1 UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON D.C. 20460
OFFICE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD
November 24, 2009
EPA-C AS AC-10-001
The Honorable Lisa P. Jackson
Administrator
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, NW
Washington, D.C. 20460
Subj ect: Review of Integrated Science Assessment for Particulate Matter (Second
External Review Draft, July 2009)
Dear Administrator Jackson:
The Clean Air Scientific Advisory Committee (CASAC) Particulate Matter (PM)
National Ambient Air Quality Standards (NAAQS) Review Panel met on October 5-6, 2009 to
review the Integrated Science Assessment for Particulate Matter (ISA), (Second External Review
Draft, July 2009). The Panel also held a public teleconference on November 12, 2009 to discuss
and finalize its draft report. In this letter, CASAC offers general comments on the ISA, followed
by CASAC's consensus responses to the Agency's charge questions. Comments from individual
panelists are also attached.
CASAC commends EPA's National Center for Environmental Assessment for its
comprehensive effort to address the concerns and comments offered by CASAC in its review of
the first draft ISA. EPA has improved the document and focused its presentation of a substantial
body of evidence so that it now more clearly highlights the findings that are relevant to the Risk
Assessment (RA) and the Policy Assessment (PA). As mentioned in its comments on the charge
questions, CASAC also commends EPA for the continued evolution of the process for evidence
evaluation. The five-level classification of strength of evidence for causal inference has been
systematically applied; this approach has provided transparency and a clear statement of the level
of confidence with regard to causation, and we recommend its continued use in future ISAs. The
implementation of the Health & Environmental Research Online (HERO) database greatly
enhances the ISA and facilitates its evaluation by providing electronic hotlinks to citations and
study data. We support its further evolution to include abstracted, key findings.
-------
With regard to the approach taken in the ISA, we find EPA's decision to focus the
discussion of causality on PM2.5 and PMio-2.5 independently to be appropriate and supported by
the scientific evidence. As stated in our earlier review of the 1st draft ISA, the handling of PMi0
as a separate pollutant was problematic because it is comprised of the fine (PM2 5) and the coarse
(PMio-2.s) fractions. CAS AC also supports EPA's changes to the causal determinations for long-
term exposure to PM2.5 and cardiovascular effects (from "likely causal" to "causal"), for cancer
and PM2.5 (from "inadequate" to "suggestive") and for cardiovascular and respiratory effects of
short-term exposure to ultrafine particulate matter (UFP) (from "inadequate" to "suggestive").
CAS AC concurs with EPA's definition of "susceptible subpopulations" as those that have a
greater likelihood of experiencing health effects related to PM exposure; and we find that the
ISA offers a careful characterization of the factors that may contribute to increased susceptibility
to PM-induced health effects.
In terms of welfare effects, the visibility valuation and preference studies help to
demonstrate that there are levels of visual air quality which are consistently judged to be
"unacceptable" at levels well below the current PM NAAQS in a number of different study areas,
an important finding that is appropriately highlighted. CASAC concurs that Chapter 9 has been
improved by the additional discussion of optical measurement methods. Added material
provides further support for the proposed use of a new "PM light extinction" indicator, which
would be both readily measurable and reflective of the human perception of visibility effects
caused by PM pollutants. CASAC also concurs that the added discussions on climate forcing
effects and on ecological effects of PM strengthen the chapter's coverage of these topics.
Ultimately, EPA should move toward developing exposure-response relationships for ecological
endpoints, perhaps in the next review cycle for PM.
We have two major suggestions for improvement:
• The ISA could be strengthened by providing more information on the correlations of
concentrations of PMio with those of PM2.5 and PMio-2.5, preferably stratified by region of
the country, in order to better understand PMio as an indicator of the two subtractions,
PM2 5 and PMio-2.5. As currently written, the ISA uses PMio findings to support PM2 5
effects whereas the relevance of PMio findings for PMio-2.5 effects is left ambiguous.
Because of the role played by the current PMio standard, PMio findings that are
particularly relevant to PMio-2.5 should also be highlighted.
• CASAC recommends "upgrading" the causal classification for PM2 5 and total mortality
to "causal" for both the short-term and long-term time frames. There are epidemiological
studies showing a positive association of all-cause mortality with PM2.5. We
acknowledge that the association is weaker for all-cause mortality (in comparison with
cardiovascular mortality), but cardiovascular deaths are the largest component of total
-------
deaths attributable to PM. If PM2.5 effects for cardiovascular mortality are "causal," they
must also be "causal" for total mortality, and the weaker association with all-cause
mortality would be anticipated.
CASAC recommends that future assessments more fully consider evidence on the
compositional differences in PM effects in relation to both health and welfare, especially given
the emerging view that metals and carbonaceous aerosols, typically referred to as elemental and
organic carbon, are of greater concern for many endpoints.
CASAC agrees that the PM ISA will be adequate for rulemaking after EPA incorporates
our major comments and recommendations as stated above. We thank the Agency for the
opportunity to provide advice on the PM ISA.
Sincerely,
/Signed/
Dr. Jonathan M. Samet, Chair
Clean Air Scientific Advisory Committee
Enclosures: A. CASAC Particulate Matter Review Panel Roster
B. CASAC Responses to Charge Questions
C. Individual Responses to Charge Questions
-------
Enclosure A
Clean Air Scientific Advisory Committee
Particulate Matter Review Panel
CHAIR
Dr. Jonathan M. Samet, Professor and Chair, Department of Preventive Medicine, University of
Southern California, Los Angeles, CA
CASAC MEMBERS (FY 2009)
Dr. Joseph Brain, Philip Drinker Professor of Environmental Physiology, Department of
Environmental Health, Harvard School of Public Health, Harvard University, Boston, MA
Dr. Ellis B. Cowling, University Distinguished Professor At-Large Emeritus, Colleges of Natural
Resources and Agriculture and Life Sciences, North Carolina State University, Raleigh, NC
Dr. James Crapo, Professor of Medicine, Department of Medicine, National Jewish Medical and
Research Center, Denver, CO
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
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
' / Unable to participate in this review.
-------
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. Helen Suh Macintosh, Associate Professor, Environmental Health, School of Public Health,
Harvard University, Boston, MA
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 Medicine; Co-Director, Air Pollution Health
Effects Laboratory; Professor of Occupational & Environmental Health, Center for Occupation
& Environment Health, University of California Irvine, Irvine, C A
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, University of Washington, Seattle, WA
SCIENCE ADVISORY BOARD STAFF
Dr. Holly Stallworth, Designated Federal Officer, EPA Science Advisory Board Staff Office,
Washington, DC
-------
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 B
Integrated Science Assessment for Particulate Matter
(Second External Review Draft, July 2009)
CASAC Responses to Agency Charge Questions
1) Evaluation of the health evidence in Chapters 6 and 7:
a) In response to the CASAC PM Panel comments about the determinations of
causality for PMio, we have refocused the evaluation of health effects resulting from
exposure to PMio. PMio studies are now included where they provide insights into
general relationships between PM and its effects, such as concentration-response
relationships. Separate causality determinations are no longer presented for PMi0;
however, PMio study results are considered in causality determinations for PMi.s
and PMio-i.s where appropriate. Please comment on this approach to evaluation of
evidence from studies
At the April 2009 review of the first PM ISA draft, CASAC noted that findings on PMio, PM2.5
and PMio-2.5 were often discussed as if these were separate pollutants, even though the PM2.5 and
PMio-2.5 fractions comprise PMio. This led to inconsistencies in causal determinations, since
these were provided for each of the three PM indicators independently. In this draft, causal
determinations are no longer provided for PMio as they were in the initial draft of the ISA.
Motivation for dealing with PMio as a separate entity, at least in the setting of short-term
exposure, came from the continuation of the 24-hour PMio standard. Given the different
sources and the differing chemical compositions of the PM2.5 and PMio-2.5 fractions of PMio, and
to some extent the evidence for differing toxicity, there is little scientific justification for dealing
with PMio as a single entity. The approach taken in the current draft is therefore more in keeping
with the state of the science.
Findings pertaining to PMio in this 2nd draft are discussed alongside those of the two component
fractions. There was obviously (and understandably) a need to consider the extensive findings
pertaining to PMio when evaluating the evidence on PM2.5 and PMio-2.5 since both fractions
comprise PMio. Findings on PMio should have some relevance to evaluating evidence on PM2.5
and PMio-2.5- However, the relevance of evidence on PMio to PM2.s or to PMio-2.5 likely varies
across the country. For example, in some regions of the country PMio is more reflective of PM2.5
whereas elsewhere PMio is more reflective of PMio-2.5. The charge question states that "PMio
studies are now included where they provide insights into general relationships between PM and
its effects, such as concentration-response relationships. PMio study results are considered in
causality determinations for PM2.s and PMio-2.5 where appropriate." For example, Figure 6-6
(p. 6-120) shows atypical concentration-response relationship for short-term PMio exposures and
is offered as providing insight into the "PM" concentration response relationship. "PM" is a
general designation and used without sufficient specificity in this context.
-------
findings are also used in evaluation of the long-term effects of exposure to specific PM
fractions (e.g., p. 7-24). For effects of long-term PM exposure on respiratory morbidity, the
specific finding is made that "studies showing associations only with PMio were conducted in
locations where PM was predominantly fine particles, providing support for associations with
long-term exposure to fine particles" (p. 7-28, line 32; p. 7-60, line 16). PMio findings are used
in support of findings on PM2 5. However, in some of the referenced studies, associations were
found for PMio only and not for PM2.5 (p. 7-43, line 6), a finding not consistent with the cited
statement.
In the respective PM fraction Summary and Causal Determinations sections, there is little
mention of PMio. However, when findings on PMio are mentioned in this context, they are only
offered in regard to PM2.s (e.g., PMio findings are cited as strengthening the evidence for PM2.s-
related respiratory effects [p. 6-244] and for mortality effects [p. 6-317]). PMio evidence is not
used in support of PMio-2.s effects. Yet in the discussion of short-term PM exposure mortality
effects (section 6.5.2), PMio studies are cited as providing "an underlying basis for the overall
pattern of associations observed when examining the relationship between PMi0-2.5 and PM2.s and
mortality." We urge more consistent handling of the PMio evidence.
In order to better interpreting PMio findings relative to those of PM2.5 and PMio-2.5, it would be
helpful, perhaps in Chapter 3, to include and/or highlight data on the relationships of PMio with
both PM2.5 and PMio-2.5, stratified by region of the country. Also, in consideration of the role of
the current 24-hour PMio standard with respect to PMio-2.5, it would be important to better
document the relevant evidence on the relationship between PMio and PMio-2.5 across the
country.
In summary, the extensive evidence on PMio is used without specificity or consistency.
evidence is used to support the occurrence of overall effects of PM, but specifically in support of
effects of PM2.5 but not of PMio-2.5- The current PMio standard, however, is in place for control
of PMio-2.5- Nevertheless, providing causality judgments for PMio-2.5 and PM2.5, but not for PMio,
as in the current draft, is more in line with the current scientific understanding.
b) In the first draft ISA, causal judgments were made for cardiovascular morbidity,
respiratory morbidity, and all-cause mortality. In the second draft PM ISA, causal
judgments were made for cardiovascular effects, and respiratory effects, using
evidence from both morbidity and mortality. New sections were included within
cardiovascular and respiratory effects for cause-specific mortality in Chapters 6
(Sections 6.2.11 and 6.3.9, respectively) and 7 (Sections 7.2.10 and 7.3.8,
respectively) that draw upon the more complete discussions of this evidence in the
later mortality sections (Sections 6.5 and 7.6). These latter sections continue to
present evidence of cause-specific mortality, as it informs the discussions and
causality determinations for all-cause mortality. Furthermore, important results of
analyses that examined potential effect modifiers, potential confounding by
copollutants, PM-mortality concentration-response relationships, and the influence
of different modeling approaches on the PM-mortality relationship remain in
Chapter 6, Section 6.5. Considering both mortality and morbidity as part of a suite
of effects, in addition to the other considerations underlying causality judgments,
-------
In general, CASAC agrees with EPA staffs decision to link the determination of causal
judgment to cause-specific cardiovascular and respiratory mortality. In addition, CASAC
concurs with including cause-specific mortality within the outcomes considered in assessing
cardiovascular and respiratory effects. New evidence, primarily epidemiological in nature,
published since the last review cycle, makes such a judgment possible and informative. Because
much of all-cause mortality is comprised of cardiovascular and respiratory deaths, it is
appropriate to show these two groupings separately. By showing these two classes of death
separately, the reader can better appreciate both the magnitude and nature of the deaths
attributable to PM2.5.
CASAC finds that the available scientific evidence supports the judgment of a "causal"
determination for short-term and long-term effects of PM2.5 on cardiovascular effects. CASAC
also supports the determination of a causal association between short-term and long-term effects
of PM2 5 exposure and total mortality, based both on the relevant scientific evidence and the
finding of a causal association for cardiovascular mortality, which makes up a substantial
proportion of total mortality. Yet, despite strong associations with these cause-specific effects,
total mortality is less strongly and more variably associated with PM2.5, perhaps because total
mortality includes a number of categories more weakly linked to PM2 5 or not all associated with
PM2.s. Consequently, long-term exposure to PM2 5 was judged "likely to be causal" for total
mortality. This determination might be justified when considered in the context of the
epidemiological data on all-cause mortality alone, but it is not consistent with the classification
of cardiovascular mortality as "causal," given the substantial contribution of cardiovascular
mortality to all-cause mortality. Specifically, on page 2-18 the ISA states that "a causal"
relationship exists between long-term exposures to PM2.5 and cardiovascular effects, yet on 7-137
the ISA states that "the relationship between long-term PM2.5 exposure and mortality is likely to
be causal." Because the impact of PM on cardiovascular effects is judged "causal," and since
cardiovascular mortality is the largest contributor to total mortality, CASAC believes that the
association with total mortality should logically be classified as "causal" rather than "likely
causal." While cardiovascular mortality is the single greatest contributor to total mortality, other
causes are less well associated with PM2 5 and some conditions not at all. Because total mortality
is strongly related to cardiovascular events, consideration should be given to explaining why the
available studies addressing all-cause mortality appear to be conclusive, by carefully explaining
the contribution of other less strongly associated conditions. CASAC concurs with the
conclusion of a "likely to be causal" association between respiratory effects and long-term and
short-term exposures to PM2 5.
The added sections included in the ISA are effective in summarizing the recently reported
evidence and in providing a basis for the causal determinations. As the ISA process matures and
-------
continues to evolve, CASAC recommends that a table be included that lists the previous review
cycle's causal judgment determinations directly beside the current cycle determinations, with a
brief comment about what has changed in the informational database to warrant a change in the
causal classification. This table would provide the reader with a valuable summary of the
historical process, and document changes in knowledge and interpretation that form a critical
part of NAAQS evaluation.
c) We reexamined the controlled human exposure and toxicological studies of fresh
diesel and gasoline exhaust and determined that while these exposure atmospheres
contain relatively high mass concentrations of PM2.s, the particle number count
distribution was predominantly in the ultrafine size range. Therefore, for the second
draft ISA, we considered the diesel and gasoline exhaust studies, in addition to the
other considerations underlying causality judgments, when making our causal
determinations for ultrafine PM exposure and cardiovascular and respiratory
effects, resulting in a change from "inadequate" to "suggestive" for both categories.
Diesel and gasoline exhaust studies also continued to be considered as part of the
evidence for PMi.s health effects, as was done for the first draft ISA. With
consideration of this rationale, please comment on the scope of evidence considered
in the causal determinations for ultrafine PM.
CASAC agreed with EPA's rationale in moving the causal determination for UFP from
"inadequate" to "suggestive". Much of the evidence provided from clinical exposures and
toxicological studies is based on gasoline or diesel exhaust studies, and combustion of these fuels
generates high levels of-UFP. Epidemiological investigations have been less specific in their
assignment of health outcomes specifically to UFP, although an increasing number of studies are
observing effects of exposures proximal to roadsides, reflecting some component of combustion
exhaust or vehicle operation. While additional research is needed, we concur with changing the
causal determination for UFP from "inadequate" to "suggestive."
d) The CASAC PM Panel recommended that we further consider the evidence related
to PM exposure and cancer. Section 7.5 (Cancer) was revised to include evaluation
of epidemiologic studies of both mortality and incidence of cancer with exposure to
PM, as well as a brief overview of the toxicological evidence conducted using
intratracheal instillation or dermal routes of exposure to better characterize the role
of PM in mutagenicity, genotoxicity, and carcinogenicity. As a result of revisions to
this section, the causal determination for PMi.s and cancer was changed from
"inadequate" to "suggestive"; the causal determination for PMi0-2.5 remained
unchanged. Please comment on the expansion of this evaluation to include a
summary of toxicological studies using routes of exposure other than inhalation, as
well as consideration of both mortality and incidence studies.
10
-------
EPA has carefully summarized a complex literature related to both the general and specific
findings of mutagenicity, genotoxicity and carcinogenicity studies on PM and cancer. The
charge question has two components. First, publications in toxicological studies to assess
mutagenicity and genotoxicity on mechanisms for mammalian cancer research have used routes
of exposure other than inhalation. For some agents, such studies have provided insights on
potential cancer risks in humans. Thus, such data can be appropriately used for determining the
degree of certainty with regard to causality. Second, with regard to the recently reported
findings of epidemiological studies, these have included large, well assessed populations and
they have shown associations with cardiovascular and respiratory outcomes. The same approach
has been used for lung cancer, although interpretation is complicated by the possibility of
potential confounding by smoking. With the exception of one relatively large European study
that was null, these studies show an association of PM exposure with cancer risk that is robust
after controlling for smoking. These studies provide support for the "suggestive" classification,
particularly in the context of other lines of evidence reviewed in the ISA.
2) Revisions to Chapters 1 (Introduction), 4 (Dosimetry), and 5 (Mode of Action): Changes
to these chapters included expansion or clarification in a number of areas, as presented
in more detail in the recent letter from EPA's Administrator. These included expansion
or addition of sections on the history of the previous PM NAAQS review in Chapter 1;
deposition and clearance of particles in Chapter 4; and epigenetics, lung development,
atherosclerosis and consideration of acute and chronic responses in Chapter 5. Please
comment on the revisions to these chapters.
We commend the organizers and authors of this document for providing the CASAC PM Panel
with well organized and readily readable texts for Chapters 1, 4, and 5 that are, for the most part,
thorough and accurate in presenting the peer-reviewed literature most relevant to an Introduction
(Ch. 1), Dosimetry (Ch. 2), and Mode of Action (Ch. 3). The amount of text on UFP could be
condensed, especially in Chapter 5, where the potential to cause adverse effects is so
prominently featured, along with the implication that particle number concentration is the most
important metric for risk. Some additional discussion of still weak evidence in support of this
hypothesis is warranted. There is inadequate discussion of the usefulness of the limited data on
ambient air UFP concentrations currently available. The data are essentially limited to particle
number concentrations, which make no distinction with respect to particle composition. The
likelihood of effects is likely to differ among UFPs formed as condensation aerosols of sulfuric
acid droplets, metal oxides, elemental carbon, and organic carbon, and with the changes in
composition and particle size with residence time in the ambient air.
3) Evaluation of susceptible population groups in Chapter 8: In response to the CASAC
PM Panel comments, we have now focused specifically on susceptible subpopulations
(defined for this ISA as those subpopulations that have a greater likelihood of
experiencing health effects related to PM exposure) in Chapter 8. The introduction of
this chapter was revised to include this new definition of susceptible subpopulations,
recognizing that the terms susceptible and vulnerable have sometimes been used
interchangeably in the literature, and in other cases have been used to represent two
different categories (i.e., biological factors [e.g., age gender, etc.] vs. non-biological
factors [e.g., SES, differential exposure, etc.]), resulting in the lack of a clear and
consistent definition. The discussion was reorganized to improve the characterization of
11
-------
factors that may contribute to increased susceptibility to PM-induced health effects.
Each section was also revised to include the different exposure durations (short- and
long-term) and PM size fractions examined in the studies discussed within each
subsection of the chapter. Please comment on the organization and presentation in
Chapter 8 of evidence regarding susceptible subpopulations.
The organization and presentation of Chapter 8 is complete, clear and well-organized. The
authors have been responsive to CAS AC's critique of the first draft of the chapter. The
definition of "susceptible" is clearly stated with a rather complete listing of how the term has
been used in the literature in the past. We recommend that the authors consider identifying those
definitions which are most timely and relevant for the purposes of this chapter.
Table 8-2 introduces all susceptibility factors that have been evaluated, as well as the particle
size fraction evaluated and the duration/length of exposures. Table 8-2 is followed by a more
detailed description of the evaluations that were made. This is a highly appropriate organization
that could be improved by the addition of a column to the table which briefly summarizes the
impact of each susceptibility factor on health impacts and/or observed health effects. The title of
the table "Susceptibility Factors," gives the impression that all factors listed determine
susceptibility, but some of those listed, such as gender, do not. Therefore, the title could be
modified to "Susceptibility Factors That Have Been Evaluated" to more accurately reflect the
contents of this table. The addition of a column to the table to indicate the results of each
evaluation with a simple plus/minus summary approach would further improve the presentation.
4) Revisions to Chapter 3 on Source to Exposure: Consistent with revisions made to the
health effects chapters, Chapter 3 was revised to clarify that PMi0 incorporates both
PM2.s and PMi0-2.s and reorganized to begin with PM2.s and PMi0-2.s, followed by PMio,
where applicable. The discussion of measurement techniques and chemistry of PMio-2.5
has been expanded in Sections 3.4 and 3.5, in response to CASAC comments. In
addition, Section 3.8 on human exposure to PM has been reorganized and expanded to
better characterize the evidence and provide useful information for interpretation of
epidemiologic studies. We would appreciate comments from the CASAC PM Panel on
these revisions.
The Chapter is a comprehensive summary of the current state of the evidence on PM sources,
ambient aerosols, monitoring and modeling methods, concentrations and exposures. The
Chapter generally reflects the state of science in a clear, comprehensive, and well organized
fashion. Further, the Chapter revision appropriately incorporates comments from the April
review. Of note, the relationships among PM2.5, PMio-2.5, and PMio were clarified, although
some additional information could be pulled forward from the Appendix (see below). The
inclusion of additional information on UFP, PM components, and PMio-2.5 monitoring methods
was also an improvement. Similarly, the organization and content of the exposure section was
greatly improved, and the relevance of exposure findings to interpretation of the results of
epidemiologic studies was more fully addressed. While significantly improved, further
improvements could be made to the exposure section to help with the organization, seeming
overlap with earlier sections, and clarity, as discussed below.
12
-------
Specific comments:
• The Chapter would benefit from a more detailed discussion of the relations among
ambient PMio, PM2.5, and PMio-2.5, particularly for the 15 cities selected for more detailed
analysis. While this information is in the Appendix, it would be helpful to bring a
summary of the findings into the main chapter.
• The distinction between primary and secondary organic aerosols should be expanded to
discuss the potential for oxidation of primary organic particles which makes them
become similar to secondary organic aerosol.
• Standard Reference Materials (page 2-20, line 20): It says "To date, there are few
standard reference materials ..." This sentence should be reworded to make clear that we
currently do not have a way to assess mass concentration measurement accuracy, as there
are no standard reference materials (SRMs) or equivalent standards (to our knowledge)
that provide a test of accuracy or airborne PM mass. There are, however, SRMs that are
useful for testing the accuracy of the analytical methods for PM composition.
• Policy Relevant Background (PRB): The EPA promulgates a health-based standard
irrespective of the background concentration of particles and thus, the PRB primarily has
implications for implementation. If the PRB increases because of sources outside of the
United States, there will be greater difficulty in attaining concentrations that are fully
protective of public health. The modeling approach to establishing PRB has some
appeal, but given the inaccuracy of emissions inventories in the US, where substantial
effort and resources have been expended to develop them, there is little likelihood that we
can adequately model emissions for the rest of the world. Consequently, the ISA might
also offer comparisons between the model-based approach and the remote site approach
to assess the certainty of the model-based estimation of the PRB.
• Exposure Section: Given the relevance of the measured PM concentrations in the earlier
sections of the Chapter, and for interpreting the epidemiological studies and the Health
RA, discussion of exposure measurements should be placed before that for exposure
modeling. Further, the exposure measurements section should include discussion of new
methods to measure (1) personal exposures to total PM, (2) PM of outdoor origin, (2)
ambient concentrations outside individuals' homes, and (3) indoor PM exposures to total
PM and to PM of outdoor origin). In this context, studies using tracers of outdoor or
regional pollution can be introduced.
o The discussion of each topic should be consolidated into single sections, with
references to this section made as necessary in later sections. As currently
written, certain topics, such as spatial variability and particle infiltration, appear
repeatedly in several subsections. The discussion on these topics would be more
focused and less confusing if consolidated.
o In Section 3.8.4.3, it would be helpful to clarify which component of PM is
discussed. Further, within each subtopic or heading, separate or otherwise distinct
discussions for PM2.5, PMio-2.5 and the PM components would improve clarity.
13
-------
o Section 3.8.4 (Exposure Assessment Studies at Different Spatial Scales) and its
subheadings are misnamed. While the heading titles do provide parallels with
earlier sections of Chapter 3, they do not accurately describe the contents. For
example, section 3.8.4.1 does not actually discuss urban scale ambient PM
exposure, but issues related to exposure error, exposure modeling, and tracers of
ambient particles.
o It would be helpful to include a subsection in the beginning of the section,
perhaps in Section 3.8.1 (General Exposure Concepts) that summarizes the range
of measured exposures to PM2.5, PMio-2.5, and PMio, their composition, and
differences by city, susceptible group, and season.
o Section 3.8.6: Other factors that could affect exposures should be noted, such as
home ventilation " and activity patterns.
5) Integrative Synthesis in Chapter 2: The CASAC PM Panel recommended expanding
Chapter 2 to include all important findings of the PM ISA. The integration of health
evidence in Chapter 2 was reorganized to focus on effects of PM2.s, PMio-i.s, and
ultrafine particles, and was expanded to include discussions of effects for which a
"suggestive" causality determination was drawn. New integration sections were added
that combine the evidence for health effects of PM2.s, PMio_2.s, and ultrafine PM across
exposure durations. In addition, these integration discussions incorporated evidence
related to mode of action, dosimetry, atmospheric chemistry, and exposure assessment
to the extent possible. When appropriate, figures were added that summarize the
overall U.S. and Canadian epidemiologic evidence for specific size fractions and
exposure durations, along with the concentrations reported in the studies or provided
by study authors. A new section was also added to Chapter 2 that contains policy-
relevant considerations, including summaries for the evidence for susceptible
subpopulations, lag structure of associations in epidemiologic studies, and the PM
concentration-response relationship. Please comment on these revisions and additions
to the integration of health effects evidence in Chapter 2.
Overall, CASAC commends EPA for an excellent draft of Chapter 2. The chapter appropriately
addresses CASAC's previous comments from the April review of the 1st draft ISA. The
revisions as described in the charge questions are appropriate and enhance the content and
structure of the chapter.
Although the current draft of Chapter 2 is excellent, it could be strengthened with selected
revisions. The main technical points that should be addressed in finalizing this chapter are: (a)
more attention is needed to the role of variability in the composition of PM by size fraction with
respect to both health and welfare effects; (b) more analysis and discussion of PM effects on
climate is needed, particularly with regard to the role of PM composition and its implications for
adverse or beneficial effects; (c) additional synthesis is needed to more clearly tie together
material related to PM characteristics and impacts on both health and welfare, including strength
of association, and the implications of uncertainties; (d) there should be more effort to move
toward exposure-response relationships for ecological effects endpoints; and (e) the conclusion
regarding only a "likely" causal relationship between long-term exposure to PM2.5 and mortality
14
-------
should be carefully revisited, since there is evidence that the relationship exists with more weight
of evidence than conveyed in the document.
Section 2.3 should have one summary table of all health effects, and the discussion should start
with those effects for which there is the strongest weight of evidence, moving to those with
weaker or no weight of evidence (rather than vice versa as in the second draft ISA). Some
portions of Chapter 2 may be too detailed, such as the text on pages 2-24. There are places
throughout Chapter 2 (e.g., page 2-13) where citation of later sections of the report would help
readers in finding more detail to support the summary statements. Although Figures 2.1, 2.2, and
2.3 are generally very useful, they would be easier to interpret if disaggregated into figure panels,
each addressing only one outcome. Furthermore, there appears to be an error in the scale at the
bottom of Figure 2.3. Although there is a lower correlation of PMi0 with PM2.5 (Page 2.5, lines
14-19), the correlation for PMio is nonetheless remarkably high and should be interpreted more
appropriately.
The discussion of PM and climate concludes with an ambiguous summary statement that "a
causal relationship exists between PM and effects on climate." This use of the term "effects" is
inconsistent with causality statements elsewhere in the ISA, where the PM effects being
considered were consistently and unquestionably adverse. In the case of climate, PM can exert
both positive and negative climate forcing influences. Some discussion should be offered to
frame and further clarify this issue, perhaps in Chapter 1.
CASAC recommends that EPA add more material regarding PM composition, an overall
synthesis of both health and welfare effects, and a synthesis of the most significant uncertainties
and their implications for the robustness of conclusions.
6) Welfare effects evaluation in Chapter 9: Several revisions were made to the evaluation
of the welfare effects evidence in Chapter 9, in response to the CASAC PM Panel
comment, to focus further on effects on climate and ecosystems and include further
evaluation of urban visibility evidence, where possible. In addition, as recommended by
the CASAC PM Panel, key findings and conclusions from this chapter were
incorporated in Chapter 2. The discussion of PM effects on climate was increased with
substantially more detail from recent publications, including discussion of specific
climate forcing effects from individual PM components and size fractions. The
discussion of ecological effects was also reorganized to focus on the types of effects and
effects of individual components. For the effects of PM on visibility, new material was
added including sections on direct optical measurements and the value of good visual
air quality. Please comment on the effectiveness of the reorganization and revisions
regarding welfare effects.
15
-------
CASAC was pleased to note the effective addition of sections on "Ecological and Welfare
Effects" in the Integrative Overview summary discussion in Chapter 2. CASAC also agreed that
the revisions to discussions of PM welfare effects in Chapter 9 strengthen the chapter, improve
its clarity and are directly responsive to previous CASAC comments on the 1st Draft ISA.
Several minor revisions would be helpful to improve accuracy in some sections (see individual
panelist comments for detail), but no major revisions are needed.
Newly added details on climate, ecological and materials effects of PM and its various
components provide more comprehensive coverage of these important subject areas; however,
concentration-response functions for climate and ecological endpoints will be needed for future
ISAs. Information in Chapter 9 on optical measurement methods provides added support for the
proposed use of a new "PM light extinction" indicator, which is readily measurable and directly
reflects the human perception of visibility effects caused by PM. The section on night-time
visibility in Chapter 9 is important and should be brought forward to the summary chapter
(Chapter 2). The information presented on visibility valuation and preference studies helps to
demonstrate that visual air quality is consistently judged to be "unacceptable" at levels well
below the current PM NAAQS in a number of different study areas. The findings in the ISA are
sufficient to propose levels of PM light extinction above which public welfare is affected
adversely. Finally, it is important to recognize, as the ISA does, the very substantial welfare
effects from airborne PM, NOX and SOX are addressed in a separate NAAQS review.
16
-------
Enclosure C
Compendium of Individual Comments
CASAC Particulate Matter Review Panel
Integrated Science Assessment for
Particulate Matter (Second External Review Draft, July 2009)
Ashbaugh Comments (Dr. Lowell Ashbaugh) 2
Avol Comments (Mr. EdAvol) 4
Brain Comments (Dr. Joe Brain) 7
Cascio Comments (Dr. Wayne Cascio) 8
Cowling Comments (Dr. Ellis Cowling) 24
Frey Comments (Dr. Chris Frey) 25
Grantz Comments (Dr. David Grantz) 27
Helble Comments (Dr. Joseph Helble) 36
Henderson Comments (Dr. Rogene Henderson) 39
Hopke Comments (Dr. PhilHopke) 40
Lippmann Comments (Dr. Mort Lippmann) 42
Malm Comments (Dr. William Malm) 47
Pinkerton Comments (Dr. Kent Pinkerton) 49
Phalen Comments (Dr. Robert Phalen) 50
Poirot Comments (Mr. Rich Poirot) 52
Russell Comments (Dr. Armistead (Ted) Russell) 56
Speizer Comments (Dr. Frank Speizer) 59
Suh Comments (Dr. Helen Suh) 63
Vedal Comments (Dr. Sverre Vedal) 66
-------
Ashbaugh Comments (Dr. Lowell Ashbaugh)
Charge Question 6:
Several revisions were made to the evaluation of the welfare effects evidence in Chapter 9, in
response to the CASAC PM Panel comment, to focus further on effects on climate and ecosystems
and include further evaluation of urban visibility evidence, where possible. In addition, as
recommended by the CASAC PM Panel, key findings and conclusions from this chapter were
incorporated in Chapter 2. The discussion of PM effects on climate was increased with substantially
more detail from recent publications, including discussion of specific climate forcing effects from
individual PM components and size fractions. The discussion of ecological effects was also
reorganized to focus on the types of effects and effects of individual components. For the effects of
PM on visibility, new material was added including sections on direct optical measurements and the
value of good visual air quality. Please comment on the effectiveness of the reorganization and
revisions regarding welfare effects.
The revisions are quite good; the authors have done an excellent job of incorporating comments
from the earlier review. The integration of welfare effects into Chapter 2 works well. I would
like to see a short paragraph in Chapter 2 summarizing the effects of PM on nighttime visibility.
I especially liked the section on nighttime visibility in Chapter 9 on page 9-8. The paragraph on
lines 22-30 of page 2-37 is not clearly written. Twice in succession it states that the use of a no-
threshold log-linear model is supported, but then cites other studies that suggest otherwise. It
would be good to revise this paragraph to more clearly state - well, I'm not sure what. Probably
that more research is needed.
Chapter 9 is also much improved. The organization is good, and small details make it easier to
read. For example, the spatial contour plots are much easier to compare now that the contour
scales are similar for related figures. The haze trends figures (9-26 & 9-27) would benefit from
using a color other than yellow, though, which does not show up clearly. Or perhaps the yellow
(up) and light green (down) arrows could have a black border. Then they would also be
distinguishable from the red and dark green when copied in B/W.
I like Section 9.3.1 but the figures would benefit from better reproduction. Most of the figures
copied from the CCSP SAP2.3 did not reproduce well (also, the caption for Figure 9-70 contains
a typo - "Arcic" instead of "Arctic"). Is it possible to obtain the originals?
The individual components discussion is well organized and easy to read. I'm familiar with one
reference that could be added to the section on smelters and roadsides (page 9-245). The
reference is "Resuspension of Soil as a Source of Airborne Lead near Industrial Facilities and
Highways," by Thomas M. Young, Deo A. Heeraman, Gorkem Sirin, and Lowell L. Ashbaugh,
Environmental Science & Technology 2002 36 (11), 2484-2490. It seems to be relevant to the
discussion. Here's the abstract:
Abstract: Geologic materials are an important source of airborne particulate matter less
than 10 um aerodynamic diameter (PMio), but the contribution of contaminated soil to
concentrations of Pb and other trace elements in air has not been documented. To
examine the potential significance of this mechanism, surface soil samples with a range
of bulk soil Pb concentrations were obtained near five industrial facilities and along
-------
roadsides and were resuspended in a specially designed laboratory chamber. The
concentration of Pb and other trace elements was measured in the bulk soil, in soil size
fractions, and in PMio generated during resuspension of soils and fractions. Average
yields of PMio from dry soils ranged from 0.169 to 0.869 mg of PMi0/g of soil. Yields
declined approximately linearly with increasing geometric mean particle size of the bulk
soil. The resulting PMio had average Pb concentrations as high as 2283 mg/kg for
samples from a secondary Pb smelter. Pb was enriched in PMio by 5.36-88.7 times as
compared with uncontaminated California soils. Total production of PMio bound Pb from
the soil samples varied between 0.012 and 1.2 mg of Pb/kg of bulk soil. During a
relatively large erosion event, a contaminated site might contribute approximately 300
ng/m(3) of PMio-bound Pb to air. Contribution of soil from contaminated sites to airborne
element balances thus deserves consideration when constructing receptor models for
source apportionment or attempting to control airborne Pb emissions.
-------
Avol Comments (Mr. Ed Avol)
1) Evaluation of Chapters 6 & 7 Health Evidence
(a) Evaluation of evidence from PMu) studies (lead discussant Vedal)
The revised approach is well-presented and understandable. The authors and staff have done a
commendable job in improving the accessibility of data, readability of the document, and
documentation of the included information over the previous version.
I did not read the entire document (so I am not certain of this), but I did not readily see any
discussion (or reference to a discussion) announcing, justifying, or explaining this decision to
"collapse" PMio into these other categories. For example, this might be justified on a scientific
basis (high inter-correlation with coarse PM, for example) or possibly even a regulatory one
(research evolving to smaller particle sizes). Did I miss this somewhere?
(b) inclusion of cause-specific mortality in the development of causal judgments (lead
discussant Cascio)
I agree with the decision to link the determination of causal judgment to cardiovascular and
respiratory mortality, since these have been very active and productive areas of research since
the last review cycle. The added sections are effective in summarizing the cumulative recent
evidence and providing a basis for the causal determinations. As the ISA process mature and
continue to evolve, it might be worth considering inclusion of a summary table that lists the
previous review cycle's causal judgment determinations directly beside the current cycle
determinations, with a brief comment about what has changed in the informational database to
warrant a change in causal determination, This would provide the reader, the Staff, and the
public with a readily trackable means of documenting changes in knowledge and interpretation
that form a critical part of NAAQS evaluation.
Response to Charge Question Ic
(c) Scope of evidence considered in causal determination of UFP (lead discussant Avol)
I agree with the rationale and thinking of the Staff and authors on moving the causal
determination for UFP from "inadequate" to "suggestive". Although there is growing research
activity in the effects of PM in this size range, there is still much that is not understood,
appreciated, or being considered. This is complicated by the difficulty in generating and/or
characterizing "reproducible" or "representative" UFP for study use at ambient exposure-
appropriate concentrations. Additional concerns include observations that particle surface
reactivity may be strongly related to fresh combustion and/or new particle formation, and
observations that smaller particles seem to have increased toxicological properties.
-------
(d) Inclusion of cancer mortality and incidence, additional routes of exposure (lead discussant
Speizer)
I agree with the expansion decision for both routes of exposure and consideration of both
mortality and incidence studies.
2) Chapters 1 (Introduction),4 (DosimetryXS (Mode of Action) revisions
Comments on Chapter 1
(I repeat this comment from la, since it is relevant to the Chapter 1 Introduction)
I did not read the entire document (so I am not certain of this), but I did not readily see any
discussion (or reference to a discussion) announcing, justifying, or explaining this decision to
"collapse" PMio into these other categories. For example, this might be justified on a scientific
basis (high inter-correlation with coarse PM, for example) or possibly even a regulatory one
(research evolving to smaller particle sizes). Did I miss this somewhere?
P. 1-2, first bullet - Three categories of exposures (short-term, chronic, and peak) are described
in this carefully crafted opening list of review considerations.. .but short-term and chronic
exposures are the focus of the documents. Should some comments, concern, consideration be
focused on peak exposures, even if just to say we don't know much about it or need to learn
more?
Comments on Chapter 4
P4-2, line 31 - Manufactured nanoparticles are not considered in this chapter (or in the review.
Presumably manufactured nanoparticles lie somewhere on the continuum from completely
unrealistic exposures to completely relevant ones.. .Are manufactured nanoparticles being
considered as "occupational" exposures? Regardless, some thought should be given to the
potential for population exposure to these particles, and what might be learnable from these
engineered exposures (which may not be all that different from laboratory-generated exposures
of TiO2 or other specifically generated artificial particles that serve an important purpose in
identification of mechanisms and pathways). Would knowledge and consideration of their
potential health effects (as a function of their particle size) be an appropriate section in an UFP
chapter?
Comments on Chapter 5
I think this chapter was very well done and especially appreciated Section 5.6, which
summarized new results AND provided references to the appropriate document section to read
more detail about the individual finding.
The final text page of the chapter (5-31) "gaps in knowledge" is worthy of a section title or
tabulation. In a larger document sense, this final summary and assessment of knowledge gaps
would be useful additions to most every chapter in the ISA.
General Comments on PM ISA. 2nd Draft. Avol
-------
In a large, data-rich, and complex document such as this, it would seem helpful to establish a
consistent presentation paradigm that is followed throughout the relevant chapters, so that the
reader can move from chapter to chapter, in or out of sequence, and still have some expectation
of finding a consistent set of validators or signposts with which to review, interpret and integrate
the presentations. Substantive progress has been made towards that objective in the ISA, but
there still remain some inconsistencies that provide unnecessary obstacles to the reader. Here are
two examples of this:
(1) A consistent PM-size-oriented summary presentation would help to establish a standardized
format and framework for review. Accordingly, it would be useful if, in each Summary and
Causal Determinations section of each chapter, health outcome, etc, a paragraph were denoted
for PM2.5, PMio-2.5, and UFP and a brief summary statement regarding causality judgment - for
each size-cut - were made. (This comment motivated by the observation that UFP summary
statements seem to be sporadically missing, such as in section 7.4.3).
(2) In Chapter 2 , there is (almost) a developed (and pragmatically useful) standardized listing of
health outcomes, in five categories (cardiovascular; respiratory; mortality; reproductive and
developmental; cancer, mutagenicity, and genotoxicity) that could be consistently evaluated
against each of the pollutants under review for a current judgment of causal determination (from
not likely to causal). These tabular summaries could then form the basis for guideposts to areas
of needed future investigations, markers for progress from review to review, etc.
I suspect there is an error in the listed range of effects estimates in Figure 2-3 (P2-29); it is
difficult to believe that the effects estimates range from -12 to +32.
7.2.11 Summary & Causal Determinations (of Cardiovascular and Systemic Effects)- Much
improved and clearer, providing a basis for what was known prior to the current review, and
what the new information shows. Summary judgments for PM2.5, PMio-2.5, and UFP all seem
reasonable, but the supporting comment for UFP ("... due to a few studies being conducted
without gaseous co-pollutants..." seems unnecessary and inaccurate, since the current breadth of
studies reporting UFP and health associations is minimal, independent of adjustments for the
presence or absence of gaseous co-pollutants. There may be some suggestions of effects and a
valid concern for additional research, but at the time of this review, the cumulative import of
observed effects in the assembled body of work is unclear and in need of additional
investigation.
There still is appears to be an (unnecessary) tendency to discuss study after study after study at a
level of detail that could be more appropriately placed in the annex. The main section of this
document would be more useful as an integration of current knowledge, not an exhaustive listing
of it; additional materials could and should be provided in the annex.
P7-27, lines 28-29 - "Evidence from these studies alone is inadequate to infer the presence or
absence of a causal relationship due to a few studies being conducted without gaseous
pollutants." This may be true, but can't the same comment be made for the studies reporting
effects of PMio, PM2.5, and PMi0-2.5? Why make this point for UFP only?
-------
Brain Comments (Dr. Joe Brain)
ISA-PM: Charge Question 5: Integrative Synthesis in Chapter 2
Overall, chapter 2 is much improved. Especially, making it a synthesis of the entire document
makes sense. This is consistent with CASAC's recommendation regarding the first draft of the
ISA. In particular, I like the delineation of the three critical size ranges of PM: coarse, fine, and
UFP. These three size ranges are now prominent throughout and adequately discussed in terms
of exposure and response. The inclusion of welfare considerations and ecological effects is well
integrated into the chapter as a whole. In toto, this is an extremely valuable component of the
ISA. It is a section which will be heavily used by policy makers, the scientific community, and
the lay public. The authors are to be congratulated on producing a summary which is both
sophisticated as well as accessible.
Section 2.4, "Policy Relevant Considerations," is done well. The role of susceptible populations
is emphasized. This is important since these individuals make a dominant contribution to the
observed health effects.
-------
Cascio Comments (Dr. Wayne Cascio)
Charge Question la
General Comments:
1) The outstanding effort of the staff at NCEA to address the concerns and comments of the
CASAC PM panel is very much appreciated, and has improved the document and
focused the information in such a way as to more clearly justify the message it is
designed to convey.
2) As discussed previously by the CASAC PM panel the determination of causality for
PMio is problematic as this fraction contains both the fine and coarse (thoracic PM)
fraction, and its risk appears to be more strongly determined by the fine fraction.
Consequently, the approach to focus the discussion of causality determination on PM2.5
and PMio-2.5 independently is appropriate, but appears to be at odds with the current
standard. How does EPA intend to reconcile this discontinuity?
3) Nevertheless, the separation of PM into PM2.5 and PMio-2.5 is strongly supported by
scientific data.
4) The concept that the abundant data on PMio might serve to selectively support the
causality determinations for PM2.5 and PMio-2.5 is problematic. As described above
contains both fractions and there will always be uncertainty as to whether effects
measured in response to PMio are related to the PM2.5 fraction, the PMio-2.5 fraction or
some combination of the two.
Charge Question Ic
General Comments:
1) The outstanding effort of the staff at NCEA to address the concerns and comments of the
CASAC PM panel is very much appreciated, and has improved the document and
focused the information in such a way as to more clearly justify the message it is
designed to convey.
2) In Chapters 6 and 7 a comprehensive list of studies reporting health effects, toxicity of
diesel emission and gasoline emission PM is provided. The rationale to include diesel and
gasoline emissions as evidence of effects of UFP is reasonable given the strong
association between PM mass from these source and particle number. Of relevance to this
question are the results of our recently completed California Freeway Study that showed
that particle number was most strongly associated with health effects endpoints in
subjects exposed to traffic related pollutants on two Los Angeles freeways for 2 hours.
3) Based on the weight of the evidence with the inclusion of diesel and gasoline emission
data the change in the causal determination for UFP from "inadequate" to "suggestive" is
appropriate.
4) The presentation of diesel, gasoline and wood smoke emission data independently
provides a means to quickly find relevant health effects data linking a specific exposure
type to health effects. Such an approach was useful in judging the strength of
associations and determination of causation.
-------
Charge Question Id
General Comments:
1) The outstanding effort of the staff at NCEA to address the concerns and comments of the
CASAC PM panel is very much appreciated, and has improved the document and
focused the information in such a way as to more clearly justify the message it is
designed to convey.
2) In the 2nd external review draft Section 7.5 provides a more detailed presentation of
epidemiological and toxicological data. The information is well integrated. This has
improved the quality of information regarding the association between PM exposure and
cancer.
3) In the first ISA the presentation of the possible link between PM exposure and cancer
was judged as "inadequate". Now with better integration of epidemiological and
toxicological data the causal determination has been changed to "suggestive". This
change is appropriate based on the presented data. While toxicological studies included in
this section were conducted with exposure routes other than inhalation these studies are
important for providing a mechanistic basis to justify the biological plausibility of a
causal determination.
Chapter 6
Specific Comments:
SEC/PAGE
PARA/LINE
REVIEW COMMENTS
6.2.1/2
1. 1 -5
One can add that decreased HRV may precede some
clinically important arrhythmias such as atrial
fibrillation, and SCD in high risk populations. (Chen
PS, Tan AY. Autonomic nerve activity and atrial
fibrillation. Heart Rhythm 4(3 Suppl):S61-4, 2007;
Thong T, Raitt MH. Predicting imminent episodes of
ventricular tachyarrhythmias using heart rate. Pacing
Clin Electrophysiol 30:874-84, 2007; Sandercok GR,
Brodie DA. The role of heart rate variability in
prognosis for different modes of death in chronic heart
failure. Pacing Clin Electrophysiol 29:892-904, 2006)
6.2.1/2
2. 5-7
The sentence, "SDDN generally reflects..." can be
rewritten, "SDDN generally reflects the overall
modulation of HR by the autonomic nervous system,
whereas rMSSD and frequency variations in HR
-------
generally reflect parasympathetic activity".
6.2.1/2
2.7-9
The sentence, "LF is predominately..." can be rewritten
as "LF is predominately determined by both
sympathetic and parasympathetic tone and increased
LF/HF indicates sympathoexcitation,..."
6.2.1/6
1. 5-6
In the study by Yeatts et al. (2007, 091266) it is true
that rMSSD but this was not statistically significant.
P=0.16. Likewise, pNNSO did increase in response to
PM2.5-10 but this result also did not reach statistical
significance. P=0.07.
6.2.1/6
2.7-9
In the study by Langrish et al. (2009, 191908). Did the
authors confirm that the facemask did not affect the
rate or depth of breathing. If not then it is possible that
the results could have been influenced by the
facemask.
6.2.1.2/13
2,2
Space between "of and "20"
6.2.2.1/21
1,5-7
The sentence, "VPBs are ..." should be rewritten.
Consider the following, "VPBs are spontaneous beats
originating from either the right or left ventricles. VT
refers to 3 or more ventricular beats in succession at a
rate of 100 beats per minute or greater, while VF is
characterized by rapid and disorganized ventricular
electrical activation incapable of generating an
organized mechanical contraction or cardiac output.
6-22
1, 10 and 2, 1
The sentence, "Pathophysiologic mechanisms..."
should be rewritten. Consider the following,
"Pathophysiologic mechanisms underlying this
established cause of sudden cardiac death include
activators and facilitators of arrhythmia such as
electrolyte abnormalities, modulation of the autonomic
nervous system, membrane channels, gap junctions,
oxidant stress, myocardial stretch and ischemia acting
on myocardium conditioned for arrhythmia by altered
by cellular and tissue architecture, fibrosis, and
intracellular calcium handling."
10
-------
6-27
6-27
6-29
6-29
6-34
6-39
6-39
6-40
6-48
6-62
Table 6-4/65
6-69
6-75
6-77
1,7
2,1
2, section
heading
1,4
1, section
heading
1,1
1,1
1, 1
1,4
3,1-5
2,5
2,1
The sentence, "Such beats. . ." should include the
conduction system as a source of arrhythmia. For
example, "Such beats can arise in the atria, AV node,
conduction system, or the ventricles.
"nonsmoking" to "non-smoking"
Change "ECG abnormalities indicating arrhythmia" to
"ECG changes associate the modulation of
repolarization".
A final statement might include, "Alternatively, novel
non-invasive electrocardiographic telemetry systems
might be used more effectively to assess the
relationship between air pollutants and arrhythmia in
patients with history of arrhythmias or at risk for
serious arrhythmia."
Change "ECG abnormalities indicating ischemia" to
"ECG changes suggestive of increased ischemia".
Write out "yr" as "year"
Write "IHD" as "ischemic heart disease"
Consider changing the first sentence to "... examine the
association between 2 measures of vascular reactivity,
non-endothelium dependent nitroglycerin mediated
reactivity and endothelium-dependent flow-mediated
reactivity, and ambient. . ."
Write "ET" as "endothelin-1"
"QA interval" should be defined.
The closed bracket after m3 is supra-scripted and
should not be.
Continue "on markers of inflammation..." directly on
the previous line without indentation.
Substitute "highway police" for "troopers"
Write "wood smoke" instead of "WS"
11
-------
6-81
6.2.10.5
6-116
6-123
6-128
6-138
6-241
References
6-336
6-337
6-338
1,2
2,7
2
1,17; 2, 7
2, last line
1, last line
1,
Write "thrombin anti-thrombin complex" instead of
"TAT"
"associate" should be written "associated"
Consider "Co-pollutant" rather than "Copollutant".
Consider "multi-city" rather than "multicity".
Add to the final sentence such that it would read, "
Furthermore, the HRV result. . .including age and pre-
exiting conditions and genetic polymorphisms
modulating responses to oxidative stress."
Consider "co-pollutant" rather than "copollutant".
"Copollutant" is used throughout the document. If such
a change is accepted then other examples should be
found.
Consider changing "nonasthmatics" to "non-
asthmatics".
Aekplakorn - add Journal title
Analitis - add Journal title
Andersen - add Journal title
Andersen - add Journal title and volume
Atiga - add Journal title
Atkinson - - add Journal title, volume and pages
Bastain - add Journal title
Bell - add Journal title
12
-------
6-339
6-340
6-341
6-342
6-343
6-344
Berger - add Journal title
Brauner - add Journal title
Brook - add Journal title
Burnett - add Journal title, volume and pages
Caligiuri - add reference. "Presented at" ?
Carlsten - add Journal title
Checkoway - add Journal title, volume and pages
Chevalier - add Journal title
Chuang - add Journal title
Ciencewicki - add Journal title
D'Ippoliti - add Journal title
Daniels - add Journal title, volume and pages
Danielsen - add Journal title
De Bruin - add Journal title
De Haar - add Journal title
Dockery - add Journal title, volume and pages
Dominici - add Journal title, volume and pages
Dusek - add Journal title
Fairley - add Journal title, vol. and pages
Fakhri - get reference
Farraj - add Journal title
Franklin - - add Journal title, vol. and pages
Gent - add volume
Gerlofs-Nijland - add volume
13
-------
6-345
6-346
6-347
6-348
6-349
Gerlofs-Nijland - - add Journal title
Gilmour - add Journal title
Godleski - add Journal title
Gong - add Journal title, volume, and pages
Hanigan - add Journal title
Happo - add Journal title
Harkema - add Journal title, vol. and pages
Harrod - add Journal title
Hogervosrt - add Journal title
Holguin - add Journal title
Host - add Journal volume and pages
Inoue - add Journal volume
Ito - add Journal title, vol. and pages
Ito - add Journal title, vol. and pages
Ito - add Journal title
Jalaludin - add Journal title
Janes - add Journal title
Johnson - add Journal title
Katsouyanni - add Journal title, vol. and pages
Klein-Patel - add Journal title
Kleinman - add Journal title
Kleinman - add Journal title
Klemm - add Journal title, volume and pages
Laupacis - add Journal title
14
-------
6-350
6-351
6-352
6-353
6-354
Le Tertre - add Journal title, vol. and pages
Lee - add Journal title
Levy - add Journal title
Li - add Journal title
Lin - add Journal title
Lippmann - add whole reference
Lippmann - add whole reference
Lisabeth - add Journal title
Liu - add Journal title
Ljungman - add Journal volume and pages
Low - add Journal title
Mar - Epidemiol. Is redundant
Mar - add Journal title
Mar - add Journal title, vol. and pages
Matsumoto - add Journal title
McCreanor - add Journal title
Metzger - add Journal title
Mills - add Journal title
Murata - add Journal title
Mutlu - add Journal title
Nadziejko - add Journal title, vol. and pages
Nadziejko - add Journal title
O'Connor - add Journal title
O'Neill - add Journal title
15
-------
6-355
6-356
6-357
6-358
6-359
Okin - add Journal title
Ostro - add Journal title, vol. and pages
Park - add Journal title
Park - add Journal title
Peel - add Journal title
Pekkanen - add Journal title
Peng - add Journal title
Pennanen - add Journal title
Pereira - add Journal title
Perez - add Journal title
Peters - add Journal title
Peters - add Journal title
Peters - add Journal title
Piedrahita - form of reference is not consistent with the
others
Pope - add Journal title
Prystowsky - add Journal title
Rodriguez - add Journal title, vol. and pages
Rundell - add Journal title
Samet - add Journal title, vol. and pages
Samet - add Journal title, vol. and pages
Sarin - add Journal title
Schwartz - add Journal title, vol. and pages
Schwartz - add Journal title
16
-------
6-360
6-361
6-362
6-363
6-364
Schwartz - add Journal title
Seagrave - add Journal title
Sheppard - add Journal title
Sillanpaa - add Journal title
Smith - add Journal title
Steinvil - add Journal title
Stolzel - add Journal title, vol. and pages
Sullivan - add Journal title
Tamagawa - add Journal title
Tankersley - add Journal title
Thurston - add Journal title
Trenga - add Journal title
Tsai - add Journal title
Tunnicliffe - add Journal volume and pages
Upadhyay - add Journal title
Van der Werf - add Journal title
Vincent - add Journal title, vol. and pages
Wellenius - add Journal title
Wellenius - add Journal title
Whitekus - add Journal title
Whitsel - add Journal volume and pages
Wichmann - add Journal title, volume and pages
Witten - add Journal title, vol. and pages
Wong - add Journal title
17
-------
6-365
Yang - add Journal title
Yue - add Journal title
Zabel - add Journal title
Zanobetti - add Journal title
Zanobetti - complete reference
Zanobetti - complete reference
Zanobetti - add Journal title, vol. and pages
Zanobetti - add Journal title
Zhang - add Journal title
Chapter 7
Specific Comments:
SEC/PAGE
PARA/LINE
REVIEW COMMENTS
7-2
L. 15-16
Consider "Atherosclerosis is a progressive disease that
contributes to several adverse outcomes, including acute
coronary syndromes such as myocardial infarction,
sudden cardiac death, stroke and vascular aneurysms."
7-3
L. 31
Consider starting line 31 with, "CAC represents the
accumulation of calcium in coronary artery
macrophages and represents an advanced stage of
atherosclerosis. As such CAC is a measure of...."
7-4
L.
Replace "AAC" with "Abdomenal aortic calcium
(AAC)" at least initially when the term is introduced.
7-5
L. 3-4
Replace "anti-hyperlipidemic drugs" with "lipid
lowering drugs".
7-5
L. 18
Replace "anti-hyperlipidemic drugs" with "lipid
lowering drugs.
18
-------
7-5
7-11
7-12
7-14
7-14
7-14 and 15
7-19
7-19
7-19
7-20
7-20
7-20
7-21
7-22
7-22
7-59
7-76
L. 24
L. 14
L. 14
L. 10
L. 26
L. 32-38 and
L. 1-4.
L. 20
L. 22
L. 27
L. 7
L. 9
L. 13 and 26
L. 5
L. 17 and 22
L. 28
L. 20
L. 16 and 17
Change font for "in 2002 (the midpoint of the baseline
exam.).
"decease" should be "decrease".
Replace "TAT" with "Thrombin anti-thrombin complex
(TAT)" at least initially when the term is introduced.
For some of these terms that are use very infrequently
throughout the text it is better to write them out and
minimize the use of abbreviations.
"non-statistically" does not meaningful. Consider
"statistically non-significant".
"nonsignificant" should be written as "non-significant".
There is no mention of renal function data in the
Calderon-Garciduenas et al. 2007 paper. Yet, in the
subsequent paragraph it is mentioned that no mention of
an assessment of renal function.
Replace "avg" with "average".
Correct font of "respectively".
Correct font of "PM2.5".
Replace "avg" with "average".
Consider "The OR for fatal MI was increased but not
significantly."
Replace "avg" with "average".
Replace "avg" with "average".
Replace "avg" with "average".
Write "hospital admissions" rather than "HAs".
Change font of "reduction" and "reduced".
Consider the following sequence, "Several age intervals
have been explored: neonatal (<1 month), infants (<1
year), and post-neonatal (1 month to 1 year)." rather
than what is present. Subsequently, on page 7-77 the
19
-------
7-77
7-121
7-139
7-140
7-141
7-142
7-143
7-144
L. 3-13
L. 14
order of presentation should follow this order.
Move the paragraph "Infant Mortality and Infant
Respiratory Mortality, <1 Year" after "Neonatal
Mortality and Neonatal Respiratory Mortality, < 1
Month". "M" in "month" should be capitalized for
consistency.
Font of "reduction" and "reduced" should be corrected.
Achenbach - Title of journal
Arlt - Title of journal
Baccarelli - Volume and pages
Binkova - Title of journal
Bobak - Title of journal
Brown - Title of journal
Craven - Title of journal
Dales - Title of journal
Dinneen - Title of journal
Dockery - Title of journal
Dugandzic - Title of journal
Eftim - Title of journal
Fedulov - Title of journal
Fujimoto - Title of journal
Gauderman - Title of journal
Gehring - Title of journal
Geroulakos - Title of journal
Gotschi - Title of journal
Geenland - Title of journal
20
-------
7-145
7-146
7-147
Grigg - Reference
Gabelova - Title of journal
Gabelova - Title of journal, "Farmer Peter B" should be
written "Farmer PB".
Ha - Title of journal
Haland - Title of journal
Hashimoto - Title of journal
Heinrich - Title of journal
Hiramatsu - Title of journal
Hollander - Title of journal
Iba - Title of journal
Ishibara - Title of journal
Izawa - Title of journal
Jacobsen - Title of journal
Jalaludin - Title of journal
Janes - Title of journal
Janssen - Title of journal
Khattar - Title of journal
Kizu - Title of journal
Knobel - Title of journal
Kunzli - Volumen and pages
Kunzli - Title of journal
Lacasana - Title of journal
Loomis - Title of journal
21
-------
7-148
7-149
7-150
7-151
Lund - Title of journal
Maheswaran - Title of journal
Matsumoto - Title of journal
Miller - Title of journal
Nordling - Title of journal
O'Leary - Title of journal
O'Leary - Title of journal
O'Neill - Volume and pages
Oftedal - Title of journal
Oftedal - Title of journal
Parker - Title of journal
Pedersen - Title of journal
Pignoli - Title of journal
Pires-Neto - Title of journal
Pope - Title of journal
Pope - Title of journal
Resnick - Title of journal
Ritz (2008) - Title of journal
Ritz (2006) - Title of journal
Ritz (2000) - Title of journal
Rogers - Title of journal
Rojas-Martinez - Title of journal
Rosenlund - Title of journal
Schatz - Title of journal
22
-------
7-152
7-153
7-155
Schindler - Title of journal
Sharma - Title of journal
Shinkura - Title of journal
Somers - Title of journal
Song - Title of journal
Sorensen - Title of journal
Sram - Title of journal
Suh - Title of journal
Sun - Title of journal
US. EPA (2005) - "Just" is hanging. What does this
mean?
U.S. EPA (2006) - Correct "AGency".
Yauk (2008) - Reference
Yauk (1996) - Reference
Yoshida - Title of journal
23
-------
Cowling Comments (Dr. Ellis Cowling)
Charge Question 1:
Welfare effects evaluation in Chapter 9: Several revisions were made to the evaluation of
the welfare effects evidence in Chapter 9, in response to the CASAC PM Panel comment,
to focus further on effects on climate and ecosystems and include further evaluation of
urban visibility evidence, where possible. In addition, as recommended by the CASAC
PM Panel, key findings and conclusions from this chapter were incorporated in Chapter
2. The discussion of PM effects on climate was increased with substantially more detail
from recent publications, including discussion of specific climate forcing effects from
individual PM components and size fractions. The discussion of ecological effects was
also reorganized to focus on the types of effects and effects of individual components.
For the effects of PM on visibility, new material was added including sections on direct
optical measurements and the value of good visual air quality. Please comment on the
effectiveness of the reorganization and revisions regarding welfare effects.
I was pleased with the adjustments in organization that were achieved in this second draft ISA
for PM and with the statements of key findings with regard to public welfare effects in Chapter
2.
I was even more pleased to find that a reasonably very well-balanced distribution of attention
was given in Chapters 2 and 9 of the PM ISA to three of the four types of welfare effects of PM
- with visibility effects being covered in the first 83 pages in Chapter 9, climate effects being
covered in the next 104 pages of Chapter 9, and ecological effects being covered in 43 pages in
Chapter 9 (recognizing, of course, that the very substantial ecosystems effects the PM forms of
airborne NOx and SOx on acidification, nutrient enrichments, and methyl-mercury enhancement
are being handled very thoroughly in the separate integrated NAAQS assessment document for
public welfare effects of NOx and SOx.
It was somewhat disappointing, however, to discover that a pretty skimpy 6 pages in Chapter 9
and only three short paragraphs in Chapter 2 is all the attention that was given to the materials
damage effects of PM. I would have expected that soiling of exposed textiles and the painted
and exposed stone and other surfaces of buildings and monuments would have deserved
somewhat more attention.
24
-------
Frey Comments (Dr. Chris Frey)
Charge Question 5: Integrative Synthesis in Chapter 2: The CASAC PMPanel recommended
expanding Chapter 2 to include all important findings of the PM ISA. The integration of health
evidence in Chapter 2 was reorganized to focus on effects ofPM2.s, PM'10-2.5, andultrafine
particles, and was expanded to include discussions of effects for which a "suggestive" causality
determination was drawn. New integration sections were added that combine the evidence for
health effects ofPM2.s, PM 10-2.5, andultrafine PM across exposure durations. In addition, these
integration discussions incorporated evidence related to mode of action, dosimetry, atmospheric
chemistry, and exposure assessment to the extent possible. When appropriate, figures were
added that summarize the overall U.S. and Canadian epidemiologic evidence for specific size
fractions and exposure durations, along with the concentrations reported in the studies or
provided by study authors. A new section was also added to Chapter 2 that contains policy-
relevant considerations, including summaries for the evidence for susceptible subpopulations,
lag structure of associations in epidemiologic studies, and the PM concentration-response
relationship. Please comment on these revisions and additions to the integration of health effects
evidence in Chapter 2.
Response: The integration of health evidence and its organization with respect to effects of
PM2 5, PMio-2.5, and UFP is extremely helpful to the reader. The discussion of health effects
endpoints for which a "suggestive" causality determination was drawn is helpful and
appropriately fits in this chapter. The material on evidence for health effects of PM2.5, PMio-2.5,
and UFP for long- and short-term exposure durations is important to the chapter.
The graphical summary (Figure 2-1, and Figure 2-4) of U.S. and Canadian epidemiologic
evidence is good. The discussion of policy-relevant considerations is helpful. Overall, the
structure and content of Chapter 2 is very good.
A few specific comments are offered.
Page 2-4, line 8. What averaging time is the basis for these "average" concentrations?
Page 2-5, line 18. "maintained" is not the right word. Could delete "maintained at"
Page 2-9, line 7. EGUs are point sources with stacks, so it would be more clear here to refer to
"non-EGU point sources," if that is the intended meaning.
Page 2-10, line 14. Use "such as" rather than "like"
Page 2-13, lines 5-20. For clarity, should explain why PMio is not separately considered, just to
avoid confusion on the part of some readers who might be expecting to see PM2.5 and PMio as
the focus, since these are the indicators of the current standard.
25
-------
Page 2-13 - last half. As a rhetorical device, it is weak to start out with negative findings or
what is not known. Also, it would help to have just one table that summarizes for each size
range the list of health effect outcomes and the causality determination. As a reader, I would
prefer to first see a list of the key outcomes that are the focus of the ISA, followed by a list of
those for which there is inadequate evidence.
Page 2-16, line 1. The word "precise" is used here and in a few other places. However, the
intended meaning might be "statistically significant" rather than "precise." Either change the
term here or define what is meant by "precise."
Page 2-24: this page seems to go into a lot of detail to a depth that is not consistent with the rest
of the chapter. Perhaps cut back on this, and just make key points.
Page 2-25. Similar comment. Reader is wondering what is the "bottom line" here? Details are
in later chapters. However, the last paragraph on this page is very good.
Page 2-30, line 14. Delete "It is also important to note that". More importantly, what are the
implications of this paragraph? This paragraph is descriptive but it is not explained as to the
significance or implications of what is being described.
Page 2-33, line 11, this is awkward, "studies" are inanimate and cannot attempt to do anything.
This section needs copy editing.
Page 2-34, line 32-35. "various" might more accurately be replaced with "particular." The term
"susceptible" is used here but is not defined. The first time that the terms "susceptible" and
"vulnerable" are used (either in Chapter 1 or this chapter) they should be defined briefly, with a
reference to the later section where they are discussed in more detail.
Page 2-42 -top of the page. Is PM size a factor?
Page 2-42, lines 15-17. This text refers to "effects" on climate. In the context of all other effects
discussed throughout the document, the presumption is that they are adverse effects. In this case,
it is not clearly stated as to whether the "effects" are adverse or beneficial, and hence the policy
implications of this statement by itself are unclear.
26
-------
Grantz Comments (Dr. David Grantz)
ISA for PM, 2nd External Review Draft
Comments by David Grantz on Charge Question #6—Welfare Effects
Overall
I was very pleased to read this second draft ISA. The authors have paid close attention to
the concerns of previous reviews. The resulting two chapters of concern here, Chapters 2 & 9,
are well organized (few comments below) and provide a useful compendium of available
knowledge from the previous CD and more recent literature. The current Chapter 2 contributes
significantly to the presentation. Very few issues of interpretation remain, and these are
addressed below.
The increased attention to climate change is appropriate and generally well done. As noted
below, there is some confusion in the discussion of PM impacts on photosynthetically active
radiation (PAR). Briefly the issue is that photosynthesis per leaf responds nearly linearly to PAR
up to some near-saturation flux. At this flux it becomes useful to spread that (excess) radiation
around to lower leaves which (due to shading of direct beam radiation by upper leaves) are
generally still on the linear portion of the PAR response curve. Diffusion by PM accomplishes
this spreading around of PAR. If PM reduces total PAR below saturation for the upper leaves,
diffusing the flux will have only minimal and second order impacts on total plant photosynthesis.
Also, once diffuse radiation has become a significant fraction of total PAR, then further increase
in PM will reduce plant photosynthesis by decreasing total PAR. The current text is unclear and
seems to imply a conflict or conundrum that does not exist. The issue is important because it
lends itself very well to the ultimate goal of modeling of response-exposure relationships to PM.
I remain disappointed that there is little attempt to link ecological endpoints to specific levels of
ambient PM, with a view towards moving the debate towards response-exposure relationships. In
several cases cited there are clear endpoints, and exposures are documented. The studies of
Regoli et al. 2006 with snails and Kuki et al. 2009 with plants, as examples, seem to demonstrate
causality and at least the beginnings of such relationships. Reference to the study of Nash, 1975
(page 9-235, line 34) seems to best approximate this goal. Overall, the current organization of
Chapter 9 by component works very well.
Specific Comments
Chapter 2—Substantive comments
page
line
Comment
27
-------
2-3
25
It is important to define UFP (ultrafme particles) at first use, with
respect to size and mode of origin. This might be accomplished by
moving all of Section 2.1.5 (on PM Formation and Removal) into this
introductory location, as a way to introduce and define PM.
2-4
2-5
2-12
14
25
14-21
The commonly understood more rapid deposition of coarse PM is
initially stated, but this contrasts with later references to the very
rapid disappearance of UFP. This should be explained (ie. is UFP
growing out of the class or depositing out of the atmosphere?)
2-6
14
Is the afternoon peak of UFP also due in part to the collapse of the
boundary layer as in line 8, just above?
2-6
32
Might want to explain why this logical association between fine PM
and O3 was not observed in these three cities.
2-6
33
Section 2.1.4 on Measurement Techniques may not be required at all
in this summary chapter.
2-9
1-4
The implications of this first half of the paragraph are not clear.
Perhaps this should be merged into Section 2.2.1
2-11
2-12
14, 27-30
23
Define first use of "ambient" in this context, is it in opposition to
regional or to indoor or to personal?
2-14
2-26
2-31
titles
These section headings might better be titled "Effects" rather than
"Exposure". Also these sections (especially 2.3.1) may have too
much detail for this summary chapter.
2-14
Table 2.1
Seems odd that many of the CV endpoints were death, but CV is
"causal" and mortality is only "likely to be causal"?
2-20
2-28
2-32
Sections
2.3.2
2.3.4
2.3.6
These sections seem redundant in an already 'overview' chapter. In
particular, section 2.3.2 contains much of the information that is
required in chapter 2. Perhaps these summaries can be blended into
their preceding sections and condensed.
2-30
22-23
Text states "is associated" when it was concluded "is suggestive.
These are not equivalent.
28
-------
2-37
Sect 2.4.3
This entire section is cumbersome and unclear. A figure illustrating
the no threshold log-linear relationship for some endpoint could be
placed here (and in section 6.2.10.10 where it is also lacking).
2-38
Add "including NOx and Sox" between "airborne PM" and "on
visibility". The distinction between the PM document and the NOx-
Sox document will confuse many readers and should be made very
clear every time it comes up.
2-43
5-72
The relationship between crop growth and total intercepted PAR, and
the role of diffuse PAR when upper leaves are light saturated, is
poorly conveyed here and in the relevant parts of chapter 9. This is an
important point, that if properly explained the first time would
obviate much later discussion and remove the apparent conflict with
data showing that increasing PM both increases plant growth due to
light scattering and decreases it.
Chapter 2—Other comments
page
2-4
2-4
2-5
2-7
2-8
2-9
line
5
8
21
28
18
23
Comment
This summary statement leaves the impression that nothing can be
known. A more positive concluding sentence here, perhaps
considering the importance of regional averages, and the existence of
local hot spots, and more accurately summarizing the subsection,
would be helpful, particularly for readers who may only read this
chapter.
Should specify the averaging period (is it as specified at 2-5, 29 or at
2-14, 5?)
need a summary statement here
"isoprene" seems to be in the wrong place.
farm equipment may be even more confounded with trucks than are
locomotives.
should indent new paragraph
29
-------
2-14
2-15
2-20
2-31
2-36
2-36
2-36
2-39
2-41
2-42
2-43
14
3
9-10
4
13
22
25-26
27-29
7-8
9-14
32-34
Reference to null findings is unclear and potentially confusing
Might define "vasomotor" for general readers
Methylation is not a change in the genome
Typographical error in "cardiovascular"
Missing word before "children"
"result(s)"
a priori should be italicized
Is carbonaceous PM largest only in the West, needs to be clarified.
There should be a formal reference to the IPPC AR4.
This run-on sentence is hard to interpret.
The sentence is poorly constructed and factually unclear. Possible
solution is to insert "which are more often deposited by wet
deposition" between "fine particles" and "are more likely to contain"
Chapter 9—Substantive comments—line numbers are approximations
page
line
Comment
9-2
21-22
"Biogenic" combustion should be "biomass" combustion. If
prescribed burns are "natural" from a regulatory perspective then this
requires a brief explanation.
9-8
1-31
The level of detail on dark visibility is not necessary here, as it is
ultimately dismissed. Perhaps this section could be moved to later in
the chapter, as a suggestion of future concerns.
9-40
9-43
31
Fig. 9-
22
The text refers to Puerto Rico values of fine soil, but the figure
appears to indicate no data for Puerto Rico.
30
-------
9-46 Fig. 9- San Joaquin Valley data are unclear. This is a nearly closed airbasin,
24 unlike South Coast. Text is needed to explain what is the source of
regional PM (i.e. what is the region), and where local PM was
measured (just in urban Fresno?). Given internal gyre based
circulation in this basin it is not clear that the Valley itself constitutes
a truly regional source, nor that offshore sources are dominant as
implied on 9-54, 28 and on 9-55, Ito29. There seems to be a general
confounding of South Coast and San Joaquin Valley. An explicit
analysis similar to that provided for the Columbia Gorge (9-57, Ito5)
would be useful.
9-93 20 Maximum sight distance of 8 km in Washington DC is outside the
range of visibilities presented (Table 9-2), does this require some
explanation?
9-102 25 on The source of NOx to support the Midwest nitrate bulge has not been
well discussed to this point, and should be here. Ammonia has been
dealt with adequately.
9-210 21 PM effects on precipitation have been omitted from section 9.3.10
and should be considered here.
9-213 6-7 The concluding sentence of this paragraph does not summarize the
paragraph, and may belong with the following paragraph. The
appropriate concluding sentence here should summarize the
preceding CD and the material to be excluded from the current ISA.
9-221 3 There is a distinction made between uptake (or maybe just response?)
at the air-plant interface, and uptake by above ground tissue. This is
hard to understand. The authors may want to consider citing a
description of these processes, and of the leap-frog behavior of SOCs
discussed later, in the review by Krupa et al., 2008, Journal of Air
and Waste Management Assoc 58: 986-993. This reference has not
been cited and perhaps should be consulted. This reference could be
useful at p. 9-226 mid page, p. 9-256, bottom, and p. 9-259, mid
page.
9-221 13 Scavenging by the snowpack implies only dry deposition. There may
also be wet deposition in the falling snow.
9-226 7-17 This paragraph takes a generally positive view of biomonitoring
using mosses etc., which contrasts with conclusions on p. 224-225.
31
-------
These sections should be more closely aligned with each other.
9-227 8-15 There are some important concepts contained in the previous CD
regarding deposition, including specific cation depletion with depth
and distance into forests, that should be reviewed here. The material
at top of 9-236 on mosses and lichens should be incorporated into this
section. These data form the basis for existing models, and for the
criticism of them posed by Pry or and discussed here. In general, this
paragraph is hard to understand and should be revised. An entirely
new idea is introduced with "condensation processes...". This
sentence should start a new paragraph, but more importantly should
be integrated with the overall theme of this section, which has to do
with the appearance and disappearance of PM in canopies.
9-227 16-22 This paragraph requires some revision to better communicate that
biomonitoring and ecosystem impact are different endpoints. The
sentence "As an ecosystem pool.. .mosses" could perhaps be moved
to the end of the paragraph.
9-231
10
Only near the coast is most salt from sea salt.
9-231 21-28 This paragraph is out of place. It might be inserted near the beginning
of 9.4.3.1. The last sentence, "Fine particles...demonstrated" should
be moved to the bottom of page 9-229.
9-234 1-8 The material currently in 9.4.5.4 could be included here. Throughout
the metals section there is some blurring of direct and indirect effects
that should be clarified.
9-235 24 It would be useful to explore how S. terebinthifolius avoided stress
since this would be a useful attribute to look for in other systems. As
currently written this statement does not communicate much.
9-246 3-6 The discussion of loss of peat and increase in Hg should be moved to
or summarized in the section on climate change.
9-247 24 First paragraph of 9.4.5.5 seems to imply that these are toxicity issues
("metals in PM deposition limit phytoplankton growth"), but actually
they are fertility issues and positive effects.
9-252 22-25 The dependence of uptake on solubilized Cd is reasonable, but this
begs the question what determines solubilization (redox, acidity, ?)
32
-------
9-253
32-36
While the effects of Cu and Ni may be solely attributed to their
toxicity as statedhere, it was implied previously that these heavy
metals displace from soil exchange sites the very cations (Mg, Ca and
Mn) that are reported here to be depleted. Both inhibition of uptake
and depleted soil stores may be at work.
9-258
20-24
The greater content of PAH in littoral than pelagic species may
reflect sequestering in sediment as indicated, or it may (also) reflect
differences in diet of the two communities.
9-259
The role of elevation does not seem to justify use of salmonids as a
biomonitor. I suspect there was some linkage to prevailing PM
concentrations (which may have also varied with elevation?) in the
original publication which could be discussed here.
9-262
1-10
Again, discussion of PAR and diffuse light requires some attention.
Also haze has been "modeled" or "calculated" or "measured" rather
than "estimated" (to diminish visible radiation).
9-267
4-8
Rust generally applies to iron oxide. Whereas such coatings on
aluminum do provide some protection against further corrosion, iron
oxide does not, due to its flaking behavior.
9-267
16
Is this really dry deposition of gypsum (CaSO4) or is it formation in
place when sulfuric acid meets limestone or marble?
Chapter 9—Other comments—line numbers are approximations
page
9-1
9-13
9-24
9-49
line
18
Eg. 9-1
Eg. 9-2
8-9
Comment
Insert "including NOx and Sox" between "PM" and "on visibility"
The RH functions shown at top of page 9-14 could be placed within
equation 9-1, as "where" statements. Analogous RH functions should
be described for equation 9-2.
The sentence, "Note that the concentration. . .trend line slope" could
be moved to the figure legend.
33
-------
9-61
9-64
9-
72,3
9-96
9-213
9-222
9-235
9-237
9-242
9-243
9-250
9-251
9-252
9-256
Fig 9-32
Fig 9-34
Figs 9-
38,39
26-33
29
20
15-17
22-26
22
8-11
7
7
1-5
25-33
It is unclear what the sulfate and nitrate concentrations in the
individual figure legends refer to.
Is top panel average of Class I areas as indicated in figure, or only
Yellowstone NP as indicated in the legend?
Scale is not technically unitless as indicated in legend (it is %).
The deciview units should be translated into visual range, as was
done above.
The reference to Odum 1985 is not complete (p. 9-292)
The distinction between combustion sources and petrogenic sources
is unclear. Does combustion source refer to biomass combustion?
Sentence beginning, "A greenhouse study. . ." is confusing. There is a
word missing ("under"?) prior to "simulated".
Second half of paragraph, "Sequential extractions... unavailable"
could be incorporated into the beginning of this section 9.4.5.2.
In line beginning with "by increasing" there is a critical typographical
error, "creasing Cd uptake". Is is increasing or decreasing? Also at
beginning of next paragraph "mycorrhizae" is misspelled.
The discussion of shallow-rooted species is not very useful, and
should be deleted as a distraction.
MADE should be AMDE. It would be useful to define the reactive
gaseous Hg species of interest here.
Should include cement factories as sources of metals, given earlier
citation of Massicote et al., 2003 to this effect.
It would be useful to describe briefly how isotopes of C and N can
demonstrate biomagnifications of metals. In the second paragraph,
the lack of biomagnifications in aquatic systems, if generally true, is
sufficiently counterintuitive that it should be explained a bit more.
The first 6 lines of this paragraph should be moved to the earlier
section on deposition.
34
-------
9-261 18 Ecological effects also include changes in soil microbiology, which
should be added to the list at the beginning of section 9.4.7. In the
second paragraph, ions should be added to the list of constituents
with ecological effects.
35
-------
Helble Comments (Dr. Joseph Helble)
Charge question 4, ISA
Revisions to Chapter 3 on Source to Exposure: Consistent with revisions made to the health
effects chapters, Chapter 3 was revised to clarify that PM 10 incorporates both PM2.5 and
PM10-2.5 and reorganized to begin with PM2.5 andPM10-2.5, followed by PM10, where
applicable. The discussion of measurement techniques and chemistry ofPM10-2.5 has been
expanded in Sections 3.4 and 3.5, in response to CASAC comments. In addition, Section 3.8 on
human exposure toPMhas been reorganized and expanded to better characterize the evidence
and provide useful information for interpretation of epidemiologic studies. We would appreciate
comments from the CASAC PM Panel on these revisions.
Summary
The expansion and reorganization of Chapter 3 now provide a generally clear, well-organized,
and comprehensive review of PM25, PMio-2.s, and PMi0. The sequencing of discussing PM2.5
followed by the successively larger size fractions provides a logical framework for addressing
the important differences among these size classes. The revised chapter also does a good job,
overall, of summarizing areas where studies suggest differing trends, and identifying areas where
data are sparse and additional research is needed.
The reorganization and expansion specifically to address deposition in more detail, measurement
techniques, and diesel emissions were generally clear and well-structured. With regard to
deposition, a minor suggestion is that Section 3.3.4 page 3-21, the general introduction, be
expanded to address wet deposition. It is meant to be a general overview of deposition, but only
discusses dry deposition and ignores wet deposition other than a passing reference in the first
sentence. Organizationally, if this is to function as the introduction to the sections providing a
more detailed discussion of both wet and dry deposition, an overview of wet deposition needs to
be added.
The importance of particle morphology was clarified in this draft (pages 3-6 to 3-7) as requested
in the last review, but the discussion remains a general one. There is some limited literature on
surface area, for example, that could be considered.
The embedded links to reference documents are a very welcome feature.
Detail Comments
Lines 19-20, page 3-7, states that "particles in ambient air are generally mixtures or agglomerates
of particles... from multiple sources..." This statement is true if one is referring to condensed or
adsorbed species on the surface of nearly all PM, but otherwise, this statement is misleading
because it is dependent upon particle size and proximity to source. This statement should be
clarified or, better, removed.
36
-------
In the discussion of county-based PM spatial distribution in section 3.5.1.1, a slightly expanded
discussion of the completeness criterion would be helpful in interpreting the results. Counties
appear in white in the associated figures when they fail to meet the 75% completeness criterion.
Failing to meet this criterion could, however, be a result of a lack of appropriate monitoring sites
within the county, or a failure to obtain adequate (complete) data from appropriate monitoring
sites. Parsing this would be helpful - i.e. how many of the counties failed to meet the criterion
because existing monitoring sites did not report sufficient data?
Organizationally, it isn't clear that the summary in section 3.5.4 is necessary. The individual
subsections in section 3.5 provide a detailed review of the data, and the summary is essentially
repeated in section 3.9. Section 3.5.4 therefore adds little. To shorten the document, this sub-
summary, as well as a few other sub-summaries at the end of individual sections, could be
deleted, with section 3.9 serving as the overall summary for Chapter 3.
Page 3-225, line 28 - it seems that two different sentences, and two different ideas, run together
in this line. The first part of line 28, and the text that precedes it, appears to focus on automobile
indoor exposure. The remainder of the paragraph, with no transition, addresses school buses.
Were parts of two different paragraphs truncated?
Minor / Editorial
Figure 3-2 additional detail needed regarding sampling location and sampling details.
Lines 10-11, page 3-16 - first sentence in the paragraph confusing. Substitute "being emitted"
for "emission" to clarify.
Lines 16 and 17, page 3-31 - impaction-based collection efficiency will decrease, not increase,
as dp decreases.
Line 5, page 3-34: "monvolatile" should be changed to "nonvolatile"
Line 14, page 3-45: change "detect" to "detecting"
Line 20, page 3-45: change "media" to "medium"
Tables 3-3:, 3-4, etc. add commas to N values so magnitude of population figures more easily
seen
Figure 3-7 and 3-8: display would be enhanced by showing PM2.5 and PM10 monitors on the
same figure
Line 6 page 3-59, fix font on PM2.5. Similar comment lines 8 and 9, page 3-60
Line 31 pg 3-70, insert "per" before cm3
37
-------
Figures 3-12 to 3-14: did EPA consider plotting these figures without using different sizes for
the different concentrations (i.e., using different colors only)? It might make the figures easier to
read.
Figure 3-19 and similar figures - adding state borders would facilitate a quicker understanding of
monitor locations
Lines 10 and 11, page 3-135: figure numbers should be changed to 3-52 and 3-52
Table 3-17 is not adequately explained in the text nor the caption. The parenthetical values in
column 2 need to be described, and the basis for the data (ng/kg of what? presumably emitted
PM) stated explicitly. Also, while it is likely that this table came directly from the cited source,
is EPA certain that 5 significant figures (seen in several of the table entries) are justifiable?
Line 10, page 3-144: "mighty" should be changed to "might"
Line 8 page 3-151, "much" should be changed to "must"
Figure 3-65, page 3-165: caption is incomplete
Figure 3-68 caption - symbol D is missing. Also missing from title of Table 3-19 page 3-187,
and from titles of Tables 3-20 through 3-23
Figure 3-73 page 3-174: caption, change "theem" to "them"
Line 10, page 3-215: "In this work," - ? Incomplete sentence
Line 20, page 3-215: complimentary should be complementary
Lines 10-11-12 page 3-237, sentence would read better as opening to section. It does not follow
as a conclusion.
Page 3-246, Low Angeles should be Los Angeles
38
-------
Henderson Comments (Dr. Rogene Henderson)
Reply to Charge Question #3; Susceptible Populations
I found Chapter 8 to be complete, clear and well-organized. The authors have been
responsive to the CAS AC critique of the first draft of the chapter. Their definition of
"susceptible" is clearly stated and there is a rather complete listing of how the term has been used
in the literature in the past.
Table 8-2 introduces all of the susceptibility factors that have been evaluated, as well as
the size fraction evaluated and the length of the exposures. The table is then followed by a more
detailed description of the evaluations that were made. This is an appropriate organization, with
one exception. The introductory table 8.2 does not have a column indicating the results of the
evaluations. Thus if one only looks at the table, which is titled simply "Susceptibility Factors,"
one might get the impression that all of the listed factors showed a susceptibility, when some
factors, such as gender, did not. I recommend at least expanding the title to "Susceptibility
Factors That Have Been Evaluated," and if possible add a column that suggests what the results
of the evaluation were. Maybe just a plus/minus approach could be used.
1)1 found Chapter 8 to be complete, clear and well-organized. The authors have been responsive
to the CASAC critique of the first draft of the chapter. Their definition of "susceptible" is clearly
stated and there is a rather complete listing of how the term has been used in the literature in the
past.
2) Table 8-2 introduces all of the susceptibility factors that have been evaluated, as well as the
size fraction evaluated and the length of the exposures. The table is then followed by a more
detailed description of the evaluations that were made. This is an appropriate organization, with
one exception. The introductory table 8.2 does not have a column indicating the results of the
evaluations. Thus if one only looks at the table, which is titled simply "Susceptibility Factors,"
one might get the impression that all of the listed factors showed a susceptibility, when some
factors, such as gender, did not. I recommend at least expanding the title to "Susceptibility
Factors That Have Been Evaluated," and if possible add a column that suggests what the results
of the evaluation were. Maybe just a plus/minus approach could be used.
39
-------
Hopke Comments (Dr. Phil Hopke)
Chapter 3
This chapter is in generally good shape. I think there are some minor changes noted below that
will improve it.
There are a number of cases where they talk about "accuracy" with respect to sampling or
determination of concentrations. In most cases, there is absolutely no knowledge of accuracy
since there is no way to provide a truly known standard to which the measurement can be
compared. There may be systems with good precision, but there are no deployed system with
know accuracy. Thus, there needs to be a careful review of the use of "accurate" with respect to
the description of any ambient measurement.
Primary vs secondary organic aerosol:
I suggest it would be very helpful to clearly distinguish the various types of organic carbon
species associated with airborne particulate matter. It is too simplistic to just classify the OC
into primary and secondary. There is also oxidized primary organic matter. The problem is the
oxidized POM will also be water soluble and have the general characteristics of SOA.
However, given the sources are different and the chemistry is entirely heterogeneous, it is
important to clearly distinguish these three different portions of the particulate organic matter.
The process of forming SOA or oxidized POM leads to the formation of oxidizing species
(reactive oxygen species). This source of exogenous ROS could be important in the induction of
adverse health effects. Thus, connecting SOA formation with high concentrations of ROS such
as seen by Docherty et al. (Docherty, K. S., Wu W., Lim Y. B., Ziemann P. J. (2005)
Contributions of organic peroxides to secondary aerosol formed from reactions of monoterpenes
with Os, Environmental Science and Technology 39, 4049-4059). Thus, we can expect ROS to
be associated with ambient particles and that has been observed by multiple groups (Hung, H-F.,
Wang, C-S.(2002) Experimental determination of reactive oxygen species in Taipei aerosols.
Journal of aerosol science 32, 1201-1211; Hasson A. S., and Paulson S. E. (2003). An
investigation of the relationship between gas-phase and aerosol-bourne hydroperoxides in urban
air, J. Aerosol Sci., 34: 459-468.; Venkatachari, P., Hopke, P.K., Brune, W.H., Ren, X., Lesher,
R., Mao, J., Mitchell, M. (2007) Characterization of wintertime reactive oxygen species
concentrations in Flushing, New York, Aerosol Science and Technology, 41, 97-111;
Venkatachari, P., Hopke, P.K. ,Grover, B.D., Eatough, DJ.(2005) Measurement of particle-
bound reactive oxygen species in Rubidoux aerosols, Journal of atmospheric chemistry, 50, 49-
58). Thus, there needs to be a short paragraph added to point to this pathway to the formation of
a potentially important class of atmosphere aerosol species that could be directly related to
observed human health effects.
Other major issues:
Although past the cut-off date for inclusion, it may be useful to note Amato F., Pandolfi M.,
Escrig. A., Querol. X., Alastuey A., Pey. J., Perez N. and Hopke P. K., Atmospheric Environ. 43:
2770-2780 (2009). This paper demonstrates how known source profiles can be included within
the PMF framework and yield an improved source resolution.
40
-------
The description of PLS on page 3145 is not fully correct. PLS is in fact a biased regression
method that is really designed for prediction and not for explication as is the goal of receptor
modeling. In addition, there is really no basis for choosing PLS over other prediction methods
like principal components regression (PCR). It is really not a receptor model and it is not clear it
really belongs in this section.
There are some minor issues to be resolved.
On page 3-4, the table uses the term "condensation of gases." Gases do not condense at normal
temperatires and pressures. The correct term here is condensation of "vapors."
Chapter 5
The discussion in section 5.1 only focuses on endogeneous ROS. It should be recognized that
there is exogenous ROS associated with the ambient particulate matter to which we are
continuously exposed. Even in the evening, there is still significant particle-bound ROS and
thus, it should be noted as part of the total dose of ROS to the tissue.
41
-------
Lippmann Comments (Dr. Mort Lippmann)
Post-Meeting Review Comments
General Comments:
1) I commend the organizers and authors of this document for providing the CAS AC PM
Panel with a well organized and readily readable text that is, for the most part, thorough
and accurate in its presentation of the peer-reviewed literature most relevant to Standard
setting for PM.
2) I have one major disagreement with the authors in terms of their characterization of
Health Effects (Section 2.3). Specifically, in Section 2.3.1.2 on Effects of Long-Term
Exposure to PM2 5, the conclusion that "a causal relationship is likely to exist between
long-term exposure to PM2.s and mortality" is insufficient. It should be changed to "a
causal relationship exists between long-term exposure to PMi.sand mortality". How
is it possible to conclude that "a causal relationship exists between long-term
exposure to PM2.sand cardiovascular effects" (page 2-18, lines 8 & 9), and that the
strongest support for that causality determination comes from the associations of ambient
PM2.5 and annual mortality, and then equivocate on the conclusion that PM2.5 causes
excess mortality. The document amply showed that cardiovascular mortality accounts for
most of the excess mortality (Pope et al. 2004). The conclusion that I challenge is
apparently drawn, for the most part, from Section 7.6.5, in which it failed to properly
consider that chronic PM2.5 exposure studies in mice caused exposure-related aortic
plague progression. This research provides a toxicological basis for supporting the
epidemiological findings of excess cardiovascular mortality and a causality conclusion.
3) I have an important organizational comment on the draft ISA. It is inappropriate to
include the summary and synthesis on associations of PM sources and components with
health effects only in Chapter 6 on short-term exposures, since it covers the literature on
both short-term exposures and long-term exposures, the latter being relevant to Chapter 7.
This topic should be a separate chapter. If this is not possible, the topic should be covered
in both Chapters 6 & 7.
4) There seems to be an overemphasis on UFP, especially in Chapter 5, where the potential
to cause adverse effects is so prominently featured, along with the implication that
particle number concentration is the most important metric for risk. Some additional
discussion of the fact that supporting evidence for this hypothesis remains weak is
warranted.
5) Some generic changes that should be made for the final version are:
a) Use a consistent criterion for "recent". In the context of this document it should refer
to papers appearing since the closure on the last PMCD;
b) Italicize in vivo, in vitro, in situ, a priori, etc.
42
-------
Specific Comments:
PAGE
1-3
1-17
1-24
1-24
2-31
3-111
4-1
4-1
4-2
4-2
4-2
LINE
1 and 2
9-12
15
34
5
1-14
26
30
2
6
6
REVIEW COMMENTS
Two lines of text were left out here
It is inappropriate to include a summary and synthesis
on associations of sources and components with health
effects in Chapter 6, since it covers both effects of
short-term and long-term exposures. It needs to be a
separate chapter, or divided into parts for Chapter 6
and parts for Chapter 7.
Insert "sizes, respiratory parameter and" before
"branching"
Citations to the concentrators developed and used by
the Harvard Group need to be added here
I don't agree with the statement: "fresh DE, which is
typically dominated by UFPs". It may be true for the
number concentration, but is usually not true for mass
concentration due to rapid coagulation of the UFPs
before they are inhaled.
Cite work of Peltier et al (in press) and Peltier and
Lippmann (in press) on intra-urban variations in N and
V in NYC
Insert "Substantial" before "exposures"
Change "frequently" to "can"
Change "Fibers, therefore, deposit largely" to "fibers
longer than 10 jim can deposit"
Insert "long" before "fibers"
Change "prevent" to "affect"
43
-------
4-3
4-3
4-5
4-6
4-6
4-14
4-15
5-1
5-2
5-2
5-3
5-8
5-27
6-4
6-17
6-17
6-95
6-103
6-250
6-251
6-252
28
29
2
25
25
17
12
18
Fig. 5-1
Fig. 5-2
21
13
15
12
7
25
15
Fig. 6-2
Sect. 6.4.1
Sect. 6.4.3
Sect. 6.4.3.1
Change "airways" to "respiratory tract"
Change "airways" to "respiratory tract"
Change "rats" to "rodents"
Change "scintigraphic" to "external measurements of
retained particles having radioactive tracers"
Change "These" to "Scintigraphic"
Insert "anatomical and" before "ethical"
Add "and nasal vs. oral breathing" after "size"
Insert "with the exception of ultrafme PM" before
"There"
Lowest oval is mis-labeled. It should refer to in-vivo
responses, rather than cellular system assays.
Insert "soluble" before "metals"
Change "PM2.5 " to "accumulation mode PM"
Delete "ultrafme"
Change "m" to "ms"
Change "Tuxedo" to "New York"
Change "of to "in"
Change "NA+" to "Na+"
Show vertical line at 0 change
Chen & Schwartz (2009) does not belong here it is
clearly a study of chronic exposure and effects
Cite Veronesi et al (2005) in this section. (See ref.
listed at end of my comments).
Calderon-Garciduenas et al. (2003) does not belong
here in the acute exposure section.
44
-------
6-253
6-320
6-320
6-321
6-325
6-326
6-326
6-327
6-327
6-331
7-2
7-6
7-8
7-10
7-25
7-30
7-31
2nd and 3rd para.
14
26
16
12
6
3rd para.
3rd para. LineS
3rd para. Line7
Table 6-17
14
22a
12a
Sect. 7.2.3
9
3
Table 7-3
These studies do not belong in the acute exposure
section.
Change "correlation between each" to "correlations
among the"
Change "inherently" to "sometimes"
Insert "and analyzed the concentrations of the
components therein" after "CAPs"
Add citation to "Patel et al (in press)". (See ref listed
at end of my comments).
Insert "of after "study"
What about the "Franklin et al (2008)" paper? It needs
to be cited in this paragraph.
Change "similar" to "simultaneous"
Cite the "Sama et al (2007)" paper on the effects of
CAPs on the liver here. (See ref. listed at end of my
comments).
Add another source category, i.e., "smelter effluent" ,
which is enriched in Ni, Cr, and Fe.
Cite Veronisi et al (2005) re: nervous system and Tang
et al (in press) re: hepatic system
Insert "PM2.5" before "CAPs"
Change "PMio" to "Ambient Air PM"
Create a new subsection on Epidemiology that cites
the study of Chen and Schwartz (2008), using text
from page 7-11, lines 4-6
Add at end of sentence: "and with PM2.5- related
plaque progression in chronically exposed mice"
"SAPALDIA" is mis-spelled
Add descriptions of "Gent et al (2009) and Patel et al
45
-------
7-115
7-131
7-137
5
Table 7-9
18
(in press) to this table
Add citation to "Xu et al (in press)
Capitalize "New York City" in the table title
As noted previously, "likely to be" needs to be deleted
References cited in there Reviewer Comments
Chen, L.C, Quan, C., Hwang, J.S., Jin, X., Li, Q., Zhong'M., Rajagopalan, S., Sun, Q.,
Atherosclerosis lesion progression during inhalation exposure to environmental tobacco smoke
in a susceptible animal model, and comparison with that occurring during exposure to
concentrated ambient air fine particles. Inhal. Toxicol. (in press).
Lippmann, M. Semi-Continuous Speciation Analyses for Ambient Air Particulate Matter: An
Urgent Need for Health Effects Studies.. J. Expos. Sci. Environ. Epidemiol. 19:235-247.
Lippmann, M. and Chen, L.C. Health effects of concentrated ambient air particulate matter
(CAPs) and its components. Crit. Rev. in Toxicol. 39(10):865-913.
Patel, M., Hoepner, L.,Garfmkel, R., Chillirud, S., Reyes, A., Perera, F., Millrt, R. (2009).
Ambient metals and elemental carbon in fine particulate matter predict wheeze and cough in very
young urban children. Am. J. Respir. Crit. Care Med. (in press).
Sama, P., Long, T.C., Hester, S., Tajuba, J., Parker, J., Chen, L-C., Veronesi, B. (2007). The
cellular and genomic response of an immortalized microglia cell line (B V2) to concentrated
ambient particulate matter. Inhal. Toxicol. 19:1079-1087.
Tan, H-H., Jinsheng, G., Fiel, M.I., Alvarez, C.E., Chen, L-C., Sun, Q., Friedman, S.I., Odin,
J.A., Alina, J. Enhancement of fatty liver disease progression and TLR-4-dependent Kupffer cell
activation by air particulate matter. J. MOLEC. MED, (in press).
Peltier, R.E., Hsu, S.i., Lall, R., and Lippmann, M. Residual Oil Combustion: A major source of
airborne nickel in New York City. J. Expos. Sci. Environ. Epidemiol. 19:603-612.
Peltier, R.E., and Lippmann, M. Residual Oil Combustion: 2. Distribution of airborne nickel and
vanadium within New York City. J. Expos. Sci. Environ. Epidemiol. (in press).
Xu, X:, Kherada'N., Hong'X., Quan' C., Zheng'L., Wang'A., Zhong'M., Lippmann, M. Chen'
L.C.. Rajagopalan' S., Sun'Q. Diesel exhaust exposure induces angiogenesis. Toxicol. Lett, (in
press).
46
-------
Malm Comments (Dr. William Malm)
Review of Integrated Science Assessment for PM
Page 9.3: The use of "transparency" is inappropriate here. It usually is reserved for a
psychophysical reference of the "perceived" transmittance of the atmosphere. One could use
"transmittance", but that would also be incorrect because contrast is dependent on other "stuff
besides transmittance. I suggest using physiochemical characteristics.
Page 9.5: The statement in the first paragraph is only true if certain scenic characteristics are
met! ".. .regardless of background conditions" seems to say that a percent change in extinction
exudes a similar perceptual response under all atmospheric conditions, which of course is not
true.
Page 9.9: How about using "imperceptible" instead of "unperceivable"? "Achromatic" what?
Should sentence read "... achromatic contrast and discoloration"?
Page 9.16: Should include some discussion of the level of uncertainty associated with
nephelometer measurements of coarse particle scattering.
Page 9.21: "The resulting underestimation of total light extinction is typically much smaller
since fine particle light extinction generally exceeds that contributed by coarse particles." Not
true - coarse particle fraction of total scattering is often comparable to fine particle scattering.
Page 9.23: "Among the issues raised is that the algorithm tended to underestimate the light
extinction for the haziest conditions that occur principally during the summer in the southeastern
U.S. and overestimate for near pristine conditions that tend to occur most often in the arid
western U.S." This statement is not generally true. In the sites where the IMPROVE algorithm
underestimated extinction for the haziest conditions, it also overestimated extinction for the
lowest extinction days. Generally, in the West the IMPROVE algorithm tended to work better
than in the hazier eastern United States.
Page 9.82: It is well established that current regional scale models underestimate the SOA rather
substantially. Model limitations should be discussed at some level before presenting modeling
results. Also, it would be helpful to have a paragraph or two highlighting limitations and
advantages in the various apportionment approaches. Something along the lines of:
Transport regression receptor models are based on the association of source signatures
with measured concentrations. Two different methods were employed in the RoMANS
study. One method, trajectory mass balance (TrMB), used the residence time over
defined source regions, estimated from RMNP back trajectories, to estimate the source
signatures, while a hybrid approach developed source signatures from modeled
concentrations of conservative tracers released from various source regions.
In TrMB the model establishes a "scaling" factor between the residence time over a
source region and the measured concentrations at the receptor site. Average emission,
dispersion, chemical transformation, and deposition processes are incorporated into this
one scaling factor. However, all of these processes vary with time, and the source
apportionments derived from this technique are more accurate on the average than on an
hourly or episodic basis. A statistical apportionment approach will inherently tend to
overestimate those source regions where endpoints over the region are better correlated
with measured concentrations, while underestimating those source regions where
47
-------
endpoints over that source region show weak-to-little correlation with measured
concentrations. Furthermore, air often arrives from two or more source regions
simultaneously, causing the source signatures to be correlated, which increases the
uncertainties in the analysis.
In the hybrid modeling approach, the apportionment analysis is more robust in that
emissions and dispersion are explicitly modeled, and only chemistry and deposition are
incorporated into the "scaling" factor. In both approaches the scaling factors account for
possible errors in the models used to develop the source signatures, such as any
systematic biases in the emissions used in the hybrid approach. Like TrMB, the
apportionment estimates are more accurate on the average than for an incrementally small
time period.
In reading the discussion of sources of ammonia and Nox that contribute to particulate nitrate
formation, it occurred to me that a discussion of the concept of partial scattering efficiency
would be helpful, as first proposed by White. The concept being that the only meaningful way to
look at scattering change response is to explicitly model the change in scattering resulting from a
change in emissions. Of course, this assumes that the model captures all the complexities
discussed in the pages presenting the nitrate apportionment work.
Page 9.89: Here, scattering as a function of emission change is presented. It would be helpful to
have a discussion, as outlined above, of the concept of partial scattering efficiencies, including
the limitations and advantages of approaching the apportionment problem in this manner.
Page 9.103: "Both are a product of incomplete combustion of fuels, including those used in
internal combustion processes (gasoline and diesel emissions) and open biomass burning (smoke
from wild and prescribed fire)." EC is more likely to be a product of incomplete combustion
than OC. I don't think lumping EC and OC together as in the above statement is accurate. Much
of measured OC is derived or SOA associated with VOC emissions, which can be natural or
manmade.
48
-------
Pinkerton Comments (Dr. Kent Pinkerton)
Response to Charge Question 3:
Susceptible Populations, chapter 8 in the Second External Review Draft of the Integrated Science
Assessment for Particulate Matter, provides an excellent overview and summary of the current
literature on this topic.
1) The majority of the definitions provided in Table 8-1 for the terms susceptible and
vulnerable are appropriate, although it would be good for the authors to make some
conclusion as to the most timely and relevant definitions for the purposes of this chapter.
2) The chapter discussion on epidemiologic, controlled human exposure, and toxicological
studies with inclusion or highlighting of only those studies with stratified results to
demonstrate susceptible populations is highly appropriate to allow for the comparison of
subpopulations exposed to similar populations and using the same study design. The
only concern for such comparisons would be the lack of personal monitoring to arrive at
the appropriate conclusion of similar exposure conditions. Never-the-less, such
comparisons of susceptibility based on factors of age, gender, race/ethnicity and pre-
existing health conditions are critical and extremely helpful for presentation in this
chapter.
3) Table 8-2 is a nice list of susceptibility factors along with exposure conditions (i.e. short-
term or long-term) and the PM size fraction evaluated, but no explanation is provided as
to the outcome and/or consequences of each susceptibility factor in terms of health
endpoints. The authors should consider the addition of a column to briefly summarize
the impact of each susceptibility factor on health impacts and/or observed health effects.
4) Inclusion of the hyperlinks to chapter 8 is a novel and extremely useful feature to the
document to provide convenient and quick access to the database of scientific literature
used by the USEPA. How often is this database (HERO) updated in view of the rapidly
changing literature on PM and health effects? This is a wonderful addition to the REA!
5) The chapter subsections for each susceptibility factor listed in Table 8-2 are well-written
and appropriate with corresponding hyperlinks to provide direct access to relevant
documents under each topic. The authors should be commended for making these
sections a pleasure to read and to comprehend.
6) Table 8.3 contains fascinating information on the percent of the US population with pre-
existing cardiovascular disease (actual numbers are 51.6 million with hypertension, 24.1
million with heart disease, and 14.1 million with coronary heart disease), respiratory
disease and diabetes. The table as well as the text for each of these conditions along with
that of obesity emphasize the increased vulnerability of the US population to potential
health effects of exposure to PM. It would be nice if the authors could provide some
overarching conclusions for the relevance of these pre-existing diseases based on type
and age for this section of the chapter.
7) Excellent summary for chapter 8.
49
-------
Phalen Comments (Dr. Robert Phalen)
Chapter 1
The revision represents a significant improvement.
pg. 1-2, line 34: The last sentence is incomplete.
pg. 1-3, lines 5 to 8: Thank you for mentioning the "non-PM-exposure factor that might
influence the association..."
pg. 1-12, lines 16-18: This looks like the cutoff date for cited literature. Is 5/09 a firm date for
citing publications?
pg. 1-16, lines 13 to 16: It is good to see that "strength" of the evidence is considered, and also in
Table 1-2 on "Aspects to aid in judging causality." The more I read on causality, the more I
believe that the strength of an association is a key indicator of a true causal factor, as opposed to
a surrogate exposure. When regulations are issued it is critical that the regulated pollutant is a
true culprit. Much of the controversy, and opposition to PM NAAQS, seems to related to
concern over whether particle mass is an appropriate indicator for health effects.
pg. 1-32, lines 5 to 11: There are several garbled words (typos) that need correcting.
pg. 1-33, line 3: The sentence is garbled.
pg. 1-33, line 9: Its "express", not "xpress."
pg. 1-34: Journal title missing from first reference:
Chapter 4
pg. 4-1, lines 23-26: Thanks for excluding fibers, nano-objects, etc. Inclusion of such unusual
particles would expand the chapter unnecessarily.
pg. 4-3, lines 24 to 25: Please insert "small particles and small" before "aggregates", as large
aggregates are not characterized by a "thermodynamic-equivalent size." Also, I would replace
"thermodynamic-equivalent size" with "diffusion-equivalent diameter", in order to clarify and
avoid confusion with the thermodynamic concepts used in describing granular flows.
pg. 4-11, line 6: Add "in part" before "because" as mechanisms other than mixing prevent
particle deposition from being zero.
pg. 4-12, line 12: True, relatively few coarse particles will pass through the nose in an "average"
person. However, some people have greater penetration, and even a few coarse particles can
have appreciable masses.
50
-------
Accordingly, I suggest adding a sentence at the end of this line, such as: "However some
individuals have less efficient coarse particle deposition in the ET region, and such particles can
carry significant mass into the tracheobronchial and alveolar airways."
pg. 4-13, line 1: Change "between" to "among".
pg. 4-15, lines 8 and 9: I don't know what "the maximum deposited incremental dose" is.
Explain or delete the sentence.
pg. 4-18, line 9: Bennett and Zeman's "children" were age 7-14 years, so the result may not apply
to young children. Insert "age 7-14 years" before "children". Infants and young children, e.g.
age under 3-4 years, are very different than adults with respect to particle deposition. We
must not assume that 7-14 year old children are representative of smaller children.
pg. 4-22, line 20: Change "humans" to "the average human", variability is significant in particle
deposition. It is important that we recognize individual differences in dosimetry.
pg. 4-27, line 11: Change "Stat" to "that".
pg. 4-27, lines 29-31: Omit this sentence. Long-term TB retention of particles was discovered in
rodents. The sentence may not be true in all cases.
pg. 4-28, line 6: Add "average" before "path length".
pg. 4-37, lines 6 to 7: Delete the sentence starting with "In contrast" as it is not always true.
pg. 4-44 to 51: Many journal titles are missing from the references.
Chapter 5
The chapter is thorough, but speculative, which is ok. The chapter does not address the
importance of dose very well. This could be stated in the beginning of the chapter.
pg. 5-11, lines 1 and 3: Add "specific" before "larger", and add "the same mass of before
"particles of larger size".
pg. 5-18, line 12: Add "realistic" before "PM exposure".
pg. 5-32 to 5-39: Many journal titles are missing.
51
-------
Poirot Comments (Mr. Rich Poirot)
July 2009 PM Integrated Science Assessment, Charge Question 6:
Several revisions were made to the evaluation of the welfare effects evidence in Chapter
9, in response to the CASAC PM Panel comment, to focus further on effects on climate
and ecosystems and include further evaluation of urban visibility evidence, where
possible. In addition, as recommended by the CASAC PM Panel, key findings and
conclusions from this chapter were incorporated in Chapter 2. The discussion ofPM
effects on climate was increased with substantially more detail from recent publications,
including discussion of specific climate forcing effects from individual PM components
and size fractions. The discussion of ecological effects was also reorganized to focus on
the types of effects and effects of individual components. For the effects ofPMon
visibility, new material was added including sections on direct optical measurements and
the value of good visual air quality. Please comment on the effectiveness of the
reorganization and revisions regarding welfare effects.
Generally, the revisions made to Chapter 9 strengthen the chapter, improve its clarity and are
directly responsive to previous CASAC comments on the 1st Draft ISA. A few minor revisions
may be warranted, but I don't think any major revisions are required. The added discussion on
direct optical measurement methods and on visibility valuation is helpful, although there is
relatively little presentation or analysis of optical measurement data, and in particular, from
measurements using combined nephelometer and aethalometer - equipped with various PM (1,
2.5, or 10 micron or "switching") size fractioning inlets. Presumably such measurements can be
shown to agree reasonably well with transmissometers, aerosol reconstruction, etc. Better
demonstration that there are currently viable, field-tested monitoring techniques to implement the
proposed optical standard would be helpful in the ISA or should at least be added to the REA.
Substantial detail has been added on aerosol effects on climate, much of it taken directly from
recent IPCC documents (which is fine - as these are current, relevant and authoritative).
Unfortunately the quality of reproduced graphic images is extremely poor - although the fault is
not with EPA, as the poor graphics appear to come directly from the original IPCC reports.
While the introduction to section 9-3 indicates that the reproduced CCSP SAP2.3 incorporated
"significant sections from EPA data and reports related particularly to U.S. emissions and
measurements", there is very little reported in the chapter directly relating to emissions from or
measurements within the US. While a number of causal relationships between emissions,
aerosol concentrations and climate effects can be identified on global scales, there doesn't seem
to be a clear, current basis for establishing specific US secondary PM standards for the purpose
of ameliorating local or global climate effects. Consequently, I don't think it would be
productive for staff to spend time adding to the aerosol/climate section in the current ISA.
In a similar way, the newly added details on ecological effects of various PM chemical
components provides a much improved and more comprehensive summary of the subject area,
but does not suggest that there's currently a strong justification for setting secondary PM
standards to protect against adverse ecological effects at the present time, or for expending much
additional effort on this section of the ISA in the current review cycle.
52
-------
Specific Comments on PM ISA, Chapter 9:
p. 9-15, 2nd para, last sentence: Why not name the 5 urban areas with transmissometers? Also,
this reminds me to suggest that if time allows, possibly the Phoenix hourly transmissometer data
could be compared with the modeled hourly reconstructed aerosol extinction estimates - to help
evaluate the quality of those estimates, and/or to suggest ways in which they might be improved.
p. 9-15, 3rd para, last sentence: A higher resolution (than what's routinely reported) version of
the ASOS data is available (and more useful). EPA used to fund STI to routinely access and
report this higher resolution data, but has discontinued support.
p. 9-49, 5th line from bottom: This very large (Title IV-related) drop in SC>2 emissions from 1994
to 1995 could be mentioned as a possible explanation for the apparent lack of post-1995
improvement in clean or hazy days at about half the Eastern IMPROVE sites in figs 9-26 & 9-27.
p. 9-88, Figure 9-50: The figure caption indicates the right figure includes effects from "nitrate"
& OMC, while the Y-scale label and above the figure both indicate "Sulfate + OMC". Change
"nitrate" to "sulfate" in the figure caption. It might also be noted that the RAIN network
continuous sulfate data (as measured by Teco 5020) has been shown to be highly correlated with,
but lower (by factor of 1.3) than collocated IMPROVE filter-based sulfate measurements. If the
sulfate-associated extinction in left & right sides of Figure 9-50 were increased by 1.3 the slopes
and R2 would increase. This effect is shown in Figure 5.3 of the referenced NESCAUM report in
which reconstructed Bsp from OMC + adjusted sulfate = 0.91 Measured Bsp + 4, R2 = 0.94.
p. 9-95, 2nd para, last line: It would be useful here to also report other aspects (besides just the
Bext level) of the Denver standard - for example the fixed 8-hour 8 am to 4 pm daylight window,
4-hour averaging time and <70% RH constraints.
p. 9-96, 3rd para, last line: As for the Denver standard, why not report other aspects (besides just
the Bext level) of the Phoenix standard - the daylight window, 4-hour averaging time and <90%
RH constraint. Also, the Phoenix "standard" is not based on an absolute threshold, but on a
required improvement across the entire visibility distribution - moving bad days to moderate,
moderate to good, good to excellent, etc. Discussing this might help emphasize that while there
appear to be fairly convergent public opinions on levels of poor visibility that are considered
"unacceptable" in the various studies, there are also adverse welfare effects from reductions in
the frequencies of periods with moderate, good and excellent visibility, and would be benefits
from improvements across the whole distribution. Conceivably this might lead toward
considering a secondary "standard" which would be "progress-based" over fixed future time
intervals, rather than "threshold-based" with an indeterminately-defined time fuse of "as
expeditiously as practicable".
p. 9-98, last 4 lines: This observation that respondents were often "confused about the role of
weather and humidity in the different visibility conditions presented in the photos" (even when a
weather and RH-neutral WinHaze approach was employed) might also lead to a position that a
good secondary NAAQS should be based on metrics that make as strong a distinction as possible
53
-------
between effects of pollution and effects of humidity and other weather. Examples include:
applying an RH filter (I think anything between the 70% used in Denver and 90% used in
Phoenix could be justified); shortening the daylight window (to exclude the few most humid
hours), increasing the averaging time (from 1 to 2, 3, or 4 daylight hours); and/or altering the
form from the xth (90th, 95th or whatever) percentile of the worst 1 hr (or 4hr or 8 hr) in a day
summed for the year to the xth (90th, 95th or whatever) percentile of the hourly (or 4hr or 8 hr)
levels in a year or season. Any/all of the above will tend to "dry out" the metric, which would
reduce differences between the West and the East, spatially and temporally focus regulatory
efforts more on times and places of maximum pollution rather than maximum humidity, and
provide the public with as clear a distinction as possible between effects of pollution and effects
of humidity and other weather.
p. 9-102, lines 3-11: It might be helpful to include a complementary paragraph describing some
of the disadvantages an optical measurement-only approach - especially as a basis for a standard.
For example: cost and difficult siting criteria for transmissometers, similar cost and complex
operational and data processing issues with nephelometers and aethalometers, absence of
information on mass, size distribution or composition of pollutants causing the extinction. I
think these measurement issues can be resolved, but it will require some effort from EPA and the
states. It might be prudent to start with a small pilot network to find and work out some of the
bugs.
p. 9-901, Figure 9-53, and elsewhere in this section: The quality of most of the graphic figures
copied from the CCSP SAP2.3 is regrettably poor, although I note in checking the original
reference its not your fault - the original is barely legible.
p. 9-131: You could add an example here on MODIS AOD vs. AIRNOW PM2.5 from the
EPA/NASA IDEA site that might be a clearer and more relevant illustration.
http://www.star.nesdis.noaa.gov/smcd/spb/aq/index.php
p. 9-194, line 6: Its not clear what "to have a factor of 2" refers to.
p. 9-208-210: This section (9.3.9.3) is about the only place in section 9.3 that refers to any
specific effects within the US or resulting directly from emissions from US sources (and which
therefore might be relevant to consideration of secondary PM standards). Given this, it would
seem important to be more clear about exactly what the PM causal mechanism(s) is (are). There
are also indications that urban "heat islands" tend to increase convective precipitation in
downwind areas (http://www.gsfc.nasa.gov/topstory/20020613urbanrain.html). See also the
counter arguments presented by Alpert et al. (2008)
http://www.tau.ac.il/~pinhas/papers/2008/Alpert_et_al_JAMC_2008.pdf Possibly there are
also other non-aerosol effects of urbanization, such as more rapid runoff and deforestation, that
would tend to alter downwind precipitation patterns.
p. 9-210, lines 17-18: This description of "planet brightening when seen from space because
aerosols scatter most of the visible spectrum light" isn't quite right. Most aerosol-scattered light
is ultimately scattered in a forward direction. The "brightening" results only from that (relatively
54
-------
small, 10-15%) fraction of visible light that is back-scattered in the direction of the hypothetical
observer in space.
p. 9-211, 2nd para, line 1: Not clear what you mean by "progress has been made with in-situ and
remotely sensed aerosols."
55
-------
Russell Comments (Dr. Armistead (Ted) Russell)
First, I note that the word concise was removed from the first sentence, as is appropriate.
When I showed my air pollution physics and chemistry students the PM ISA, they said "Can't
they shorten it?", to which I replied this is shortened from the old Criteria Document approach,
and it should demonstrate how active is the field. (I also noted to them how much more is
known now then when I took the course about 29 years ago when taking a one semester course
pretty much put you near the front of the field.) On that note, however, I do think that future
versions (in coming years, not in the revision to this one) should be shortened. I found each
section to be extremely thorough and providing more information than needed to adequately
inform the process. This was particularly true for Chapter 3 (and this view may be due, in part,
to being most familiar with that area), where a part of the material will likely not have much role
in further review of the NAAQS. It almost appeared as though it was reverting back to the old
AQCD approach.
In spite of the call to shorten future versions, overall, I think the ISA has been improved
significantly. I still view that it does not bring enough attention to compositional differences in
PM effects (both health and welfare). There is significant information that carbonaceous
aerosols (typically referred to elemental and organic carbon) are of greater concern for many
endpoints. The ISA should lay out this information more completely.
I continue to worry about the treatment of climate. While the inclusion of parts of the
CCSP provides the foundation for discussing PM impacts on climate, given the importance of
this issue, and how it could impact how a secondary NAAQS might be formulated, a bit more
analysis would have been beneficial. Composition and size is very critical here, but is not fully
transmitted in the synthesis chapter.
Chapter 2 does a reasonably good job of integrating the findings of the rest of the ISA,
though a few improvements can be made. The major concern I have, at present, echoes the
comment above that there is not enough emphasis on compositional differences and their
implications. I would add to the Section 2.4," Policy Relevant Considerations" (or possibly a
new section "2.6. Composition Considerations" since it synthesizes beyond just health effects),a
subsection on composition that integrates what we understand and suspect about compositional
differences on health and welfare effects. A particular finding in such a synthesis could be that
control of organic and elemental carbon (EC/OC) PM from sources such as automobiles and
biomass burning appears to be more beneficial than controlling other components of PM (or it
could be couched that a EC/OC has greater deleterious impacts than the same amount of other
components).
I think in prior ISA's, there was a summary table of those impacts that were causal, likely to be
causal and suggestive right at the start of the section on effects. This should be followed here. It
is a bit odd that you find out the "inadequate" relationships first. Thus, add a summary table at
the beginning of 2.3 of the health effects examined. I find it a bit unnatural to suggest that the
evidence for the relationship between PM and visibility to be "strong and consistent" in that it is
well beyond strong and consistent, and if this were not the case it would be in direct
contradiction to our basic understanding of physics. While such terminology is appropriate for
the health effects where such a direct linkage is not so established, here it just seems odd.
56
-------
A final synthesis section might be useful that puts it all together, saying that PM with certain
characteristics/sources has the following impacts on what aspects of BOTH health and welfare
(and strength of association). This section would also synthesize uncertainties as to how they
impact our overall understanding and how to interpret results.. At present the chapter does not
really synthesize health and welfare and how the various uncertainties impact our overall
understanding, and ends rather weakly. Might there be a section synthesizing uncertainties
(possibly in the section discussed above or separately?
Minor:
2-7-28: should be ."...oxidation of isoprene, terpenes and..."
2-8-1 This is still in the research phase and its importance is still being explored.
2-9-28: Explain the variation in PRB levels.
2-9-33: Weak beginning of a section.2-11-28/30: I would hesitate to state this as such as it may
impact the use of such information in other assessments and it is not apparent you are using this
information in this assessment.
2-14-10: FontonPM2.5
2-14-14: You might provide a bit of explanation as to why this is consistent with other null
findings in the West.
2-19-12 (and related section in the report). In the section above, it appears as though the
information for this endpoint is very similar to that for acute CVD where the relationship was
found to be causal. Why the difference?
2-25-14/29 Much of what this section appears to be trying to explain can be explained by
compositional/source differences as opposed to the items discussed here, and I think
compositional differences provide a better explanation.
Section 2.3.5: It should be noted that UFP effects can not be separated from compositional
effects. Indeed, there are results that might suggest that sulfate-dominated UFP do not have the
same effects seen for CAPS and DE.
2-34-29: Status is not protective... I would say nutritional status can affect response.
Section 2.4.2: Lag structure is not so important here as the other two, and these parts could be
integrated in to the appropriate sections discussing the specific health endpoints.. It may be
important to interpreting the epi studies, and how they were conducted, but as presented this
section does not have that much to say as to whether the causal/likely causal... decisions were
appropriate, nor guide what the level of the standard might be chosen to be. For this section to
stand alone, here, the discussion of lag structure between the endpoints needs to be synthesized.
Chapter 3
As noted above, this chapter is very thorough, and is likely more exhaustive than necessary. I
still might have provided a bit more on the evaluation of exposure models, but if that is done as
part of the Risk Assessment, that is fine. In general, I found it to be balanced and correct, with a
few exceptions.
As an example of where the chapter really goes in to unnecessary detail, and where this leads to
an error, on page 150, they note that three dimensional CTMs typically use finite difference or
finite element approaches. This level of detail is not overly informative to reviewing the
NAAQS. Further, many models use finite volume approaches. Further, the approach used for
57
-------
deposition is incorrect for particulate deposition (there is no sedimentation velocity in the
formulation given), and the whole discussion of deposition could be shortened.
I was a bit taken aback by the paragraph on page 3-225 saying "Clearly there are variations...",
concluding that ""too poorly characterized on a broad scale to allow conclusions to be drawn.
I thought NARSTO did just that. While there are variations, there are typically consistent
features as well. This paragraph strikes me as a bit negative.
The last paragraph on page 3-226 seems out of place, since it is talking about impacts on the
health effects studies discussed later, and I did not pick this up from the chapter (I may have
missed it).
Minor:
Figure Numbering: It appears as though figures were removed, leading to figure numbers later
in the document being off.
3-13-22: You need to be very careful in how you interpret AMS measurements in terms of what
is secondary or not. They measure spectra and hypothesize what fraction is secondary.
3-15-8: The 2.8% contribution to OC is a bit misleading in that it represents only the two
diastereomisomers, and more SOA from isoprene can be present.
3-18-15: Should just be Maricq. (also line 20)
3-21-6: (eqn 3-1) This is not truly correct, particularly for particles, and wet deposition is
usually parameterized differently as well. This section is probably longer than needed here in
that what is of importance is to note the typical lifetimes and loss processes.
3-33-21: What is meant by reliable, noting that the prior discussion also suggests it is
consistently impacted by artifacts?
3-34-5: nonvolatile.
3-37-25: "AMS RESULTS did not..."
3-82-4: Move "only" about 5 words down.
3-82-21: I would remove "slightly".
3-83-5 ".. .where atmospheric nucleation"
Figs. 3-42/43: Usually the sizes are given small-big.
3-186: I would be a bit more tentative about deriving PMio-2.s from the PM2.5 and PMio values.
It is almost assuredly true that there is some correlation in the PRB PMio and PM2.5 values
calculated by Trijonis, and the current approach would actually indicate that in some
circumstances, PMio-2.5 is -90% of the PMio, which is almost assuredly not true. Use typical
PM2.s/PMio ratios for the two regions to get a better estimate of the range in the values.
Tables 3-19-23...: In many cases, the micro sign is given as a box.
It is sometimes the case that the rest of the sentence after an equation is given is indented (e.g., 3-
205-14).
3-223-18: "higher in clear"
3-223-19: remove the";"
3-224-3: Use -34 times more
3-225-25:"... 1-h commute...'
58
-------
Speizer Comments (Dr. Frank Speizer)
Chapter 2
General Comments:
EPA Staff are to be commended for doing an excellent job in pulling together and
presenting in a very readable format the summary of the data contained in the ISA. By placing
the conclusions here with appropriate documentation of where to find the primary discussed data,
the whole document becomes more useful.
Specific Comments:
Page 2.5, lines 14-19. Although accurate as stated, what seems to be left out in the comparison
of PM 2.5 to PMio is that the PMio correlations in themselves (.70) are really remarkably high and
suggest, although not quite as consistent as PM2.5 that they can be usefully used to model
exposure. It might be worth mentioning this.
Pge 2.6, line 30. Isn't there also the potential for greater production in the winter of NO2
(though less efficient winter burning or greater production) to lead to increased quenching of
O3?
Page 2.13, lines 21-25 bullets. It might be useful to add to the end of each of these lines the
specific sections in Chapter 6 or 7 where each topic is discussed. Although I can agree with each
of the conclusions, I can imagine some "stakeholders" wanting to raise questions and getting
hung up on not having the specific background data to support the statement. By providing the
specific paragraph reference it would make documentation easier.
Page 2.14, Table 2.1. This may be quibbling over semantics. I like putting the summary table
up front. However, by describing the effects as indicated, the suggestion is that short term
exposure is causing a chronic condition. These effects are for the most part acute events
occurring among persons with chronic conditions. (If the table occurred after the descriptions
that follow there would be no problem). Perhaps a modest change in the Table title or a footnote
would help.
Page 2.15, first two paragraphs. I would suggest reversing the order of these paragraph (with
minor editorial changes). Offering the fresh diesel exhaust data before the PM2.5 data seems to
be mixing the consistency argument by suggesting that UFP are explaining PM2.5 effect when
they are really supplementing the findings of the PM2.5 effects reported in the second paragraph.
Page 2.21 and 2.22, Figures 2.1 and 2.2. Ordering the points in these figures by level of
exposure estimate has made them less useful. They mix up outcomes that really cannot be
looked at as a result of gradient of exposure. It would be better if they were divided into small
figures by specific outcomes, and then look at gradients of exposure.
Pages 2.261-2.30, whole section. Not enough emphasis is made of the fact that for each of the
outcomes discussed there are only a limited number of studies, thus the designation of
59
-------
suggestive. Again breaking out the kinds of outcomes in Figure 2.3 would be better than
lumping them all together.
Page 2.29, Figure 2.3: There is clearly something wrong here. The units in this Figure are
wildly different from Figure 2.1 and 2.2, and are simply not believeable. I just may be a labeling
problem, but it needs to be fixed.
Page 2.33, paragraph beginning on line 14. Some editing needed. The first sentence says
controlled study effects seen in susceptible subjects with CVD and/or respiratory diseases the
last sentence says respiratory diseases. Suggest change the last to cardiorespiratory diseases.
Responses to Charge Questions (paraphrased and shown in italics) on ISA Draft 2
1. Evaluation of health evidence in Chapters 6 and 7
a. The role of the studies ofPM10 in consideration in causality determinations for
PM25 andPM10-25.
Although this seems perfectly appropriate, it would seem that leading into this
argument there should be some material (which I looked for in Chapter 3, but could
not find, maybe just my inability to find) that could be referenced here or perhaps
even summarized to justify the correlative nature of the relationships. This would
help justify using the PMio studies when neither PM2.5 or PM 10-2.5 were available to
further assure the consistency of the data.
b. The inclusion of cause-specific mortality as part of suite of CVD and
respiratory effects, in the development of causality judgments.
The use of cause-specific mortality provides a refinement of the data that is
logical and appropriate. Since the bulk of the mortality is dominated by
cardiovascular deaths it is appropriate to show these separately, and similarly since
respiratory deaths are the fourth leading cause of deaths these also provide important
information. By showing these two classes of death separately and then showing the
total (which is clearly dominated by these two, but less subject to reporting or disease
classification error) the reader gets a better understanding of both the magnitude and
nature of the deaths that are occurring.
c. Scope of evidence considered in the causality determination for ultrafine PM.
The change of the designation from inadequate to suggestive appears to be
dominated by the interpretation of the relatively greater number of clinical and
lexicological studies of diesel and gasoline exhaust studies, that are clearly dominated
by UFP. This is in contrast to the few epidemiological studies, that are for the most
part null or poorly defined as UFP exposure studies. Thus, the designation of
suggestive is rational and probably right, but clearly the designation itself will need
additional work in the future to either sustain or refute the designation. This probably
should be indicated.
Charge Question Id:
d. With regard to the role ofPM in mutagenicity, genotoxicity and carcinogenicity
comment upon the expansion of this evaluation to include a summary of
toxicological studies using routes of exposure other than inhalation, as well as
consideration of both mortality and incidence studies.
60
-------
Staff has done an excellent job in summarizing a complex literature related to
both the general and specific associations of using mutagenicity, genotoxicity and
carcinogenicity studies in assessing the PM effects on cancer. The charge question is
really in two parts. With regard to using routes of exposure other than inhalation, to
assess mutagenicity and genotoxicity, these are standard procedures used in the field
and have formed the basis of much of the general literature basis of mechanistic
understanding in mammalian cancer research. There are examples where such studies
have resulted in leads to where to look for cancer risks in humans. Thus it is
appropriate to use such data in determining where along the spectrum of certainty
with regard to causality these studies should be considered. Secondly, with regard to
the population studies that are reported in the last few years, these have been large,
well assessed groups, in which the same exposure parameters have been used for
cardiovascular and respiratory findings (that appear to be defined as causal
associations) with an outcome that is well defined, albeit it dominated by the potential
confounder of smoking. The relatively consistent finding of lung cancer excesses as
an exclusively associated cancer, along with the attempts to control for smoking are
suggestive, and thus the appropriate interpretation and designation appears to have
been made.
2. Revisions to Chapters 1, 4, 5.
Chapter 1, expansion or additions to sections on history of previous PMNAAQS
reviews.
It seems to me this is just about right. There is a slight shift of tone from the early
historical discussion of setting the NAAQS to a more "legalistic" description of what has
gone on in the last 10 years or so. I found one issue (described in a separate memo) with
regard to table 1.2 on page 1.26. The category of Experimental Evidence seems
inappropriately defined.
Chapter 4, and 51 leave to others to comment upon.
3. Chapter 8 Susceptible populations
Please comment on the organization and presentation in Chapter 8 of evidence
regarding susceptible subpopulations
Although the Chapter summarizes definitions used by others (and EPA) for susceptibility,
the definitions provided are not very useful. It might have been better to talk about host vs
environment more formally. In addition, in discussing each of the "susceptibility factors"
much of what is reported is really not much different that what is contained in Chapters 6 &
7. However, the discussion is better organized in defining and using susceptibility factors as
related to PM exposure. What is clearly left out, and perhaps should have a brief discussion
is a contrast between susceptibility and vulnerability and how these factors play into a
discussion of risk and margin of safety.
4. Chapters.
Comments on revisions
61
-------
This chapter read rather well. I leave to others specific comments on Sections 3.4 and
3.5. In section 3.8 the Staff has reviewed the factors potentially influencing exposure levels
for health studies. I would suggest a summary table at the end of this chapter that takes all
the factors and provides qualitative directionality for the potential factor in biasing away or
toward the null (or not at all). This I think would identify more readily for the reader what
to worry about and what not to be concerned about.
5. Chapter 2.
Please comment on the integration.
I believe this is an important chapter and I have already submitted comments on this
chapter separately. I repeat them here simply to put all my comments together. Clearly I
have done this in greater detail than what I have done to answer other questions.
62
-------
Suh Comments (Dr. Helen Suh)
Charge Question 3:
Revisions to Chapter 3 on Source to Exposure: Consistent with revisions made to the health
effects chapters, Chapter 3 was revised to clarify that PM 10 incorporates both PM2.5 and
PM10-2.5 and reorganized to begin with PM2.5 andPM10-2.5, followed by PM10, where
applicable. The discussion of measurement techniques and chemistry ofPM10-2.5 has been
expanded in Sections 3.4 and 3.5, in response to CASAC comments. In addition, Section 3.8 on
human exposure toPMhas been reorganized and expanded to better characterize the evidence
and provide useful information for interpretation of epidemiologic studies. We would appreciate
comments from the CASAC PM Panel on these revisions.
The Chapter is a massive and comprehensive summary of the current state of the PM sources,
monitoring and modeling methods, concentrations and exposures. The Chapter was generally
clear, comprehensive, and well organized, which is notable given the amount of information it
contains. It clearly and appropriately incorporates comments from the April review. Of note, the
relation among PM2.5, PMio-2.5, and PMio was greatly clarified and the inclusion of additional
information on UFP, PM components, and PMio-2.5 monitoring methods was welcomed.
Similarly, the organization and content of the exposure section was greatly improved, especially
with regards to the relevance of exposure findings to epidemiologic studies. While significantly
improved, further improvements could be made to the exposure section to help with the
organization, seeming overlap with earlier sections, and clarity, as discussed briefly below.
Given the importance of measured PM in the earlier sections of the Chapter, in the
epidemiological studies, and also in the Health REA, it makes sense to place the discussion
of exposure measurements before that for exposure modeling. Further, the exposure
measurements section should include discussion of new methods that have been made to
measure (1) personal exposures to total PM, (2) PM of outdoor origin, (2) outdoor
concentrations outside individuals' homes, and (3) indoor PM exposures to total PM and to
PM of outdoor origin). In this context, studies using tracers of outdoor or regional pollution
can be introduced.
The discussion of each topic should be consolidated into single sections, with references to
this section made as necessary in later sections. As is currently written, certain topics, such
as spatial variability and particle infiltration, appear repeatedly in several subsections. The
discussion on these topics would be more focused and less confusing if their discussions
were consolidated.
In Section 3.8.4.3, it would be helpful to clarify which component of PM is being discussed.
Further, within each subtopic or heading, separate or otherwise distinct discussions for PM2.5,
PMio-2.5 and the PM components would help with clarity.
Section 3.8.4 (Exposure Assessment Studies at Different Spatial Scales) and its subheadings
are misnamed. While the heading titles do provide parallels with earlier sections of Chapter
3, they do not accurately describe their contents. For example, the section 3.8.4.1 does not
actually discuss urban scale ambient PM exposure, rather issues related to exposure error,
exposure modeling, and tracers of ambient particles.
- It would be helpful to include a subsection in the beginning of the section, perhaps in Section
3.8.1 (General Exposure Concepts) that summarize the range of measured exposures to
63
-------
PM2.5, PMio-2.5, and PMio, their composition, and differences by city, susceptible group, and
season.
Section 3.8.6: Other factors that could affect exposures are worth noting, such as home
ventilation patterns and activity patterns.
Response to Charge Question 4
Charge Question 4:
Revisions to Chapter 3 on Source to Exposure: Consistent with revisions made to the health
effects chapters, Chapter 3 was revised to clarify that PMIO incorporates both PM2.5 and
PMIO-2.5 and reorganized to begin with PM2.5 andPM10-2.5, followed by PMIO, where
applicable. The discussion of measurement techniques and chemistry ofPM10-2.5 has been
expanded in Sections 3.4 and 3.5, in response to CASAC comments. In addition, Section 3.8 on
human exposure to PMhas been reorganized and expanded to better characterize the evidence
and provide useful information for interpretation of epidemiologic studies. We would appreciate
comments from the CASAC PM Panel on these revisions.
The Chapter is a comprehensive summary of the current state of the PM sources, ambient
aerosols, monitoring and modeling methods, concentrations and exposures. The Chapter
generally reflects the state of science in a clear, comprehensive, and well organized fashion,
which is notable given the amount of information the Chapter contains. Further, the Chapter
revision appropriately incorporates comments from the April review. Of note, the relation
among PM2 5, PMio-2.5, and PMio was greatly clarified and the inclusion of additional
information on UFP PM components, and PMio-2.5 monitoring methods was welcomed.
Similarly, the organization and content of the exposure section was greatly improved, especially
with regards to the relevance of exposure findings to epidemiologic studies. While significantly
improved, further improvements could be made to the exposure section to help with the
organization, seeming overlap with earlier sections, and clarity, as discussed briefly below.
Specific comments:
• Standard Reference Materials (page 2-20, line 20): It says "To date, there are few standard
reference materials ..."
This sentence should be reworded to make it clear that we currently do not have a way to
assess mass concentration measurement accuracy, as there are no materials (to our
knowledge) that provide a test of accuracy or airborne PM mass. There are, however, SRMs
that are useful for testing the accuracy of the analytical methods for PM composition.
• Policy Relevant Background (PRB)
There are the usual concerns regarding the policy relevant background (PRB). The EPA has
to set a health-based standard irrespective of the background concentration of particles and
thus, the PRB is really an implementation issue and not a regulation setting issue. As the
PRB increases because of activity outside of the United States, there is decreasing flexibility
in attaining concentrations that are fully protective of public health. The modeling approach
has some appeal, but given the problems with the accuracy of the emissions inventories in the
US where substantial effort and resources are expended to develop them and there are still
problems, the likelihood that we can adequate model emissions in the rest of the world seems
very low. It would seem that there could be some comparisons between the model-based
64
-------
approach and the remote site approach to provide some basis for confidence in the model
estimation of the PRB.
Exposure Section
o Given the importance of measured PM in the earlier sections of the Chapter, in the
epidemiological studies, and also in the Health REA, it makes sense to place the
discussion of exposure measurements before that for exposure modeling. Further, the
exposure measurements section should include discussion of new methods that have been
made to measure (1) personal exposures to total PM, (2) PM of outdoor origin, (2)
outdoor concentrations outside individuals' homes, and (3) indoor PM exposures to total
PM and to PM of outdoor origin). In this context, studies using tracers of outdoor or
regional pollution can be introduced.
o The discussion of each topic should be consolidated into single sections, with references
to this section made as necessary in later sections. As is currently written, certain topics,
such as spatial variability and particle infiltration, appear repeatedly in several
subsections. The discussion on these topics would be more focused and less confusing if
their discussions were consolidated.
o In Section 3.8.4.3, it would be helpful to clarify which component of PM is being
discussed. Further, within each subtopic or heading, separate or otherwise distinct
discussions for PM2.5, PMio-2.5 and the PM components would help with clarity.
o Section 3.8.4 (Exposure Assessment Studies at Different Spatial Scales) and its
subheadings are misnamed. While the heading titles do provide parallels with earlier
sections of Chapter 3, they do not accurately describe their contents. For example, the
section 3.8.4.1 does not actually discuss urban scale ambient PM exposure, rather issues
related to exposure error, exposure modeling, and tracers of ambient particles.
o It would be helpful to include a subsection in the beginning of the section, perhaps in
Section 3.8.1 (General Exposure Concepts) that summarize the range of measured
exposures to PM2.5, PMio-2.5, and PMio, their composition, and differences by city,
susceptible group, and season.
o Section 3.8.6: Other factors that could affect exposures are worth noting, such as home
ventilation patterns and activity patterns.
65
-------
Vedal Comments (Dr. Sverre Vedal)
la. UseofPMio.
Findings pertaining to PMio in this 2nd draft are discussed in conjunction with two component
fractions (PMio-2.5 and PM2 5); causal determinations are no longer provided for PMio itself.
There is value in this approach, since both fractions comprise PMio. It does not, however, seem
logical that evidence on PMio would be equally relevant to both PM2.5 and PMio-2.5- In some
regions of the country PMio is more reflective of PM2.5 whereas elsewhere PMio is more
reflective of PMio-2.5.
In the respective PM fraction Summary and Causal Determinations sections, there is little
mention of PMio. Where PMio is mentioned in this context, it only comes up in support of PM2.5
(e.g., PMio evidence is cited as strengthening the evidence for PM2.s-related respiratory effects
[p. 6-244] and for mortality effects [p. 6-317]). PMio evidence is not used in specific support of
PMio-2.5 effects. Yet in the discussion of short-term PM exposure mortality effects (section
6.5.2), PMio studies are cited as providing "an underlying basis for the overall pattern of
associations observed when examining the relationship between PMio-2.5 and PM2.5 and
mortality." There is inconsistency then in the way PMio findings are used.
findings are also used in evaluation of long-term PM fraction exposure effects (e.g., p. 7-
24). For respiratory morbidity effects of long-term PM exposure, a more precise argument is
made that "studies showing associations only with PMio were conducted in locations where PM
was predominantly fine particles, providing support for associations with long-term exposure to
fine particles" (p. 7-28, line 32; p. 7-60, line 16). However, in some studies effects are only seen
for PMio and not for PM2.5 (p. 7-43, line 6), which is not consistent with that statement.
In order to better interpret PMio findings relative to those of PM2.5 and PMio-2.5, it would have
been helpful, perhaps in Chapter 3, to include and/or highlight data on the relationship between
PMio and both PM2.5 and PMio-2.5, stratified by region of the country. Also, in light of the role of
the current 24-hour PMio standard with respect to PMio-2.5, it would be important to bring
together the evidence relating to PMio that allows it to be used for this purpose.
In short, the way in which PMio is used is vague and inconsistent. PMio evidence is used
generally to support overall effects of PM, but specifically only in support of PM2.5 effects. The
role of the current PMio standard, however, is in controlling PMio-2.5- Better justification for the
utility of PMio findings is needed from a scientific and policy perspective. Nevertheless,
providing causality judgments for PMio-2.5 and PM2 5, but not for PMio, as is done in the current
draft, is more in line with the current science.
66
------- |