v UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
\ WASHINGTON B.C. 20460
OFFICE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD
June 10, 2014
EPA-C AS AC-14-001
The Honorable Gina McCarthy
Administrator
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, N.W.
Washington, D.C. 20460
Subj ect: C ASAC Review of the EPA's Integrated Review Plan for the Primary National
Ambient Air Quality Standards for Nitrogen Dioxide (External Review Draft)
Dear Administrator McCarthy:
The Clean Air Scientific Advisory Committee (CASAC) Oxides of Nitrogen Primary National Ambient
Air Quality Standards (NAAQS) Review Panel met on March 12, 2014, and May 7, 2014, to peer
review the EPA's Integrated Review Plan for the Primary National Ambient Air Quality Standards for
Nitrogen Dioxide (ExternalReview Draft)., hereafter referred to as the Draft IRP. The Chartered CASAC
approved this report during a public teleconference on May 7, 2014. The CASAC's consensus responses
to the agency's charge questions and the individual review comments from the CASAC Oxides of
Nitrogen Review Panel are enclosed.
Overall the CASAC finds the Draft IRP to be well written, well organized, and the topics are clearly
presented. There are several recommendations for strengthening and improving the document
highlighted below and detailed in the consensus responses. With the recommended revisions, the Draft
IRP should serve its intended purpose in presenting the review plan, schedule, and process as well as the
key policy-relevant science issues that will guide the Primary NAAQS for Nitrogen Dioxide (NO2)
review. A key purpose of the IRP is to present the key policy-relevant science issues that will inform the
Integrated Science Assessment (ISA); thus, completion of the IRP should precede the completion of the
Draft ISA, which was not the case for this review. Although the EPA did seek consultative advice on a
portion of the IRP document from the CASAC Oxides of Nitrogen Panel in June 2013, the CASAC
recommends that for future reviews, the IRP be completed and reviewed in its entirety prior to the
completion of the First Draft ISA.
The CASAC recommends providing more detail and clarity regarding data from the new near-road
monitoring network. The Draft IRP should indicate whether any of the near-road monitoring data will be
obtained prior to the development of the Risk and Exposure Assessment (REA) document.
The Draft IRP includes many of the key policy-relevant questions for consideration in the current
NAAQS review. The CASAC identified a few additional policy-relevant questions that should be
-------
included regarding important biological mechanisms and related modes of action, the assessment and
use of exposure measurement and modeling error, and information on at-risk populations that live near
NO2 sources.
In the ambient environment, NO2 exposures occur with those of copollutants. More attention should be
given to NO2 as a component of multi-pollutant mixture exposures. The Draft IRP needs to clearly
frame the issue of copollutant exposures and clearly describe the plan for separating effects associated
with NO2 from effects associated with copollutants. The Draft IRP contains a list of questions that could
be expanded to better assess the degree to which environmental pollutants, such as ozone or combustion-
related copollutants (e.g., carbon monoxide, metals, black carbon, some organic species, fine particles
and ultrafine particles) might act as confounders in epidemiological studies used in the ISA.
Consideration also should be given to how these pollutants might act as effect modifiers in evaluations
of mode of action or mechanisms and how the pollutants might interact in the atmosphere to alter spatial
distributions and exposures. The near-road monitoring data may help inform these issues and there
should be a discussion of how the data will be used.
The CASAC recommends including sensitivity analyses and formal uncertainty analyses (quantitative,
when possible) in the plan for the Risk and Exposure Assessment (REA). These analyses would
strengthen both the transparency and confidence in the REA and also might identify key gaps in the
science related to oxides of nitrogen and inform both the research and regulatory community on future
priorities.
The CASAC appreciates the opportunity to provide advice on the Draft IRP and looks forward to the
EPA's response.
Sincerely,
/Signed/
Dr. H. Christopher Frey, Chair
Clean Air Scientific Advisory Committee
Enclosures
-------
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. The 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. The CASAC reports are posted on the EPA
website at: http://www.epa.gov/casac.
-------
U.S. Environmental Protection Agency
Clean Air Scientific Advisory Committee (CASAC)
CHAIR
Dr. H. Christopher Frey, Distinguished University Professor, Department of Civil, Construction and
Environmental Engineering, College of Engineering, North Carolina State University, Raleigh, NC and
Visiting Professor, Department of Civil and Environmental Engineering, Adjunct Professor, Division of
Environment, Hong Kong University of Science and Technology
MEMBERS
Mr. George A. Allen, Senior Scientist, Northeast States for Coordinated Air Use Management
(NESCAUM), Boston, MA
Dr. Ana Diez-Roux, Dean, School of Public Health, Drexel University, Philadelphia, PA
Dr. Jack Harkema, Professor, Department of Pathobiology, College of Veterinary Medicine, Michigan
State University, East Lansing, MI
Dr. Helen Suh, Interim Chair, Director of Population Health Doctoral Program, Department of Health
Sciences, Northeastern University, Boston, MA
Dr. Kathleen Weathers, Senior Scientist, Gary Institute of Ecosystem Studies, Millbrook, NY
Dr. Ronald Wyzga, Technical Executive, Air Quality Health and Risk, Electric Power Research
Institute, Palo Alto, CA
SCIENCE ADVISORY BOARD STAFF
Mr. Aaron Yeow, Designated Federal Officer, U.S. Environmental Protection Agency, Science
Advisory Board (1400R), 1200 Pennsylvania Avenue, NW, Washington, DC
-------
U.S. Environmental Protection Agency
Clean Air Scientific Advisory Committee
Oxides of Nitrogen Primary NAAQS Review Panel
CASAC CHAIR
Dr. H. Christopher Frey, Distinguished University Professor, Department of Civil, Construction and
Environmental Engineering, College of Engineering, North Carolina State University, Raleigh, NC and
Visiting Professor, Department of Civil and Environmental Engineering, Adjunct Professor, Division of
Environment, Hong Kong University of Science and Technology
CASAC MEMBERS
Mr. George A. Allen, Senior Scientist, Northeast States for Coordinated Air Use Management
(NESCAUM), Boston, MA
Dr. Jack Harkema, Professor, Department of Pathobiology, College of Veterinary Medicine, Michigan
State University, East Lansing, MI
Dr. Helen Suh, Interim Chair, Director of Population Health Doctoral Program, Department of Health
Sciences, Northeastern University, Boston, MA
Dr. Ronald Wyzga, Technical Executive, Air Quality Health and Risk, Electric Power Research
Institute, Palo Alto, CA
CONSULTANTS
Dr. Matthew Campen, Associate Professor, College of Pharmacy, University of New Mexico,
Albuquerque, NM
Dr. Ronald Cohen, Professor, Chemistry, College of Chemistry, University of California, Berkeley,
Berkeley, CA
Dr. Douglas Dockery, Professor and Chair, Department of Environmental Health, School of Public
Health, Harvard University, Boston, MA
Dr. Philip Fine, Assistant Deputy Executive Officer, South Coast Air Quality Management District,
Diamond Bar, CA
Dr. Panos Georgopoulos, Professor, Environmental and Occupational Medicine, Rutgers University -
Robert Wood Johnson Medical School, Piscataway, NJ
Dr. Michael Jerrett, Professor and Chair, Division of Environmental Health Sciences, School of Public
Health , University of California, Berkeley, Berkeley, CA
-------
Dr. Joel Kaufman, Professor, Department of Environmental Health & Occupational Health, University
of Washington, Seattle, WA
Dr. Michael T. Kleinman, Professor, Department of Medicine, Division of Occupational and
Environmental Medicine, University of California, Irvine, Irvine, C A
Dr. Timothy V. Larson, Professor, Department of Civil and Environmental Engineering, University of
Washington, Seattle, WA
Dr. Jeremy Sarnat, Associate Professor of Environmental Health, Rollins School of Public Health ,
Emory University, Atlanta, GA
Dr. Richard Schlesinger, Associate Dean, Dyson College of Arts and Sciences, Pace University, New
York, NY
Dr. Elizabeth A. (Lianne) Sheppard, Professor, Biostatistics and Environmental & Occupational
Health Sciences, School of Public Health, University of Washington, Seattle, WA
Dr. Junfeng (Jim) Zhang, Professor of Global and Environmental Health, Division of Environmental
Sciences & Policy, Nicholas School of the Environment & Duke Global Health Institute, Duke
University, Durham, NC
SCIENCE ADVISORY BOARD STAFF
Mr. Aaron Yeow, Designated Federal Officer, U.S. Environmental Protection Agency, Science
Advisory Board (1400R), 1200 Pennsylvania Avenue, NW, Washington, DC
-------
Consensus Responses to Charge Questions on
EPA's Integrated Review Plan for the Primary National Ambient Air Quality Standards
for Nitrogen Dioxide (External Review Draft)
Overall Organization and Clarity
To what extent does the Panel find that the draft IRP clearly and appropriately communicates the plan
for the current review of the primary NO2 NAAQS and the key scientific and policy issues that will guide
the review? To what extent are the decisions made in the last review, including the rationales for those
decisions, clearly articulated?
The Draft IRP is very well organized and the plan for the current review is clearly presented. The history
of the NO2 NAAQS and the summary and rationale for the decisions made in the previous NO2 NAAQS
review are clear and concise. The summary of the previous review in section 1.1 may benefit from
moving information related to legal aspects elsewhere, perhaps in footnotes, to improve readability.
Chapter 1 (Introduction) and Chapter 2 (Schedule)
To what extent does the Panel find that Chapters 1 and 2 clearly communicate the NAAQS legislative
requirements, summarize the steps in the review process, summarize the history of the NO2 NAAQS, and
present the anticipated schedule for the current review?
Chapters 1 and 2 are well written and clearly communicate the NAAQS legislative requirements, the
steps in the review process, the history of the NO2 NAAQS, and the anticipated schedule for the current
review. It would be helpful if Chapter 1 could note the start date for new literature considered for this
review cycle (early 2008 as noted in section 3.1.1) and the anticipated end date of studies considered
based on the current review schedule presented in Chapter 2.
There is some concern about the timing of the availability of sufficient NO2 data from the near-road
monitoring network relative to the need for these data to be used in the Risk and Exposure Assessment
(REA) document (assuming an REA is developed for this review). An estimate of the number of near-
road sites and amount (duration) of data from these sites that is expected to be available for use in the
REA would be useful.
For near-road sites where the anticipated NO2 design value is more than half of the NAAQS, there is a
need for site descriptive information such as the proportion of truck traffic. Other traffic-related
information, such as time-resolved traffic counts, and distance from the road would be helpful for a
better understanding of the traffic factors that impact a specific site.
-------
Chapter 3 - Key Policy-Relevant Issues
Building on key considerations and issues addressed in the last review, Chapter 3 presents a set of
policy-relevant questions that will serve as a focus in this review. To what extent does the Panel find
that these questions appropriately characterize the key scientific and policy issues for consideration in
the current review? Are there additional issues that should be considered?
Chapter 3 presents a set of policy-relevant questions that will serve as a focus for the review of the
NAAQS. Chapter 3 is well written and Figure 3-1 is particularly useful, as is the historical context
summarized at the beginning of the chapter. In addition, the questions appropriately characterize many
of the key scientific and policy issues for consideration in the current review.
The set of questions at the end of the chapter are overarching questions. The more specific questions in
the other chapters of the IRP should be mapped to these overarching questions.
There are a few missing questions that should be added, including:
1. If new near-road monitoring data become available, how will they be used in the ensuing risk
assessment?
2. What are the important biological mechanisms and related modes of action?
3. What are the important issues related to the assessment and use of exposure measurement and
modeling error?
4. What information exists on at-risk populations near NO2 sources?
Chapter 4 - Science Assessment
Chapter 4 describes the plan for the Integrated Science Assessment. Chapter 4 reflects revisions made to
the Draft Plan for Development of the ISA for Nitrogen Oxides - Health Criteria, with consideration of
comments received during the June 2013 consultation with CASAC.
To what extent does Chapter 4 clearly and adequately describe the scope, specific issues to be
considered, and organization of the ISA ? Please provide suggestions for any other issues that should be
considered.
Chapter 4 provides the framework for the Integrated Science Assessment (ISA), highlighting the prior
areas of identified uncertainty, and proposing an approach to assessing the new literature (and totality of
evidence to date) related to oxides of nitrogen. The chapter is well written and provides a useful
framework for the scientific questions that need to be addressed. There is some concern about the
description of criteria for studies to be included in the literature review, and how decisions will be made
with regard to the pertinence of studies to the questions at hand. More attention should be given to NO2
as a component of multi-pollutant mixture exposures. The chapter contains a list of questions that could
be expanded to better assess the degree to which environmental pollutants, such as ozone or combustion-
related co-pollutants (for example, carbon monoxide, metals, black carbon, some organic species, fine
particles and ultrafine particles) might act as positive or negative confounders in epidemiological studies
used in the ISA. Consideration should also be given to how these pollutants might act as effect modifiers
-------
in evaluations of mode of action or mechanisms and how the pollutants might interact in the atmosphere
to alter spatial distributions and exposures.
With regard to animal research, will recent articles be held to new standards for reporting research, such
as the Animals in Research: Reporting In Vivo Experiments (ARRIVE) guidelines (Kilkenny et al.,
2010)?
Specific attention should be paid to the role of exposure assessment with regard to spatial and temporal
variability and measurement error and how they relate to health effects. The discussion of measurement
error starts well but does not discuss how measurement error will be handled in the ISA. A very
important feature of error in exposure assessment is with respect to its adequacy for inference about
health effects. This needs to be clearly discussed in the Draft IRP.
Please comment on the adequacy of the expanded discussion in Section 4.4 of issues that will be
considered in the ISA related to: (a) spatial heterogeneity in ambient concentrations of oxides of
nitrogen, particularly near- and on-road gradients, and implications for human exposures and (b)
various factors to consider in the evaluation of health effects associated with ambient NO2 exposure,
including traffic, noise, indoor NO2 exposures, and copollutant exposures.
In general the expanded discussion in Section 4.4 is adequate. However, the language concerning
"multi-pollutant" exposures is inconsistent in the document. In non-experimental settings, it is difficult
to disentangle effects associated with NO2 exposure from effects associated with other near-combustion-
source exposure. Moreover, source control measures that change NO2 exposure are likely to have
impacts on a suite of traffic-related air pollutants. The IRP should more clearly lay out the framework
for how the EPA will separate out the effects associated with NO2 from the effects associated with
copollutants.
Other general concerns for this chapter
Chapter 4 would be further improved with an enhanced discussion of at-risk populations, including
identification of additional at-risk groups that may have enhanced exposures to NO2, including groups
that have higher exposures as a result of their occupation, commuting exposures, and physical activity.
Further, because exposures to oxides of nitrogen may be heavily influenced by time-in-traffic, the IRP
should call for a specific focus on characterizing the exposures and response within the on-road
microenvironment.
This section of the Draft IRP also lacks a discussion of socio-demographics of exposure, which may be
an important consideration.
Chapter 5 - Quantitative Risk and Exposure Assessment
Chapter 5 summarizes the key risk and exposure analyses from the last review, including associated
uncertainties, and discusses our planned approach to considering the potential for additional analyses
in the current review. To what extent does Chapter 5 clearly and adequately describe the scope and
specific issues, including the identification of the most important uncertainties, to be considered in
developing the REA Planning Document for this review? To what extent is there additional information
-------
that should be considered or additional issues that should be addressed in considering the potential for
risk and/or exposure analyses in the current review?
Generally, Chapter 5 is well written and presents a reasonable series of approaches for conducting an
REA for NO2. The CASAC appreciates the manner in which the uncertainties from the 2008 REA are
identified. Many of these uncertainties will serve as important topics for discussion in the upcoming
REA planning document. In particular, a proper characterization of NO2 spatial heterogeneity remains a
key uncertainty and challenge in developing a future REA.
The CASAC recommends including sensitivity analyses and formal uncertainty analyses (quantitative,
where possible) in the plan for the REA. Including sensitivity analyses of the primary modeled input
parameters would strengthen both transparency and confidence in the REA. For example, the use of
alternative or a range of plausible concentration-response functions for the epidemiology-based human
health risk assessments would be useful (Page 5-17). Similarly, for both the air quality and human
exposure components of the assessment, conducting formal uncertainty analyses or presenting
propagation of error findings might identify key gaps in the science related to oxides of nitrogen and
inform both the research and regulatory community on future priorities.
It is not clear how a future REA will utilize new methodological approaches, specifically with regard to
hybrid and fused methods for estimating NO2 spatiotemporal distributions. Models based on the
AERMOD and Community Multi-scale Air Quality (CMAQ) models would generate refined near-
roadway estimates, while accounting for potential chemistry and meteorological influences. For more
details on new approaches that could be considered in an REA, see the individual comments by Dr.
Georgopoulos.
The Draft IRP correctly cites the new near-road monitoring network as an important source of future
information concerning NO2 spatial distributions near roadways. Over time, these data hopefully will
provide key information for use in exposure and risk analyses. Even in the absence of the near-road
network data, however, more information currently exists concerning near-road NO2 concentration
gradients that could be utilized in a future REA. Individual comments from Dr. Sheppard note
specifically that the Multi-Ethnic Study of Atherosclerosis and Air Pollution (MESA Air) study team
has assembled geographic information on Air Quality System (AQS) sites that could be leveraged to
inform questions on near-road NO2.
The CASAC stresses the importance of the on-road microenvironment in contributing to total oxides of
nitrogen exposure. Thus, modeling approaches used in the REA (i.e., Air Pollutants Exposure Model
[APEX]) should pay specific attention to accurate characterizations of on-road and commuting
exposures.
Although the focus of the REA should be on health risks associated with exposures to ambient NO2,
caution should be taken not to diminish the contribution to total exposure from indoor NO2 sources, as
well as the potential human health risks from these sources.
In the epidemiology-based REA approach, a major issue is whether the observed effects are from
exposure to NO2 alone, from other pollutants, or from the pollutant mixture. Although it is discussed
within the context of uncertainty, this issue should receive greater emphasis. See individual comments
from Dr. Zhang for further details.
10
-------
Chapter 6 - Ambient Air Monitoring
To what extent does Chapter 6 clearly and appropriately communicate, for the purposes of this plan, the
key aspects of measurement methods and surveillance network requirements for the NO2 NAAQS?
The discussion in Chapter 6 about NO2 measurement is useful and to the point with respect to NAAQS
measurement of NO2. The chapter can be strengthened by:
• Reviewing it for consistency with Chapter 4 (pp. 4-11 to 4-12 and the questions at the top of p.
3-15);
• Giving greater attention to an evaluation/discussion of other measurements that are key to
interpreting the near-road network data; and
• Discussing two distinct purposes of observations (and their implications for network design):
o to assess compliance with a NO2 standard;
o to better understand exposures and health effects of near road pollutants. This requires
measurements in addition to NO2 including nitric oxide, carbon monoxide, PIVh.s, and
other non-NAAQS pollutants, such as black carbon and particle number concentration.
Chapter 7 - Policy Assessment and Rulemaking
To what extent does Chapter 7 clearly summarize the general process for the policy assessment and
rulemaking phase of this review?
Chapter 7 is well written and covers the relevant topics. The CASAC has no recommended changes to
this chapter.
11
-------
References
Kilkenny, C., Browne, W. I, Cuthill, I. C., Emerson, M., and Altman, D. G. (2010). Improving
bioscience research reporting: The ARRIVE guidelines for reporting animal research. PLoS Biol,
8(6):el000412+.
12
-------
Appendix A
Individual Comments by CASAC Oxides of Nitrogen Primary NAAQS Review Panel Members on
EPA's Integrated Review Plan for the Primary National Ambient Air Quality Standards
for Nitrogen Dioxide (External Review Draft)
Mr. George A. Allen A-2
Dr. Matthew Campen A-4
Dr. Ronald C. Cohen A-5
Dr. Philip M. Fine A-6
Dr. Panos G. Georgopoulos A-7
Dr. JackHarkema A-10
Dr. Michael T. Kleinnian A-ll
Dr. Timothy V. Larson A-12
Dr. Jeremy Sarnat A-14
Dr. Richard Schlesinger A-15
Dr. Elizabeth A. (Lianne) Sheppard A-16
Dr. Ronald E. Wyzga A-19
Dr. Junfeng (Jim) Zhang A-20
A-l
-------
Mr. George A. Allen
These comments focus on Chapters 1 and 2 (Introduction and Schedule) and Chapter 6 (Ambient Air
Monitoring)
Charge Questions
Overall organization and clarity: To what extent does the Panel find that the draft IRP clearly and
appropriately communicates the plan for the current review of the primary NO2 NAAQS and the key
scientific and policy issues that will guide the review? To what extent are the decisions made in the last
review, including the rationales for those decisions, clearly articulated?
This draft of the IRP is very well organized. The plan for this current review is clearly presented. The
history of the NO2 NAAQS and the summary of the decisions and the rationale for them in the last
review are clear and concise.
Introduction (Chapter 1) and Schedule (Chapter 2): To what extent does the Panel find that Chapters 1
and 2 clearly communicate the NAAQS legislative requirements, summarize the steps in the review
process, summarize the history of the NO2 NAAQS, and present the anticipated schedule for the current
review?
These two chapters are well written, and meet the goals noted in this charge question.
Ambient Air Monitoring (Chapter 6): To what extent does Chapter 6 clearly and appropriately
communicate, for the purposes of this plan, the key aspects of measurement methods and surveillance
network requirements for the NO2 NAAQS?
Section 6.1, Consideration of Sampling and Analysis Methods, provides a clear summary of existing and
new methods for measurement of NO2. Of most interest is the recent commercial availability of direct
NO2 measurement methods using the cavity attenuated phase shift (CAPS) technique. One commercially
available CAPS instrument has an FEM designation, and a second is in the final stages of FEM approval
at ORD. These instruments are expected to be a practical alternative (in terms of cost and operational
effort) to the traditional CL-moly converter FRM monitor.
This section raises an important question regarding the potential of routine network deployment of
CAPS or any other method that only measures NO2 (e.g., does not measure NO). The potential loss of
NOx data is of concern, since NOx is often the only widely available exposure surrogate for on-road
pollutants. In addition, section 2.6.4.3 of the draft NO2 ISA discusses the development of an "Integrated
Mobile Source Indicator" to improve exposure assessment to on-road air pollutants. This "Indicator"
uses CO, EC or BC, and NOx as input parameters. The loss of NOx data in routine ambient
measurement networks would have a substantial impact on the performance of this indicator approach.
Section 6.2, Consideration of Air Monitoring Network Requirements.
This section is a concise summary of the existing and planned NO2 monitoring networks. The Area-
Wide and "susceptible and vulnerable communities" components of the monitoring network required by
A-2
-------
the 2010 NO2 rule are in place, since existing NO2 monitors met these requirements. Section 6.2 also
summarizes the design and requirements of the new near-road network component that has just recently
begun to be deployed. This summary covers the number of sites and time line for deployment, but could
benefit from additional detail on what other measurements are required at these sites. Some near-road
sites require CO and PM2.5 along with (NO and) NO2, and usually have optical black carbon and
meteorological measurements also (some have particle number concentration too). These sites generate
all the measurement inputs needed for the "Integrated Mobile Source Indicator" approach noted above
and in the ISA. Other sites require only NO2 measurements.
Most of the "NO2 only" sites are in the third and final deployment phase, scheduled to be operational by
January 2017. One of the near-road network goals is to support research; sites with only NO2
measurements have minimal value in this context.
Section 6.2 ends with a sentence (Pg 6-5, lines 3-5) that suggests the minimum near-road network
requirements promulgated in the 2010 rule could be re-evaluated during this review:
"Considering the availability of new near-road NO2 monitoring data, the EPA may be in a position to
re-evaluate the analyses underlying the minimum monitoring requirements promulgated in the 2010
revisions in this review."
Since it is unlikely that EPA would increase the minimum requirements, this sentence could be taken to
mean that when sufficient data is available, EPA may be able to justify reducing the final near-road
network size by reducing or eliminating the third phase of near-road site deployment. This would be
consistent with the continued downward trend of primary on-road NO2 emissions due to both new
(2010) controls on diesel emissions and the Tier 3 standards for both fuels (lower S gasoline) and
automotive emissions controls that go into effect in 2017.
Thus, it is important that EPA commence to conduct analysis of NO2 data from the near-road network as
soon as it becomes available (later this or early next year).
A-3
-------
Dr. Matthew Campen
Comments on Chapter 4
To what extent does Chapter 4 clearly and adequately describe the scope, specific issues to be
considered, and organization of the ISA ? Please provide suggestions for any other issues that should be
considered.
Literature search - are all search-retrieved documents recorded and rationale for why they are
considered/not considered also recorded?
Comment on page 4-8: "In addition, consideration will be given to studies that investigate exposure to
oxides of nitrogen separately and in combination with other pollutants such as ozone, PM, and sulfur
dioxide."
It is generally understood that ozone and NOx are mutually exclusive. That is, they will react out and
thus tend not to co-exist. I would consider just dropping "ozone" from this sentence. Certainly, it is a
sentence of hypothetical options, but all the same...
In 4.3.3, at the end regarding in vitro studies, I think a statement to anatomical relevance would be nice
to see - that is, we should not be studying direct exposures of NO2 on neurons or endothelial cells.
Pretty much lung epithelia and other airway cells.
Comma after NAAQS review, bottom of 4-10 (very long sentence)
Please comment on the adequacy of the expanded discussion in Section 4.4 of issues that will be
considered in the ISA related to:
(a) spatial heterogeneity in ambient concentrations of oxides of nitrogen, particularly near- and on-road
gradients, and implications for human exposures and
There is strong language related to concerns about inadequacies of central site monitoring and an
apparent appreciation for the roadway-associated nature of these exposures. I feel the approach is
appropriate.
(b) various factors to consider in the evaluation of health effects associated with ambient NO2 exposure,
including traffic, noise, indoor NO2 exposures, and copollutant exposures.
Certainly there are profound covariates to consider, but such is the case with most NAAQS pollutants.
NOx should be treated in a consistent manner with other recent ISAs.
A-4
-------
Dr. Ronald C. Cohen
Comments on Chapter 6
The Chapter should be more strongly connected to the overview of "Atmospheric science and ambient
concentrations" in CH4 pgs 4-11-4-12
The Chapter overemphasizes the question of loss of NO measurements and underemphasizes the
benefits of new FEM methods that are specific to NO2.
The substantial positive bias of the FRM NO2 should be explicitly acknowledged on pg 6-1 near line 30
on pg 6-23 lines 1-4 and 6-13
The lack of positive bias from higher oxides of N (PAN, RONCh and FINO3) should be explicitly
mentioned for all three new instruments.
pg 6-2 line 18-20 The tone of the question presumes a negative. It would be more appropriate to ask
what the balance between the benefits of having interference free measurements and the costs of losing
NO measurements. Also, it would be appropriate to ask whether losing NO measurements is necessary.
pgs 6-3-6-5
some discussion of what concurrent measurements are needed to support isolating exposure effects to
NO2 as separate from other traffic related emissions would be pertinent here.
A-5
-------
Dr. Philip M. Fine
Comments on Chapter 6 - Ambient Air Monitoring
To what extent does Chapter 6 clearly and appropriately communicate, for the purposes of this plan, the
key aspects of measurement methods and surveillance network requirements for the NO2 NAAQS?
The chapter provides a concise summary of the NOx measurement methods and the existing and future
national NOx monitoring network. The last line of Chapter 6 anticipates that EPA may re-evaluate the
analysis that underlies the minimum monitoring requirements for the NOx near road network. A more
detailed discussion may be appropriate regarding the criteria for evaluating the adequacy of the network
relative to various monitoring objectives (NAAQS, health studies, public information). There is a
potential divergence of opinions among stakeholders on this topic, and there should be plan for
analyzing this issue carefully to help support any potential monitoring regulation changes in the future.
Page 6-1. Line 27
It is stated that the catalytic converter reduces ALL oxidized nitrogen species to NO. It may not be true
that all oxidized nitrogen species are reduced, and the ISA discussion on this topic points to varying
conversion efficiencies for different species, depending on temperature on other factors. The discussions
in the two documents should be reconciled.
A-6
-------
Dr. Panos G. Georgopoulos
Comments on Chapter 5
Chapter 5 (Quantitative Risk and Exposure Assessment) of "Integrated Review Plan for the Primary
National Ambient Air Quality Standards for Nitrogen Dioxide" focuses on and describes the approach
pursued during the prior review of 2008 (Section 5.1, pages 5-2 to 5-11). The consideration of
"quantitative assessments for this review" is the subject of Section 5.2 (pages 5-11 to 5-12 and Tables 5-
1 and 5-2). Section 5-1 includes a discussion of the uncertainties involved in the approach of the prior
review, while Table 5-1, "Information (data, methods, models, etc.) identified as potentially important
and/or newly available to inform the air quality characterization for the current review," and Table 5-2,
"Information (data, methods, models, etc.) identified as potentially important and/or newly available to
inform the exposure assessment for the current review," summarize, in the rightmost column, potential
approaches for addressing components of the above uncertainties with this new information. However, it
is not clear whether any of the methods for modeling air quality and exposure, which are applicable to
oxides of nitrogen, that have evolved since the prior review, are going to be utilized in this context. For
example, Ozkaynak et al. (2013) summarized the findings of a series of presentations that took place at
the International Society of Exposure Science 2011 Conference in Baltimore, MD. Symposium
presenters considered a range of "alternative exposure metrics, including: central site or interpolated
monitoring data, regional pollution levels predicted using the national scale Community Multiscale Air
Quality model (CMAQ) or from measurements combined with local-scale (AERMOD) air quality
models, hybrid models that included satellite data, statistically blended modeling and measurement data,
concentrations adjusted by home infiltration rates, and population-based human exposure model
(Stochastic Human Exposure and Dose Simulation, and Air Pollutants Exposure models) predictions."
(See also Ozkaynak et al., 2014.) In another study that also used complementary air quality models
(Beevers et al., 2013) employed both KCLurban, which gives source apportionment information, and the
Community Multi-scale Air Quality model (CMAQ)-urban to characterize NOx and NO2 and evaluate
the performance of the modeling approach. Given the fact that in recent years long-term (annual and
multi-year) CMAQ simulations are becoming more commonly available for North America (e.g.
Civerolo et al., 2010; Zhang et al., 2009), that can provide hourly estimates of NO and NO2
concentrations at "background" level (typically 12x12 km resolution), it is strongly recommended that
USEPA at least evaluate a hybrid modeling approach that would use a model such as AERMOD to
"downscale" CMAQ estimates at point level (near-road, neighborhood, etc.) and use these estimates in
conjunction with an exposure model such as APEX. It would also be useful, in such an enterprise, to
consider dispersion models alternative to AERMOD, specifically CALPUFF which was used in the
study of (Yu & Stuart, 2013), as this model may have more flexibility than AERMOD for applications
relevant to the NOx system, where nonlinearity of photochemical interactions poses particular
challenges.
Another issue that should be addressed carefully in the context of exposure characterization is the issue
of indoor NOx emissions. The IRP document states on page 5-9 that"... in a limited set of targeted
exposure analyses, exposures were also modeled considering indoor source emissions. The
characterization of indoor source emissions of NO2 and estimated air exchange rates used to simulate
indoor microenvironments were considered an important uncertainty." However, in footnote number 53
on the same page it is stated that "While potentially important in understanding health effects and the
total exposure/health risk from NO2, exposures resultant from indoor sources of NO2 have limited
A-7
-------
relevance in understanding health risk associated with ambient concentrations." This statement should
be modified/clarified in the context of the new review as it may be misleading with respect to the
significance of indoor NOx sources and exposures. In fact, improving indoor emission inventories of
NOx is needed in order to better characterize overall exposure and risk to these air pollutants.
References:
Beevers, S.D., Kitwiroon, N., Williams, M.L., Kelly, F.J., Ross Anderson, H., and Carslaw, D.C. 2013.
Air pollution dispersion models for human exposure predictions in London. J Expo Sci Environ
Epidemiol 23 (6):647-53. DOI:10.1038/jes.2013.6
Civerolo, K., Hogrefe, C., Zalewsky, E., Hao, W., Sistla, G., Lynn, B., Rosenzweig, C., and Kinney,
P.L. 2010. Evaluation of an 18-year CMAQ simulation: Seasonal variations and long-term temporal
changes in sulfate and nitrate. Atmospheric Environment 44 (31):3745-3752.
DOI:10.1016/j.atmosenv.2010.06.056
Ozkaynak, H., Baxter, L.K., Dionisio, K.L., and Burke, J. 2013. Air pollution exposure prediction
approaches used in air pollution epidemiology studies. J Expo Sci Environ Epidemiol 23 (6):566-72.
DOI:10.1038/jes.2013.15
Ozkaynak, H., Isakov, V., Baxter, L., Graham, S., Sarnat, S., Sarnat, J., Mulholland, J., Turpin, B., Rich,
D., and Lunden, M. 2014. Evaluating Alternative Exposure Metrics Used for Multipollutant Air Quality
and Human Health Studies. In Air Pollution Modeling and its Application XXII, edited by D. G. Steyn,
P. J. H. Builtjes andR. M. A. Timmermans: Springer Netherlands.
Yu, H.F., and Stuart, A.L. 2013. Spatiotemporal distributions of ambient oxides of nitrogen, with
implications for exposure inequality and urban design. Journal of the Air & Waste Management
Association 63 (8):943-955. DOLDoi 10.1080/10962247.2013.800168
Zhang, Y., Vijayaraghavan, K., Wen, X.-Y., Snell, H.E., and Jacobson, M.Z. 2009. Probing into regional
ozone and particulate matter pollution in the United States: 1. A 1 year CMAQ simulation and
evaluation using surface and satellite data. Journal of Geophysical Research 114 (D22).
DOI:10.1029/2009jd011898
A-8
-------
Supplementary References for Consideration by the USEPA
Levy, I, Mihele, C., Lu, G., Narayan, J., and Brook, J.R. 2014. Evaluating multipollutant exposure and
urban air quality: pollutant interrelationships, neighborhood variability, and nitrogen dioxide as a proxy
pollutant. Environmental Health Perspectives 122 (l):65-72. DOI:10.1289/ehp. 1306518
Lin, W., Brunekreef, B., and Gehring, U. 2013. Meta-analysis of the effects of indoor nitrogen dioxide
and gas cooking on asthma and wheeze in children. International Journal of Epidemiology 42 (6): 1724-
37. DOI:10.1093/ije/dytl50
Matsui, E.G., Hansel, N.N., Aloe, C., Schiltz, A.M., Peng, R.D., Rabinovitch, N., Ong, M.J., Williams,
D.L., Breysse, P.N., Diette, G.B., and Liu, A.H. 2013. Indoor pollutant exposures modify the effect of
airborne endotoxin on asthma in urban children. American Journal of Respiratory and Critical Care
Medicine 188 (10):1210-5. DOI:10.1164/rccm.201305-0889OC
Montagne, D., Hoek, G., Nieuwenhuijsen, M., Lanki, T., Pennanen, A., Portella, M., Meliefste, K.,
Eeftens, M., Yli-Tuomi, T., Cirach, M., and Brunekreef, B. 2013. Agreement of land use regression
models with personal exposure measurements of particulate matter and nitrogen oxides air pollution.
Environmental Science and Technology 47 (15):8523-31. DOI:10.1021/es400920a
Paulin, L.M., Diette, G.B., Scott, M., McCormack, M.C., Matsui, E.C., Curtin-Brosnan, J., Williams,
D.L., Kidd-Taylor, A., Shea, M., Breysse, P.N., and Hansel, N.N. 2013. Home interventions are
effective at decreasing indoor nitrogen dioxide concentrations. Indoor Air. DOI: 10. Ill 1/ina. 12085
Wang, K. 2012. Application, Evaluation, and Process Analysis of the US EPA's 2002 Multiple-Pollutant
Air Quality Modeling Platform. Atmospheric and Climate Sciences 02 (03):254-289.
DOI: 10.4236/acs.2012.23025
A-9
-------
Dr. Jack Harkema
No critical comments for Chapters 3 and 7. The plan is clearly stated with ample thoughtful and
appropriate guiding questions for the review. The flow chart is helpful and could come earlier in the
Chapter if so desired.
A-10
-------
Dr. Michael T. Kleinman
The Integrated Review Plan provides an excellent overview of the process. For this preliminary
evaluation I will focus on the charge questions related to Chapter 4.
Chapter 4 describes the scope and specific issues that will be considered in the ISA. The expanded
discussion provides a very comprehensive listing of the topics that will be considered in the ISA and
appropriately focuses on new information that can fill knowledge gaps identified in the previous 2008
ISA. The expanded discussion in section 4.4 appropriately identifies key issues related to spatial
heterogeneity near heavily trafficked roads and near emission sources. While it is implied, it would be
useful to discuss the direct consideration of demographics (i.e. what new information is available
regarding the size of the population, its age distribution or its socioeconomic distribution) of the people
residing or working near heavily trafficked roads or near other sources.
A-ll
-------
Dr. Timothy V. Larson
Overall organization and clarity:
The February 2014 draft IRP is well organized and clearly written. Overall, the questions posed cover
the main issues. Tables 5-1 to 5-3 are especially informative. Along those lines, I have suggested a few
additional questions that follow.
Schedule:
One issue not specifically mentioned in this particular section is the timing of the roll-out of the new
near road monitoring network relative to the timing of the REA (It is discussed later in the document).
We struggled with this issue last time and made decisions based on a limited set of near road
measurements in a limited set of cities. A comprehensive data set of both traditional and near road
monitoring would greatly enhance the final decisions on the form of the standard. Is there a more
detailed roll out plan that optimizes the choice of sites and therefore the relevant monitoring information
used in the REA?
Policy Relevant Issues:
(1) The relevant averaging time is an issue given the results of the epidemiology. Will the
consequences of such an averaging time(s) be examined?
(2) In the 2008 REA analysis, possible alternative standards were evaluated in part by predictions of
the resulting 1-hr on-road concentration estimates. Is this still the plan and, if so, will a uniform
gradient be used across all sites to make such predictions?
Risk and Exposure Assessment:
(1) Emphasis on U.S. and Canadian studies would presumably down weight the reported health
associations from a number of European studies. Is this also true for the exposure information to
help establish near-road gradients?
(2) What information exists on at-risk populations who live near NO2 sources?
(3) The 2008 analysis pointed out the importance of on-road exposure estimates. Recent studies (e.g.
Hudda et al Atmos. Environ. 59: 578-586, 2012; and Hudda and Fruin ES&T 47(19): 11048-
11055, 2013) have shown, not surprisingly, that in-vehicle concentrations relative to those on the
roadway are a strong function of the state of the cabin air circulation system (indoor air vs
outdoor air setting) when windows are rolled up. Are the APEX model predictions consistent
with the models derived from these recent studies?
A-12
-------
Policy Assessment and Rulemaking:
The important issues related to the assessment and use of exposure measurement and modeling error
need to be more clearly stated. In addition, the role of panel studies with respect to co-pollutant
confounding needs to be given more emphasis.
A-13
-------
Dr. Jeremy Sarnat
Comments on Chapter 4
Generally, I feel that the Science Assessment does an adequate job of describing the most important
questions and uncertainties related to NO2 exposure and health. The rewritten section 4.4 is strong, and
properly stresses the importance of understanding the near-road environment. Data generated from the
new near-road monitoring network should be able to address several key questions including gradients
around roadways, associations between specific traffic species and NO2 near traffic sources, and the
relationship between noise and NO2 in this microenvironment.
Understanding the specific role of NO2 as either an independent predictor of health response or as a
marker for a suite or source of pollution, is adequately recognized within the IRP. I believe statements
similar to the following reflect major issues that should be considered in the ISA:
" What new information exists regarding oxides of nitrogen measurements in a multipollutant
context? To what extent do NO2 measurements serve as surrogates of exposure to other gaseous
pollutants (e.g., carbon monoxide, nitrous acid), particle phase pollutants (e.g., ultrafine particles,
black carbon, organic carbon, transition metals) generated by traffic or other combustion sources,
or a mixture of traffic-related pollutants!"
As noted in my review of the ISA, there appears to be inconsistency with regard to the particular role of
NO2 as a traffic pollution surrogate. A more cohesive message should be developed linking the IRP
directives to the ISA message.
Since total exposure to NO2, for many people, occurs while commuting, there should be greater attention
paid to characterizing exposures and response occurring within the on-road microenvironment (page 1-9
of ISA mentions this briefly).
Comments on Chapter 5
One approach for strengthening the risk assessment would be to include greater amount of sensitivity
analyses of the primary modeled input parameters to enhance the robustness of the findings. Using
alternative, realistic C-R functions for the epidemiologic-based human health risk assessments would be
useful (Page 5-17). Similarly, for both the air quality and human exposure components to the
assessment, conducting formal uncertainty analyses, or presenting propagation of error findings would
potentially inform the research and regulatory community on the largest sources of uncertainty.
A-14
-------
Dr. Richard Schlesinger
Comments on Chapter 4
p.4-5, lines 33-36. While clearly studies that reduce uncertainty need to be evaluated, this suggests that
studies that may show novel results that may add some additional information about health effects that
may not be totally consistent with other studies will not be evaluated. The description here needs to be
retooled.
p.4-8, lines 23-26. It is not clear what is meant by intake dose. Is that exposure concentration? If not,
intake dose is not necessarily available in these studies, so perhaps an additional focus that should be
listed here is "exposure concentration."
p.4-14, lines 1-2. NOx is also a direct acting irritant that can produce adverse health outcomes without
production of secondary products.
p.4-14, line 18. Replace "internal NCh" with "endogenous NCh"
p.4-14, line 25. Should read, "...can be qualitatively and quantitatively compared...."
p.4-15, line 9 et seq. Many of these bullets are redundant and the list can be made more concise while
not losing any of the concerns.
p.4-16, line 16. What is meant by other "disciplines?"
p. 4-15 to 4-17. Almost all of the bullets for short and long term exposures are the same, so the question
is whether they need to be listed separately?
p.4-19, line 29. Is it not important to distinguish among risk due to intrinsic, acquired or extrinsic factors
in determining relative susceptibility to exposure in different groups?
A-15
-------
Dr. Elizabeth A. (Lianne) Sheppard
Overall this document is well-written and clear. I have some specific suggestions for the ISA (Chapter
4) and REA (Chapter 5) sections.
Comments on Chapter 4 - Science Assessment
I suggest that for evaluation of toxicological studies adherence to the ARRIVE guidelines also be
considered.
The human exposure section is focusing on questions about human exposures with particular emphasis
on epidemiological inference. The questions should be reviewed in the context of how relevant they are
for epidemiology. This should be carried forward into the ISA itself (see my related comments on the
ISA w.r.t. exposure assessment and exposure measurement error).
Regarding specific questions:
How have ambient models been merged with stochastic population exposure models recently to improve
estimates of exposure?
Use of stochastic models is not appropriate for epidemiological inference. However, they are very
valuable for risk assessment.
What new information exists regarding characterization of error in exposure assessment of oxides of
nitrogen and how it influences personal-ambient exposure relationships?
I suggest that a very important consideration for error in exposure assessment is with respect to its
adequacy for inference about epidemiological health effects. I think the question should be either
rephrased or expanded to incorporate inference in epidemiological studies. (My related ISA comments
may also help with rephrasing this question.)
What information is available regarding differences in exposure patterns for oxides of nitrogen and
personal-ambient exposure relationships among various lifestages and populations?
I suggest the more relevant question is whether there are differences that might affect epidemiologic
study inference or possibly risk assessment. Consideration of this question with respect to epidemiologic
inference and risk assessment are distinct topics.
Comments on Chapter 5 - Quantitative Risk and Exposure Assessment
Scope and specific issues:
1. I agree overall with EPA's plan for the REA, specifically noting the plan to decide later how
justified it is to conduct an exposure and risk assessment.
A-16
-------
2. Dr. Sarnat's comments on sensitivity analysis and other approaches to uncertainty analysis are well
taken and should be included in the REA.
Identification of most important uncertainties:
3. I agree that spatial heterogeneity of NOx remains an important uncertainty.
4. Exposure occurs both on roads and near roads. Is there a way to incorporate both into the
characterization? See below for some summaries of AQS data that may contribute to this.
5. Would evaluation of AERMOD in conjunction with existing AQS sites that are near roads but in
cities other than Atlanta help address the AERMOD uncertainties?
Additional information that should be considered:
6. The document states (p 5-6): "One of the most important uncertainties overall regards the spatial
representation of the ambient monitors, ..." The existing AQS data may be able to address this
uncertainty better than is appreciated. I insert below some data summaries based on compilation of
AQS monitoring data that has been done at the University of Washington as part of our work for the
MESA Air and NPACT studies. This (and additional potential data/analyses that could be provided)
may help inform some of the existing uncertainties related to the ambient monitoring network. The
following table characterizes how two geographic features are associated with all NOx monitors in
the US. Pasted below is a map of the locations. In separate research, we have recent on-road data
from a mobile monitoring network that might help summarize NOx.
Table: Characterization of 368 AQS monitoring locations that reported NOx in 2012 with respect to
proximity to nearest road (Al, A2, A3) or nearest truck route. (Note: There are 30 additional
locations that reported NO2 monitoring that are not included in this summary.)
Meters To Road To Truck Route
< 50 59 (16%) 5 ( 1%)
50-100 38(10%) 7(2%)
100-150 44(12%) 6(2%)
150-300 78(21%) 18(5%)
> 300 149 (40%) 332 (90%)
7. I think the use of APEX is a reasonable approach to estimating population exposures for risk
assessment. Are there ways to use characterization of the current national network to address the
biases identified in the past review from AERMOD? (p 5-16) Information is available that will help
better characterize the existing national network (see above). Could this information be used to
improve the AERMOD estimates?
A-17
-------
A-18
-------
Dr. Ronald E. Wyzga
I have relatively few comments on the plan as I believe that the devil is in the detailed implementation
of the plan. In general it provides a good outline.
The Agency should use the 2011 NEI in its risk and exposure calculations. The significant reduction in
NOx emissions since 2008 merits this change.
Some specific comments:
Comments on Chapter 3
p. 3-16: at the conclusion of the last review, CASAC recommended and EPA implemented a program to
undertake monitoring near roadways. My understanding is that results from these monitors are not yet
available; however, should there be any discussion about how these data are to be used in
interpreting/extrapolating from health studies or in the risk assessment to be undertaken.
Comments on Chapter 4
p. 4-6:11. 30-31: In the case of NOx it is important to learn whether NOx itself is responsible for the
associated health effects or whether NOx is a surrogate for another pollutant. As such it is important that
studies address this issue by considering copollutants as well. In particular the co-pollutants that appear
to be of greatest interest are PM, CO, EC, and OC.
p. 4-7,11. 15-27: This is important, but the document needs to address how or what it will do with
respect to the exposure error issue. Although it is not the end-all, statistical significance is noteworthy
and should be a factor that is noted in presenting study results.
Comments on Chapter 5
p. 5-3,11. 19-21: This statement needs further elaboration.
p. 5-7,1. 4: My understanding is that AERMOD does not incorporate any chemistry; is there an
alternaitve model that considers chemistry that could be used to replace/supplement AERMOD?
A-19
-------
Dr. Junfeng (Jim) Zhang
The overall structure of the Integrated Review Plan for the Primary NAAQS for Nitrogen dioxide is well
organized. The tabular and graphical presentations of information are particularly useful. For example,
Figure 1-1 and Table 2.1 are very useful. I think the information presented in tables and figures along
with figure/table captions are most easily conveyed to the reader. For this reason, I encourage the EPA
authors to check whether they can further explain the tables and figures (in the captions) so the tables
and figures can be readily understood even without the need to read the text.
My review focuses on Chapter 5 - Quantitative Risk and Exposure Assessment. Below are my
comments on this chapter.
1. It is appropriate to use the three approaches to estimating exposures and health risks or REA.
Each approach is well described in the chapter. I suggest that a section be added to explicitly
describe how the information from the three approaches can be integrated to support an overall
REA.
2. In the epidemiology-based REA approach, a major issue is whether the effects are of NO2 or of
other pollutants or the pollution mixture. This is discussed within the context of uncertainty. I
think this issue shall be emphasized more strongly. It is well known that many epidemiological
studies have used NO2 concentrations as a surrogate of traffic-related pollution exposure. It
would be useful to evaluate NO2 correlations with PM and other pollutants in cited
epidemiological studies. This should be independent from evaluating studies that have used two-
pollutant models (presented in ISA).
3. In the evaluation of personal NO2 exposure and the contribution from indoor sources, a useful
parameter would be indoor-to-outdoor concentration (I/O) ratio. A table to summarize I/O ratios
from the available literature will be useful. If I/O is smaller than 1, concentration-response
relationships or risks derived from using outdoor concentrations may not be significantly
affected by indoor source confounding. Then the issue will be how good the outdoor NO2
concentration (measured or modeled) represents personal exposure. Approaches to addressing
this issue are described and discussed well in this chapter.
A-20
------- |