primary Standards

       The Policy Assessment recommends consideration of NAAQS that cover both fine
particles, using PM2 5 as the indicator, and thoracic coarse particles, using either PMio-2 s or the
combination of PM2 5 and PMi0. At the present time, CASAC agrees that these indicators are the
best available given scientific knowledge to date. The evidence presented in the Integrated
Science Assessment, Quantitative Health Risk Assessment and Policy Assessment shows that health
effects have been observed in areas that meet current standards for PM2 5- Thus, CASAC supports
EPA's conclusion that consideration should be given to revising the current PM2 5 primary
standards to provide increased public health protection. Since the last NAAQS review, findings
from additional epidemiological studies support the association between PMio-2 s and
cardiovascular and  respiratory morbidity and mortality. While the certainty of the associations
between health effects for the association with PMio-2 5 exposure does not match that for PM2 5,
CASAC also agrees with EPA's conclusion that there is sufficient evidence to consider an
adjustment of the PMio standard.

       The choice  of indicator is driven by findings of epidemiological studies that have drawn on
monitoring data for PMio and PIVbs as exposure measures, as well as biological and source-related
considerations. For the fine particle standard, the Policy Assessment draws heavily on the analyses
in the Quantitative  Health Risk Assessment for Paniculate Matter, which are based on the
assumption of a linear no-threshold model for risk.  The Risk Assessment covers the evidence
supporting this model, drawing on the findings of the ISA, particularly the finding of associations
of PM indicators with morbidity and mortality at concentrations measured in the United States. In
commenting on the Second External Review Draft of the  Quantitative Health Risk Assessment,
CASAC recommended that EPA develop criteria for setting the lower bound for the scenarios
considered in the risk analysis.  That same issue emerges  with the Policy Assessment as lower
bounds are considered for the range for the annual and 24-hr PM2 5 standards.  The approach could
be statistically based, using, for example, interquartile range, standard deviation, 10 percentile
level, or lower-bound confidence interval, but also based  on broader consideration of uncertainty
and of the level of health protection to be achieved. In fact, selection of the lower bound of the
inter-quartile range may not be sufficiently health protective. Specifying the criterion by which a
lower bound concentration is chosen becomes central in decision-making under a no-threshold
model, since any suite of NAAQS above policy-relevant  background leaves  residual morbidity and
premature mortality.

       Secondary  Standard

       With respect to EPA's consideration of a secondary PM standard, CASAC supports the
recommendations for secondary PM standards to protect  public welfare as presented in the first
draft Policy Assessment.  The focus on protecting visibility  is appropriate given the advanced
scientific understanding of the direct effects of PM on visual air quality, as set out in the Integrated
Science Assessment. The Policy Assessment establishes that the current secondary PM standards
are inadequate to protect visibility, as their levels are too  high, their averaging times are too long,
and their (dried) mass-based indicators do not fully represent PM effects on visual air quality in the
ambient atmosphere.

-------
       CASAC supports the proposed "PM light extinction" indicator, as it represents an actual
and measurable consequence of particles in the ambient air and it relates directly to the
demonstrated harmful welfare effect of ambient PM on human visual perception.  PM light
extinction also captures the physical and chemical properties of ambient aerosols and associated
meteorological conditions. The Policy Assessment also proposes an optional PM25 mass indicator,
which the Panel finds to be a less direct measure of light extinction, but an adequate indicator for a
secondary PM standard, especially if the mass scattering/absorption efficiencies are adjusted on
seasonal, site-specific and/or regional  basis to reflect differences in aerosol composition and
relative humidity.  Moreover, a mass-based standard could be considered on an interim basis while
methods to measure the preferred PM  light extinction indicator are being developed, tested and
deployed.

       At most locations, a majority of PM light extinction is caused by fine particles. Coarse
particles are less of a contributing factor for light extinction, while posing measurement problems
for instruments measuring light scattering or adsorption.  For these reasons, an initial focus on
PM2 5 light extinction may be appropriate.  However, coarse particle visibility effects can be very
important at some  locations, and a variety of approaches could be employed to measure or estimate
coarse particle effects, regardless of whether light extinction or mass-based indicators are
employed (see responses to questions  14,17 and individual panelist comments for more detail).

       CASAC supports the use of a one-hour averaging time for a secondary visibility standard,
as this time span approximates the nearly instantaneous nature of human perception of visual air
quality, which itself can change rapidly over the course of a day. An hourly averaging time would
also be well within the  instrumental response capabilities of currently available light extinction
monitoring instruments, although if a PMzs mass indicator were used, a somewhat longer
averaging time  may be necessary to accommodate the larger response variability of some
continuous PM2 5 monitors.  The current focus on daytime visibility is also appropriate, given the
fact that nighttime  visibility, while important, has a complex relationship with PM and other
factors such as light pollution, for which scientific understanding is less advanced than for daylight
conditions.

       The Policy Assessment presents a large  number of optional combinations of levels and
forms, including levels for PM light extinction in the range of 20 to 30 deciviews, with forms
ranging between the 90th and 98th percentiles, which would be calculated based on either the single
worst hour in a day (with 90th or 95th percentile) or based on all daylight hours (with 95th or 98th
percent!le).  These many options are further complicated by consideration of PM2 5 mass as  a
possible alternative indicator.  While the range of options proposed appears reasonable, CASAC
recommends that the next draft of the Policy Assessment provide more detailed illustrations of the
similarities and differences in the many optional combinations, to provide a more informed  view of
the policy implications of these options.  Specifically, it  would be useful to describe the levels of
visibility protection and/or PM25 mass concentration, the PM species, the sources and event types
that would be most affected, and the temporal and  spatial variations in these factors that would
result from different combinations of optional indicators, levels and forms.
       In addition  to its principal focus on visibility protection, the Policy Assessment also
includes very brief discussions of effects of PM on climate, ecosystems and materials. CASAC

-------
concurs with the Policy Assessment's conclusions that while these effects are important, and should
be the focus of future research efforts, there is not currently a strong technical basis to support
revisions of the current standards to protect against these other welfare effects.  Because different
PM components that affect visibility can have differential (cooling or warming) effects on climate,
CASAC recommends revising the discussion of climate effects to better inform policy makers that
reducing certain aerosol components could lead to increased radiative forcing and regional climate
warming while having a beneficial effect on  visibility. However, the causal relationship between
PM and climate is established, and the risk of aerosol impact on climate is high; therefore further
research on a regional basis is urgently required.

       Finally, CASAC reiterates its expectation that the revised Policy Assessment will be
accompanied by a delineation of key changes from the first draft. This will enhance the efficiency
and targeting of our review, and will provide a transparent record of the basis for these changes.
As always, we appreciate the opportunity to  provide advice on this important subject and look
forward to continuing our dialogue with the Agency on PM.
                                   Sincerely,

                                         /Signed/

                                   Dr. Jonathan M. Samet
                                   Chair
                                   Clean Air Scientific Advisory Committee
Enclosures A: CASAC Paniculate Matter Review Panel Roster
           B: CASAC Responses to Charge Questions
           C: Individual Panelists' Responses to Charge Questions

-------
                                        NOTICE


This report has been written as part of the activities of the EPA's Clean Air Scientific Advisory
Committee (CASAC), a federal advisory committee independently chartered to provide extramural
scientific information and advice to the Administrator and other officials of the EPA. CASAC
provides balanced, expert assessment of scientific matters related to issues and problems facing the
Agency. This report has not been reviewed for approval by the Agency and, hence, the contents of
this report do not necessarily represent the views and policies of the EPA, nor of other agencies
within the Executive Branch of the federal government. In addition, any mention of trade names or
commercial products does not constitute a recommendation for use. CASAC reports are posted on
the EPA Web site at: http://www.epa.gov/casac.

-------
Enclosure A
                        Clean Air Scientific Advisory Committee
                            Participate Matter Review Panel
CHAIR
Dr. Jonathan M. Samet, Professor and Chair, Department of Preventive Medicine, University of
Southern California, Los Angeles, CA

CASAC MEMBERS
Dr. Joseph Brain, Philip Drinker Professor of Environmental Physiology, Department of
Environmental Health, Harvard School of Public Health, Harvard University, Boston, MA

Dr. H. Christopher Frey, Professor, Department of Civil, Construction and Environmental
Engineering, College of Engineering, North Carolina State University, Raleigh, NC

Dr. Donna Kenski, Data Analysis Director, Lake Michigan Air Directors Consortium, Rosemont,
IL

Dr. Armistead (Ted) Russell, Professor, Department of Civil and Environmental Engineering,
Georgia Institute of Technology, Atlanta, GA

Dr. Helen Suh, Associate Professor, Harvard School of Public Health, Harvard University, Boston,
MA

Dr. Kathy Weathers, Senior Scientist, Gary Institute of Ecosystem Studies, Millbrook, NY
CONSULTANTS
Dr. Lowell Ashbaugh, Associate Research Ecologist, Crocker Nuclear Lab, University of
California, Davis, Davis, CA

Prof. Ed Avol, Professor, Preventive Medicine, Keck School of Medicine, University of Southern
California, Los Angeles, CA

Dr. Wayne Cascio, Professor, Medicine, Cardiology, Brody School of Medicine at East Carolina
University, Greenville, NC

Dr. David Grantz, Director, Botany and Plant Sciences and Air Pollution Research Center,
Riverside Campus and Kearney Agricultural Center, University of California, Parlier, CA

Dr. Joseph Helble, Dean and Professor, Thayer School of Engineering, Dartmouth College,
Hanover, NH

-------
Dr. Rogene Henderson, Senior Scientist Emeritus, Lovelace Respiratory Research Institute,
Albuquerque, NM

Dr. Philip Hopke, Bayard D. Clarkson Distinguished Professor, Department of Chemical
Engineering, Clarkson University, Potsdam, NY

Dr. Morton Lippmann, Professor, Nelson Institute of Environmental Medicine, New York
University School of Medicine, Tuxedo,  NY

Dr. William Malm, Research Physicist, National Park Service Air Resources Division,
Cooperative Institute for Research in the Atmosphere, Colorado State University, Fort Collins, CO

Mr. Charles Thomas (Tom) Moore, Jr., Air Quality Program Manager, Western Governors'
Association, Cooperative Institute for Research in the Atmosphere, Colorado State University, Fort
Collins, CO

Dr. Robert F.  Phalen, Professor, Department of Community & Environmental Medicine; Director,
Air Pollution Health Effects Laboratory;  Professor of Occupational & Environmental  Health,
Center for Occupation & Environment Health, College of Medicine, University of California
Irvine, Irvine, CA

Dr. Kent Pinkerton, Professor, Regents of the University of California, Center for Health and the
Environment, University of California, Davis, CA

Mr. Richard L. Poirot, Environmental Analyst, Air Pollution Control Division, Department of
Environmental Conservation, Vermont Agency of Natural Resources, Waterbury, VT

Dr. Frank Speizer, Edward Kass Professor of Medicine, Channing Laboratory, Harvard Medical
School, Boston, MA

Dr. Sverre Vedal, Professor, Department of Environmental and Occupational Health Sciences,
School of Public Health and Community Medicine, University of Washington, Seattle, WA

SCIENCE ADVISORY BOARD STAFF
Dr. Holly Stallworth, Designated Federal Officer, EPA Science Advisory Board Staff Office
                                            8

-------
Enclosure B

 CASAC Responses to EPA Charge Questions on the First Draft Policy Assessment for the
                        Review of the Particulate Matter NAAQS

 1.  Organizational Structure
   a.  Does the Panel find the introductory and background material, including that
       pertaining to previous reviews of the PM standards and the current review, to be
       clearly communicated and appropriately characterized?
       The introductory and background material in the Policy Assessment for the Review of the
       Particulate Matter NAAQS (PA), including the summary of previous reviews, is
       appropriate and clearly communicated; however, the discussion could be condensed or
       summarized, thereby improving clarity.
   b.  To what extent does the Panel find that the questions posed appropriately reflect the
       important policy-relevant issues for this review? Are there additional policy-relevant
       questions that should be addressed?
       The questions posed in the Policy Assessment reflect important policy-relevant issues
       relevant to this consideration of the PM NAAQS. As discussed in our letter to the
       Administrator, we recommend the Agency begin to consider the evidence needed to
       continue the refinement of the PM indicator for such variables as chemical composition
       and optimal size ranges.
   c.  What are the Panel's views regarding the level of detail used to address these policy-
       relevant questions? Please comment on the extent to which the discussions are
       appropriately focused in addressing the questions.
       The level of detail is excessive. It should be cut back to provide succinct summaries of the
       information extracted from the Integrated Science Assessment (ISA) and Quantitative
       Health Risk Assessment (RA) documents. An Executive Summary would be very helpful,
       and could be based on the summaries offered throughout the Policy Assessment.

 2.  Technical Content
   a.  Do the discussions accurately reflect and clearly communicate the currently available
       health and welfare effects evidence, including important uncertainties, as
       characterized in the final ISA?
       Yes, the Policy Assessment captures the  currently available health and welfare evidence, as
       well as important uncertainties as characterized in the final ISA; however, much of the
       evidence could be summarized and referenced to the ISA, making the Policy Assessment a
       more "reader friendly" document. The Policy Assessment satisfactorily captures the range
       of proven and possible health effects literature that has expanded since previous PM
       reviews.
       The discussion of the welfare effects on  visibility accurately reflects the currently available
       scientific evidence. The rationale for selecting the 1 -hour averaging time is particularly
       clear and convincing. The summary of the differences between nighttime visibility and

-------
      daytime visibility, and the basis for considering the human impact of short-term visibility
      vs. all daylight hours, is clearly stated.
   b.  Do the discussions accurately reflect and clearly communicate policy-relevant
      information from the human health risk and visibility assessments, including
      important uncertainties, as characterized in the second draft REAs?
      Yes, the Policy Assessment captures the policy-relevant information and important
      uncertainties as characterized in the second draft RA. However, more of the evidence could
      be summarized and referenced to the RA. The policy-relevant information for the visibility
      assessment is clearly communicated, particularly the discussion of the nighttime visibility
      effects and the rationale for selecting the 1 -hour averaging time over all daylight hours.

3.  Approach (section 2.1.3): Various approaches are described for translating different air
    quality metrics from epidemiological studies into the basis for reaching preliminary staff
    conclusions on the adequacy of the current PM2.s standards and on alternative
    standards for consideration. What are the Panel's views regarding the appropriate
    weight to place on these various approaches; should specific approaches be afforded
    more or less weight? Are there additional approaches that should  be considered?

   The selection of both evidence- and risk-based approaches is sound, as  is the consideration of
   the utility of both annual and 24-hour concentration limits, and of various pairings of the two
   in exploring the degree of public health protection offered under possible NAAQS. The
   balance between evidence-  and risk-based approaches is appropriate.

   Charge question #3 also pertains to approaches used to translate different air quality metrics
   from epidemiological studies  into the basis for identifying concentration ranges (levels)
   relevant to assessing the adequacy of the  current PM2 5 standards and to suggesting alternative
   standards for consideration. These approaches included using concentrations somewhat below
   the aggregate mean, at one standard deviation below the mean, and at the  lowest quartile of the
   interquartile range. A more direct approach would use confidence bounds on the
   concentration-response relationships, such as those found in time-series studies. Attention
   could then be focused on the region where these bounds and other sources of uncertainty
   indicate an unacceptable degree of uncertainty about the concentration-response relationship.
   The less direct approaches used in the Policy Assessment instead make use of the
   concentrations observed in the epidemiological studies without consideration of the estimates
   of effect. The confidence bounds along a concentration-response  relationship and
   distributional measures of concentrations are intrinsically related,  since the estimated
   exposure-response relationship is most certain in a concentration range  where most of the data
   reside. We recommend  that EPA consider exploring the use of confidence bounds on the
   concentration-response relationships, where they are available, rather than the distributional
   measures of concentrations.

4.  Adequacy of current suite of PM2.s standards (section 2.2):
   What are the Panel's views on the following:
   a.  The appropriateness and characterization of the sets of long- and short-term PM2.s
      exposure studies highlighted? (section 2.2.1)
                                           10

-------
    •   Are there specific studies that should be given more or less emphasis?
    The Policy Assessment features appropriate studies and they are given appropriate
    emphasis. The discussion would be enhanced if there were an added indicator in Figure 2.1
    to specify which of the studies were multicity vs. single-city studies.  This would visually
    emphasize that more confidence should be given to the multicity studies (since the
    confidence intervals would be tighter) and provide justification for preferentially relying on
    multi-city studies. The Krewski et al. (2009) study receives greater emphasis in comparison
    to the other highlighted studies of annual mortality, however, because the study population
    was a relatively high SES group of volunteers, the estimates from the study may
    underestimate the risk for the US population in general. The Policy Assessment should
    acknowledge this potential underestimation, and present the findings of this study in the
    context of the other US cohort studies.

    •   Are there additional studies that should be focused on?
    Yes, the section concludes that older adults and children along with those with pre-existing
    cardiopulmonary diseases should be considered as part of the susceptible population. In
    addition, CASAC suggests that additional groups (as identified in the ISA) should be
    mentioned, including diabetics, and people with specific genetic polymorphisms that were
    reported in the ISA.
    •   Does the Panel generally support placing greater weight on multi-city vs. single-
       city short-term- exposure studies?
    Yes, see above in response to the first bullet.
    •   Does the Panel agree with the characterizations of PMks air quality at which
       associations have been observed?
    Yes, the presentation of results indicating PMj 5 effects below the current 24 hour and
    annual standard is appropriately summarized from multiple studies, adds to the
    observations previously made, and supports the conclusion about the inadequacy of the
    current two primary PM standards, However, the selection of the lower bound of the inter-
    quartile range may not be sufficiently health protective. Further consideration should be
    given to using the 10th percentile as a level for assessing various scenarios of levels for the
    PM NAAQS.
b.  The focus on persons with lower socioeconomic status as a susceptible population in
    addition to consideration of newly available evidence for susceptible populations
    identified in previous reviews (e.g., children, older adults, persons with pre-existing
    cardiac and  respiratory disease)? (section 2.2.1)
    The data presented fully justifies consideration of lower socioeconomic  status (SES) people
    as a susceptible group. This consideration would build from the integration of recent
    findings in the ISA  that expand the identification of susceptible groups.
c.  The preliminary staff conclusion that the estimated risks remaining upon simulation
    of just meeting the current suite of standards can reasonably be judged to be
    important from a public health perspective? (section 2.2.2)
                                        11

-------
      A substantial body of empirical evidence is summarized indicating that significant health
      effects are reported at or just below the current 24 hour and annual standards, justifying the
      conclusion in the Policy Assessment that the observed effects are important from a public
      health perspective. Thus, the discussion of estimated public health risks remaining if
      ambient air quality levels were to just meet the current standard is valuable and should
      contribute to the deliberations made by the Administrator in setting a PM NAAQS to
      protect the public with an adequate margin of safety.
   d.  The relative roles of the annual and 24-hour standards in providing public health
      protection as informed by the quantitative risk assessment, specifically in focusing on
      simulation of estimated risks remaining upon just meeting the current suite of
      standards?  (section 2.2.2)
      The approach used to characterize  the relative contributions of the annual and 24-hour
      concentrations is clearly presented. The Policy Assessment clearly explains why greater
      confidence can be placed on the roll-back estimates when the annual standard estimates are
      controlling, as compared to  when the 24-hour values are controlling. The link to the
      characteristics of the data available is made clear. As CASAC wrote in discussing the RA,
      we would have preferred that EPA use Total Cardiovascular Mortality for both
      comparisons, although the conclusions would not likely change substantially. The RA and
      Policy Assessment do explain their use of ischemic heart disease (IHD) mortality.
   e.  The integration of evidence-based and  risk-based considerations to inform the
      preliminary staff conclusion that the available information clearly calls into question
      the adequacy of the current suite of PIVh.s standards and provides strong support for
      giving consideration to revising the current standards to provide increased public
      health protection? (section 2.2.3)
      With regard  to the integration of evidence-based and risk-based considerations, CASAC
      concurs with EPA's conclusion that the new data strengthens the evidence available on
      associations previously considered in the last round of the assessment of the PM2 5 standard.
      CASAC also agrees that there are significant public health consequences at the current
      levels of the standard that justify consideration of lowering the PM2sNAAQS further.
      More text might be provided to describe specifically the uncertainties that have been
      addressed by the new evidence since the last review.

5.  Indicator (section 2.3.1):
   What are the Panel's views on the following:
   a.  The preliminary staff conclusion that it is reasonable to retain PM2.sas an indicator
      for fine particles?

      This conclusion is reasonable for this round of the PM NAAQS review. There was
      insufficient peer-reviewed literature to support any other indicator at this time. However, it
      is essential that EPA continue to develop the ambient air quality database so that the suite
      of PM indicators considered in epidemiological studies can be expanded.  Population-based
      studies will be needed to begin to assess more refined PM indicators, including chemical
      speciation of fine particles and measurements of ultrafine particles (UFP) that may more
      directly reflect the risk of PM2sto  human health.
                                           12

-------
   b.  The preliminary staff conclusion that the currently available information is too
      limited to support a distinct PM standard for ultrafine particles?
      This conclusion was unavoidable in that there was insufficient peer-reviewed literature on
      ultrafine particle (UFP) exposures and exposure-response relationships to support a UFP
      indicator at this time. As for fine PM (above), EPA needs to implement a network of UFP
      monitors to support broader research on associations between UFP exposures and health.
   c.  The preliminary staff conclusion that the currently available information is not
      sufficient to support consideration of a separate indicator for a specific PM2.5
      component or for the mix of fine particles from any specific source category, and that
      data are too limited to consider eliminating any component or source category from
      the mix of fine particles included in the PIVh.s mass-based indicator?

      See answer to 5.a. above.

6.  Averaging Times (section  2.3.2):
  What are the Panel's views on the following:
  a.  The preliminary staff conclusion that it is reasonable to retain annual and 24-hour
      averaging times?

      This conclusion is reasonable in view of the current limited evidence pointing towards
      other averaging times.
  b.  The preliminary staff conclusions that the currently available evidence is too limited
      to support additional averaging times to address sub-daily exposures  or seasonal
      exposures?

      This conclusion is reasonable. To the extent that EPA supplements or replaces 24-hour fine
      PM samplers with continuous monitors, it may become possible to conduct studies that
      demonstrate that other averaging times for acute responses may be more appropriate than
      24-hour averages, for at least some effects.

7.  Forms (section 2.3.3):
    What are the Panel's views on the utility of additional analyses to inform consideration
    of eliminating provisions  included in the current form of the annual standard that allow
    for spatial averaging across monitors, specifically with regard to the potential for
    disproportionate impacts on susceptible populations with lower socioeconomic status?

    Given mounting evidence showing that persons with lower SES levels are a susceptible group
    for PM-related health risks, CASAC recommends that the provisions that allow for spatial
    averaging across monitors be eliminated for the reasons cited in the Policy Assessment.
    Moreover, it should be recognized that monitors are located specifically to reflect exposures
    to surrounding populations, which may vary spatially.
                                         13

-------
    Given this recommendation and its basis, additional analyses are not needed to inform a
    decision about spatial averaging. Should EPA decide that additional analyses are warranted,
    however, a description of the aims, methods, and the decision criteria to apply these analyses
    are needed to assess their utility. CASAC notes the need for additional analyses and models
    to assess local heterogeneity in PM concentrations, both in terms of assessing population
    exposure, especially for low SES populations, and as a guide in future monitor siting
    decisions.

8.  Levels (sections 2.3.4, 2.3.5, and 2.3.6):
   What are the Panel's views on the following:
  a.  The preliminary staff conclusions regarding alternative standard levels that are
      appropriate for consideration, and  the rationale upon which those conclusions are
      based?

      CASAC believes that the document provides a clear and appropriate rationale for using the
      annual PM2 5 standard as the controlling standard and the 24-h standard to provide
      additional  protection. As discussed in more detail in our response to Charge Question 3,
      the text could be clarified substantially, specifically with regard to the rationale used to
      select the two sets of alternate standards considered in the Policy Assessment.  This
      explanation should  include the rationale behind the use of a particular lower bound to
      define annual PIVh 5 standards and the relation of the alternative standards to policy-relevant
      background levels.
  b.  The insights that can be gained from the quantitative risk assessment to inform our
      understanding of the roles that the  annual  and 24-hour standards play in  providing
      public health protection, specifically in focusing on simulations of estimated risks
      remaining upon just meeting alternative suites of standards? (sections 2.3.4.2 and
      2.3.5.2)
      CASAC supports the use of the quantitative risk assessment to illustrate the impact of
      alternative standards on health risk. The value of the risk assessment to this process would
      be increased substantially if the discussion was clarified and condensed.
  c.   The policy implications of the uncertainties associated with estimating risks, including
      potential underestimation of risk, in reaching conclusions regarding standards that
      would provide public health protection with an adequate margin of safety?
      CASAC recommends that the impacts of model choices on long- and short-term risk
      estimates should be expanded to assess whether and how estimated risks would differ if:
      (1) comparisons were made to the LML (as compared to the PRB); and (2) alternative
      concentration-response (C-R) functions were selected. The selection of C-R functions is
      relevant to margin of safety considerations. For example, use of C-R functions from
      Krewski et al. (2009) may underestimate risks for lower SES groups.
  d.  A policy approach for identifying a suite of standard levels in which the annual
      standard would be the "generally controlling" standard to provide  protection for both
      long- and  short-term PM2.s exposures, in conjunction with a 24-hour standard set to
      provide supplemental protection against days with  high peak concentrations
                                          14

-------
      associated with localized "hotspots" and risk arising from seasonal emission that
      might not be controlled by a national annual standard? (section 2.3.6)
      The policy approach for identifying the suite of standard levels was clearly described.
  e.  Additional considerations that could inform our conclusions related to alternative
      suites of fine particle standards?
      CASAC identified a few areas for further consideration, including potential alternative
      lower bounds for estimating risk, margin of safety issues (as described above) with regard
      to the impacts on estimated risks of model choices, particle speciation and size (e.g., coarse
      and ultra-fine).

9.  Adequacy of the current 24-hour PMio standard (Sect. 3.2):
   What are the Panel's views on the following:
   a.  The preliminary staff conclusion that the available scientific evidence supports
      maintaining a standard that provides protection against health effects associated with
      exposure to  thoracic coarse particles?
      CASAC concurs that the evidence since the last review is stronger in linking adverse health
      effects  to the thoracic course fraction.  One complexity is the finding that after taking into
      account the component of the PMio that is made up of PMz 5, urban thoracic coarse
      particles appear to be more toxic than rural PMio. Thus, no single measure of thoracic
      coarse fraction can be applied uniformly across all sites.
   b.  The various approaches described for translating the evidence and related
      uncertainties into the basis for conclusions on the adequacy of the current PMio
      standard?
      The Policy Assessment describes the uncertainty coming from having insufficient PMio-2 s
      monitors and limited geographic coverage, particularly for the purpose of comparing the
      thoracic coarse fraction in rural and urban areas. See the answer to question 13. Further
      insight could be gained by using the data from all locations with co-located PM samplers
      so that  PM 10-25 could be calculated as PMio concentration minus PM2 5 concentration.
   c.  The adequacy of the public health protection afforded by the current PMio standard
      against exposures to thoracic coarse particles?
      The relatively toxicity of the total PMio fraction depends on several factors. The
      proportion of the PMio that is in fact PM25 is different between urban and rural sites. In
      addition, although the limited evidence is only suggestive, there are independent
      differential potential health effects of the PM 10-25 portion of PM)0 in both urban and rural
      areas that can be separated from the health impact of PM25- Since such effects have been
      reported in both urban and rural areas at concentrations below the current standard, it is
      clear that the current PMio standard is not adequate to protect the public health. It is also
      the case that  there is no evidence to guide the choice between PMio or PM 10-2 5 that would
      be appropriate as a standard for both urban and rural regions as a single national standard.
      Since there are insufficient spatial PMio-2 5 measurements at this time, CASAC suggests
      that we continue with the PMio measure as the indicator, but encourage EPA to work on a
                                          15

-------
      national monitoring scheme for a range of particles including both smaller and larger
      fractions.
10. Indicator (sections 3.3.1, 3.3.2, and 3.3.3):
      What are the Panel's views on the following:
   a.  The preliminary staff conclusion that it is appropriate to maintain a standard that
      provides protection against all thoracic coarse particles, regardless of their source of
      origin or composition?

      This conclusion is justifiable at this time, in view of the evidence from the PM ISA
      showing that coarse thoracic PM exposures are associated with acute respiratory system
      responses that are not accounted for by fine PM. However, it is important to recognize that
      there is a lack of data on the influence of coarse thoracic PM composition. While it is
      reasonable to not consider chemical composition in the current round of review of a coarse
      thoracic NAAQS, it is essential that a suitable database for PMio-2 5 concentrations and
      composition be developed over the next several years so that a more satisfactory judgment
      on the need for composition-specific indicators can be made in the next review cycle.
   b.  The appropriateness of maintaining an indicator that allows lower PMio-2.s
      concentrations in urban areas than non-urban areas?
      Because PM2 5 concentrations are generally higher in urban areas than rural areas,  particles
      in this size range constitute a greater proportion of PMio in such areas.. In view of the
      evidence that there is a greater proportion of road dust, incinerator effluents, and other
      coarse PM that contain known toxicants in urban areas than in windblown soil dusts in
      rural areas, it is appropriate to maintain a coarse thoracic PM indicator that requires lower
      concentrations in urban areas than non-urban areas. In this review cycle, PMio can serve as
      a prudent index of coarse thoracic PM. However, as noted (in lO.a. above) chemical
      speciation of PM 10-2 5 may provide a more robust basis for future coarse thoracic PM
      NAAQS.
   c.  The appropriateness of either a PMIO or PMio-2.s indicator for a standard meant to
      protect  against exposures to thoracic coarse particles?
      While a PMio-2 5 indicator for a standard meant to protect against exposures to thoracic
      coarse particles would be preferable, a PMio indicator is an acceptable substitute in this
      round of review. However, in light of the implications of different source profiles  leading
      to differences in the current PMio nonattainment areas in the U.S., CASAC notes potential
      problems with  use of either the PMio or PMio-2 5 fractions as  the  indicator of coarse PM. In
      the short term, information from the ISA on PMio-25 health effects and regional differences
      (in particular, section 6.5.2.3) could be brought forward into the Policy Assessment to
      motivate consideration of alternative coarse PM indicators that could allow for a more
      focused  and  health relevant coarse PM standard.  Further analyses based on this
      information might also be considered. In the long term, information on chemically
      speciated coarse PM health effects would allow consideration of a coarse PM standard that
      reflects differential toxicity related to composition.

11. Averaging Time (section 3.3.4):
                                           16

-------
    What are the Panel's views on the preliminary staff conclusions that the currently
    available evidence supports maintaining a 24-hour standard to protect against health
    effects associated with short-term exposures to thoracic coarse particles, and does not
    support consideration of a long-term thoracic coarse particle standard?

    This conclusion is reasonable.

12. Form (section 3.3.6):
    What are the Panel's views on the preliminary staff conclusion that it is appropriate to
    consider a 98th percentile form for a revised 24-hour standard meant to protect against
    short-term exposures to thoracic coarse particles?

    CASAC finds the 98th percentile approach to be appropriate, but some members questioned
    the approach used to calculate the 98th percentile, noting that in the absence of additional
    information, a 98th-percentile calculation from the multi-year distribution of values was more
    defensible than a year-by-year calculation over three successive years.

13. Level (section 3.3.7):
    What are the Panel's views on the appropriateness of considering 98th percentile, 24-
    hour PMioor PM]o-2.s concentrations in epidemiological study locations for identifying a
    range of potential alternative standard levels for consideration, recognizing the more
    limited air quality information available for PMio-2.s?

    This portion of the Policy Assessment provided only a promissory note that additional
    analyses would be done, dependent upon the conclusions as to whether CASAC would
    recommend staying with PMio or moving to PMio-2 s as the indicator for coarse thoracic
    particles. The proposal to use PMio for alternative levels parallels what was done with PN/h 5
    and seems appropriate. Because of the potential lack of sufficient spatial data available for
    PMio-2 s, it is not clear that similar analyses can be accomplished.  However, there may be
    sufficient data coming from locations with co-located PM2 5 and PMio monitors, so that by
    subtraction an estimate of PM10-2 5 could be made.  It would be valuable to conduct some
    additional analyses if there is a smaller subset with sufficient data to carry out such estimates
    of PM 10-2 s-  In addition there may be data to make direct comparisons of the estimated levels
    with directly measured PM 10-2 5 in selected sites for purposes of testing the sensitivity of the
    subtraction estimates. This comparison would serve the purpose of demonstrating relative
    differences (or lack thereof) of the different indicators.

14. Adequacy (section 4.2):  What are the Panel's views on the preliminary staff conclusion
   that the currently available information clearly calls into question the adequacy of the
   current suite of PMz.5standards to provide public welfare protection from unacceptable
   levels of visibility impairment, primarily in urban areas, and supports consideration of
   alternative standards to provide appropriate protection?

    CASAC agrees that the Policy Assessment clearly describes how the form and level of the
    current PM standard are inadequate to protect the public from unacceptable levels of
    visibility impairment. Specifically, the levels of the current PM25 and PMio standards are too
                                          17

-------
    high, and their averaging times are too long, to guard against levels of visual air quality
    considered adverse over the short (hour or less) time periods during which changes in visual
    air quality are perceptible. The current mass-based indicators are also not ideal, and could be
    made much more directly responsive to visibility effects by using a PM light extinction
    indicator. Although an alternative PMi 5 mass indicator applied over shorter daylight hourly
    averaging times and at lower levels is not ideal to protect visibility, it would represent a
    substantial improvement over the current 24-hour and annual means.

    The relationship between the maximum daily 1-hour 90th percentile and the all daylight hours
    98  percentile is clearly and convincingly presented. The reasoning for excluding hours with
    relative humidity above 90% is also readily understood. The box-and-whisker plots of PM
    light extinction clearly show that rolling back the PM2 5 concentrations to meet the current
    standards would not protect visibility in most locations, regardless of which candidate
    protection level (CPL) or percentile form is considered.  It appears from Figure 4-3 that
    Phoenix (along with Tacoma) would also not exceed the least restrictive CPL more than 2%
    of the time. If so, Phoenix should also be mentioned on line 17 of page 4-26.

    The summary in section 4.2.2 is very clear in describing why the current PM standards are
    not protective of visibility. An addition that could strengthen the chapter in future drafts
    would be to show - with data examples - how high a 1-hour PM25 mass concentration or a 1-
    hour PM light extinction level could be on (a) day(s) when the 24-hour PM2 5 mass
    concentration does not exceed the current 35 ug/m3 standard. In addition, some photographic
    illustrations using actual photos or Winhaze images would be helpful to show what visual air
    quality looks like at levels of the current standards (and at the higher  1-hour levels that are
    possible without exceeding the 24-hour average standards).

15. Indicator (section 4.3.1): What are the Panel's views on the following:
    a.  The appropriateness of considering PM light extinction and PIVh.s mass as indicators
       for a distinct secondary PM standard for visibility protection, and the relative
       advantages and disadvantages associated with each indicator?
    b.  The appropriateness of considering the contribution of coarse particles with respect
       to a potential PM light extinction indicator?

      CASAC agrees it would be preferable to use PM light extinction as the indicator since  it is
      directly  related to visibility and can be  directly measured. Light extinction is a
      fundamental physical property of the ambient aerosol and thus,  it represents a very
      appropriate indicator for a secondary visibility standard.  While PM light extinction is
      much preferred, PM2 5 mass (with appropriate averaging time, level and form) could also
      be an adequate indicator for PM visibility effects, and may be more reflective of other
      possible welfare outcomes.

      Since a continuous PM25 mass monitoring network is already in operation and robust
      methods for widespread PM light extinction measurements will take several years to
      develop, evaluate and deploy, the Agency might consider a phased approach in which a
      standard is set based on PM light extinction, with compliance measurements initially made
      using Federal Equivalent Method PM2s mass as a surrogate for light extinction, until such
                                           18

-------
      time as a more direct PM light extinction network is in place. To facilitate a better
      agreement between mass and extinction some consideration should be given to adjusting
      the mass scattering efficiencies that relate mass to extinction in such a way as to reflect
      spatial and seasonal variability in the relative composition of ambient aerosols and regional
      differences in relative humidity. Setting a standard to any specific level of PM2S mass
      concentration would still allow a wide range of visibility effects at different times and
      locations and would provide much less efficient protection than could be attained with a
      light extinction indicator. CASAC still encourages the Agency  to develop suitable methods
      for measuring PM light extinction, and to evaluate these methods in a small pilot field
      network.

      The bulk of the light extinction at a majority of locations is driven by fine particles (PM2 5).
      There are significant problems with the measurement of both light scattering and light
      absorption from larger particle sizes.  Consequently, it is appropriate to focus initially on
      measurement of PM2 5 light extinction or its components. Another approach that would
      improve the quality of measurements of extinction from both fine and coarse particles
      would be to deploy a system that periodically switches between PM2 5 and PMio inlets for
      the same measurement device (nephelometer or aethalometer).  A third approach would be
      to limit light extinction measurements to instruments with PNhs inlets (which would also
      reduce instrumental maintenance requirements), and estimate coarse particle extinction
      from continuous PMio-2 s mass measurements, based on the difference method.

16. Averaging times (section 4.3.2): What are the Panel's views on the following:
   a.  The preliminary staff conclusion that consideration should be given to a 1-hour
      averaging time?
      CASAC concurs that a one-hour averaging time is most appropriate for a secondary
      standard to protect visibility, since PM effects on visibility can  vary widely and rapidly
      over the course of a day, and such changes are almost instantaneously perceptible to human
      observers.  An hourly averaging time is also well within the instrumental response times of
      the various currently available and developing optical monitoring methods.  If a PlVh 5 mass
      indicator is used, the Agency may wish to consider somewhat longer averaging times - 2 to
      4 hours - to assure a more stable instrumental response, although  obtaining reliable hourly
      data from currently available continuous PM2 5  monitors should be both feasible and a
      highly desirable objective for a variety of other applications.
   b.  The advantages and disadvantages of focusing on a  1-hour daily maximum or each 1-
      hour average concentration during all daylight hours?
      As noted in the second draft Urban-Focused Visibility Assessment (UFVA), there is
      relatively little difference in the levels of visibility protection that would be provided by a
      PM light extinction standard based on considering the single worst daylight hour in a day
      combined with a relatively low 90th percentile form and one based on considering all
      daylight hours combined with a higher 98th percentile form. It  is  likely that these two
      approaches may tend to emphasize different kinds of visibility events and source
      influences, and in its recent review of the  UFVA, CASAC  requested that the Agency
      prepare additional analyses to better illustrate and evaluate what these differences may be.
                                          19

-------
      We expect that the results of these analyses will be helpful to Staff in recommending the
      best approach and to CASAC in commenting on it more objectively.
      While it may be that the single worst daylight hour is judged to be a better approach,
      CASAC did not think that the stated rationale of protecting the most "sensitive" or
      "susceptible" individuals who may only have a short period of outdoor time each day to
      view scenery was an especially strong basis for this selection. We suggest that a different
      justification needs to be provided if this approach  is considered superior to one based on all
      daylight hours. In addition, CASAC recognizes that there are other  complexities associated
      with evaluating and protecting nighttime visibility - such as the combined influence of PM
      and light pollution.  The Policy Assessment should make clear that the current focus on
      daytime visibility is not because nighttime visibility is unimportant,  but rather because the
      scientific understanding of PM visibility effects under daylight conditions is currently
      much more advanced than for nighttime effects.

17. Levels and Forms (section 4.3.3):  What are the Panel's views on the following:
   a.  The preliminary staff conclusions regarding  ranges of options with varying levels and
      forms in combination with different indicators (light extinction  and PIVh.s mass-based
      indicators) and averaging times (1-hour daily maximum or each 1-hour average
      concentration during all daylight hours)?

      There is general agreement that the standard should be based on  1-hour averaging time.
      Furthermore, CASAC found that a standard based on extinction is preferable to a mass-
      based standard. However, because of potential issues in implementing monitoring for light
      extinction, a mass-based indicator may be acceptable, at least on an  interim basis, given the
      right averaging time and percentile level.  However, the consensus of the committee was
      that not enough information is provided in the first draft Policy Assessment to clearly
      evaluate the differences in protection associated with the various combinations of levels,
      percentiles, and 1 hr worst visibility vs. all daylight hours. The Policy Assessment more
      specifically does not address how the protection afforded by these various combinations
      would be further modified by the choice of an extinction vs. a mass  indicator.  For the
      limited comparisons illustrated in the Policy Assessment, CASAC agreed that 20 to  30
      deciviews was an appropriate range of levels for a PM light extinction indicator, but
      concludes that the upper bound of 40 - 60 ug/m3 was unreasonably  high for a  PM2 5 mass
      indicator and that a range more along the lines of  15-35 ug/m3 should be evaluated in
      future drafts.

      In considering the implications of using a mass vs. an extinction  indicator, it is critical to
      understand what level of average or median mass  concentration corresponds to a given
      level of extinction as a function of atmospheric characteristics in representative eastern and
      western cities. A given level of visibility protection will correspond to differing mass
      concentrations around the country, while a given level of PM2S mass will  correspond to
      differences in associated visibility protection. More detailed comparisons should be
      included in the Policy Assessment to better illustrate the differences  in visibility protection
      as the degree of protection varies with mass concentration and species composition.
      Regional and seasonal variation in protection should also be explored.
                                           20

-------
      While it has been demonstrated that a 90th percentile form combined with the worst
      daylight hour each day would generally result in similar average levels of visibility
      protection as the 98th percentile based on all daylight hours, these two options may tend to
      focus on different kinds of visibility impairing events, PM species compositions and
      emissions sources. The Policy Assessment should provide clearer illustrations of the
      differences (or similarities) between these approaches, and how they might be further
      modified  if a PM2s mass indicator is selected rather than the preferred light extinction
      indicator.

   b.  Additional approaches that could help inform our consideration of alternative levels
      and forms?

      The 90% RH screen may not be needed if a PM2 5 indicator is used.  Some consideration
      should be given to using a nephelometer that cycles between a PM25 and a PMio inlet.
      This approach would allow for the very accurate measurement of PM2 5 scattering with a
      reasonable estimate of coarse particle scattering. The  same approach could be used for a
      continuous mass or light absorption monitoring system.

      As indicated in previous comments on the UFVA, a "progress-based" attainment test rather
      than the traditional "threshold-based" attainment test would be preferable from a scientific
      perspective.  While the Agency has indicated this is not an option in the standard-setting
      process, this approach could and should be considered in the implementation phase (where,
      for example, the Agency could provide guidance on what would be considered reasonable
      rates of progress). Possibly an example of how this approach might work could be
      provided  in the next draft Policy Assessment.

      It would be helpful if the next draft of the Policy Assessment included a section  identifying
      and prioritizing key near-term monitoring and research needs to support refinements and
      revisions  to secondary PM standards in the future.  This could include establishing a small
      pilot urban visibility network to field-test and evaluate alternative PM light extinction
      measurement methods, additional visibility preference studies to better define levels
      (and/or frequencies) of visual air quality considered acceptable over a wider range of
      locations  and viewing conditions, and chemically speciated measurements of fine particles
      to support the linkage between PM mass and visibility impairment.

18. Climate (section 5.2): What are the Panel's views on the preliminary staff conclusion
   that there is insufficient information to base a standard on climate impacts associated
   with current ambient PM concentrations?

    CASAC agrees that the preliminary  staff conclusions on the climate impact of aerosols are
    correct - there is insufficient evidence on which to base a national standard. However, the
    causal relationship is established, and the risk of aerosol impact on climate is high so further
    research on a regional basis is urgently required. This need should be strongly voiced in the
    Policy Assessment, and research should be undertaken sooner rather than later. If possible,
    research should be designed and begun now to be included in  future assessments of the
    NAAQS.
                                          21

-------
  While there may be insufficient information to base a secondary standard on potential climate
  change, CASAC emphasizes that policy makers should be better informed that reducing certain
  aerosol components could lead to increased radiative forcing and regional climate warming
  while having a beneficial effect on visibility. There is abundant information demonstrating that
  aerosols that scatter visible light (e.g., nitrates and sulfates) are cooling, and can lead to
  increased/longer lived/more reflective cloud formation.  The Integrated Science Assessment
  (based largely on other assessments) demonstrates that light scattering aerosols lead to a net
  decrease in radiative forcing, and while the uncertainty is significant, it suggests that the effect
  may be quite large. Further, there is sufficient evidence that light absorbing aerosols (e.g.,
  black carbon) can lead to warming.  To go forward with a secondary standard based on
  reducing light extinction, the Policy Assessment needs to more directly .weigh the potential that
  regional warming could accompany improved visibility if light scattering aerosols are
  preferentially targeted. There also may be significant regional heterogeneity to controls that
  would address a visibility-based secondary standard which should be addressed. For example,
  the areas that currently have lower PM concentrations may benefit most from maintaining or
  improving visibility, but the areas with regionally high sulfate levels (and reduced visibility)
  would likely benefit most from the reduced radiative forcing. At a minimum, EPA needs to
  address how a secondary standard might impact regional radiative forcing and climate, assess
  the uncertainties, and provide a reasoned analysis weighing the two  issues. Also, on Page 5-11,
  lines 10-11 and lines 31-32, it is noted that "aerosols that are warming are co-emitted with
  aerosols that are cooling."  While this statement is correct in some instances, it does not
  accurately represent the aerosols with these competing effects. Most of the warming aerosols
  in the US are emitted by biomass burning and internal engine combustion.  Much of the
  cooling aerosol is formed in the atmosphere by oxidation of 862 or VOCs. Thus, a set of
  controls to control warming PM would not necessarily have that much impact on cooling PM
  and vice versa. These statements should be modified accordingly.

  Finally putative difficulties in relating PM to climate effects are laid out in sections 5.2.3 and
  5.2.4 (pages 5-11 to 5-13) and again in section 5.3.4 (page 5-24). However, these apply to all
  other potential effects of PM, and it  is not appropriate to emphasize them here, as if they were
  uniquely related to climate. For example, other PM-effects relationships would be improved by
  using chemically-speciated mass measurements; other relationships  are expected to vary
  locally and regionally; others have offsetting positive and negative effects (warming vs.
  cooling, radiation attenuation vs. radiation scattering); others are confounded by effects at a
  distance from sources, by the difficulty of distinguishing anthropogenic from natural
  constituents (PRB issues should be explicitly mentioned here - section 5.2.4); by spatial and
  temporal heterogeneity; etc. In addition, many of the caveats in section 5.3.4 are equally true of
  nitrogen oxides, sulfur oxides, and ozone, but have not presented insurmountable obstacles to
  revising NAAQS for those contaminants.

19. Ecological effects (section 5.3): What are the Panel's views on  the preliminary staff
   conclusion that data are insufficient to support establishing a distinct standard for
   ambient PM based on ecosystem  effects not addressed in the ongoing NOx/SOx
   secondary NAAQS review?
                                           22

-------
  CASAC agrees with the conclusions of the Policy Assessment that the published literature is
  insufficient to support a national standard for PM effects on ecosystem services. The best
  established effects are related to particles containing nitrogen and sulfur, which are no longer
  considered in this assessment of paniculate matter. CASAC considers that some causal
  relationships are known, but that sensitivities are not well characterized. Further research
  should be advocated, including progress  toward a monitoring network that will yield time
  resolved information on chemically speciated particulate mass that is likely to be most closely
  related to  ecosystem endpoints.

  The references in the Policy Assessment  to welfare valuation associated with end use, taken
  from the ozone assessment documents, and to ecosystem services as specific and individual
  endpoints, taken from the Millennium Ecological Assessment, may prove useful in structuring
  this research.

  CASAC recommends incorporation of text into Section 5.3 that is similar to that in the current
  draft of Section 5.4, regarding the importance of maintaining an appropriate degree of control
  of both fine and coarse particles to address ecosystem effects attributed to them, even in the
  current absence of sufficient information to develop a standard.

20. Material effects (section 5.4): What are the Panel's views on the preliminary staff
   conclusion that no new evidence calls into question the adequacy of the protection
   afforded by the current suite of PM standards, and that there continues to be support
   for retaining an appropriate degree of control for both fine and coarse particles to
   provide  protection against materials damage and soiling?

  CASAC agrees with the conclusions of the Policy Assessment that the published literature,
  including  literature published since the last review, is insufficient either to call into question the
  current level of the standard or to support any specific national standard for PM effects on
  materials. However, causal relationships are well established, particularly with regard to soiling
  of buildings and monuments. CASAC believes this end point is particularly  amenable to
  demonstration of relationships between PM mass (particularly chemically speciated mass) and
  levels of soiling, because the non- linear effects that characterize ecosystem  endpoints may be
  less significant. Soiling effects of PM on materials may be linked to atmospheric burdens
  simply through an appropriate deposition velocity for specific components.

  We believe it is important to retain the text in the current draft of Section 5.4 regarding the
  importance of maintaining an appropriate degree of control of both fine and  coarse particles to
  address materials damage and soiling attributed to them, even in the absence of sufficient
  information to develop a new standard.
                                           23

-------
Enclosure C
                      Compendium of Individual Comments
                on Particulate Matter Policy Assessment (March 2010)
Comments from Dr. Lowell Ashbaugh	2
Comments from Mr. Ed Avol	6
Comments from Dr. Joe Brain	12
Comments from Dr. Wayne Cascio	14
Comments from Dr. Chris Frey	17
Comments from Dr. David Grantz	18
Comments from Dr. Joseph Helble	21
Comments from Dr. Rogene Henderson	22
Comments from Dr. Mort Lippmann	23
Comments from Dr. Phil Hopke	31
Comments from Dr. Robert Phalen	37
Comments from Mr. Rich Poirot	38
Comments from Dr. Armistead Russell	46
Comments from Dr. Frank Speizer	48
Comments from Dr. Helen Suh	53
Comments from Dr. Sverre Vedal	55

-------
Comments from Dr. Lowell Ashbaugh

Question 2a:  Do the discussions accurately reflect and clearly communicate the currently
available health and welfare effects evidence, including important uncertainties, as
characterized in the final ISA?
I believe the discussions of the health effects evidence accurately reflect the currently available
evidence, but I'll focus more on the welfare effects and leave it to the other reviewers to provide
more detail on the health effects evidence.
The discussion of the welfare effects on visibility is quite complete and it accurately reflects the
currently available science. The rationale for selecting the 1-hour averaging time is particularly
clear and convincing. The summary of the differences between nighttime visibility and daytime
visibility, and  the reasoning for human impact of short-term visibility vs. all daylight hours is
clearly stated.
The term 'deciview' is introduced on page 4-18, but it is not defined. A simple description of its
relationship to extinction, along with guidance on how to understand it, would be useful. For
example, it would help to explain that a one deciview change is just noticeable, and that higher
values correspond to reduced visibility.
The discussion of the CPL values and how they were arrived at is very good.  It is particularly
useful to point out that the greater uncertainty, represented by the more scattered data points, for
the Denver and British Columbia results may arise from the use of actual photos taken under
varying conditions, while the better fit of the Phoenix results could be due to the use of WinHaze
to alter the same photograph to display different visibility conditions.
The distinction between the Washington, DC results and the results from western cities is
accurately and clearly described. The results presented by Smith (2009) on a new study at
Washington, DC are not discussed here, though, and probably  should receive some attention.
Although I believe there are good reasons to discard the results for the Policy Assessment, they
should be mentioned.
Question 2b:  Do the discussions accurately reflect and clearly communicate policy-relevant
information from the human health risk and visibility assessments, including important
uncertainties, as characterized in the second draft REAs?
The policy-relevant information for the visibility assessment is clearly communicated. I
particularly like the discussion of the nighttime visibility effects and the rationale for selecting the
1-hour standard over all daylight hours. The selection of CPLs is described well and the
justification for selecting them is consistent with the evidence presented. The only caveat I would
make is that the results of the Smith (2009) study presented at the April 2009 CASAC meeting
should be discussed. The results of that study provide an opportunity to explain uncertainties in
visibility studies and why these results should not be included in the final assessment.

-------
Question 14:  Adequacy (section 4.2): What are the Panel's views on the preliminary staff
conclusion that the currently available information clearly calls into question the adequacy of
the current suite of PM2.5 standards to provide public welfare protection from unacceptable
levels of visibility impairment, primarily in urban areas, and supports consideration of
alternative standards to provide appropriate protection?
 The Policy Assessment clearly describes how the form and level of the current PM standard is not
 adequate to provide public welfare protection from unacceptable levels of visibility impairment.
 The relationship between the maximum daily 1-hour 90th percentile and the all daylight hours 98th
 percentile is clear and convincing. The reasoning for excluding hours with relative humidity above
 90% is also clear. The box-and-whisker plots of PM light extinction are clear evidence that the
 current levels of PM standards are not protective of visibility, particularly combined with the
 estimated light extinction if PM levels are rolled back to just meet the current standard. It appears
 from Figure 4-3 that Phoenix (along with Tacoma) would also not exceed the least restrictive CPL
 more than 2% of the time, if so, it should be included in line 17 of page 4-26.
 The summary in section 4.2.2 is very clear in describing why the current PM standards are not
 protective of visibility.
 Question 15: Indicator (section 4.3.1): What are the Panel's views on the following:
      a.   The appropriateness of considering PM light extinction and PIVh.s mass as
          indicators for a distinct secondary PM standard for visibility protection, and the
          relative advantages and disadvantages associated with each indicator?
      b.   The appropriateness of considering the contribution of coarse particles with respect
          to a potential PM light extinction indicator?
 The issues associated with selecting PM light extinction as a standard are described well. The
 limitations of using 24-hour PM25 mass to predict 1-hour visibility impacts could be better
 described. In particular, the discussion of visibility impacts clearly points out that short-term
 visibility is most clearly representative of the welfare effects of visibility. PM mass, though, is not
 currently measured on a 1-hour time scale, so it is not able to adequately represent visibility
 impacts. Figure 4.3-1 shows the relationship between 24-hour visibility and 1 -hour light extinction.
 The uncertainty associated with this relationship could be discussed in more detail to illustrate the
 shortcomings of PM25 mass as an indicator of short-term visibility.
 It's appropriate to consider the contribution of coarse particles with respect to a potential light
 extinction indicator. Although coarse particles are not a significant contributor in some parts of the
 country, they are significant in other parts - notably the southwestern U.S. It's important that EPA
 acknowledge these differences by including coarse particles in the indicator.
 Question 16: Averaging times (section 4.3.2): What are the Panel's views on the following:
      a.   The preliminary staff conclusion that consideration should be given to a 1-hour
          averaging time?
      b.   The advantages and disadvantages of focusing on a 1-hour daily maximum or each
          1-hour average concentration during all daylight hours?
 The discussion of visibility effects on welfare presents clear and convincing evidence that a I-hour
 averaging time is appropriate for a visibility  indicator. The assessment points out that a significant
 fraction  of the population is exposed to visibility conditions for only brief periods during the day,
 and their welfare is affected by these brief exposures. The relationship between the 1-hour daily

-------
maximum and 1-hour average daylight visibility indicators is clearly explained and the protective
benefits of using the 1-hour maximum indicator is justified in the Policy Assessment.
Question 17: Levels and Forms (section 4.3.3): What are the Panel's views on the following:
     a.  The preliminary staff conclusions regarding ranges of options with varying levels
         and forms in combination with different indicators (light extinction and PM2.5
         mass-based indicators) and averaging times (1-hour daily maximum or each 1-hour
         average concentration during all daylight hours)?
     b.  Additional approaches that could help inform our consideration of alternative
         levels and forms?
The staff conclusions regarding the levels and forms of the light extinction indicator are related
well to the evidence presented earlier in the Policy Assessment. The difficulty in relating PM levels
to light extinction is also described well, and the selection of alternative PlVhs mass concentration
levels is well justified. In the discussion on page 4-34 it is not clear why the 98th percentile form is
not considered. A brief comment on why it is not included would be helpful. I infer from the
discussion that EPA staff considers it too restrictive, but this could be explained more explicitly.
The discussion of the maximum daily 1-hour indicator versus the average daylight 1-hour visibility
is well done. I like the characterization of people who have infrequent access to daytime visibility
as a susceptible population, and the explanation that the 1-hour daily maximum indicator would be
also protective of people who have access to visibility throughout the day is effective.
The box-and-whisker plots and the tables showing the effect of "just meeting" the alternative
standards are very effective in conveying the anticipated results. The use of color-coding in Figure
4.3-2 is very helpful in comparing the alternative standards. The blue shading in the table of
Appendix A is too light; it should be  darker to more effectively highlight those conditions.
Question 18: Climate (section 5.2): What are the Panel's views on the preliminary staff
conclusion that there is insufficient information to base a standard on climate impacts
associated with current ambient PM concentrations?
The preliminary staff conclusions on the climate impact of aerosols are accurate — there is
insufficient evidence on which to base a national standard. The risk of aerosol  impact on climate is
high, though, so this issue needs further development. It's important to state that research into the
effects of aerosols on climate is needed and should be undertaken sooner rather than later. If
possible, research should be designed and begun now to be included in future assessments of the
NAAQS.
Question 19: Ecological effects (section 5.3): What are the Panel's views on the preliminary
staff conclusion  that data are insufficient to support establishing a distinct standard for
ambient PM based on ecosystem effects not addressed  in the ongoing NOx/SOx secondary
NAAQS review?
The preliminary staff conclusion is appropriate for ecological effects. While acknowledging that
ecological effects occur, it is not possible at this time to establish a PM level that would be
protective of ecosystems.
Question 20: Material effects (section 5.4):  What are the Panel's views on the preliminary
staff conclusion  that no new evidence calls into question the adequacy of the protection
afforded by the current suite of PM standards, and that there continues to be support for

-------
retaining an appropriate degree of control for both fine and coarse particles to provide
protection against materials damage and soiling?

This preliminary staff conclusion is also appropriate. The wording of this conclusion could be
applied to the climate effects and the ecological effects sections. In particular, a version of the
following sentence from the materials effects section could be applied to the other two sections:
"However, in the absence of information that provides a basis for establishing a different level of
control, observations continue to support retaining an appropriate degree of control on both fine
and coarse particles to help address materials damage and soiling associated with PM."

-------
Comments from Mr. Ed Avol

General Comments: The objectives of the Policy Assessment (PA) document are laudatory - to
provide the Administrator with a review of the policy implications arising from the conclusions
drawn from the ISA and RA. The Policy Assessment has all of the elements to make a persuasive
and compelling presentation of the assembled data. However, in my opinion, the current Policy
Assessment is too repetitive, too wordy, and devotes too many pages to  re-visiting previous
discussions presented in the ISA and RA documents, rather than summarizing the findings and
moving on to the policy implications.

An inordinate amount of space is afforded to the 2004 AQCD, with regular discussion of what was
presented and/or concluded in the 2004 review document. The ISA and RA presented the relevant
science and evaluated it in the context of risk estimates and levels of uncertainty; what are the
policy implications of these findings?  The current document should focus on what the cumulative
information, through the current review, tells us now about PM exposure, effects, levels, and
susceptibility, so that an informed judgment can be made at this point in time.  It may be
interesting to compare what was known several years ago with what is known now, but that is not
the central issue at hand. As a result of this approach, this Policy Assessment document is much
too long, making it frustrating to read. This style of presentation results in weakening the impact of
the document, and masking its key messages. I personally found it very difficult to wade through,
even though I realize it  contains important information that will be critical to the Administrator's
determinations. This needs to dramatically improve in the next draft, in order to make this
document accessible and worthwhile.

Specific Comments:
P2-2, line 19 to P2-9, line 8 (sections 2.1.1 and 2.1.2)-This is historically interesting, helps to lay
the groundwork for how EPA got to where it is with the fine particle standard.. .and is arguably
completely irrelevant to the current document and discussion. The issue for the current policy
document is interpretation of the currently available information for re-examining the current
standard. Details about previous cycles of review and their outcomes could be placed in an
appendix, but do not centrally address the focused theme of this document - the basis for review of
the current standard.

P2-9, Iines24 & 25 - ".  ..the annual and 24-hour standards taken together, rather than to consider
each standard separately..." But didn't the RA clearly demonstrate that the annual standard was
"controlling" (i.e., set at a given annual level, it effectively achieved a number of 24-hour standards
being considered in conjunction with it?

P2-10 through2-13 - all this preparatory introduction about what is going to be discussed is
unnecessary meandering (in my opinion); the document should be a focused application of the data
presented in the ISA and RA, presented clearly and then clearly summarized; these pages could
(and should) be distilled down to a paragraph or two. The Policy document should be brief,
focused, and clear; this  document has the potential to become that, but isn't there yet.

-------
P2-10, line 9 - Declaring a focus on US and Canadian studies side-steps the issue that careful and
well-performed science (or poorly-performed science, for that matter) can be performed in other
countries, just as they can be performed in the US and Canada.

P2-10, lines 26 & 27 - This conclusion line is not needed here; first present the information and
then summarize the document's position in a succinct integrative final chapter or section.

P2-13, lines 30 to 35 - not necessary, can be deleted without impacting presentation.

P2-14, lines 11 to 18 - not necessary, can be deleted without impacting presentation.

P2-17, lines 25 to 34 - The text is slipping into a listing and re-discussion of studies and findings,
which was done quite well in the ISA and re-visited in the RA. This document should be a crafted,
directed summary, rather than a re-listing of the studies...so I think this section could be edited to a
line or two.  The policy-relevant comment is in lines 25 through 27 (".. .locations with the largest
reductions in PM2.5 saw the largest improvements in reduced mortality rates, while those with the
smallest decreases in PM2.5 concentrations saw the smallest improvements in reduced mortality
rates..."); the rest is repetitive carryover from the ISA.  Figure 2.1  (which is on an un-numbered
page that might be P2-23) could be refered to and used as the basis for comments about what was
reported in the ISA.. .without going through and talking about several of the studies in the text.

P2-19, lines 9-17 - Here, again, I think the point has been made and re-listing ISA references does
not add to the presentation; I think these can be deleted. The current presentation seems to have
slipped back into a review and listing of the body of evidence that  supports a given conclusion
(which was already  done in the ISA), rather than presenting the conclusions brought forward from
the ISA and RA and presenting a discussion that begins with those ISA and RA conclusions being
carried forward.

P2-24, lines 33 to 35 — The notion of a summary is great and much-appreciated, but this is too
wordy and meandering.  Delete lines 33 to 35 and get to the point!

P2-25, lines 1 and 2 - same as  previous comment, delete this and get to the point.

P2-25, line 26 to p2-26, line 2 - this is yet another example of unnecessary meandering and  re-
opening an issue that does not need to be re-visited in this document.  Earlier versions of ISAs, and
the PM ISA, discussed and evolved  towards a functional and useful definition of susceptibility, and
this document should begin from that point, not re-visit what the evolution was.

P2-26, line 29 - change the word "factors" to "observations".

P2-27, lines 8 to 11  - Is this consistent with the previous comment (P2-10, Iine9) that the focus
would be on US and Canadian studies (a focus with which I respectfully disagree)?

P2-43, line 4 - replace the phrase "peak shaving" with a more accurate one that describes the
approach.

-------
P2-44, lines 3 to 7 - This statement about the importance of annual average PIVb 5 concentrations
on interpreting risk results for both short and long-term exposures is a key observation that should
be highlighted, bolded, underlines, or otherwise clearly identified.

P2-45, line 9 and line 21 - The comments about lung cancer here (when the argument was
previously made about basing the presentation on "causal" levels of determination) raises the issue
of staff confidence and belief in the strength of association between Plvh 5 long-term exposure and
lung cancer. What is the determination on this matter that Staff wants to transmit to the
Administrator?

P2-45, line 11 to 14 - Does this comment about 10 of the 15 study areas having ".. .design values
which result in the current 24-hour standard controlling..."  contradict the previous one (p2-44,
lines 3-7), which asserted that annual averages were controlling for both short and long-term
exposures?  Together with a later comment (p2-49, lines 27 to 29, ".. .This pattern is generally
characteristic of the larger set of urban areas across the US that do not meet the current suite of
standards..."), this leaves the reader somewhat confused -which standard is controlling? What is
the policy implication?

P2-51, line 27 on to P2-52, line 33 - This discussion of the  representativeness of the study areas
seems out of place here, 50+ pages into the discussion...Shouldn't this be established early in the
presentation so that the credibility of all that follows is increased?

P2-52, line 34 to p2-55, line 3 - This seems like an obvious and unnecessary comment - Does
Staff really think that someone might conclude that these 15 areas in the presentation are the
ONLY 15 areas in the country for which this discussion is relevant?

P2-52, line 4 to p2-55,  linelO - This question and response  seems unnecessary and out of place
here. The purpose of the NAAQS review is to periodically  assess the current information relevant
to the current standard and make  a judgment about the adequacy of the standard to protect the
public health. The determination as to whether there are "risks remaining upon simulation of the
current suite of standards..." will be addressed when the body of information is presented and the
Administrator makes that determination.

P2-54, lines 24 to 28 - This is yet another example of meandering monologue - rephrase this to
present the conclusion clearly and succinctly - the data suggest that level of the current PM2 5
standard needs to be re-considered.

P2-57, lines 6 & 7 - Change this  line to read, "These techniques, while widely used in aerosol
research, have not yet been widely used in health effects studies, or systematically deployed for
extensive time periods in local or regional air monitoring networks."

P2-57, lines 23 to 28 - Unnecessary and can be deleted...

P2-58, lines 25 to 29 - This is certainly true and potentially a  self-fulfilling prophecy (in that there
will never be national-level ultra-fine data until and unless there is directed interest to support the

-------
collection of such data). Could an appendix of "Research Needs To Advance Future Reviews" be
added?

P2-62 - The presentation has slid back to a review of data, which has already been done in the ISA
(where it rightfully belongs, not here).

P2-64, lines 24 to 34 - This is a useful summary comment. Based on the history of PM research
(from mass-based assessments to particle size, constituent, and biological toxicity) and the PM
NAAQS (gradually moving to smaller size fractions), it seems reasonable to anticipate that a
component of PM might be the biological actor of interest. With the STN network and large-scale
chemical characterization studies, we are on the cusp of perhaps collecting the information needed
to identify the important chemical constituents of biological significance. This could and should
be added to the "Research Needs" list.

(The same generic comments apply for most of the next 50+ pages- too wordy, unnecessary detail,
substantial editing needed!)

P2-107, section 2.4 - This is a well-constructed, focused presentation and should serve as the
template for the previous 100 pages. These comments could be supported by references to the ISA
or RA, but the specific supporting materials do not need to be here to make a convincing
presentation (and to the authors' credit, the supporting materials do not appear here).

P2-1 09, Section 2.5 - Key Uncertainties - This section can and should be a focused, terse, and
brief presentation. . .a few pages at most, possibly only a page or less. It is somewhat surprising
that in this first draft, given  the excellent discussion in the RA, there is only a title and placeholder
here...

Chapter 3, Thoracic Particles - The presentation scheme is analogous to the previous chapter, and
in doing so, suffers from the same generic shortfalls (too much detail, too much repetition, only a
placeholder for the uncertainty discussion, etc).

Chapters 4 and 5 - defer to my visibility colleagues on these sections...

Specific Charge Question Assignments:
Question 4: Adequacy of current suite of PM^.s standards (section 2.2):
    What are the Panel's views on the following:
    a.  The appropriateness and characterization of the sets of long- and short-term PM2.s
       exposure studies highlighted? (Section 2.2.1). Too-long-winded a discussion but studies
       are the appropriate ones.

       •  Are there specific studies that should be given more or less emphasis? No, and the
          studies' discussion should be referenced to ISA and RA, not re-visited in the manner
          and extent here.

       •  Are there additional studies that should be focused on? No.

-------
   •   Does the Panel generally support placing greater weight on multi-city vs. single-
       city short-term exposure studies?  Favor multi-city studies to better substantiate
       findings and make findings that much more representative.

   •   Does the Panel agree with the characterizations of PIVh.s air quality at which
       associations have been observed? Yes.
b. The focus on persons with lower socioeconomic status as a susceptible population in
   addition to consideration of newly available evidence for susceptible populations
   identified in previous reviews (e.g., children, older adults, persons with pre-existing
   cardiac and respiratory disease)? (section 2.2.1). This is appropriate and could arguably
   be included in those with compromised health, because lower SES carries with it a host of
   related increased risks (access to health care, poor diet, increased smoking rates, increased
   exposure (roadway proximity of residences and occupational exposures), increased
   violence and stress, etc).
c. The preliminary staff conclusion that the estimated risks remaining upon simulation
   of just meeting the current suite of standards can reasonably be judged to be
   important from a public health perspective? (section 2.2.2). Based on the referenced
   health risk estimates, this is a logical conclusion.
d. The relative roles of the annual and 24-hour standards in providing public health
   protection as informed by the quantitative risk assessment, specifically in focusing on
   simulation of estimated risks remaining upon just meeting the current suite of
   standards? (section 2.2.2). Agreed and acceptable.
e. The integration of evidence-based and risk-based considerations to inform the
   preliminary staff conclusion that the available information clearly calls into question
   the adequacy of the current suite of PM:.5 standards and provides strong support for
   giving consideration to revising the current standards to provide increased public
   health protection? (section 2.2.3). Agreed and acceptable, but raises a policy issue as to
   whether PM speciation is potentially more informative rather than total mass.

Question 8: Levels (sections 2.3.4, 2.3.5, and 2.3.6):
What are the Panel's views on the  following:
 a. The preliminary staff conclusions  regarding alternative standard levels that are
    appropriate for consideration, and the rationale upon which those conclusions are
    based? The conclusions are fine - appropriate for the foundational arguments made. The
    presentation, however, is too wordy, too long, too repetitious, and in need of substantial
    editing.
 b. The insights that can be gained from the quantitative risk assessment to inform our
    understanding of the roles that the annual and 24-hour standards play in providing
    public health protection, specifically in focusing on simulations of estimated risks
    remaining upon just meeting alternative suites of standards? (sections 2.3.4.2 and
    2.3.5.2) Useful and warranted.  Presentation-wise, the figures and the summaries, with
    reference to the RA, would have been sufficient to make the point.
 c. The policy implications of the uncertainties associated with estimating risks, including
    potential underestimation of risk, in reaching conclusions regarding standards that
                                       10

-------
       would provide public health protection with an adequate margin of safety? My
       preference would have been for a clearer and concise section directly discussing
       uncertainty here, rather than comments about "increased variability" scattered throughout.
    d. A policy approach for identifying a suite of standard levels in which the annual
       standard would be the "generally controlling" standard to provide protection for
       both long- and short-term PM2.s exposures, in conjunction with a 24-hour standard
       set to provide supplemental protection against days with high peak concentrations
       associated with localized "hotspots" and risk arising from seasonal emission that
       might not be controlled  by a national annual standard? (section 2.3.6) The question
       itself is indicative of the systematic problem permeating the document - lack of clear,
       succinct and focused presentation.
    e. Additional considerations that could inform our conclusions related to alternative
       suites of fine particle standards? Clarity, focus, and brevity....

Question 12: Form (section 3.3.6):  What are the Panel's views on the preliminary staff
conclusion that it is appropriate to consider a 98th percentile form for a revised 24-hour
standard meant to protect against short-term exposures to thoracic coarse particles? The
argument for utilizing a concentration-based approach rather than an exceedance-based approach is
compelling, and the presentation is terse but clear.

Typos:
P2-1, line 16 - missing a period.

P2-43, Iine32 - change "references" to "reference5.

P2-51, line 23 - insert the word "to" after ".. .24-hour standard is controlling due..."
                                            11

-------
Comments from Dr. Joe Brain

1.  Organizational Structure
   a.   Does the Panel find the introductory and background material, including that
       pertaining to previous reviews of the PM standards and the current review, to be
       clearly communicated and appropriately characterized?

       The panel views the Policy Assessment (PA) document as critically important to the
       overall process of standard setting. It is the culmination of a deliberate process leading
       from the ISA to the REA, and ultimately to an informed decision by the Administrator.  In
       toto, the panel approves of the current versions of the ISA and the REA, and feels that key
       ideas from them now appear in the Policy Assessment. In particular, the panel feels that
       the discussion of previous reviews of the PM standards is adequate, appropriate, and
       clearly communicated.

    b. To what extent does the Panel find that the questions posed appropriately reflect the
       important policy-relevant issues for this review? Are there additional policy-relevant
       questions that should be addressed?

       The panel feels that the questions posed (questions 2-20) are pertinent to an overall
       assessment of the evidence for the regulation of PM. The questions - and the answers -
       focus on policy-relevant issues. Other than minor aspects addressed by answers to other
       charge questions, the panel feels that there are no major additional questions that should be
       addressed.

   c.   What are the Panel's views regarding the level of detail used to address these policy-
       relevant questions? Please comment on the extent to which the discussions are
       appropriately focused in addressing the questions.

       The panel was  divided regarding the level of detail present in the current Policy
       Assessment. Some thought the scope of the present Policy Assessment was appropriate. It
       may be necessary to include the considerable detail which characterizes the current
       document. However, a substantial minority felt that an ideal policy assessment should be
       far shorter and more focused.  Key staff and the administrator can turn to the REA and the
       ISA as necessary. The Policy Assessment could be one third its current length and be a
       brief summary of the evidence and how that relates to alternative PM  levels. Moreover, a
       concise description of how the form of the standard is important would also be useful.

2. Technical Content
   a.   Do the discussions accurately reflect and clearly communicate the currently available
       health and welfare effects evidence, including important uncertainties, as
       characterized in the final ISA?

    The panel welcomes the expansion of material relating to health and welfare effects.  The range
    of proven and possible health effects has been expanded since previous PM reviews. We
    welcome a discussion of how PM affects visibility and is involved in climate change.  We
                                           12

-------
believe these topics are important, and we applaud the authors' attempt to concisely inform the
Administrator regarding the importance of these relevant outcomes.

 b.  Do the discussions accurately reflect and clearly communicate policy-relevant
    information from the human health risk and visibility assessments, including
    important uncertainties, as characterized in the second draft REAs?

The panel believes that the current Policy Assessment discussion does clearly communicate
information relating to human health risks and visibility. There is a good discussion of
uncertainties and how these  uncertainties impact quantitative risk assessments.
As noted earlier, some panel members thought that the Policy Assessment would be more
effective if it were shorter. More of the evidence contained in it could be summarized and
references to the REA or ISA - as appropriate - could be included to compensate for some of
the  missing or truncated detail.
                                        13

-------
Comments from Dr. Wayne Cascio

Question 4: Adequacy of current suite of PM2.s standards (section 2.2): What are the Panel's
views on the following?
   a.  The appropriateness and characterization of the sets of long- and short-term PM2.s
      exposure studies highlighted? (Section 2.2.1). Are there specific studies that should be
      given more or less emphasis? Are there additional studies that should be focused on?
      Does the Panel generally support placing greater weight on multi-city vs. single-city
      short-term exposure studies? Does the Panel agree with the characterizations of
      PM2.5 air quality at which associations have been observed?

Comment: EPA staff in authoring the Policy Assessment has achieved an appropriate and balanced
emphasis between the strengths and weaknesses of the studies, and weighing their relative
contribution to the assessment. There are no additional studies that need to be introduced from the
ISA in support of the Policy Assessment. Multi-city studies do merit greater emphasis (or weight)
for the purpose of establishing the association between PM2 5 exposure and health effects. Table 2-
1 provides an informative overview summary of the causality determinations for PM2 5. While the
present ISA does not specifically address long-term neurocognitive effects of Pfohs, it is likely to
be a factor to be evaluated in the next ISA. As such it might be worthwhile placing such an
outcome in the Table.  One could use "Central Nervous System Effects" as is already present in the
Short-term category and indicating that the causal determination is "inadequate". Figure 2-1
provides an informative summary of the effect estimates and air quality distributions for long-term
exposure studies. An added designation that would be useful would be to identify each study as
single- or multi-city study and note the number  of cities included in the multi-city studies.  The
value of multi-city studies is emphasized in the  ISA, RA and Policy Assessment therefore it would
be useful to see the specific number. Have the region of the country present in the table as has
already been done is useful.

   b. The focus on persons with lower socioeconomic status as a susceptible population in
      addition to consideration of newly available evidence for susceptible populations
      identified in previous  reviews (e.g., children, older adults, persons with pre-existing
      cardiac and respiratory disease)? (Section 2.2.1)

Comment: The identification of individuals having a lower socioeconomic status as a susceptible
population is appropriate. Two noticeable omissions are the absence of a description of the
association with certain gene polymorphisms associated with enzymes modulating oxidant stress,
and diabetes.  While more data needs to be obtained to better establish the magnitude of these risks
and define the mechanism accounting for the increased risks, there appears to sufficient evidence
(Chapters 4, 6 and 8) in the ISA to comment in the Policy Assessment.

   c. The preliminary staff conclusion that the estimated risks remaining upon simulation
       of just meeting the current suite of standards can reasonably be judged to be
       important from a public health perspective? (Section 2.2.2)

Comment: The conclusion that simulations of just meeting the current suite of standards
demonstrates persist risk to the health of the public is justified and important. Yet, it must be noted
                                           14

-------
that these estimates of excess risk are extrapolated to PIVh 5 levels for which there are only limited
real world data and higher degrees of uncertainly. Consequently, the rationale for further adjusting
the PM standard rests with assuring a margin of safety.

   d.  The relative roles of the annual and 24-hour standards in providing public health
       protection as informed by the quantitative risk assessment, specifically in focusing on
       simulation of estimated risks remaining upon just meeting the current suite of
       standards? (Section 2.2.2)

Comment:  Comparisons of short-term and long-term cardiovascular risk based are challenging
based on Tables 2-2 and 2-3 that describe different health effects end-points, incidence of ischemic
heart disease mortality in Table 2-2 and total cardiovascular mortality in Table 2-3.  Providing total
cardiovascular mortality in both tables or in the text would allow a more direct comparison of the
relative roles of the annual and 24-hour standard.

   e.  Does the current suite of PM2.s standards provide adequate protection of public
       health? (Section 2.2.3)  EPA staff in the Policy Assessment has provided a  logical
       approach and justification of the conclusion that the current PM^s standard is
       insufficient to protect health and should be revised.

Comment: Taken in total the data presented in the ISA and summarized in the Policy Assessment
provide strong justification for reconsidering the current PlVhs standard. Convincing evidence
continues to be present indicating that serious cardiovascular and respiratory health effects and
adverse heath outcomes  occur at the level of the current standard, and that a lower standard will
result in greater protection of health.

Question 9: Adequacy of the current 24-hour PM standard (section 3.2): What are the
Panel's views on the following?
    a. The preliminary staff conclusion that the available scientific evidence supports
       maintaining a standard that provides protection against health effects associated  with
       exposure to thoracic coarse particles?

Comment:  While data from epidemiological studies are suggestive of short-term cardiovascular,
respiratory and mortality effects, at the present time significant uncertainties exist with respect to
the health impact of coarse PM (PMjo-2 5). Limitations exist with respect to exposure assessment,
and non-human studies addressing biological plausibility. Since the last review new information
has generally supported an association but an important multi-city Canadian study did not find
association between PM 10-25 statistically significant, thus emphasizing the persistent uncertainty of
the impact of this PM fraction. On page 3-12 (L. 26-28) it is stated that "Considered as a whole, the
ISA notes that epidemiologic  studies that have evaluated thoracic coarse particles have reported
consistent, positive associations between PM 10-2sand mortality (US EPA, 2009a, section 6.5.2.3)".
It should further be stated,"..., yet, in many cases the associations did not achieve statistical
significance."

    b. The various approaches described for translating the evidence and related
       uncertainties into the basis for conclusions on the adequacy of the current
                                            15

-------
       standard?

Comment: Evidence appearing since the last review shows that mortality is increased even in some
cities where the current PMio standard is met. This suggests that the PMio standard should be
reconsidered. Issues related to the uncertainty of air quality estimates in the epidemiological
studies reporting the health effects of PMio-2 5 remain, as does the limitation of few studies
addressing confounding by co-pollutants. For these reasons in contrast to PM2 5 significant
uncertainties persist regarding association between PMio-25and overall mortality as well as cause
specific mortality.

    c.  The adequacy of the public health protection afforded by the current PMIO standard
       against exposures to thoracic coarse particles?

Comment: Since the last review additional epidemiological studies support the association between
PMio-2 s and cardiovascular and respiratory morbidity and mortality. While the precision of the
associations between health effects for the association with PMio-2 5 exposure does not match that
for PM2 5 there is sufficient evidence to consider an adjustment of the PMio standard.  Because
evidence is presented in the ISA that shows that health effects have been observed in areas that
meet current standards there is reason to expect that a downward adjustment of the standard will
benefit the health of the public. Importantly, these associations at lower than the current standard
levels are preserved even in two-pollutant models. And there is limited but supportive evidence
from human exposure studies and non-human toxicological studies. While the available evidence
continues to be less robust than the PM2 5 by the limitations noted, one can argue that the standard
should be lowered to provide a greater margin of protection.
                                            16

-------
Comments from Dr. Chris Frey

Initial comments on the draft Policy Assessment

       It is unusual that C-R estimates based on the low end of a plausible range of values would
       be used.  In public health applications, it is more typical to user upper bound estimates of
       dose-response relationships in order to be protective of public health. Thus, one could
       argue that EPA has biased the risk characterization to be very low and, therefore, is on a
       path that will not be adequately protective of public health.
       P 2-73:  The draft Policy Assessment seems to say that the form of the standard will not
       be assessed until the 2nd draft of the Policy Assessment. This would seem to effectively
       disengage CASAC from being able to comment on a fully specified alternative standard
       that takes into account indicator, level, averaging time, and form.  All of these must be
       considered jointly in order to assess a regulatory alternative.
       The Policy Assessment is very very long.  While it may be useful to provide background
       for people who have not read the ISA or the REA, the text is simply too verbose. Avoid
       being so tentative with statements. Avoid passive phrases such as "we note that...".  The
       text will have more impact over all if things are said just once and not repeated. While it
       is useful to summarize material from the ISA and REA, there does not need to be an
       extensive verbose discussion of these. The interested reader can access these documents
       for more detail.
       the summary paragraphs at the end of each major section are very useful. These
       summary paragraphs should be the basis of creating an executive summary or synthesis
       chapter, that would enable a reader to get the main points without having to labor through
       hundreds of pages of material.
       The objective in  evaluating alternative standards should not be risk reduction but rather
       seeking a standard that is protective of public health with an adequate margin of safety.
       (e.g., page 2-87)
   -   P 2-97 - should also take into account that sensitivity analysis of the C-R functions
       generally lead to higher effects estimates that the core estimates.  Furthermore, the
       assessment does  not adequately take into account that the C-R functions are likely to  be
       underestimated for reasons not quantified, such as exposure error.
       Section 2 does not seem to  deal with the number of adverse effects cases associated with
       each alternative standard. This is necessary to consider in order to satisfy the legal
       objective/statutory mandate for the NAAQS.
                                            17

-------
Comments from Dr. David Grantz

Overall, this Policy Assessment is well laid out and presented. It moves smoothly and clearly
through the relevant issues and provides abundant information on each. There is some repetitive
introductory text which might be condensed. The amount of information presented is large, and
may be excessive since it has been already carried through the ISA and one or both of the Risk
Assessments. What needs to be highlighted in the Policy Assessment are staff conclusions, their
implications, and the type of information on which they are based.

Question 14. Adequacy. The Policy Assessment does a nice job of demonstrating that, for a
visibility endpoint, the current standard is not protective of public welfare. In the text describing
Fig. 4-3 it should be noted that Pheonix as well as Tacoma would meet the candidate level 98% of
the time.

Question 15. Indicator. The Policy Assessment adequately presents the advantages of a visibility-
based indicator. It also, however, demonstrates that a mass-based PM 2.5 indicator would work
almost as well (subject to modest additional spatial variability in achieved visibility).

There are advantages to retaining a mass-based indicator for the Secondary Standard:

    1.  Choosing a visibility-based indicator would be analogous to choosing one of the several
       health endpoints as the indicator for the Primary Standard. In either case, this
       inappropriately diminishes the importance of other endpoints, whether or not there is
       sufficient evidence to use them in the current review cycle.
    2.  Mass is more conducive to future development of concentration-effects relationships for
       other endpoints. Research should be advocated in the Policy Assessment to  develop
       appropriate deposition velocities for PM to link atmospheric concentrations to loads on
       ecosystems and materials, and to develop relationships between loading and diverse
       effects. A similar advantage of the mass approach holds for effects on cloud nucleation
       (climate) and radiation scattering (penetration into plant canopies).
    3.  Nothing is sacrificed with respect to visibility protection if the level of the mass-based
       standard  is derived from known physical relationships between visibility and PM2.5 mass
       (accepting some variability) and the human preference data (also with considerable
       regional variability).

A goal to be advocated in the Policy Assessment is movement towards a chemically speciated,
mass-based standard. This appears to be equally true for the Primary and Secondary standards. The
Policy Assessment should endorse further research on relationships between chemically speciated
mass and various endpoints, and further proliferation of monitoring networks to support such a
standard in the future. There is insufficient information to consider such an indicator during this
review cycle.

Because the coarse PM fraction contributes significantly to visibility impairment in some locations,
it is appropriate to consider a visibility-based standard  for coarse PM mass.
                                            18

-------
Question 16.  Averaging time.  A visibility-based criterion is reasonably based on a worst one
hour averaging time, since the Policy Assessment demonstrates that visibility is essentially
instantaneously perceived.

It is not appropriate to apply the concept of a "sensitive sub-population" to individuals who only go
outside occasionally (page 4-15; line 22). It is equally reasonable to postulate that people who
spend their whole day outside are exposed to repeated instantaneous updates regarding poor
visibility, and that they are the true "sensitive sub-population".

Any standard based on a one hour averaging time will require a new monitoring protocol. This
could be for visibility, or it could be used to push towards a real-time (continuous) mass
measurement network. The latter would be useful for more purposes, has been advocated
previously by CASAC, and therefore should be given serious consideration at this time.

While visibility and corresponding mass levels could be assessed during daylight hours, the
standard could apply at all hours if a mass indicator were selected. This could have potential
benefits, for example  in control of "dumping" of PM emissions at night when visibility is not
monitored.

Question 17.  Levels and Forms. The options are well laid out, in the Policy Assessment and
Appendix A. These are presented relative to the Candidate Protection Levels, based on the
visibility preference studies. The underlying data regarding visibility preferences are somewhat
weak, as appropriately noted in the Policy Assessment, and further research will be required to
determine whether there is real regional variability or these relationships can be collapsed into a
population mean.

It is appropriate to consider a <100% form, allowing some hours of exceedance. This is more to
stabilize the statistic than, as stated in the current draft Policy Assessment, because this is
appropriate for an aesthetic standard. A relatively high percentile form, with a correspondingly
lower level, would support the significance of the chosen level by seeking to minimize its
exceedance.

It is appropriate to consider a screen at 90% RH, to remove non-PM associated visibility
impairments.

Question 18.  Climate. The Policy Assessment appropriately concludes that there is insufficient
information to proceed toward a  national standard for PM effects on climate. However, the causal
relationship is established, and further research on a  regional basis is urgently required. This should
be advocated in the Policy Assessment.

A number of putative difficulties in relating PM to climate effects are laid out in sections 5.2.3 and
5.2.4 (pages 5-11 to 5-13) and again in section 5.3.4 (page 5-24). However, these apply to all other
potential effects of PM, and  it is not appropriate to emphasize them here as if they were uniquely
related to climate. For example, other PM-effects relationships would be improved by using
chemically-speciated  mass measurements; other relationships are expected to vary locally and
regionally; others have offsetting positive and negative effects (warming vs.cooling, radiation
                                             19

-------
attenuation vs. radiation scattering); others are confounded by effects at a distance from sources, by
the difficulty of distinguishing anthropogenic from natural constituents (PRB issues should be
explicitly mentioned here (section 5.2.4); by spatial and temporal heterogeneity; etc. In addition,
many of the caveats in section 5.3.4 are equally true of N, S and ozone, but have not prevented
those effects from being used in evaluation of the NAAQS for those contaminants.

Question 19. Ecological Effects. The Policy Assessment appropriately concludes that there is
insufficient information to proceed toward a national standard for PM effects on ecosystem
services. However, causal relationships are known, and  further research should be an Agency
priority. This should be advocated in the Policy Assessment, incorporating the desirability of
chemically speciated mass determinations.

The framing of the discussion (section 5.3.1) in terms of the end use of the product, taken from the
ozone documents, and in terms of ecosystem services, following the Millennium Assessment, is
very good. Structuring future research in terms of effects on individual ecosystem services may
lead to these effects becoming useful in development of a NAAQS for PM.

Question 20. Materials. The Policy Assessment appropriately concludes that there is insufficient
information to proceed toward a national  standard for PM effects on materials. However, causal
relationships are well established. Further research should be advocated in the Policy Assessment,
though other potential welfare effects may be more significant. Soiling effects on materials may,
however, be the most straightforward to relate to atmospheric PM, since the linkage is only
through deposition velocity for specific components.

Two editorial comments
    1.  On page 5-16, line 36, it should be added that plant genotype also determines uptake of
       metals and organics.
   2.  On page 5-19, line 32,1 assume this is "peroxisomal" rather than "peroxidomal".?
                                            20

-------
Comments from Dr. Joseph Helble

Charge Question 3:  Approach (section 2.1.3)

The discussion in section 2.1.3 is very general, and not particularly well-addressed by charge
question 3.  The text in this section, while clearly written, is longer than necessary for making the
general points that 24 hour and annual standards should be considered together, since controls
directed towards one will affect the other, but that the Agency feels that "more consistent risk
reductions are likely to be accomplished through strategies aimed at meeting an annual standard"
(p 2-11).  This is clear. The balance of the section (p. 2-12 and 2-13) is an overview of the
approach taken in the rest of the document to developing potential standards. My recollection is
that this was addressed adequately in the RA; a one paragraph summary with a reference back to
the RA would be better here.

Small point, but one that arose earlier and persists: the scientific rationale for focusing on studies
conducted primarily in the U.S. and Canada (p 2-10, line 9) is not clear, and not addressed in the
Policy Assessment.

Charge Question 7:  Forms (section 2.3.3)

The question addresses the appropriateness of allowing spatial averaging in determining
compliance with the annual standard. The discussion both in this section and elsewhere in the
Policy Assessment indicates that EPA is moving away from the use of spatial averaging in this
context.

The utility of spatial averaging depends upon the heterogeneity or "spikiness" of the PM
concentration distribution relative to the local population distribution.  It will therefore strongly
depend upon monitor placement, with respect to population density, source location, and local
airflow patterns. EPA notes in the Policy Assessment that it will  address these questions through
"additional analysis" (section 2.3.4.1) that will be addressed in the 2nd draft Policy Assessment.
Some description and discussion of the planned "additional analysis" would be helpful, as it was
not addressed here, nor in the RA.

Further, this is one area where detailed modeling of PM concentration distributions could be
helpful in assessing local heterogeneity of PM concentrations, both in terms of assessing
population exposure and as a guide in future monitor siting decisions.

Additional  Comment

In the footnote on p 1-5, is this backwards? The wording is not clear, but  particles with dp > dso
are those captured with higher efficiency than those at dso.  Smaller particles (particles with dp <
    are collected with lower efficiency.
                                            21

-------
Comments from Dr. Rogene Henderson

Charge Questions 1 and 2, Organizational structure and technical content: I found the
organizational structure to be helpful in following the logic of the conclusions of the staff. The
technical content was excellent. I especially liked the summary paragraphs. It would be helpful to
reduce the amount of verbiage by referring back to the ISA or the RA more often, rather than
repeating the content of those documents.

Charge Question 8. Levels: Based on the lack of ability to detect a threshold level for adverse
health effects in epidemiology studies, and the need to protect sensitive sub populations with a
margin of safety, we  are beginning to arrive at risk estimates that drive our estimates of health-
protective levels down almost to background levels. I think this must be discussed in the Policy
Assessment document.  As a science advisory body, it is not our place to make policy
recommendations as  to where to draw the line. However, the Agency should discuss the criteria
they will use to determine where to draw the line. This is their policy choice,  but it would be wise
to have their choice informed by discussions from CASAC..

Charge Question 9.  Coarse Particles
a. I agree with the staff conclusion  that the available scientific evidence supports maintaining
standard for coarse particles.
b. I agree with the staff approaches  to considering the adequacy of the current PM 10 standard.

Charge Question 8:  Levels for Coarse Particle Standards
1 agree with the described approach.
                                            22

-------
Comments from Dr. Mort Lippmann

Major Comment and  Concern:
On the basis of the scientific evidence and risk assessment, the OAQPS Staff has recommended
that a revised PM2 5 annual NAAQS in the range of 13 - 12 wg/m3 is needed to provide better
public health protection, and furthermore that consideration should also be given to a PlVfes annual
NAAQS in the range of 11 - 10 t/g/m3.1 find that these options were well considered, and
scientifically justifiable on the basis of the analyses cited in the PM Policy Assessment draft. Their
implementation would  be consistent with the Administrator's obligations under the Clean Air Act
with respect to setting NAAQS that would protect the public health.
However, I also believe that 2nd draft of the PM Policy Assessment needs to more explicitly address
the policy implications of the selection of a concentration limit that, even at 13 t/g/m is
considerably more stringent than the current limit of 15 ug/m3. Promulgation of this concentration
limit would require: a broad array of new controls on emissions of PMasand its precursors; would
result in much more non-attainment over long periods of time; and would only be achievable in a
time frame measured in years in most urban areas in the eastern and Midwestern States and
California. For the even lower alternate concentration limits (12 -10 ug/m3), the requisite controls
on emissions of fossil fuel combustion effluents would require huge capital investments in new
power plants and/or in alternate energy sources. These considerations lead me to conclude it is now
time to begin the transition to alternate indicators that better target the health-risks  that have been
associated with PM2 5 mass concentration. I come to this conclusion on the basis that:
    1.  There is a rapidly growing body of scientific evidence, much  of it cited in the PM ISA, that
       indicates that PM2 s-associated cardiovascular mortality and morbidity risk coefficients
       vary widely across the US, as  does the PM2 5 chemical composition.
   2.  The establishment and maintenance of a chemical speciation network (CSN) over the past
       decade has made it possible for well-targeted epidemiological studies to demonstrate that
       the urban areas with the highest health-based risks are those most heavily exposed to PMzs
       components originating from the combustion of coal and residual oil, as well as the
       pollutants associated  with motor vehicle traffic. Thus, emission controls focused on
       reductions of effluents from these source categories are highly likely to reduce PM25-
       associated health risks more than emission reductions from other source categories or for
       PM2 5 as a whole.
   3.  Unfortunately, the current limitations of the CSN, in terms of temporal and spatial
       coverage, have  greatly limited the opportunities for epidemiologists to establish
       associations between PM2 s-components and source categories and health effects. Having
       data for only every third or sixth day severely limits the ability to study acute responses and
       their lag structure, while having only one or a few sampling sites in an urban area severely
       limits studies of both the acute and chronic effects of components and source-related
       mixtures that are not uniformly distributed an urban area.
   4.  Moving toward the designation of better-targeted indicators for primary PM NAAQS is not
       a new strategy.  In 1987, PM penetrating into the thorax (PM 10) replaced total suspended
       paniculate matter (TSP), and in 1997, fine PM (PM2s), which differs substantially from the
       coarse particle fraction within PMio, and which penetrates in the gas exchange region of the
       lung, was adopted to supplement the protection provided by PMio.
   5.  While the scientific evidence to support a designation of one or more new indicators of
       PM2 s-related health risks is inadequate for this PM NAAQS  review round, EPA needs to
                                            23

-------
       initiate actions to ensure that an adequate scientific basis will exist within the next 4 to 5
       years for the designation of better and more efficient indices of PM2 s-related health risks
       for the next round of PM NAAQS review.
    6.  The most urgent need for data to support new indicators of health risks is to enhance the
       temporal and spatial coverage of the CSN network for PlVhs, and if possible, for PMio or
       PMio.25, and ultrafine PM as well. An expanded CSN network can be supplemented by
       arranging for the creation and maintenance of an archive for daily FRN Teflon filters that
       can be made available for epidemiological studies focused on the health effects of inhaled
       trace metals and black carbon. The archived filters could support retrospective as well as
       prospective studies.

General Comments:
    1.  The OAQPS Staff have done  an excellent job in summarizing the most relevant findings
       reported in the PM ISA and REA, and translating them into a suitable policy framework to
       facilitate and enable rational and defensible decisions by the Administrator on the
       promulgation of the next suite of primary PM NAAQS.
    2.  The OAQPS Staff warrants commendation for its straightforward and well-reasoned
       interpretations of the health effects evidence and risk estimates for PM2 5, and their
       implications to the choice of the most suitable indicators, forms, levels for further
       consideration, and monitoring site data, as well as the optimal combinations of 24h and
       annual limits for public health protection. They point to the  need for selection of a much
       more stringent annual limit, which will result in exceedances in most US urban areas.
    3.  Chapter 2 of the Policy Assessment document, while comprehensive in terms of the issues
       that needed to be addressed, has not yet been sufficiently edited to remove the
       redundancies. In too many places, there are repetitions of the same words, phrases, and
       whole sentences (e.g., in the introductory paragraphs, discussion of the long-term exposure-
       related evidence and risk estimates, the discussion of the short-term exposure-related
       evidence and risk estimates, and the options for suites of long- and short-term exposure
       limits). These common issues and themes should be presented only once in full, and cited
       or summarized briefly, where they apply, in the later sections.
    4.  The content of this first draft of Chapter 3 on a NAAQS for PMio or PMio-2 5, represents a
       place holder, and contains only boiler-plate wording adapted from the more complete
       Chapter 2 that deals with  PM2 5NAAQS. Critical comments on Chapter 3 will  have to wait
       until we have a chance to see  the second draft.

Specific Comments:
Page Line Comment
2-33 Fig. 2-1 Separate listings of mortality effects from those of other effects
2-25 17 change "fine particles at least as" to "PM2 5 NAAQS as, or more"
2-25 23 insert "the extensive" before  "interindividual"
2-25 24 insert "likely to be" before "at"
2-26 25 change "to" to "for"
2-26 29 insert "of outdoor origin" after "PM"
2-27 18 change "which" to "that"
2-31 14 insert "chemical composition, which varies" after "along"
2-36 17 change "This reduction was" to " Reductions were"
                                           24

-------
2-37 19 The ACS and 6-cities based risk estimates were not similar to and 6-cities or those of the
other cited studies
2-41 16 The entry for the brackets is missing
2-43 32 change "references" to "reference"
2-48 23 change "effected" to "affected"
2-59 7 change "larger particle sizes" to "PM2 5 particle mass"
2-60 33 insert "metals," before "sulfate" for consistency with other text entries
2-74 32 "MCAPS" has not been defined in the text
2-76 30 change "which" to "that"
2-85 35 What city, other than New York, should be cited within the bracket?
2-86 3 correct the 3.2% entry

Question 1: Organizational Structure
    a.  Does the Panel find the introductory and background material, including that
       pertaining to previous reviews of the PM standards and the current review, to be
       clearly communicated and appropriately characterized?
              Yes
    b.  To what extent does the Panel find that the questions posed appropriately reflect the
       important policy-relevant issues for this review?
              Yes, but not all of the policy-relevant issues.
       Are there additional policy-relevant questions that should be addressed?
              Yes. Specifically, the mismatch between associations between PM2smass
       concentrations and adverse health effects, and the public health need to control exposures
       to PM2 5 components that account for the effects. As the PM2 5 level approaches background
       concentrations, the PM2 5 indicator becomes much less defensible.
    c.  What are the Panel's views regarding the level of detail used to address these policy-
       relevant questions? Please comment on the extent to which the discussions are
       appropriately focused in addressing the questions.
              The level of detail is excessive. It should be cut back to provide succinct summaries
of the information extracted from the ISA and REA documents.

Question 2: Technical Content
          a.  Do the discussions accurately reflect and clearly communicate the currently
              available health and welfare effects evidence, including important
              uncertainties, as characterized in the final ISA?
              Yes.
          b.  Do the discussions accurately reflect and clearly communicate policy-relevant
              information from the human health risk and visibility assessments, including
              important uncertainties, as characterized in the second draft REAs?
              Yes.
                                           25

-------
Question 3: Approach (section 2.1.3):  Various approaches are described for translating
different air quality metrics from epidemiological studies into the basis for reaching
preliminary staff conclusions on the adequacy of the current PM2.s standards and on
alternative standards for consideration. What are the Panel's views regarding the
appropriate weight to place on these various approaches; should specific approaches be
afforded more or less weight? Are there additional approaches that should be considered?

   The selection of both evidence- and risk-based approaches is sound, as is the consideration of
   the utility of both annual and 24 hour concentration limits, and of their combinations in
   providing public health protection.

Question 4: Adequacy of current suite of PMi.s standards (section 2.2): What are the
Panel's views on the following:
          a. The appropriateness and characterization of the sets of long- and short-term
             PM2.5 exposure studies highlighted? (section 2.2.1)
             The highlighted data sets are appropriate selections.

       •   Are there specific studies that should be given more or less emphasis?
             The Krewski et al. (2009) study is overemphasized in relation to the other
          highlighted studies of annual mortality in that it relies on a higher SES population than
          the others and thereby underestimates responses for the US population as a whole.

       •   Are there additional studies that should be focused on?
             The 6-cities study, having a more representative population, should be given greater
          emphasis, at least for Eastern and  Midwestern populations, where the coefficients of
          response are greater than for Western populations. If the question refers to additional
          studies to be highlighted, the answer is no.

       •   Does the Panel generally support placing greater weight on multi-city vs. single-
          city short-term exposure studies?
             Yes.

       •   Does the Panel agree with the characterizations of PM2.s air quality at which
          associations have been observed?
             Yes, but the selection of the lower bound of the inter-quartile range may not be
          sufficiently health protective. Further consideration should be given to using the 10th
          percentile as a level at which associations have been observed.
          b. The focus on persons with lower socioeconomic status as a susceptible
             population in addition to consideration of newly available evidence for
             susceptible populations identified in previous reviews (e.g., children, older
             adults, persons with pre-existing cardiac and respiratory disease)? (section
             2.2.1)
             This was appropriate.
                                           26

-------
           c.  The preliminary staff conclusion that the estimated risks remaining upon
              simulation of just meeting the current suite of standards can reasonably be
              judged to be important from a public health perspective? (section 2.2.2)
              This was appropriate.
           d.  The relative roles of the annual and 24-hour standards in providing public
              health protection as informed by the quantitative risk assessment, specifically
              in focusing on simulation of estimated risks remaining upon just meeting the
              current suite of standards? (section 2.2.2)
              This was appropriate.
           e.  The integration of evidence-based and risk-based considerations to inform the
              preliminary staff conclusion that the available information clearly calls into
              question the adequacy of the current suite of PM2.5 standards and provides
              strong support for giving consideration to revising the current standards to
              provide increased public health protection?  (section 2.2.3)
              This was appropriate.

Question 5: Indicator (section 2.3.1): What are the Panel's views on the following:
       a.   The preliminary staff conclusion that it is reasonable to retain PM2.s as an
            indicator for fine particles?
            It was reasonable only because there was insufficient peer-reviewed literature to
            support any other indicator at this time. However, it is essential that EPA develop the
            database, over the next four years, to support the identification of one or more
            indicators that are better suited to represent the risks of exposure to PM2 5, especially
            of its components that, based on emerging evidence, represent the greatest risks to
            public health.
       b.   The preliminary staff conclusion that the currently available information is too
            limited to support a distinct PM standard for ultrafine particles?
            This conclusion was unavoidable in that there was insufficient peer-reviewed
            literature to support a UFP indicator at this time.
       c.   The preliminary staff conclusion that the currently available information is not
            sufficient to support consideration of a separate indicator for a specific PM2.s
            component or for the mix of fine particles from any specific source category, and
            that data are too limited  to consider eliminating any component or source
            category from the mix of fine particles included in the PM^s mass-based
            indicator?
           See answer to 5.a. above.

Question 6: Averaging Times (section 2.3.2):
   What are the Panel's views on the following:
    a.  The preliminary staff conclusion that it is reasonable to retain annual and 24- hour
       averaging times?
                                           27

-------
            It was appropriate.
   b.  The preliminary staff conclusions that the currently available evidence is too limited
      to support additional averaging times to address sub-daily exposures or seasonal
      exposures?
            It was appropriate.

Question 7:  Forms (section 2.3.3):
  What are the Panel's views on the utility of additional analyses to inform consideration of
  eliminating provisions included in the current form of the annual standard that allow for
  spatial averaging across monitors, specifically with regard to the potential for
  disproportionate impacts on susceptible populations with lower socioeconomic status?
            Such analyses are warranted.

Question 8:  Levels (sections 2.3.4, 2.3.5, and 2.3.6):
  What are the Panel's views on the following:
   a.  The preliminary staff conclusions regarding alternative standard levels that are
      appropriate for consideration, and the rationale upon which those conclusions are
      based?
            They were appropriate.
   b.  The insights that can be gained  from the quantitative risk assessment to inform our
       understanding of the roles that the annual and 24-hour standards play in providing
       public health protection, specifically in focusing on simulations of estimated risks
       remaining upon just meeting alternative suites of standards? (sections 2.3.4.2 and
       2.3.5.2)
            The insights gained were useful and informative.
   c.  The policy implications of the uncertainties associated with estimating risks, including
      potential underestimation of risk, in reaching conclusions regarding standards that
      would provide public health protection with an adequate margin of safety?
            Uncertainties need to be defined as well as current information allows. On the other
      hand, using uncertainty as an excuse for choosing a minimally stringent level is not
      consistent with having an adequate margin of safety.
   d.  A policy approach for identifying a suite of standard levels in which the annual
       standard would be the "generally controlling" standard to provide protection for
       both long- and short-term PM;.; exposures, in conjunction with a 24-hour standard
       set to provide supplemental protection against days with high  peak concentrations
       associated with localized "hotspots" and risk arising from seasonal emission that
       might not be controlled by a national annual standard? (section 2.3.6)
            This approach is most appropriate.
       e.    Additional considerations that  could inform our conclusions related to
            alternative suites of fine particle standards?
            The alternatives that were selected were very  reasonable choices.
                                         28

-------
Question 9: Adequacy of the current 24-hour PMio standard (Sect. 3.2):
   What are the Panel's views on the following:
   a. The preliminary staff conclusion that the available scientific evidence supports
      maintaining a standard that provides protection against health effects associated with
      exposure to thoracic coarse particles?
             This preliminary conclusion was warranted.
   b. The various approaches described for translating the evidence and related
      uncertainties into the basis for conclusions on the adequacy of the current PMio
      standard?
             The various approaches that were described were appropriate.
   c. The adequacy of the public health protection afforded by the current PMio standard
      against exposures to thoracic coarse particles?
             The adequacy is uncertain in view of the very limited database. While it is
      reasonable to maintain the current PMio NAAQS in this round, it is essential that a suitable
      database for PMio-2 5 concentrations be developed over the next four years so that a more
      satisfactory judgment can be made in the next review cycle,

Question 10: Indicator (sections 3.3.1,3.3.2, and 3.3.3):
   What are the Panel's views on the following:
      a.  The preliminary staff conclusion that it is appropriate to maintain a standard that
         provides protection against all thoracic coarse particles, regardless of their source
         of origin or composition?
             This conclusion is justifiable at this time in view of the absence of data on the
      influence of coarse thoracic PM composition. While it is reasonable to not consider
      chemical  composition in the current round of review of a coarse thoracic NAAQS, it is
      essential that a suitable database for PMio-2 5 concentrations and composition be developed
      over the next four years so that a more satisfactory judgment can be made in the next
      review cycle.

      b.  The appropriateness of maintaining an indicator that allows lower PM 10-2.5
         concentrations in urban areas than non-urban areas?
             On the basis that PMis is higher in urban areas than rural areas, it will constitute a
      greater proportion of PMio in urban areas. Therefore, one can expect that for the same PMio
      concentration, that the concentration of PMio-2swill be lower in urban areas. Thus, it is
      appropriate to maintain an indicator that allows lower PMio-2 5 concentrations in urban areas
      than non-urban areas.
      c.  The appropriateness of either a PMio or PM 10-2.5 indicator for a standard meant to
         protect against exposures to thoracic coarse partkles?
             While a PMio-2 s indicator for a standard meant to protect against exposures to
      thoracic coarse particles would be preferable, a PMio indicator would be an acceptable
      substitute in this round of review.
                                          29

-------
Question 11: Averaging Time (section 3.3.4):
  What are the Panel's views on the preliminary staff conclusions that the currently
  available evidence supports maintaining a 24-hour standard to protect against health
  effects associated with short-term exposures to thoracic coarse particles, and does not
  support consideration of a long-term thoracic coarse particle standard?
             We agree.

Question 12: Form (section 3.3.6):
  What are the Panel's views on the preliminary staff conclusion that it is appropriate to
  consider a 98th percentile form for a revised 24-hour standard meant to protect against
  short-term exposures to thoracic coarse particles?
             We agree.

Question 13: Level (section 3.3.7): What are the Panel's views on the appropriateness of
considering 98th percentile, 24-hour PMioor PMio-2.s concentrations in epidemiological
study locations for identifying a range of potential alternative standard levels for
consideration, recognizing the more limited air quality information available for PMio-2.5?
             We consider it to be appropriate.
                                         30

-------
Comments from Dr. Phil Hopke

 Question 5: Indicator (section 2.3.1):  What are the Panel's views on the following:
    a.  The preliminary staff conclusion that it is reasonable to retain PM2.sas an indicator
       for fine particles?
    b.  The preliminary staff conclusion that the currently available information is too
       limited to support a distinct PM standard for ultrafine particles?
    c.  The preliminary staff conclusion that the currently available information is not
       sufficient to support consideration of a separate indicator for a specific PIVh.s
       component or for the mix of fine particles from any specific source category, and that
       data are too limited to consider eliminating any component or source category from
       the mix of fine particles included in the PMis mass-based indicator?
These all represent self-fulfilling prophecies. Given'the very limited number of measurements of
PM1 or UFP, there is clearly no basis for extensive epidemiology and thus, no basis for an
indicator.  If we drop PM 10 and move to a coarse particle indicator as the companion standard to
PM2.5, it would really make more sense based on the nature of the atmospheric aerosol and the
sources that would  need to be controlled to provide attainment of one or both sized standards to
switch from 2.5 u,m to 1 urn (or possibly 1.5 urn) as a better separation point and eliminate most of
the tail from coarse particle sources penetrating into the fine particle indicator.
In spite of significant toxicological evidence of the role of UFP through the work of two PM
centers, the Agency has been unwilling to deploy adequate UFP monitors to provide the data
needed for epidemiology. Thus, although it seems reasonable to assume some level of toxicity
associated with these particles beyond that which is directly related to their mass, there is no basis
for setting an effective number based standard. At the moment, there seems to be no  indication that
monitoring will be  initiated in time for the data and epidemiological results to be available in the
next round of review.  There is clearly a major disconnect between the laboratory results and
activities in the field that would permit an informed basis for deciding if an UFP standard were
needed.
We measure limited composition data, but with varying relationships to observed health effects.
There may be some "toxic" metals (Ni) in the PM, but we have no significant data on organic
components and particularly on reactive organic components. There will need to be  an expanded,
but targeted monitoring based on clearly defined hypotheses to develop initial composition - health
effects links to begin to serious consider component specific
A bigger issue with respect to the indicator is the measurement method since PM mass is
operationally defined. PM mass is defined by the sampling system, filter, and post-sample
treatment process.  This results in a precise result with totally unknown accuracy.  We now have
continuous monitors that can provide a better measure of the actual airborne particle  mass. These
measurements are almost uniformly higher than the corresponding FRM mass. It would be useful
to consider changing the FRM to a more direct, time resolved system that better reflects the actual
airborne mass concentrations.
                                            31

-------
 Question 6: Averaging Times (section 2.3.2):  What are the Panel's views on the following:
a.  The preliminary staff conclusion that it is reasonable to retain annual and 24- hour
   averaging times?
b.  The preliminary staff conclusions that the currently available evidence is too limited to
   support additional averaging times to address sub-daily exposures or seasonal exposures?
Again, we have only started to collect a reasonable amount of more highly time resolved particle
mass data. However, there may not be sufficiently well time marked health data to begin to look at
whether shorter, higher exposures (peaks) are more important than the 24 hour average value that
are now available for health effects models.  There are a few hints of short, high exposures
inducing acute cardiac effects, but clearly we need far more data and analyses to permit the
exploration of time intervals less than 24 hours. Thus, for the moment, we are stuck, but there
should be more emphasis on using continuous monitors and not forcing them to reproduce the
known negative, bias in mass measurements inherent in the FRM values.

Question 7: Forms (section  2.3.3): What are the Panel's views on the utility of additional
analyses to inform consideration of eliminating provisions included in the current form of the
annual standard that allow for spatial averaging across monitors, specifically with regard to
the potential for disproportionate impacts on susceptible populations with lower
socioeconomic status?

Monitors are to be placed to provide population exposure rather than exploring "hot spots." Thus,
it does not make sense to average values across an area since that provides an uneven level of
protection to the  populations that live in high versus low areas.

There should be consideration of looking at three years data together rather than on a year by year
basis. The idea of averaging the 98th percentile values determined for each of the three year period
does not make a lot of statistical sense. We should be looking at the complete distribution of
values measured in that three  year period (assuming an adequate number of samples in each
period) to determine the 98th percentile value.

Question 8: Levels (sections 2.3.4,2.3.5, and 2.3.6):
   What are the Panel's views on the following:
a.  The preliminary staff conclusions regarding alternative standard levels that are
   appropriate for consideration, and the rationale upon which those conclusions  are based?
b.  The insights that can be  gained from the quantitative risk assessment to inform our
   understanding of the roles that the annual and 24-hour standards play in providing
   public health protection, specifically in focusing on simulations of estimated risks
   remaining upon just meeting alternative suites of standards? (sections 2.3.4.2 and 2.3.5.2)
c.  The policy implications of the uncertainties associated with estimating risks, including
   potential underestimation  of risk, in reaching conclusions regarding standards that
   would provide public health protection with an adequate margin of safety?
d.  A policy approach for identifying a suite of standard levels in which  the annual standard
   would be the "generally  controlling" standard to provide protection for both long- and
                                           32

-------
e. Additional considerations that could inform our conclusions related to alternative suites
   of fine particle standards?
As noted above, the level cannot be separated from the measurement scheme since PM mass is
NOT a fundamental property of the aerosol as defined by the standard. The PM mass would be a
fundamental property IF it included the particle-bound water. Thus, we have an operational
definition of "mass" that is known to have a significant negative bias resulting from the loss of
semivolatile species (nitrate and SVOC). It would be better to make it clear that there is such an
interdependency and move toward  more accurate measurement technologies.
I find the evidence adequate to lower the levels somewhat (13 to 12 ug/m3 and 35 to 30 ug/m3) but
I do not believe there  is a strong basis from going below these values. Again an additional
measure of protection would be afforded by moving to monitoring methods that do not have the
known negative bias inherent in the current FRM approach.

Question 9: Adequacy of the current 24-hour PMio standard (Sect. 3.2):
   What are the Panel's views on the following:
a. The preliminary staff conclusion that the available scientific evidence supports
   maintaining a standard that provides protection against health effects associated with
   exposure to thoracic coarse particles?
b. The various approaches described for translating the evidence and related uncertainties
   into the basis for conclusions  on the adequacy of the current PMio standard?
c. The adequacy of the public health protection afforded by the current PMio standard
   against exposures to thoracic  coarse particles?
There is certainly adequate evidence to suggest the need for a standard to protect health against
coarse particles. However, it is clear that PMIO is not adequate given the wide variation in the
PM2.5 to PMIO ratio.

Question 10: Indicator (sections  3.3.1,3.3.2, and 3.3.3):
   What are the Panel's views on the following:
a. The preliminary staff conclusion that it is appropriate to maintain a standard that
   provides protection against all thoracic coarse particles, regardless of their source of
   origin or composition?
b. The appropriateness of maintaining an indicator that allows lower PM 10-2.5
   concentrations in urban areas than non-urban areas?
c. The appropriateness of either a PMio or PM 10-2.5 indicator for a standard meant to
   protect against exposures to thoracic coarse particles?
                                           33

-------
The arguments in favor of retaining PM10 as the indicator are unpersuasive. It makes much more
sense to measure coarse particles directly. There is currently a disparity in the level of protection
afforded by the present standard depending on the nature of the fine to coarse ratio in any given
location. It does not appear to me that there is significantly less information on PM coarse now as
there was on PM2.5 in 1996 when we moved ahead with the new indicator.  Thus, it is time to
move to a direct indicator of coarse PM.  We have as much basis to set a PM coarse standard as we
did to set the initial PM2.5 standard and it is time to move away from a corrupted measure like
PM10.
Question 11: Averaging Time (section 3.3.4):  What are the Panel's views on the preliminary
staff conclusions that the currently available evidence supports maintaining a 24-hour
standard to protect against health effects associated with short-term exposures to thoracic
coarse particles, and does not support consideration of a long-term thoracic coarse particle
standard?

Same arguments as made above regarding short term averaging periods can be made for the coarse
particle indicator.  We have no basis for looking at anything else and we  need to  initiate adequate
measurements of PM coarse on a continuous basis so that we can actually have the data to
ascertain the adequacy of the various possible averaging times.

Question 12: Form (section 3.3.6): What are the Panel's views on the preliminary staff
conclusion that it is appropriate to consider a 98lh percentile form  for a revised 24-hour
standard meant to protect against short-term exposures to thoracic coarse particles?

I suggest the same approach as noted above with respect to the form of the PM2.5 standard.

Question 13: Level (section 3.3.7): What are the Panel's views on the appropriateness  of
considering 98th percentile, 24-hour PMjoor PMio-2.5 concentrations in epidemiological study
locations for identifying a range of potential alternative standard levels for consideration,
recognizing the more limited air quality information available for  PMio-2.s?

There are no specific ranges discussed. At this point, it would make sense to do what was done
with PM2.5 and set a standard based on the smaller end of the range of PM10-PM2.5 differences to
provide a first estimate on the standard. Such a value will be similar to the protection offered by
the PM10 standard, but will apply uniformly.

Question 14: Adequacy (section 4.2): What are the Panel's views  on the preliminary staff
conclusion that the currently available information clearly calls into  question the adequacy of
the current suite of PIVh.sstandards to provide public welfare protection from unacceptable
levels of visibility impairment, primarily in urban areas, and supports consideration of
alternative standards to provide appropriate protection?

I thought there was adequate justification for a visibility secondary standard  in the last round  and
there remains a strong justification for setting such a standard particularly one based directly on
measured light extinction. There is strong public perception that environmental quality has
degraded rather than improved as the data clearly demonstrate. Thus, visibility provides a graphic
illustration of improved air quality and is a useful part of the improvement of the public welfare.
                                           34

-------
Question 15: Indicator (section 4.3.1):
What are the Panel's views on the following:
   a.  The appropriateness of considering PM light extinction and PIVh.s mass as indicators
       for a distinct secondary PM standard for visibility protection, and the relative
       advantages and disadvantages associated with each indicator?
   b.  The appropriateness of considering the contribution of coarse particles with respect
       to a potential PM light extinction indicator?

As indicated in the discussion of PM coarse, it is best to have an indicator that can be directly
measured. Light extinction is a fundamental physical property of the ambient aerosol and thus, it
represents a very appropriate indicator for a secondary visibility standard. The bulk of the
extinction in the vast majority of locations is driven by fine particles (PM.2.5). There are
significant problems with the measurement of the light extinction from all particle sizes. Thus, the
initial secondary standard should use light extinction by PM2.5 particles as the indicator.  It could
be possible to extend the size up to 4 to 5 urn without extensive increasing the measurement
difficulty and that could provide some additional measure of protection to those locations that are
more prone to high coarse particle mass concentrations.

Question 16: Averaging times (section 4.3.2): What are the Panel's views on the following:
a. The preliminary staff conclusion that consideration should be given to a 1-hour averaging
   time?
b. The advantages and disadvantages of focusing on a 1-hour daily maximum or each 1-
   hour average concentration  during all daylight hours?
At this point, there is no really sound scientific basis for picking an averaging time. An hour is a
reasonable starting point.

Question 17: Levels and Forms (section 4.3.3):  What are the Panel's views on the following:
a. The preliminary staff conclusions regarding ranges of options with varying levels and
   forms in combination with different indicators (light extinction and PM2.5 mass-based
   indicators) and averaging times (1-hour daily maximum or each 1-houraverage
   concentration during all daylight hours)?
b. Additional approaches that could help inform our consideration of alternative levels and
   forms?
The range of proposed levels is overly large. There is no point in an alternative standard unless it
is going to be significantly different from the primary standard. Thus, the level should be set
within the lower third of the proposed range.

Question 18: Climate (section 5.2):  What are the Panel's views on the preliminary staff
conclusion that there is insufficient information to base a standard on climate impacts
associated with current ambient PM concentrations?
                                           35

-------
There are both cooling and warming effects of particles as well as the indirect effects related to
cloud formation and brightening. At this point there are far too great uncertainties in the climate
system and the role of particles to propose a standard with a strong scientific foundation.

Question 19: Ecological effects (section 5.3):  What are the Panel's views on the preliminary
staff conclusion that data are insufficient to support establishing a distinct standard for
ambient PM based on ecosystem effects not addressed in the ongoing NOx/SOx secondary
NAAQS review?

In reviewing the Policy Assessment and the ISA, there does not seem to be adequate information
available to conclude that there is need for a separate standard to provide ecosystem protection.
Thus, the staff conclusion appears justified.

Question 20: Material effects (section 5.4): What are the Panel's views on the preliminary
staff conclusion that no new evidence calls into question the adequacy of the protection
afforded by the current suite of PM standards, and that there continues to be support for
retaining an appropriate degree of control for both fine and coarse particles to provide
protection against materials damage and soiling?

Neither the ISA nor the Policy Assessment provides evidence to support the consideration of an
alternative or new secondary standard to protect against materials degradation.

Other Comments
On page 4-11, it says "The amount of light extinction contributed by PM depends on the particle
size distribution and composition, as well as its concentration." However, particle shape also
matters. Not all particles are spherical.

Please do not use "particulate" when you mean "particle"
                                           36

-------
Comments from Dr. Robert Phalen

Regarding sections 2.3.1 (Indicator) and 2.3.2 (Averaging Times), I have a few comments.

    1.  Both sections are well-written and adequately thorough in describing the many
       uncertainties in relation to the indicator and averaging times for PM2.5.

    2.  It is clear that additional research is required before specific components, and/or specific
       co-pollutant interactions, can be targeted for health-related policy assessments. For the
       present time, PM2.5 is the only feasible default indicator for health-effect considerations.
       Sadly, this will place some U.S. regions at a disadvantage in protecting public health using
       a mass-based indicator (PM2.5).

    3.  Similarly, as the available health-associated data are robust for 24 hour and annual
       averaging times, the conclusion to retain these averaging times is appropriate.

    4.  The lucid discussions of the limitations and uncertainties associated with the indicator and
       averaging times, will be very helpful for stimulating the needed research. The EPA staff
       has provided useful guidance in this respect.

    5.  As a minor editing suggestion, I would consider eliminating the word "strong" on line 6,
       pg. 2-70, as it conflicts with the lucid discussion of the limitations and uncertainties
       previously presented. This is not a strong suggestion, but it should be considered.
                                             37

-------
Comments from Mr. Rich Poirot

Overall, Chapter 4 of the Policy Assessment represents an excellent first draft presentation of
justifications and policy options for setting secondary PM standards to protect visual air quality.
While there are advantages to using an optical indicator (PM light extinction) and to using the
single worst daylight hour for a regulatory metric, I think there are also certain advantages to using
a PM2 5 mass indicator, and/or to considering effects in all daylight hours, and that plusses and
minuses associated with these various alternatives could be presented in a somewhat more
balanced way.

 14. Adequacy (section 4.2):  What are the Panel's views on the preliminary staff conclusion
     that the currently available information clearly calls into question the adequacy of the
     current suite of PM2.s standards to provide public welfare protection from unacceptable
     levels of visibility impairment, primarily in urban areas, and supports consideration of
     alternative standards to provide appropriate protection?

The Policy Assessment does a good job of demonstrating that current PM standards (for which 24-
hour PM2 5 is most relevant) are not adequate for protecting visibility in most, if not all sections of
the country.  Specifically, the levels of current Plvhs and PMio standards are too  high and their
averaging times are too long to guard against levels of visual air quality considered adverse, over
the short (hour or less) time periods over which changes in visual air quality are perceptible.  The
current mass-based indicators are also not ideal and could be made much more directly responsive
to visibility effects by  using a PM light extinction indicator.  An optional PM2S mass indicator,
applied over shorter daylight hourly averaging times and at lower levels would be adequate, and
would represent a substantial improvement over the current 24-hour and annual means.

A possible addition which might strengthen the chapter in future drafts would be to show - with
data examples - how extreme a 1 -hour PM2 5 mass concentration or 1 -hour PM light extinction
level could be on (a) day(s) when the 24-hour PM2 5 mass concentration does not exceed the
current 35 ug/m3 standard. In addition, you might include some photographic illustrations - using
actual haze cam photos or Winhaze images showing what visual  air quality looks like at levels of
the current standards (and at the higher 1 -hour levels that are possible without exceeding the 24-
hour average standards).

 15. Indicator (section 4.3.1): What are the Panel's views on the following:
a.  The appropriateness of considering PM light extinction and PM2.s mass as indicators for
    a distinct secondary PM standard  for visibility protection, and the relative advantages
    and disadvantages associated with each indicator?

The Policy Assessment does a good job of demonstrating the appropriateness and benefits of a PM
light extinction  indicator, while showing that a (daylight only, hourly) PMis mass-based indicator
would also be adequate, and a substantial improvement over the status quo.  To a certain extent,
arguments for use of a light extinction indicator can be strengthened by emphasizing the
shortcomings of a PM2 5 mass-based indicator. However, I think the Policy Assessment could
provide a somewhat more balanced view of some of the advantages and disadvantages of both
alternative indicators.
                                            38

-------
A PM light extinction indicator has the major obvious advantage that the indicator would
essentially be the effect. PM light extinction is an actual and measurable characteristic of particles
in the atmosphere, which can be directly related on the one hand to the size and composition
(including aerosol water) of airborne particles which cause it, and on the other hand to the levels of
visual air quality considered unacceptable by human observers. There are however also
disadvantages to a PM light extinction indicator (and advantages to a PlV^s mass indicator) that
could be presented more clearly in the Policy Assessment.

A disadvantage of a PM light extinction indicator is that there is currently no monitoring network
in place to assess compliance, nor have standard methods been developed which have been
deployed in anything other than research-grade applications. At a minimum, at least 2 separate
instruments (nephelometer and aethalometer) would be required using currently available
technology, and these will function best if coarse particles are excluded. Thus the measured
indicator would actually be PM? 5 light extinction, although this could be enhanced where need be
with measurements of and estimates of extinction from continuous PM 10-2 5 mass (requiring two
additional instruments). Accurate RH measurements will also be more important for this indicator.
Conversely, there is already a large network of continuous PVh 5 instruments in place that could be
used immediately to assess compliance with an hourly PM2S mass indicator, and the quality of that
data - and its utility for assessing human health or other public welfare effects - could be
substantially enhanced by forcing closer scrutiny of the accuracy and precision on an hourly basis.
A PM2s mass indicator would allow somewhat higher extinction in the more humid (and generally
less scenic) East, and provide relatively better protection in the less humid (and generally more
scenic) West - which is not necessarily a disadvantage. It would also tend to place relatively more
emphasis on organic carbon and elemental carbon aerosols, which are more negatively associated
with other public welfare (soiling and climate) effects, and which are also generally more
influenced by local emissions than are the sulfate and nitrate compounds which are emphasized by
the PM light extinction indicator.

b. The appropriateness of considering the contribution of coarse particles with respect to a
   potential PM light extinction indicator?

Since coarse particles make relatively minor contributions to light extinction at most locations, and
also present both measurement uncertainties as well as increased measurement maintenance needs,
it may be reasonable to either:
•  Exclude coarse particles from the regulatory metric (either re-name it "PM25 light extinction"
   or keep the same name but exclude coarse particles from  monitoring requirements) or,
•  Add (either on an all-site basis, or at sites where PM|o:PM2 5 > X) continuous mass
   measurements of PMio-25 by the difference method, estimating the associated PM light
   extinction increment by multiplying by 0.6 m2/g as in the IMPROVE formula.
•  Conceptually a similar treatment of PMio-2 5 effects could be added to a PM2 5 mass indicator,
   where the revised mass indicator might be something like PM25 + 0.1 x PMio-25-

 16. Averaging times (section 4.3.2): What are the Panel's views on the following:
                                            39

-------
a. The preliminary staff conclusion that consideration should be given to a 1-hour averaging
   time?

The one-hour averaging time (during daylight hours) is appropriate, given the nearly instantaneous
nature of perceived impairment of visual air quality. At the same time, 1 think somewhat longer
averaging times (of say 4 daylight hours) could also be justified for several reasons, including:
•  To minimize the influence of measurement errors - which could be important if multiple
   monitors are used for an optical indicator, and also because some of the continuous PM^s
   monitors produce "noisy" data over short 1 -hour averaging times.
•  To minimize the influence of short-term, localized meteorological  extremes - such as the
   concurrence of high RH, low wind speed and low mixing heights on  cold, clear winter
   mornings.
•  To intentionally focus on visibility-impairing events which last for more than a few hours,
   and are arguably more objectionable, and effect more people, than events of short duration.
•  To follow the reasoning used and precedents set in locations like Denver and Phoenix where
   visibility standards have been established, and based  on 4-hour averages in both cases.
b. The advantages and disadvantages of focusing on a 1-hour daily maximum or each 1-
   hour average concentration during all daylight hours?

I much prefer a metric based on using all available daylight hours (combined with a higher
percentile like 98th) over one based on only the single worst daylight hour (combined with a low
percentile like 90th) for several reasons:
•  1 think "sensitive" or "susceptible" are poor choices of terms for those who care so much
   about viewing scenery that they chose to work in windowless indoor environments, stay
   inside during lunch, and (returning home from their 8-hour shifts)  further choose not to go
   outside or look out their windows during the other 2 to 6 hours of daylight that would be
   available, during 10 to 14-hour days in different seasons at a mid-latitude location in the US.
   Do they also stay inside and away from windows on weekends, holidays and vacations? 1
   doubt these are the folks who indicate that only the very clearest levels are acceptable in the
   urban visibility preference studies.
•  There's an implied, but illogical assumption that the single worst hour will coincide with the
   one hour when the poor "shut-ins" are let outside, but that often won't occur.  At best, you're
   disregarding their views on either the to-work or from-work halves of their commutes. What
   if they work "9 to 5" (as many  do), but the worst hour comes at 6 to 7 AM (or PM) or 7 to 8
   AM or PM (as it often will), or in mid morning or mid afternoon, or what if they work the
   second or third shift? By considering only the single worst hour, you are specifically not
   affording protection to all the non-shut-ins for most of the day, and providing no protection at
   all for that fraction of the shut-ins whose hour outside doesn't coincide with the worst hour of
   the day.
•  It puts the maximum possible emphasis on uncontrollable natural meteorological influences,
   including the convergence of maximum RH, minimum wind speed and lowest mixing height
   (along with whatever weather influences a 90% RH screen might miss) in the hour just after
   sunrise.
•  This early morning met effect will be most extreme in mountain valley locations in winter,
   but these coldest, calmest winter days often offer the  best visibility later in the day when the
                                           40

-------
    inversion bums off (or at slightly higher elevations just above the inversion), so benefits of
    the extreme pollution controls needed to reduce pollutant concentrations at winter dawn, are
    likely to be spatially and temporally limited.
•   There are not really any viable 1-hour control strategies - especially for the hygroscopic SO4
    and NOs species which are emphasized by the PM extinction indicator. Conversely, when
    only the single worst hour Bext is considered, there will be the most minimal knowledge of
    the pollutants and sources responsible.
•   By emphasizing early morning RH and hygroscopic species, it tends to emphasize East/West
    differences and deemphasize the importance of organic carbon and elemental carbon
    aerosols, which have other welfare effects such as on soiling and climate.
•   For a PM light extinction indicator, measurements might include nephelometer and
    aethalometer (preferably with PM2 5 heads), plus an RH sensor, and could conceivably also
    include continuous PMio and PMz 5 mass at some locations to estimate coarse mass
    extinction. That's a lot of things that can potentially go wrong, for which a single worst
    hour/day measurement will be much more sensitive to measurement errors or failures in any
    of the above than if a more robust metric based on all daylight hours were considered.
•   It seems especially illogical when combined with a low percentile: "even a single bad hour
    will ruin my day, but I don't mind a bit until it happens more than 37 days a year."

 17. Levels and Forms (section 4.3.3): What are the Panel's  views on the following:
a.   The preliminary staff conclusions regarding ranges of options with varying levels and
    forms in combination with different  indicators (light extinction and PM2.s mass-based
    indicators) and averaging times (1-hour daily maximum  or each 1-hour average
    concentration  during all daylight hours)?

I think the ranges of options combining various levels, forms, indicators and averaging times are
reasonable, with focus on mid-range combinations while considering modest upper and lower
bounds. For the PM2 5 mass indicator, I think the upper bound concentration levels considered (40
ug/m3 and 60 ug/m3) are rather unreasonably high and could be  replaced in future drafts with
smaller 5 ug/m3 increment calculations within the range of 15 to 35 ug/m3. As indicated
previously, I much prefer consideration of all daylight hours, rather than focusing on the single
worst hour, and would like to see a full set of calculations based on that metric for both the optical
and fine mass indicators included in the next draft. Along similar lines, the Policy Assessment
indicates (p. 4-44, lines 18, 19) an intended focus on the more lenient 90th and 95th percentiles -
which I think should specifically pertain only to the single worst hour/day  indicator. Higher
percentiles (95 to 98th) would be appropriate if all daylight hours are considered.  1 suggest
rewording this to "... 90th or 95th percentile forms is appropriate if based on the single worst
daylight hour, while consideration of the 95th or 98th percentiles  is more appropriate if based on all
daylight hours."

Would it be possible to show estimates of what attaining the various hourly optical or PMjs-based
secondary standards would do to (or mean in terms of) changes  in annual and 24-hour average
PM2 5 levels in the test cities?  Along similar lines, would it be possible to show the "incremental
% reductions" in PM2 s that would be needed to, or associated with efforts  to meet the various
optical or PM2s-based secondary standards after the current and proposed  new primary PM25
standards were first attained?
                                           41

-------
I've become confused about the differences that would result if PM light extinction levels or PM25
concentrations were rolled back to just meet standards based on single worst hour compared to
using all hours (and a higher percentile). At any given  level of PM light extinction or PM2 5 mass,
it would seem like rollback to attainment would produce similar results, since once the worst hour
is "fixed", the second worst hour takes precedence, and then the third, etc. until all hours above
threshold are reduced - in which case, the single worst hour approach becomes very similar to the
all hours approach. Can this be explored/explained in more detail?

b.   Additional approaches that could help inform our consideration of alternative levels and
    forms?
As indicated above under Q 16a, I think a 4-hour averaging could be considered and justified,
especially if the occasionally noisy Plvh 5 indicator is used. I think the need for a 90 % RH screen
becomes less urgent, and might possibly be dropped  if a PM2 5 indicator is used (see other
comments on p. 4-44 below). It is definitely needed  if an optical indicator is used, and
consideration needs to be given to whether an hour would be eliminated if the average RH exceeds
90%, or if any or many  (1-minute) RH levels within the hour exceed 90%.
Because of measurement issues with coarse particle extinction, consideration might be given to
using (and measuring) Plvhs  light extinction as an indicator. Alternatively, measurements of
continuous PM 10-2 5 mass (by the difference method), combined with an assumed extinction
efficiency of say 0.6 m2/g could be used to supplement PM2 5 light extinction measurements -
either at all sites or at sites with relatively high coarse particle concentrations.  Conceptually,
coarse mass  by difference could also be used to enhance a PM2 5 mass indicator, using a
relationship of something like PM25 + 0.1 x PM 10-2 5 (assumes a generic PM2 5 extinction efficiency
of about 6 m2/gram - which would represent ammonium sulfate or nitrate at 66% RH - or a mix of
these and organic matter at somewhat higher RH).
PM light extinction is an actual and measurable characteristic of particles in the atmosphere, which
can be directly related to the levels of visual air quality considered unacceptable by human
observers and to the size and  composition (including aerosol water) of airborne particles which
cause it. Given this, a level of a secondary standard could be selected, even in the absence of- or
in advance of a monitoring network to determine compliance. Along similar lines, a  PM (or PM2 5)
light extinction standard could be set, and PM2 5 mass measurements specified as an FEM to
determine compliance - until such time as an appropriate PM light extinction measurement
network can be established.
As indicated in previous comments on the UFVA, a "progress-based" attainment test rather than
the traditional "threshold-based" attainment test would be preferable from a scientific perspective.
While the Agency has indicated this is not an option  in the standard-setting process,  it is something
that could  and should be considered in the implementation phase (where, for example, the Agency
could provide guidance on what would be considered reasonable rates of progress).  Possibly an
illustrative example of how this might work could be provided in the  next draft Policy Assessment.
I think it would be appropriate if the next draft of the Policy Assessment included a section
identifying and prioritizing key near-term monitoring and  research needs. This could include
things like establishing a small pilot urban visibility network to field-test and evaluate alternative
PM light extinction measurement methods, and should also include recommendations for
                                           42

-------
additional visibility preference studies to better define levels (and/or frequencies) of visual air
quality considered acceptable over a wider range of locations and viewing conditions, to support
refinements and revisions to secondary PM standards in the future.
Other Comments on Chapter 4

P 4-5 line 23:  You could add "most" at the end of the line.

P 4-5 lines 34-35:  You could add "(all but one)" after "most CASAC Panel members".

P 4-15 lines 15-31: As indicated above, I'm not a fan of the single worst hour approach, and think
it's a stretch to claim its needed to protect the most "sensitive" individuals...since they are neither
sensitive, nor does it protect them most of the time.

P 4-16, lines 15-23: This is generally true, but especially during the winter. In warmer months
(when people spend more time outdoors), RH still peaks at dawn, but morning rush hour occurs
several hours later, nocturnal inversions are weaker, SOi plumes aloft don't mix down until mid
morning, and secondary formation rates for sulfates and organics typically peak in mid-afternoon.

P 4-20 line 17: The "24.3 dv" level cited for the 50th percentile acceptability response in Phoenix is
different from the "24.15 dv" 50% level for Phoenix in Table 4-1.

P 4-21 lines 16-31: You might also add something like "It should also be noted that the current
version of the  Winhaze model does not include clouds in the base photos. The visibility of clouds
and their contrast with the background sky would be sensitive to changes in light extinction, even
in urban  city scenes which otherwise lack distant objects in  the background."

P 4-22 line 13: You could add "and Eastern urban scenes with more distant vistas or with
background clouds" after "...Western urban scenes."

P 4-29 lines 9-13: As previously indicated, I think the argument to protect the "sensitive" shut-ins
is not the strongest. Better arguments are the instantaneous nature of the perceived visibility effect,
and you might also add an illustration, based on data, showing how high a one-hour PM light
extinction (or PM2 5 mass concentration) could be on a day when the current 35 ug/m  secondary
PM25 standard is not exceeded.

P 4-30 lines 21-32: Maybe you could set this point up a bit by zooming in from the concept of light
extinction to PM light extinction (excluding Rayleigh and N02) — which I assume would be the
preferred indicator. Then you might consider but exclude the very small fraction of PM extinction
from ultra-coarse particles (> 10 HID) - not because they contribute nothing, but because their
contribution is so minor. Then make the point that PMio-2s particles also usually contribute very
little (and their extinction effect is more difficult to measure), but can occasionally be important.
This might lead to  an optional approach in which "PM light extinction" could be the indicator,
measured as "PM2 5 light extinction" in  most areas, but with supplemental methods to estimate
PMio-2 5 extinction in areas where coarse particles are  important.

P 4-30 line 23: Change "many" to "may".
                                            43

-------
P 4-30 line 34: I think "one to two years" may be optimistic, and would constitute "rushing" with
all of its various associated pitfalls.

P 4-31  line 2: Use of the term "promulgation/implementation" implies that the two things need to
be concurrent, but I don't think they need to be so. Conceivably, a PM light extinction standard
could be promulgated, with an implementation period delayed until a future time period (perhaps
to allow a reasonable time frame to develop, test and deploy appropriate monitoring methods).

P 4-31  line 14: No space needed in "widespread".

P 4-31  line 17: Add something like: "24-hour chemical composition data and" between "produce"
and "hourly", or otherwise correct this sentence.

P4-31  line 19: Delete "produces".

P 4-33 lines 1-4: Presumably the frequency of visibility-impairing "weather" is very small, but not
zero, when RH is less than 90%.. It might be informative to know that there may still be some
weather influence and that 90% RH is a reasonable compromise, which eliminates most of the
weather events.

P 4-33, line 28: I think 60 ug/m3 is a very extreme upper bound, and that its nearly impossible to
have PM extinction < 191 Mm"1 with PM25 as high as 60 ug/m3 (and extinction of 191 Mm"1 with
PM2 5 of 40 ug/m3 would be relatively rare as well) . In the entire IMPROVE network for the 5-
year period from 2000 through 2004, there were 73,410 samples with reconstructed  PM extinction
of less than 191 Mm"1. None of these had PM2s mass concentrations as high as 60 ug/m3, and only
2 samples and 9 samples (of 73,410) with reconstructed extinction < 191 Mm"1 had fine mass
concentrations above 50 and 40 ug/m3 respectively.

P 4-35, lines 8-9: As indicated before, I don't buy the logic that a focus on the worst one hour is
needed to protect the "susceptible" shut-ins.  For this to be true, the single hour of poorest visibility
would  need to correspond to all the shut-ins' hour outside.

P 4-36 line 4: Remove "could" or the "s" from "affords". Also, I'm not sure that providing a
uniform degree of visibility protection nationwide is such a great advantage. You could add  to the
end of this sentence ", although efforts to attain it would require substantially greater pollution
control efforts in the humid East, where scenic vistas are typically more limited.  By comparison, an
"equally stringent" nationally uniform PM2s mass indicator would allow somewhat higher levels
of visibility impairment in the East, but would provide somewhat better visibility protection in
dryer, and typically more scenic, Western  urban areas."

P 4-38: I don't understand why, for Tacoma and Phoenix, there are changes from column a to b
(90th to 95th percentiles) when both are below the 191 Mm"1 threshold?

P 4-41: This table - with the color-coding - is very helpful!  Possibly future versions might focus
on a more realistic range of say 35 to 15 ug/m3, with smaller (5 ug/m3) increments between bins.
                                            44

-------
Would there be a way to estimate the changes in and remaining concentrations of in 24-hr average
PM25 mass that would result if these hourly levels (or the hourly Bext levels in Table 4.3-1) were
obtained? It will be a lot more work, but you should include all the same calculations based on the
98th percentile using all daylight hours.  Without these you will be stuck with illustrations that only
apply to your (questionably justified) worst 1 hour metric.

P 4-42, lines 11-13: Although presumably there would be substantial overlap between the worst
10% fine mass days (hours) and the worst 10% Bext days (hours).  Could this be explored?

P 4-43. line 21:  Is there an intent to specifically emphasize "urban areas", or could this be
rephrased to "primarily in those rural, suburban and urban areas which are not protected by the
1999 Regional Haze Rule".

P 4-44 lines 18,19: By focusing on the 90th or 95th percentile, it sounds like you've already made
the (questionable) decision to use the single worst daylight hour.  Otherwise, you could change this
to "... 90  or 95  percentile forms is appropriate if based on the single worst daylight hour, while
consideration of the 95th or 98th percentiles is more appropriate if based on all daylight hours."

P 4-44, line 19:  Conceptually, a 90% RH screen could be used for the same weather-screening
reasons if a PM2 5 mass, rather than a PM light extinction indicator was used. As a practical matter,
it may be less critical for a mass indicator for several reasons. Higher RH will tend to "push"
hours into the high percentile light extinction bins, but not so for high PM mass hours. In fact any
weather (rain,  fog) concurrent with high RH is likely to push high PM hours toward lower PM
percentiles.  Also, with a PM25 mass indicator, there's no chance of requiring reductions  in very
low pollution concentrations that happen to concur with extremely high RH conditions, or of
having any public perception, however misguided, that EPA is trying to  regulate the weather.
There is also no benefit from or need to specifically exclude higher RH conditions from (dried)
PM2 5 mass measurements, nor would errors in the RH sensor measurements have large influences
on whether the "right" or "wrong" hours were excluded or included from the regulatory metrics.

P 4-44 lines 29-30:  1 assume this is a typo and that you meant to say "focus on a level of 20
ug/m3as well as levels down to 10 ug/m  and up to 30 ug/m3".
                                            45

-------
Comments from Dr. Armistead Russell

Question 7: Forms (section 2.3.3):  What are the Panel's views on the utility of additional
analyses to inform consideration of eliminating provisions included in the current form of the
annual standard that allow for spatial averaging across monitors, specifically with regard to
the potential for disproportionate impacts on susceptible populations with lower
socioeconomic status?

 I am not sure that there was sufficient justification for spatial averaging in the past, and do not see
 that there is much reason to continue the practice. It is likely that a monitor with elevated levels is
 being impacted with fresh emissions, and unless that monitor is directly adjacent to the source(s),
 there are areas with even higher levels.  Further, those elevated levels may be due to a source
 whose emissions are more harmful than the regional background PM more evenly distributed
 across a city.  Allowing spatial averaging could lead to a specific area being significantly and
 consistently adversely impacted beyond the region as a whole.

 Question 10: Indicator (sections 3.3.1,3.3.2, and 3.3.3):
    What are the Panel's views on the following:
    a.  The preliminary staff conclusion that it is appropriate to maintain a standard that
       provides protection against all thoracic coarse particles, regardless of their source of
       origin or composition?   Agree.
    b.  The appropriateness of maintaining an indicator that allows lower PMio-2.5
       concentrations in urban areas than non-urban areas?   See below.
    c.  The appropriateness of either a PMio or PMio-2.5 indicator for a standard meant to
       protect against exposures to thoracic coarse particles?
       While the association of the observed health effects with larger particles is viewed as being
       only suggestive of a causal relationship, sufficient information is available to continue to
       have a standard that includes particles above 2.5 um.  While certain sources may primarily
       emit particles that are found in the coarse size range (2.5-10 um), the tail of the size
       distribution of those emissions contains PM in the sub-2.5 um range, and those particles
       presumably contribute to the observed health effects in studies using either PMio or PMio-
       2 5.  Using a PMio-2 5 indicator would allow dismissing how this tail contributes. Further, as
       discussed in the Policy Assessment, using a PMio standard would address the (rather small
       amount of) evidence that urban coarse particles appear to have greater health impacts than
       wind blown dust. The epidemiology can support either range, and controls aimed at
       controlling PMio (beyond those used to reduce Plv^s) would reduce levels of PM 10-2 5 and
       vice versa. Given the above, a PMio standard would appear to be preferred. Question (b)
       should be rephrased as to allowing higher PM concentrations in non-urban areas.

 Question 11: Averaging Time (section 3.3.4): What are the Panel's views on the
 preliminary staff conclusions that the currently available evidence supports maintaining a
 24-hour standard to protect against health effects  associated with short-term exposures to
 thoracic coarse particles, and does not support consideration of a long-term thoracic coarse
 particle standard?  Agree.
                                           46

-------
 Question 18: Climate (section 5.2): What are the Panel's views on the preliminary staff
 conclusion that there is insufficient information to base a standard on climate impacts
 associated with current ambient PM concentrations?

 While there may be insufficient information to base a secondary standard on potential climate
 change, the current discussion does not adequately inform policy makers as to the potential that a
 standard limiting light extinction would do more harm than good. There is sufficient information
 demonstrating that some aerosols lead to cooling, in particular aerosols that scatter visible light
 and can lead to increased/longer lived/more reflective cloud formation.  The REA (based largely
 on other assessments) demonstrates that light scattering aerosols lead to a net decrease in radiative
 forcing, and while the uncertainty is significant, it suggests that the effect may be quite large.
 Further, there is sufficient evidence that light absorbing aerosols can lead to warming. To go
 forward with a secondary standard based on reducing light extinction, the Policy Assessment
 needs to more directly weigh the potential dis-benefits of regional warming with improved
 visibility unless light absorbing aerosols are preferentially targeted. There may be significant
 regional heterogeneity to controls that would address a visibility-based secondary standard which
 should be addressed.  For example, it is the areas that have the lower PM concentrations that may
 benefit most from maintaining/improving visibility, but the areas with regionally high sulfate
 levels likely benefit most from the reduced radiative forcing. At a minimum, EPA staff needs to
 address how a secondary standard might impact regional radiative forcing and climate, assess the
 uncertainties, and provide a reasoned analysis weighing the two issues.  Also, on Page 5-11, lines
 10-11 and lines 31-31, it is noted that "aerosols that are warming are co-emitted with aerosols that
 are cooling."  While this is true to some degree in some cases, it skews one's perception of the
 actual case. Most of the warming aerosols in the US are emitted by biomass burning and internal
 engine combustion. Much of the cooling aerosol is formed in the atmosphere by oxidation of
 SO2 or VOCs.  Thus, a set of controls to control warming PM would not necessarily have that
 much impact on cooling PM and vice versa. These statements should be modified.

 Question 20: Material effects (section 5.4):  What are the Panel's views on the preliminary
 staff conclusion that no new evidence calls into question the adequacy of the protection
 afforded by the current suite of PM standards, and that there continues to be support for
 retaining an appropriate degree of control for both fine and coarse particles to provide
 protection against materials damage and soiling? Agree.

Minor:

P 5-22,1 13. Replace "impossible" with "difficult"
                                            47

-------
Comments from Dr. Frank Speizer

Charge Question 3:  Approach (Section 2.1.3)
Page 2.10-2.12: With regard to the weight given to the various approaches Staff appears to
indicate that the lower level of the interquartile range for both the long term and short term
empirical evidence should be considered, in spite of indicating no threshold. This may be
absolutely the correct approach, but some further discussion as to why this decision was made
might be warranted. Indicate why not the 90th or 95th %tile level. What does "notably wider"
mean in discussing confidence intervals? How much wider? At what point do they cross zero?
Page 2.12-13:  In discussing the use of the risk estimates Staff indicates that the uncertainties are
taken into account by considering the sensitivity of the "core" risk. May need to define "core" risk
more explicitly. In addition it is not clear how the two separate approaches are to be combined and
whether the weighting of one over the other will dominate or be the determining factor. If I read it
one way I might conclude that the evidence based values will be used and the risk assessment is
simply a test of the sensitivity of reaching a conclusion. Alternatively (and most likely) somehow
the use of the total information will be used to reach a judgment.

Charge Question 4: Adequacy of current suite of PM2.s standards (Section 2.2)
   a.  Appropriateness and characterization of sets of studies. It would appear that the specific
       studies chosen are appropriate and there are none other that need to be added. However,
       Figure  3.1  seems not to be adequately discussed in terms of the presentation of differences
       in weights to be put on any given set of studies.  (There are also some specific questions
       about the way the figure is presented). Clearly the nature of the confidence intervals
       indicated in the  figure provide as estimate of both the nature and size of each study. A
       more explicit indication on the figure as to which were single city vs multicity studies
       would be helpful. This would provide further evidence of why relying on the multicity
       studies seems justified.  Specifically with regard to the figure and characterizing the PM
       exposure there are some questions raised by what is in the figure. For example for a
       number of the studies the Lower IQR value matches the lower bound of the Range. How
       does this occur?
   b.  Page 2.34 The question of susceptible groups seems to conclude with the standard group of
       older adults and children along with pre existing cardiopulmonary diseases. The paragraph
       justifying lower SES is appropriate. However, no mention  is made of diabetics (perhaps
       much of the  data is post ISA, but not all). This is an important group, the data are at least
       as firm as some other groups, and it is a growing issue in the face of the obesity epidemic.
       I would therefore suggest a brief discussion of the risk data and it be added to the list.
   c.  Use of risk assessment, judge correct. Page 2.34, line 14:  Minor point—top of page 2.37,
       line 1:  Not sure where the conclusion that the WHI represented a "healthier cohort
       population" comes from.  If I had noted it in the ISA I would have objected. These women
       were selected as post menopausal and thus were aged ~45+. In the ACS cohort I believe
       the lower bound age was 35 and the cohort was probably of a generally higher social class,
       and in Six Cities it was a random sample of households and generally younger (and
       probably healthier). Before getting to the question of risk based consideration  the several
       pages leading up to page 2.42 provide substantial evidence that the various studies
       considered demonstrate empirically that significant and substantial health effects have been
       reported by multiple investigators at or just below the current standards. It would seem that
                                            48

-------
       the conclusion that the current standards are inadequate to protect the public and should be
       so stated before making the obvious concluding sentence on page 2.42, starting on line 5.
       In fact, it is an interesting contrast that one could take the position that since the current
       standards fail to protect the public and since there are not a lot of studies that show effects
       below the current standard levels, the very reason for doing the risk assessment is to help
       the administrator find an appropriate level that includes a margin of safety (rather than
       simply trying to add the risk assessment information to the empirical data to come to some
       "combined judgment" I admit this may just be semantics but because risk assessment
       carries the implication of uncertainty it would seem better to put it into this context.
    d.  The argument used for justifying the relative contributions of the annual and 24 hour
       concentrations is logically presented and clearly stated by summarizing the results of table
       2.2. and text on page 2.45.  As we indicated in discussing the RA, although the conclusions
       would not have changed, we understood why LHD mortality was used but would have
       preferred Staff using Total Cardiovascular mortality.  The problem is further complicated
       by trying to compare directly the results as presented in table 2.3 with table  2.2 where the
       contrast is between the annual and the 24 hour.  Since table 2.2 is IHD results and table 2.3
       is cardiovascular and the magnitudes of effect are -3-10 times larger for IHD the tables do
       not appear to be directly comparable. The argument as to why there is greater confidence
       in the roll back estimates from where the annual standard is controlling than where the 24
       hour is controlling is straight forward and clearly related to the differences in the volume of
       data for each scenario from which to make the estimates.  (The argument presented on page
       2.52 for the use of the  15 urban study areas for estimating the contrast between annual and
       24 hour values is persuasive and provides the relative limitations of the sites as a fair
       assessment.)
    e.  With regard to the integration of the evidence presented in section 2.2.3, Staff is correct in
       asserting that "newly available information strengthens the associations between ..."  We
       also agree that there are significant public health consequences at the current levels of the
       standard that justify consideration of lower standards in the further. However, what is not
       is this section is the fact that this is not different from the conclusions of the previous
       review, but only strengthened by additional data obtained  in the last five years. Remember
       it was the Administrator that when outside the recommendation of both CASAC and Staff
       the during the last review. Further without specifying the nature of the uncertainties that
       were reduced since the last review (page 2.54, line 30) the text does a disservice to the
       previous conlusions that the standards should have been lowered. What are the specific
       uncertainties that have been lowered?  What  are the mechanisms that were not known 5
       years ago that are better articulated now? It is true that there are a greater suite of health
       effects, however, they do not meet a higher degree of certainty and thus add little to the
       quantitative argument available at the last round. (One cannot avoid the obvious
       conclusion that the Administrator in the last round simply ignored the available scientific
       data.)

Charge Question  5—Indicator
       a.  With regard to maintaining PM25 Staffs argument that the data have expanded over the
           last several years is correct. They are also correct in indicating that in spite of there
           being significant advance in being able to better characterize both in terms of size,
           number (and perhaps chemistry) these latter measures have generally not be applied
                                             49

-------
          sufficiently in health based research to be able to change the indicator (at least for the
          fine fraction from the current one. Of PM25.
       b.  With regard to Ultrafines (UFP) the data are certainly suggestive of health effects.
          Staff rightfully points out (Page 2.58, Paragraph beginning line 25) there is currently no
          network of UFP samplers and thus only limited site-specific data sets. These data,
          however, are compelling and point to the need for the establishment of a network of
          samplers under some federal guidelines to position EPA to be more effective in making
          judgments on future particle standards.
       c.  With regard to speciation and  sources perhaps the most important summary statement
          is that on Page 2.64, sentence  beginning line 17, ".. .evidence not sufficient to support
          eliminating any component or source from the mix..."  However, the suggestion that
          could find a specific component that might be eliminated is probably a wrong way to
          conceive of the problem. PM is obviously a summary measure of a mixture of both
          complex air chemistry and  source components. The likelihood of being able to
          identify or eliminate one of those components is small. More important is the use of
          this complex chemistry to more accurately identify and label sources that need to be
          better controlled.  Thus far the data point most specifically to oil burning and traffic
          pollution sources, and less conclusively to wood burning. This obviously is important
          information for considerations beyond the Standard to the implementation phase.
          Rather than simply indicating that the work to date does not justify expanding the
          Indicator to be more specific in terms of speciation, a more positive statement of the
          sources could be made.

Charge Question 6—Averaging times
       a.  Agree with Staffs position that the annual and 24 hour average need to both be
          retained as the standards.
       b.  Also agree that that there is insufficient information to consider a shorter than 24 hour
          average or a seasonal average at this time. However, there is mounting clinical
          evidence that short term exposures (substantially less than 24 hours) are associated with
          reversible cardiovascular changes. The degree that these changes either from acute or
          repeated exposures trigger significant adverse events or result in more sustained
          adverse effects  is as present highly suggestive, They certainly point to the need to
          consider from a policy perspective related to susceptible groups how the 24 hours
          standard should be modeled to protect the population with a margin of safety. Are
          there additional analyses that could be done to inform the Administrator on the
          distributions of the frequencies at which any specific level  of exposure for , say up to 2
          hours, would occur at specific levels of the 24 hour standard?

Charge Question 7—Form
       The argument that there may be difference in the protective effects of a given standard
based upon spatial averaging vs 'community-wide air quality" recording the highest value only is
sufficiently well presented  to justify further analyses.

Charge Question 8—Level
       a-c. The rationale presented regarding alternative standard levels is well presented,
although it is not clear how Staff intends will eventually focus of the which of the two very logical
                                            50

-------
alternatives (range 13-10ug/m3, 30-35ug/m3 vs. 11-]0ug/m3, 30-25ug/m3) will become the focus
of discussion in the second draft.  Perhaps it would be useful for CASAC to weigh in more heavily
on these alternatives. (Obviously, the latter provides a greater degree of safety and is more
consistent with an "adequate margin of safety"
       d. Clearly the idea that the alternative annual standard will shift the distribution of 24 hour
measures is well presented. However, it is also well presented that "hot spots" would remain and it
may be these very hot spots that put the most vulnerable parts of the population at greatest risk.
Thus, it becomes appropriate to continue to also regulate 24 hour levels, and thus logical to
consider and maintain an alternative 24 hour standard.
       e. Other considerations. Because one national standard may not deal effectively with the
seasonal or spatial variations that might occur (nor the potential localization of susceptible sub-
populations at greater risk) it would seem more logical to consider more conservative levels  for the
standards. Are there additional analyses that could be undertaken to provide more quantitative
estimates of the magnitude of the populations  that would be left at risk for any given suite of
standards? This might be one of the topics to be covered in the second draft of the Policy
Assessment.

Charge Question 9 -Adequacy of current 24 hour PMio standard.
       a.  Maintaining a Course Fraction particle standard.  If anything the evidence since the last
           review is stronger,  although still only suggestive that there are significant adverse
           health effects associated with the thoracic course fraction. It would appear that
           uncertainty issues remaining relate mostly to either lack or sufficient numbers of
           monitors from which to make appropriate statistical associations and adequacy of
           spatial estimates of exposure, potential for co-pollutant effects (mostly separating the
           effects of fine particles from course particles) or difficulties in technically being able to
           carry out toxicological studies of enough variety to provide a foundation  for the
           potential mechanisms (different from those that have already been confirmed for
           PM2.5).  Clearly the public health  concern remains high (potentially related more to
           morbidity effects rather than mortality as evidenced by the relation to hospitalization
           and emergency room associations). Thus, in spite of the uncertainties maintaining a
           thoracic course particle standard is warranted. Further to the degree that  the
           associations reported are at least as significant as in the last round of review, and I
           believe they are even more suggestive, strengthening the standard can also be defended.

Charge Question 10 Indicator.
       a-b. Appropriateness of maintaining a  standard that protects against particles regardless of
source of origin and composition. The answer to this question seems obvious. The issue is that the
potential fraction of total course fraction that is combustion generated vs. dust generated may or
may not be important.  The wind blown dust studies that show impacts at far distances  are
suggestive that irrespective of source and thus potential toxicity (from what we currently know)
may not make as much difference. Therefore  accepting that some fraction of the PM10 may be
related to the amount of combustion source (read PM2.5) and thus a portion of the toxicity noted
does not negate the  importance of reducing total particles. Staff seems to be arguing that being
able to reduce PMIO from any source will work to the benefit of both urban and rural sources. 1
think I agree with this, but  it still would be worth while a full discussion at CASAC.
                                             51

-------
       c.  With regard to which measure (PM10 or PM10-2.5) is appropriate, I believe we will
need to discuss this fully.  We cannot leave the question open to the Administrator to decide.
Therefore what we need is some estimates of exposure fractions with which to make a tentative
judgment of what the alternatives would predict.

Charge Question 11—Averaging time.
       Staff concludes that the 24 hour average provides the best estimate of exposure to course
fraction and is the exposure value with the most data suggesting an association. I would agree. It
probably make the most biological sense and the suggestion that control of accumulated repeated
24 hour exposures would reduce the potential for longer term averaging time effects is reasonable.

Charge Question 12. Form.
       Concur with the use of the 98th percentile form averaged over 3 years.

Charge Question 13—Level
       We currently have a promissory note that Staff will conduct some analyses of alternative
levels of the 24 hour standard for PM10 and PM10-2.5. More specific criteria of a "stopping
point" are needed. Will the lower bound be at 140,120, 100ug/m3 for PM10, and if so on what
basis? For PM10-2.5 will it be at 75%, 60%, 50%, 40%, lower(?) of PM10 and on what basis? (If
it were at all possible to see some of these analyses before the Draft 2 it would be useful to include
in our discussion.) by indicating that the focus will be on the upper end of the distributions of daily
PM will Staff will need to take into consideration the differences between total PM urban and rural
and may have to run two sets of analyses to compare differences.
                                           52

-------
Comments from Dr. Helen Sun

       The first draft Policy Assessment is a good start at presenting and discussing the policy-
relevant issues in the review of the PM NAAQS.  It is comprehensive in its inclusion of relevant
evidence and risk based information. The summaries (e.g., sections 2.3.6,2.4, 3.2.2,3.4) were
particularly notable, as they were generally succinct and thoughtful, clearly and cogently
presenting the staff conclusions. Other sections of the Assessment, however, suffered from too
much detail, seemingly recounting much of information presented in the ISA and REA.  This
rehashing of information obscured the key policy  related issues and  brought a lack of focus and
clarity to the document. Further distillation and tightening of the document is needed. While not
yet written, the sections on "Key Uncertainties and Needs for Further Research" (for both PMa 5
and PMio-2s) are welcomed and important. These sections should certainly include discussions of
PM composition, with respect to whether particulate mass is the appropriate indicator. Finally, at
its current length, the document would benefit from a chapter either at the beginning or end of the
Assessment that summarizes staff conclusions for both the primary and secondary standards.

Question 7:  Forms (section 2.3.3):  What are the Panel's views on the utility of additional
analyses to inform consideration of eliminating provisions included in the current form of the
annual standard that allow for spatial averaging across monitors, specifically with regard to
the potential for disproportionate impacts on susceptible populations with lower
socioeconomic status?

The rationale to conduct additional analyses to assess whether spatial averaging provisions should
be eliminated is sensible and appropriate. Further, the consideration of alternate forms for the
annual PM2 5 standard, specifically the highest community monitor level, is appropriate.  However,
the process by which such considerations (and from this the appropriate standard form) will be
made is unclear and should be clearly set forth in the second Policy Assessment. In particular, the
second Policy Assessment should describe how likely regional and temporal variation in the
comparisons and the relationship between highest monitor levels and low SES populations was
considered.

c.  Levels (sections 2.3.4,2.3.5, and 2.3.6):  What are the Panel's views on the following:
   d.  The preliminary staff conclusions regarding alternative standard levels that are appropriate
       for consideration,  and the rationale upon which those conclusions are based?

       The alternate standard levels to be considered and their rationale are appropriate.

   e.  The insights that can be gained from the quantitative risk assessment to inform our
       understanding of the roles that the annual and 24-hour standards play in providing
       public health protection,  specifically in focusing on simulations of estimated risks
       remaining upon just meeting alternative suites of standards? (sections 2.3.4.2 and
       2.3.5.2)
       The quantitative risk assessment is a useful and illustrative tool to illustrate the impact of
       alternative standards on health risk. The usefulness of the risk assessment to this process
                                            53

-------
       would be increased if the discussion of the risk assessment results were condensed even
       further to focus on the major points.
   f.  The policy implications of the uncertainties associated with estimating risks, including
       potential underestimation of risk, in reaching conclusions regarding standards that
       would provide public health protection with an adequate margin of safety?
       As noted in discussions and comments from the March 2010 meeting, the impacts of model
       choices on long- and short-term risk estimates should be expanded (perhaps on page 2-43,
       paragraph beginning on line 9), as this information is central to determinations of margin of
       safety.  As an example, the potential under- (and/or over-) estimation of long-term
       mortality risks from reliance on C-R functions from Krewski et al. (2009) should be
       discussed.
   g.  A policy approach for identifying a suite of standard levels in which the annual
       standard would be the "generally controlling" standard to provide protection for both
       long- and short-term PMis exposures, in conjunction with a 24-hour standard set to
       provide supplemental protection against days with high peak concentrations
       associated with localized "hotspots" and  risk arising from seasonal emission that
       might not be controlled by a national annual standard? (section 2.3.6)
       The policy approach for identifying the suite of standard levels was well written and clear.
   h.  Additional considerations that could inform our conclusions related to alternative
       suites of fine particle standards?
       With regard to a margin of safety, it may be important to include a statement in the risk
       assessment consideration of results (page 2-104) regarding the potential implications of
       model choices regarding under- or over-estimation of risks.
Review of the Primary Standard for Thoracic Coarse Particles (Chapter 3)

[As a general note, the chapter on thoracic coarse particle seems to be less formed and more
preliminary than corresponding sections for PM2 5.]

   6.  Form (section 3.3.6):  What are the Panel's views on the preliminary staff conclusion that it
       is appropriate to consider a 98th percentile form for a revised 24-hour standard meant to
       protect against short-term exposures to thoracic coarse particles?

       This conclusion seems appropriate for practical reasons.  However, it is not clear that the
       logic used to support a 98th percentile form  for NC>2 and PN/h 5 pertains  to thoracic coarse
       particles given that the high end of the PMio concentration distribution  may reflect higher
       contributions of PM25 and not PM10-25 in many parts of the US.  Although perhaps this
       issue can not be addressed, this uncertainty should probably be mentioned.
                                           54

-------
Comments from Dr. Sverre Vedal

Charge Question 1:  Organizational structure.
a. Intro and Background.
       These are well-organized and informative for both naive and informed readers.

b. Questions posed.
       All questions are appropriate -1 can't think of additional ones that would be useful.

c. Level of detail and focus.
       The balance between too little and too much detail is good, in general. There is
repetitiveness as parallel arguments are followed through in dealing with long-term and short-term
exposure risks, but I think this is largely unavoidable. The historical review in the coarse PM
section (3-26 through 3-30) is perhaps too extensive.

Charge Question 2:  Technical content.
a. ISA evidence and uncertainties.
       The Policy Assessment adequately and faithfully summarizes the relevant points from the
ISA.  I still maintain a minority  opinion that the link between long-term exposure and
cardiovascular disease is appropriately classified as causal but that with total mortality should be
classified as likely causal (p. 2-16, table 2-1).  Regardless, the Policy Assessment is faithful to the
last version of the ISA, which has the links with both cardiovascular and total mortality classified
as causal.
       The section on susceptibility is somewhat poorly focused,  including a discussion of lung
function deficits in adulthood (p. 2-27) of limited relevance.  Also, studies of a single stratum of
the population (e.g., post-menopausal women, p. 2-28) that are unable to compare risk across
different strata provide little evidence of enhanced subgroup susceptibility. The bottom line,
however, is appropriate (p.2-29).

b. Risk assessment policy relevant information and uncertainties.
       As  in by review of the RA, I take issue with the statement that the magnitude of both long-
term and short-term risk depends primarily on annual-average PM2 5 concentrations (p. 2-44, line 4
and p. 2-48).  I know this is intended to refer to short-term effects being driven by concentrations
in the middle part of the distribution, but the statement as stands is not strictly correct.

Charge Question 3:  Fine PM  primary standard approaches (Section 2.1.3).  Approach.

       EPA staff is using evidence- and risk-based approaches, as before, to guide consideration
of the adequacy of the current standards and the choice of alternative standards.  Both have their
roles and are appropriate to use  for these purposes.
       There is asymmetry in the way EPA staff uses evidence from long-term and short-term
exposure studies in assessing protection afforded by the annual and 25-hour standards: evidence
from long-term exposure studies is used only in assessing protection afforded by the annual
standard (p. 2-10, line 18), whereas evidence from short-term exposure studies is used in assessing
protection afforded by both the annual and 24-hour standards (p. 2-11, line 10). Is there a rationale
for also considering the evidence from long-term studies in assessing the 24-hour standard?  It has
                                            55

-------
been argued that it is the cumulative effect of repeated daily PM concentration increases that is
responsible for the chronic effect of PM exposure observed in long-term exposure studies (Brook
RD, Rajagopalan S. Air pollution and cardiovascular events. N Engl J Med 2007;356:2104-5). It is
therefore possible that chronic effects could be better controlled through the 24-hour standard than
the annual standard.

The relative weighting.
       I agree with using both the annual and 24-hr standard together in providing protection
against PM effects. The balance between evidence- and risk-based approaches is fine.

Additional approaches?
       None to suggest at this time.

Charge Question  8: PM2.s level.
a. Conclusions on alternative standard levels and rationale.
       I find the use of the  1SD below the mean or at the lower IQR to define annual mean
concentrations of concern to be less appealing that the use of "somewhat below the long-term
mean" concentrations, although either is somewhat arbitrary.  I know that we are to limit ourselves
here to scientific considerations and ignore issues of feasibility and cost-benefit, but as ever lower
standards are being considered, this becomes more difficult.

b. Insights from the quantitative RA on annual and 24-hour standards, especially simulations of
remaining risks with alternative standards (sections 2.3.4.2 and 2.3.5.2").
       The quantitative RA results are very important. The information presented is very dense,
for example, in detailing and contrasting effects in individual cities- but, I think this is preferable
to being vague. The two scenarios used for alternative 24-hour standards (13/30 and 12/25) may
provide most of the information needed, but additional options may be helpful unless it can be
justified that they aren't.

c. Policy implications of the uncertainties in estimating risks.
       Use of the LML rather than PRB as the lower level for estimating risk is appropriate, in my
opinion.

d. Approach with annual standard "generally controlling" standard for both long- and short-term
PMrsexposures (section 2.3.6).
       I believe the arguments made  in favor of the annual standard as the controlling standard are
well laid out and ultimately persuasive.

e. Additional considerations related to alternative standards?
       It isn't clear how the selected studies were chosen to be included in Fig. 2-4 (p. 2-91).

Charge Question  13:  PMio.2.5 level. Appropriateness of 98th %ile, 24-hour PMio or PMio-2.s
in epi study locations for identifying range of standards, in light of limited monitoring data
for PMio-2.5 (Section 3.37)?
       I am disposed to emphasizing the specific PMio-25 indicator/measure  in assessing level
rather than PM]0.1 find that the extent to which PM)0 reflects PM2s (ie, more so in urban areas)
                                            56

-------
may make it insurmountably problematic in using it for reviewing evidence and deliberating on the
level of a standard to protect against effects of PMio-2 5. I remain open to attempts to accomplish
that, however.
                                           57

-------