\
I UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
?* WASHINGTON D.C. 20460
OFFICE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD
May 21,2009
EPA-CASAC-09-010
The Honorable Lisa P. Jackson
Administrator
U.S. Environmental Protection Agency
1200 Pennsylvania Ave., NW
Washington, D.C. 20460
Subj ect: Consultation on EPA's Paniculate Matter National Ambient Air Quality
Standards: Scope and Methods Plan for Urban Visibility Impact Assessment
Dear Administrator Jackson:
The Clean Air Scientific Advisory Committee (CASAC) Particulate Matter (PM)
National Ambient Air Quality Standards (NAAQS) Review Panel met on April 2, 2009 to
provide a consultation on EP A's Draft Particulate Matter National Ambient Air Quality
Standards: Scope and Methods Plan for Urban Visibility Impact Assessment (February 2009).
The CASAC uses a consultation as a mechanism for technical experts to provide comments on
an EPA draft plan for developing technical assessments as part of CAS AC's responsibilities in
the NAAQS review process. In this letter, CASAC offers some general thoughts on the visibility
plan. More specific written comments by the individual members are enclosed.
The Plan could be improved by providing more background and justification regarding
the importance of visibility as a health endpoint. CASAC also suggests that EPA consider ways
of developing and implementing a visibility monitoring approach based on direct measurement
of visibility using low cost instruments such as a size selective integrating nephelometer in
combination with an aethalometer. EPA might also consider ways to analyze and present
network design and operation plans to use direct measurement of visibility to characterize PM
effects for both the primary and secondary PM NAAQS. Consistent with our comments on the
Integrated Science Assessment for Particulate Matter (First External Review Draft, December
2008), we would also recommend that EPA identify and briefly characterize additional metrics,
other than visibility, for ecosystem and environmental impacts of PM that could be the
-------
foundation for future assessments even if there are insufficient data to support the current
assessment as a basis for a new secondary NAAQS.
As this is a consultation, we do not expect a formal response from the Agency. We thank
the Agency for the opportunity to provide advice early in the PM NAAQS review process and
look forward to the First Draft PM NAAQS Urban Visibility Impact Assessment in October
2009.
Sincerely,
/Signed/
Dr. Jonathan M. Samet, Chair
Clean Air Scientific Advisory Committee
Enclosures: Enclosure A: CASAC Particulate Matter Review Panel Roster
Enclosure B: Compendium of Individual Comments
-------
NOTICE
This report has been written as part of the activities of the EPA's Clean Air Scientific
Advisory Committee (CASAC), a federal advisory committee independently chartered to
provide extramural scientific information and advice to the Administrator and other officials of
the EPA. CASAC provides balanced, expert assessment of scientific matters related to issues and
problems facing the Agency. This report has not been reviewed for approval by the Agency and,
hence, the contents of this report do not necessarily represent the views and policies of the EPA,
nor of other agencies within the Executive Branch of the federal government. In addition, any
mention of trade names or commercial products does not constitute a recommendation for use.
CASAC reports are posted on the EPA Web site at: http://www.epa.gov/casac.
-------
Enclosure A
Clean Air Scientific Advisory Committee
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
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
Dr. Douglas Crawford-Brown, Professor Emeritus, Department of Environmental Sciences and
Engineering, University of North Carolina at Chapel Hill, Chapel Hill, NC
-------
Dr. David Grantz, Director, Botany and Plant Sciences and Air Pollution Research Center,
Riverside Campus and Kearney Agricultural Center, University of California, Parlier, CA
Dr. Joseph Helble, Dean and Professor, Thayer School of Engineering, Dartmouth College,
Hanover, NH
Dr. Rogene Henderson, Senior Scientist Emeritus, Lovelace Respiratory Research Institute,
Albuquerque, NM
Dr. Philip Hopke, Bayard D. Clarkson Distinguished Professor, Department of Chemical
Engineering, Clarkson University, Potsdam, NY
Dr. Morton Lippmann, Professor, Nelson Institute of Environmental Medicine, New York
University School of Medicine, Tuxedo, NY
Dr. 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 Community & Environmental Medicine;
Director, Air Pollution Health Effects Laboratory; Professor of Occupational & Environmental
Health, Center for Occupation & Environment Health, College of Medicine, University of
California Irvine, Irvine, CA
Dr. Kent Pinkerton, Professor, Regents of the University of California, Center for Health and the
Environment, University of California, Davis, CA
Mr. Richard L. Poirot, Environmental Analyst, Air Pollution Control Division, Department of
Environmental Conservation, Vermont Agency of Natural Resources, Waterbury, VT
Dr. Frank Speizer, Edward Kass Professor of Medicine, Channing Laboratory, Harvard Medical
School, Boston, MA
Dr. Sverre Vedal, Professor, Department of Environmental and Occupational Health Sciences,
School of Public Health and Community Medicine, University of Washington, Seattle, WA
SCIENCE ADVISORY BOARD STAFF
Dr. Holly Stallworth, Designated Federal Officer, EPA Science Advisory Board Staff Office
-------
Enclosure B
Compendium of Individual Comments
CASAC Particulate Matter Review Panel on
PMNAAQS: Scope and Methods
Plan for Urban Visibility Impact Assessment (Feb. 2009)
Comments from Dr. Lowell Ashbaugh 2
Comments from Dr. David A. Grantz 4
Comments from Dr. William Malm 8
Comments from Mr. Rich Poirot 16
Comments from Dr. Ted Russell 21
-------
Comments from Dr. Lowell Ashbaugh
Ashbaugh Review of Scope and Methods Plan for Urban Visibility Impact Assessment
Chapter 1 - Introduction
Charge questions
1. What are the Panel members' views on the general structure and overall design of the
planned analyses?
2. Is the plan clear and transparent in its description of the planned approaches? Are the
various assumptions and judgments that must be made in carrying out the planned
assessments clear and transparent?
3. Given the goals set forth for the planned analyses, has the plan appropriately drawn from
the existing scientific and technical information in developing the overall approach? Are
there relevant features that should be added or modified in the planned approach?
4. In addition to the sub-daily PM2.5 alternative standard considered in the last PM NAAQS
review and summarized in this chapter, an alternative standard structure is being
considered in this review. This alternative structure would use daylight hourly PM light
extinction, which can be measured either by a combination of instruments (nephelometer
- PM light scattering and aethalometer - PM light absorption) or calculated from PM
speciation and concurrent relative humidity data using a linear algorithm.
a. What are the Panel members' views regarding this alternative structure and its
utility in the context of this PM NAAQS review?
b. What are the Panel members' views regarding advantages and disadvantages of
this alternate structure compared to the sub-daily PM mass concentration
approach?
The general structure and overall design of the planned approaches to the "standard" assessment
of the public preferences for and value of urban visibility are good. I particularly like the idea of
assessing whether these preferences vary across the country in cities with differing backdrops.
However, it's not clear what the EPA is planning to do regarding the alternative secondary
standard. The discussion of the "standard" assessments in sections 1.3.1 and 1.3.2 uses
declarative sentences that start out "we are planning to ...", but the alternative standard
discussion in section 1.3.3 is much less clear. The foundation for the alternative standard is
presented well but the section does not contain a clear plan. It contains phrases such as "could
include" or "would then be specified."
The plan for the "standard" assessment draws appropriately from the existing scientific and
technical information, building upon prior studies of visibility impact in remote areas to
investigate urban visibility impact. The discussion of the alternative standard also draws upon
earlier work to lay out the foundation, but does not go further to set forth a plan.
Overall, I like the concept presented in the alternative secondary standard structure. PM mass
measurements for a sub-daily 4-hour period, and especially speciation of such measurements,
pose difficult technical challenges. The amount of PM collected in such a short time is hard to
-------
measure gravimetrically and even harder to analyze for speciated components. I am concerned
about the use of "known relationships between PM mass and speciated components" to relate
PM mass to PM light extinction. This concept assumes the speciation fractions during the
daylight hours (or the 4-hour period) are the same as those measured during a 24-hour period.
This is clearly not the case with nitrate particles, and is likely not the case for other species,
either. Further investigation or review of existing information is needed here.
-------
Comments from Dr. David A. Grantz
This plan provides a good distillation of the newly released Integrated Science Assessment for
PM, First External Review Draft (December 2008). It therefore reflects both the strengths and
weaknesses of that document. The following represent suggestions for revision of the Plan, not in
order of importance nor priority, and focusing on Section 1 (Introduction) but not limited to that
section.
1. Better define the meaning and mode of application of the "policy relevant" filter to
be applied in all aspects of this review.
The recurring restatement of the "policy relevant" caveat (beginning with three mentions on page
1-1) is an indicator of a problem throughout this document (and the ISA upon which it is based).
Once the focus on visibility as the (only) surrogate for welfare impacts of PM was established,
other information seems to have become not "policy relevant". The previous Criteria Document
for PM (US EPA 2006), and related agency actions, did not find sufficient data on non-visibility
parameters to inform a secondary standard (page 1-4, line 8-13). This might have initiated
focused research in the intervening period but did not.
The current Plan, in its approach to these other impacts, does not ascribe to them sufficient
gravity to encourage the research needed for the next PM assessment cycle. This contrasts with
similar material related to human health endpoints of PMio-2.5, some of which are appropriately
to be addressed in the current review cycle (see 2009 Health Plan, Section 3.6, first paragraph).
It would be useful near the beginning of the discussion to insert a comprehensive definition of
"policy relevant", its mode of application in the current review process, and a brief consideration
of the likely consequences of doing so. This filter is an important behind the scenes driver of the
review, is a new addition to the process, and should receive some discussion.
2. Combine and expand the discussion of "other" welfare effects of PM (not related to
visibility), pointing to data and approaches that while outside the current scope of
work may be useful in the future.
The minimal discussion of non-visibility impacts in the text (page 1-4 and pages 1-10 to 1-11), is
augmented by a slightly more complete presentation in Appendix A. I suggest that these be
combined, and placed within the text. This section should be expanded, to more closely parallel
the visibility sections, to call attention to data that are feasible to obtain but currently lacking,
that might usefully inform the secondary standard setting process. These data could address a)
ecosystem impacts of locally relevant particles (analogous to the local approach taken with
respect to humidity/PM relationships to visibility), b) effects on climate (considerable data now
reaches the stage of likely causality, and more is known about climate forcing and PM than is
indicated (pages A7-A9), c) fouling of leaf surfaces and possible interference with gas exchange
-------
and radiation interception, particularly in urban locations, d) interception and scattering of
incident radiation at the vegetated landscape scale.
The text (1-12, line 16), refers to an "initial qualitative approach" to non-visibility impacts, to be
found in Appendix A. However, there is no approach, nor goal, nor suggestion of needed data, to
be found in Appendix A. In combining the Appendix with the main text, such approaches and
goals should be added.
In the summary of ecosystem affects (page 1-4, line 3-5), the review of the previous CD for PM
is missing the previous emphasis on transport. While the previous review concluded that
chemical composition dominates ecosystem effects of PM, the transport to and within
ecosystems depends strongly on particle size and distance between source and receptor. This
should be included in the discussion.
3. The focus on visibility, an aesthetic parameter, should be linked more closely to
other potential impacts of PM that are more representative of damage, otherwise
reliance on visibility introduces a number of unintended consequences.
The exclusive emphasis on visibility may have unintended consequences in the future. Proposed
solutions to the temporal mismatch (page 1-19, lines 20-29) between health related and visibility
related averaging periods, will require development of FRM hourly monitors and lead to the
discounting of nighttime PM values. The suggested focus on daylight hours is primarily to bring
cohesion to the visibility (rather than PM) data in the face of varying RH. This leaves the dark
period as a potential window of opportunity for dumping of PM-rich emissions, to avoid a more
restrictive potential secondary standard. The skewing of the averaging period by use of visibility
as a surrogate for welfare effects should be resisted as much as possible. An approach is to use
calculated visibility, once fully validated (see #3, below), in daylight and nighttime hours, or
over 24 hour averaging periods e.g. page 1-20, lines 1-7).
The acceptance of a secondary standard by regulated communities will depend on its being
relatively simple and intelligible to these communities and to the public. The proposed regulation
of calculated visibility, rather than of PM concentration, is problematic in this regard. If visibility
is to be regulated, then appropriate FRMs and appropriate PRB levels must be established. As
PRB levels likely vary across the country, it must be established that regionally valid PRBs for
visibility reflect similar concentrations of PM. The alternative (page 1-20, line 8-22), to set a
uniform national visibility standard may prove untenable because PRB visibility is expected to
differ regionally. This situation is further complicated by seasonal trends in RH, which will lead
to seasonality of NAAQS for PM. It appears unlikely that regulated communities will embrace a
standard that implies regionally differing levels of PM, to achieve differing levels of visibility.
Given the likely regional variability in PRB for visibility, the use of locally relevant scenes for
valuation should be given greater consideration (page 1-19, lines 5-12), despite the complexity of
this approach. Fewer total scenes might be evaluated, but with greater precision and local
-------
relevance, than is suggested in the plan. This might allow scope for locally relevant "valuation"
(page 1-19, line 10) which could be a powerful metric.
4. Define UVA and its relationship to the REA and ISA, at the point of its first
introduction.
It is not clear from the text what the relationship between the urban visibility assessment (UVA)
and Risk and Exposure Assessment (REA) is. This may only be semantic and can be easily
addressed with a few additional words on page 1-1, near line 21.
5. Greater attention should be paid to linking reconstructed (calculated) light
extinction to measurements of visibility.
The generality of the relationship between reconstructed visibility parameters, (calculated as
functions of PM2.5 and RH), is documented by reference to a strong correlation of such
parameters with PM2.5 (page 1-7, lines 5-12). The reasoning is somewhat circular. Better to focus
on relationships between measured visibility parameters and PM concentrations to demonstrate
these relationships, and on relationships between measured and reconstructed visibility to
support the methodology. To this end, an appropriate goal of this review cycle (page 2-7) is the
proposed development and validation of an urban optimized IMPROVE algorithm.
The discussion of needed visibility parameters, use of models to obtain them, and suitable
averaging periods, is spread over much of the existing text. This should be consolidated into the
text preceding introduction of the urban optimized IMPROVE algorithm, possibly prior to
discussion of appropriate PRB for visibility (section 2.1.1). This would allow determination of
what is needed for each purpose, and then how best to achieve it.
6. Maintain effort to provide quantitative valuation of welfare effects, including in this
case visibility impairment.
The welfare effects and associated secondary standards have not been as influential as primary
health standards for a variety of reasons. Valuation can be an important tool for welfare impacts.
It is a mistake for the current plan to dismiss this goal as too complex (page 3-6, line 11-12),
even if only incremental progress can be made in the current review cycle.
7. Minor error.
-------
It is stated that both average radius and number of cloud droplets declines with increasing PM
derived CCN (page A-9, lines 9-10). Droplet number is likely to increase with decreasing radius
under these conditions.
8. The summary of previous rulemaking should be revised to include the recent court
decision of 24 February 2009.
The discussion (page 1-5, lines 1-10) is incomplete due to court action after preparation of this
draft Plan. This should be addressed for completeness, as the current treatment is now somewhat
misleading.
9. Revise text to condense and focus arguments.
The lengthy discussion on visibility impairment and its relationship with fine PM (page 1-6, line
16 to page 1-8, line 4) is somewhat diffuse and does not lead to a clear conclusion. The objective
of the section should be laid out initially, and the key conclusion summarized at the end. This
might be facilitated by bringing it all into a single subsection under 1.1.1.
10. Elaborate upon key elements that are mentioned in passing.
Important other protections of visibility contained in the Regional Haze Rule and elsewhere
receive passing mention (page 1-11, lines 13-20). These are key to understanding the goals of
this Plan and should be expanded upon.
Certain states have enacted visibility rules (page 1-8, line 18). As above, it would be helpful to
know which states, what rules, and some indication of successes and shortcomings of the
approaches taken.
-------
Comments from Dr. William Malm
Chapter 1 - Introduction
The introductory section presents an overview of the goals and planned approaches for
conducting urban visibility assessments.
1. What are the Panel members' views on the general structure and overall design of the
planned analyses?
Very energetic for the assumed time frame to complete the assessment.
2. Is the plan clear and transparent in its description of the planned approaches? Are the
various assumptions and judgments that must be made in carrying out the planned
assessments clear and transparent?
The plan is transparent in it description; however, the assumptions and associated uncertainties
are yet to be determined.
3. Given the goals set forth for the planned analyses, has the plan appropriately drawn from the
existing scientific and technical information in developing the overall approach? Are there
relevant features that should be added or modified in the planned approach?
/ suggest that the approach is overly complex and may go beyond the information that is
currently available, or at best easily available.
4. In addition to the sub-daily PM2.5 alternative standard considered in the last PM NAAQS
review and summarized in this chapter, an alternative standard structure is being considered in
this review. This alternative structure would use daylight hourly PM light extinction, which can
be measured either by a combination of instruments (nephelometer - PM light scattering and
aethalometer - PM light absorption) or calculated from PM speciation and concurrent relative
humidity data using a linear algorithm.
a. What are the Panel members' views regarding this alternative structure and its
utility in the context of this PM NAAQS review?
/ think this is a great idea; set a standard based on extinction derived from direct
measurements of scattering and absorption. If the standard is violated, there are a
number of ways to estimate which of the major species is primarily responsible. For
-------
instance, it might be a nearby (or distant) wildfire whose effect might be assessed in a
variety of ways including measurements, models, and just plain old human observations.
b. What are the Panel members' views regarding advantages and disadvantages of this
alternate structure compared to the sub-daily PM mass concentration approach?
A standard based on direct measurement of extinction is much superior.
5. The public's preferences for urban visual air quality (VAQ) levels were assessed in the last
review in part by considering results from past western U.S. urban preference studies that used
scenes with distant mountain backdrops.
a. What are the Panel members' views on the usefulness of the planned approach to
conduct a focus group study in different regions of the country with different scene types
to enhance our understanding of the applicability of this earlier body of work to U.S.
urban areas in general?
This is an excellent idea, especially if one can pick representative scenes with realistic
distributions of haze.
b. What are the Panel members' views regarding the scope and approach envisioned for
this effort, given the time available in this review?
A lot to get done within the time frame specified.
Chapter 2 - Assessment of Urban Visibility Conditions
1. Visibility impairment is caused by both PM2.5 and PM10-2.5, though the latter is less
effective on a per unit concentration basis and there is less available PM10-2.5 data available in
urban areas with which to conduct an assessment. PM in the atmosphere includes liquid water
which contributes to light extinction but is removed when filter samples are desiccated prior to
mass and composition analysis.
c. What are the Panel members' views regarding to what degree and using what
approaches EPA should assess the role of PM10-2.5 in urban light extinction?
Could set the standard on only visibility impairment associated with PM2.5 and fit
nephelometers with 2.5 inlets. Could use adjacent IMPROVE monitors to estimate
coarse PM. Could design nephelometers to cycle between PM2.5 and PM10 and directly
measure coarse PM scattering; there would be little extra cost to do this.
Could assume a constant seasonal background for coarse PM, although the background
could vary significantly between urban areas. I would favor a direct measurement with
nephelometers.
-------
d. What are the Panel members' views regarding the advantages and disadvantages of
using direct measurements of ambient PM light extinction (e.g., nephelometerplus
aethalometer, or transmissometer) compared to using a linear algorithm that includes
particle composition and concurrent relative humidity to calculate (reconstitute) PM light
extinction?
/ think a direct measurement is the obvious and best way to go; there is less uncertainty
as to whether there is a standard violation and would be cost effective in that only two
easily calibrated and operated instruments are used. If there is a standard violation, all
data analysis tools can be brought to bear in assessing why the standard was violated as
opposed to relying on a subset of those tools to measure the variable of concern.
e. What are the Panel members' views regarding the importance of refining the
IMPROVE algorithm so that it is optimized for urban areas in order to estimate PM light
extinction for use in this assessment?
Assume some much simpler algorithm to estimate extinction from PM. There will always
be a large degree of uncertainty or variability in directly linking bext to individual
species, and the data requirements to do this as "right" as possible are significant (real-
time particle number size distribution along with continuously spedated data would be
desirable).
2. This plan anticipates using the same approach to estimate policy-relevant background PM
levels as in the health risk assessment. What are the Panel members' views on this planned
approach and its use in the context of the urban visibility conditions assessment?
Given all the uncertainties in emission inventories (like NH3, which is key to understanding
transport of secondary aerosols), deposition rates (given the dependence on surface
characteristics), gross under estimations for SOA production and so forth, why does one think a
model can develop reliable estimates ofPRB? At least CTM estimates and background data
should be combined to come up with a best estimate with uncertainty bounds (+/-factors of ?
probably 2 or more). These issues are even more important if it is anticipated that CTM results
can be meaningfully used for hourly concentrations of relevant species.
3. The planned approach includes consideration of high time resolution PM mass and component
concentrations as well as PM light extinction data, together with relative humidity values, to
assess recent air quality as well as air quality simulated to just meet current and alternative
standards under consideration. However, such high time resolution data are not broadly
available.
10
-------
a. What are the Panel members' views on this planned approach or on other approaches
or data sources that might be explored?
All data and past modeling efforts should be brought to bear to make this assessment.
The approach should be simple and transparent, with defined, broad objectives so that
large uncertainties are recognized and accepted. The goal should be to develop
methodologies that are unbiased, in that on the average you get it right even if there are
large and significant errors on short time scales.
b. What are the Panel members' views regarding useful approaches for characterizing the
uncertainties associated with the urban visibility conditions assessment?
/ didn 't see where there was a detailed plan on assessing uncertainties.
Chapter 3 - Quantitative Visual Air Quality Impact Assessment
1. EPA plans to conduct public preference studies to supplement the information from past
studies concerning VAQ levels that impact public welfare. The planned approach is to conduct
an investigative focus group in one location to develop the approach that will then be used in
group interviews in four urban areas. The proposed approach would ask participants to view an
urban-specific iconic scene and a generic scene in each study location, all selected to be sensitive
to changes in PM light extinction using WinHaze superimposed haze levels.
a. What are the Panel members' views regarding the appropriateness of EPA's preference
study approach and its adequacy to accomplish the goals outlined in the plan?
/ think the approach is appropriate and adequate.
b. What are the Panel members' views with respect to alternative approaches that could
be considered in this review to acquire this and/or additional relevant information?
2. EPA is not planning to include monetary valuation study questions (e.g. willingness to pay,
conjoint analysis) in the studies conducted for this review.
a. What are the Panel members' views regarding the usefulness of adding valuation
questions to the proposed four urban-areas group interview surveys and the adequacy of
such limited information to inform estimate of the value of improved visibility in the
broader nationwide context?
Because this is a welfare-based standard, I suggest that adding a monetary value to a
visibility level is not necessary and even inappropriate.
b. What are the Panel members' views with respect to possible approaches for estimating
monetary values associated with improved visibility that should be considered?
11
-------
Page 1.4: It would be helpful here to have a discussion, a listing of what the standards are,
annual versus 24 hr, primary versus secondary.
Page 1.5: The statement, "Visibility which can be defined as the degree to which the atmosphere
is transparent to visible light...," is patently wrong. The discussion in the following sentences of
the document is also incongruent with this statement. "Transparent" (of transmittance) is only a
small (part significant) part of visibility impairment.
Page 1.6: The statement, "Visibility is often described in terms of visual range, extinction, or
haziness," is correct. However, a sentence should be added to state that none of these metrics
really captures "visibility". Probably the best metric would be contrast transmittance weighted in
proportion to the wavelength sensitivity of the human eye. Even this metric would not take into
account color shifts and human perceptual responses. One has to take some approach to
characterizing visibility, but the EPA should show that it recognizes all the associated
limitations.
The statement, "Direct relationships exist...," is not true. There are approximate relationships
between species and extinction. Apportionment of particle chemistry (species) to extinction is an
ill-defined problem and in principle cannot be done (White, 1986). Briefly discuss the problem
of internal mixtures. (This discussion is lacking in the PM assessment write up as well)
".. .reasonable estimates..." What does reasonable mean? Each person has his own
interpretation. Estimating extinction from PM measurements can easily be in error by a factor of
2 or more; is that reasonable? State the uncertainties that might be expected.
Wording of the paragraph, "Due to regional...," is cumbersome. Consider rewriting.
Page 1.7, first paragraph: The whole discussion of correlation doesn't seem to be relevant. One
can have a high degree of correlation between variables that are different by orders of magnitude.
So if you have a correlation between PM and extinction, it doesn't mean you know anything
about the level of extinction based on the PM value.
Second paragraph again has this meaningless discussion about correlations. The Schmidt et al.
(2005) reference keeps coming up; was this peer reviewed?
Page 1.8: I think the rationale that a 12-4 visibility standard has greater importance than a 24-hr
standard is spurious. Most people drive to and from work early in the morning and later in the
afternoon. It is this time period when most people have an opportunity to see and appreciate
their surrounding landscape, not when they are sitting in their offices in the middle of the day.
Second paragraph: There is one reference to VAQ and public awareness. I think that a proper
literature review and summary of all the public awareness/well being should be carried out. The
3 or 4 urban visibility studies, I would guess, are only a subset of one kind of study that has been
carried out.
12
-------
Page 1.9: What are the "nonpollution" effects that could affect visibility?
Page 1.14: In some place, maybe in a discussion of Figure 1.1, there needs to be a discussion of
just what is meant by urban visibility: what views, what times, etc., the whys and wherefores of
why urban visibility is important, and under what circumstances. Some relevant discussion
appears in Hyslop (2009).
Page 1.19, second paragraph: "...similarly sensitive to small increments...." I agree that the
scenes should be similarly sensitive but not necessarily to small increments. They should be
similar but only to realistic, expected increments and for some representative scene. As written
it seems that the scene may be picked to be overly sensitive and not representative of a typical
urban scene.
Last paragraph: Add some discussion as to what nighttime visibility is of importance. Some
think that seeing the lights of an urban setting is what is being addressed when discussing
nighttime visibility, when I think that what may be of concern is the visibility of the night
skyscape.
Page 2.1: It is stated that "We will estimate PM light extinction from measured or estimated PM
mass, composition, and RH, using a refined, urban-optimized linear algorithm" First, the
equation will not be linear in that the f(RH) term is very nonlinear with RH. Second, because of
all the limitations of apportioning extinction to mass composition in a defensible way, I suggest a
very simplified version of the IMPROVE equation. Even the IMPROVE equation, which has
been optimized for nonurban environments, can for any given 24-hr period, much less any
shorter time increments, be in error by factors as much as 2 or more. Significant errors are also
associated with the multiplier that is used to estimate POM from OC, as well as sampling bias
associated with OC artifacts filter artifacts or with loss of SVOC particles that are dependent on
sampling face velocities. All these uncertainties suggest that one should not try to create a false
sense of understanding and accuracy by having an unduly complicated algorithm to estimate
extinction from mass measurements, speciated or not.
Should the figure reference be 1.2 instead of 2.2?
Page 2.2: How will PRB be used in visibility standard setting - why does one need to know
PRB?
The first paragraph under 2.1.1 needs either more background or a more clear presentation. "In
the last staff paper... .a possible secondary PM standard were (or was - or more possible
secondary standards) compared to % tilepoints in the estimated distribution of 24 hr average PM
PRB." Now, what am I suppose to take away from that statement? What is a policy relevant
background? "... standards were compared to..." - to what end, for what purpose?
13
-------
I am amazed at the statements concerning using CTM to estimate backgrounds because "This
approach can provide more spatial and temporal resolution for estimating PRB compared to the
use of measurements given the sparse nature of remote measurement sites and the concern the
remote sites are affected by non local anthropogenic sources." I agree that one can derive more
spatial and temporal resolution using a CTM but, again, to what end? Given all the uncertainties
in emission inventories (like NH3, which is key to understanding transport of secondary
aerosols), deposition rates (given the dependence on surface characteristics), gross
underestimations for SOA production, and so forth, why does one think a model can develop
reliable estimates of PRB, whatever that is? At least CTM estimates and background data should
be combined to come up with a best estimate with uncertainty bounds (+/- factors of ? probably 2
or more).
Why not just make best estimates of PRB for broad areas of the country and go with that?
Again, I believe all this modeling is giving a false sense of understanding. Make a best estimate
and go with it, incorporating all current knowledge, which can include CTM runs.
Page 2.4: Here the discussion goes to PRB on an hourly basis. The model performance only
deteriorates as one goes to shorter time frames. It would be interesting to estimate model
uncertainty with the uncertainty in using a very simplified PM to bext algorithm. I would bet that
the model uncertainty is far greater than any uncertainty in estimating bext from PM.
Under recent conditions I would suggest relying on available nephelometer and TEOM (or some
other estimate of hourly PM) measurements to come up with a range of hourly bext values rather
than estimate them from 24-hr-average PM filter-based measurements. Knowing the hourly
distribution of bext is really what is needed for the assessment. It will not make much difference
which species are responsible for the extinction. Of course, for any control strategy, bext will
have to be linked back to the relevant species.
Page 2.6: Where is Table B-l? Is Table B-2 the relevant table?
Page 2.9, section 2.5: In am not certain there is sufficient available data to characterize
uncertainties at the level suggested.
Page 3.4: ".. .that have long sight paths...." Better stated as sensitive view sheds. "... selected
to have sensitive scenic elements..." might be changed to also state that not only will they be
sensitive, but typical and not unusual.
Page 3.5: I don't think the strategy for selecting extinction ranges under the section assessment
scenarios makes sense. Why would the maximum hourly value be restricted to something less
than the current 24-hr-average NAAQS level? If one is using hourly values, then it should be
restricted to the max hourly value that would typically make up the 24-hr-average value.
Depending on the distribution of hourly numbers that make up the 24-hr average, one would
have significantly greater values than 35 |ig/m3, maybe as high as 100 |ig/m3 or more!
14
-------
REFERENCES
Hyslop, N. P. 2009. Impaired visibility: the air pollution people see. Atmospheric
Environment, 43, 182-195.
White, W. H. 1986. On the theoretical and empirical basis for apportioning extinction by
aerosols: a critical review. Atmospheric Environment, 20, 1659-1672.
15
-------
Comments from Mr. Rich Poirot
Visibility impairment is caused by both PM2.s and PMi0-2.s, though the latter is less
effective on a per unit concentration basis and there is less available PMio_2.s data
available in urban areas with which to conduct an assessment. PM in the atmosphere
includes liquid water which contributes to light extinction but is removed when filter
samples are desiccated prior to mass and composition analysis.
c. What are the Panel members' views regarding to what degree and using what
approaches EPA should assess the role of PMio-i.s in urban light extinction?
Coarse mass typically accounts for less than 10% of extinction in the eastern US, but is likely to
be an important contributor to visibility impairment in arid southwestern urban areas. Phoenix
should be a fairly extreme example, where IMPROVE data indicate coarse mass contributes
about 25% of extinction on average, about 15% on the haziest 20% days and exceeds the effects
of fine mass about 2% of the time. I don't see many reasons to exclude coarse mass in an
assessment of urban extinction (and for that matter as a potential component of a secondary
NAAQS regulatory metric, which might conceivably be composed of something like fine mass
plus some small fraction of coarse mass) if data are available. If coarse data aren't available, a
reasonable approximation might be made using fine crustal elements, or PMio data from nearby
sites. If an optical regulatory metric is being considered - such as measured by nephelometer,
nephelometer plus aethalometer, or transmissometer - coarse particles will contribute to the
measured effects, and may in some cases offer potential control options.
Disadvantages of including coarse mass include that it adds complexity and monitoring cost
without providing a substantially better indicator in most locations. Coarse mass measurements
are also likely to be less spatially representative than fine mass, and highest coarse mass
concentrations may often be associated with natural windblown dust events. Light scattering by
coarse mass is also not efficiently quantified by nephelometry, and complex data adjustments
may be needed if an optical indicator is being considered for a secondary standard.
d. What are the Panel members' views regarding the advantages and
disadvantages of using direct measurements of ambient PM light extinction (e.g.,
nephelometer plus aethalometer, or transmissometer) compared to using a linear
algorithm that includes particle composition and concurrent relative humidity to
calculate (reconstitute) PM light extinction?
16
-------
For conducting an assessment of visibility effects, preferences, valuation, etc., the optical
measurement data are just not likely to be available in non-class-1 areas (except cities like
Denver and Phoenix where they are used to assess compliance with visibility standards).
Conceivably, and with much work (which would be interesting and productive), airport ASOS
measurement data, with appropriate screens and/or adjustments for RH effects, might be used as
the basis of an assessment of current effects, but the required effort would be substantial.
Whereas, speciated PM2.5 data from CSN (or urban IMPROVE) sites, with collocated or nearby
PMio in most cases would allow calculations of (every 3rd day, 24-hour) reconstructed extinction
- similar to what's done for the Regional Haze Rule - with existing data. Conceivably, the 24-
hour filter data might be combined with lower daytime-only RH data to produce extinction
estimates more in line with the times when people are awake and outdoors experiencing visual
air quality effects. As a rough approximation, the speciated filter data might be disaggregated to
hourly values based on continuous mass measurements - either assuming the speciation remains
constant throughout the day (which we know is wrong) or using some knowledge of the typical
diurnal cycles of sulfate, nitrates EC, etc. Alternatively, you could make pretty reasonable hourly
extinction estimates just using continuous fine mass and a generic (or regionally adjusted)
aerosol extinction efficiency. A generic (or regional) aerosol f(RH) function would further
improve these estimates.
In considering future regulatory metrics, using optical measurements would have the major
advantages that the perceptible effect would be the indicator, and could reflect the much shorter
(than 24 hour) averaging times and daylight conditions over which effects are perceived. The
combination of nephelometer plus aethalometer would provide valuable continuous information
on separate light scattering (of the actual ambient aerosol as people see and breathe it - not some
hypothetical construct with some of the water and other volatile species artificially removed) and
light adsorption (which, with a dual wavelength aethalometer, would add other visibility and
health-relevant information on source influences like diesel exhaust and wood smoke).
Disadvantages of optical indicators include the costs and logistical problems of adding
substantial new instrumentation, difficult (in some cases impossible) siting criteria for long-path
transmissometers, inefficient measurement of coarse particle scattering by nephelometers, and
need for (much) better standard operating and data processing procedures for aethalometers.
Also with optical indicators, you know if and when an exceedance occurs, but not necessarily
much about what caused it, for which much supplemental information may be needed. Whereas
with a continuous mass -related indicator, you have a better sense of pollutant causality, but less
certainty about the actual perceived effect at any given time. A species reconstructed extinction
indicator provides a good link to pollutant causes, but has poor and intermittent time resolution
e. What are the Panel members' views regarding the importance of refining the
IMPROVE algorithm so that it is optimized for urban areas in order to estimate
PM light extinction for use in this assessment?
17
-------
It seems logical that some revisions could be developed for the IMPROVE algorithm so that it is
optimized for urban areas. However, I doubt that such revisions will end up being very
substantial, and I'm not sure sufficient data exist to do this empirically for more than a few
locations, in which case it will be hard to know whether the change is generic urban or city-
specific. Possibly, some quick revisions could be based solely on the slightly different nature of
the IMPROVE and CSN speciation data (for example to accommodate differences in the carbon
measurement methods). We might also expect urban OC to be generally less aged (less
oxygenated and smaller sizes) than rural counterparts. Possibly urban coarse particles will be
"darker" and have higher extinction efficiencies than rural counterparts. If there are sufficient
data, possible revisions to the equation should be considered, but if such data aren't available,
this shouldn't be a major concern.
2. This plan anticipates using the same approach to estimate policy-relevant background
PM levels as in the health risk assessment. What are the Panel members' views on this
planned approach and its use in the context of the urban visibility conditions
assessment?
I don't have a strong opinion on this. I assume the PRB estimates will be species-specific, which
may be important, given that the most efficient light scattering and absorbing species (sulfates,
nitrates, and EC) are predominantly anthropogenic. It might be interesting to compare the
modeled PRB concentrations at IMPROVE sites with the haze reg estimates of "natural
background". Theoretically, the difference between the two is from non-North American
anthropogenic sources.
3. The planned approach includes consideration of high time resolution PM mass and
component concentrations as well as PM light extinction data, together with relative
humidity values, to assess recent air quality as well as air quality simulated to just meet
current and alternative standards under consideration. However, such high time
resolution data are not broadly available.
a. What are the Panel members' views on this planned approach or on other
approaches or data sources that might be explored?
As indicated previously, I don't think its unreasonable to disaggregate the species data to hourly
values using the continuous mass data - either by assuming uniform temporal composition or by
weighting the species disaggregation by what we understand about photochemical production of
sulfates, nitrates and secondary OC, as well as what we know about diurnal and day of week
patterns in traffic emissions of EC, primary OC, road dust, etc. An alternative reasonable
approach would be to just use the continuous fine mass data combined with generic dry
scattering efficiencies and generic aerosol f(RH) growth functions. Possibly these functions
could be varied on a regional or seasonal basis.
18
-------
b. What are the Panel members' views regarding useful approaches for
characterizing the uncertainties associated with the urban visibility
conditions assessment?
I think the proposed approaches in the welfare assessment plan for characterizing uncertainty are
reasonable. Many of the uncertainties associated with relating specific PM mass and/or species
concentrations to optical effects are associated with the varying water content of the ambient
aerosol. In terms of conducting assessments or specifying regulatory metrics, those uncertainties
can be substantially constrained by limiting the time periods for consideration to those when
water effects are least important. This uncertainty in the "aerosol to optics" relationship can also
be reduced by use of an optical indicator (in which case we increase the uncertainty in the
contributions from pollutant mass or species vs. water). This uncertainty can also be constrained
by limiting the time periods for consideration to those when water effects are least important, at
which times the effects of different species are also minimally important. Further, the visual air
quality at any given PM2.5 concentration (of typical composition) can always be described as: as
least as bad as X , but getting worse, in a very predictable way, as humidity increases.
In addition to uncertainties in the relationship between aerosol and optical effects, different
scenes can vary widely in their inherent sensitivities to changes in light extinction, with effects
being cumulative and greatest over longer sight paths. There are also substantial variations in
individual perception and preferences, and subsequently in the levels or ranges of visual air
quality considered to be adverse. Much of this variability can and should be quantified using
standard survey techniques, but it may be difficult to develop visual survey instruments which
are representative of the kinds of views characteristic of individual urban areas which are
"equivalent" to or exchangeable with views in other areas. A generic urban view without
recognizable local landmarks may be inherently less scenic than the actual views characteristic
of most individual urban areas.
One area of uncertainty relating to the regulatory metrics of secondary and primary standards
that I believe is typically given insufficient attention is the form of the standard. Time ran short
during the last PM NAAQS review cycle, and there was not time (or a sufficient technical basis)
to consider and recommend a clear form of the proposed secondary standard. Staff & CAS AC
recommendations were for a sub daily 4 to 8-hour PM2.5 indicator in the range of 20 to 30 ug/m3.
But the recommended range was so broad - 90th or 92nd to 98th percentile - that it spanned a huge
range, from being slightly more protective than the new 24-hour primary standard to being much
less protective than the new primary standard. I think the lower bound (90th percentile) of that
recommended form was much too lenient and based on flawed rationale. In any event, I suggest
the agency put some effort into better understanding the frequencies at which the public is
willing to accept various degradations of visual air quality. Some consideration might also be
given to including a temporally varying form as a component of a secondary standard. For
example, in an analog to the progress-oriented regional haze regulations, maybe there could be a
19
-------
requirement that the frequency of exceedances of a subdaily optical or PM threshold decrease
over time. Possibly this concept of "improving visibility" - or a shift in the frequency
distribution - could be explored by survey instrument, as I don't think its very likely that people
have an absolute sense that it's the 36th or 18th or 7th haziest day a year that breaks the camel's
back...
20
-------
Comments from Dr. Ted Russell
Though generally pleased with the PM NAAQS Scope and Methods Plan for Urban Visibility
Impact Assessment (hereafter, SM-Welfare) for how they plan to consider welfare impacts, they
should go beyond urban visibility to include climate. Indeed, climate concerns might trump
other concerns, and one could conceive of controls that might actually exacerbate climate
forcing. As noted in the SM, there are significant uncertainties as to how PM will impact
climate, but climate impacts have the potential to significantly impact our assessment of potential
risks, particularly in terms of how specific components impact climate versus other components.
As such, I am currently concerned that the SM-Welfare may miss a major piece of the necessary
analysis. On the other hand, if one is just dealing with urban visibility, the current SM does
provide a good roadmap to developing the information for providing advice on the potential
revision to the secondary PM NAAQS. On the other hand, without consideration of the potential
climate impacts, I would worry that the advice may not be fully informed. Dealing with the
issue more fully here can help lay the foundation for the next review as well.
Like the SM-Health, I do note a few deficiencies, both in the document as well as the plan, e.g.,
(as noted in the SM-Health) it would have been very nice if the document had a section
summarizing criticisms by CASAC and others on the prior risk and exposure assessments, and
how they have responded. This could be done by grouping the types of comments made, and
how they plan to address them, and where in the current document the planning takes on those
criticisms, very much like a typical response to review document. This should become standard
in the process.
Chapter 1
While I think that Chapter 1 does a fine job of laying out the issue, and noting that there is a
relationship between PM levels and light extinction, it is written too much as the same chapter
for the SM-Health in that it is oriented towards saying that strong relationships have been
observed between PM and visibility. Unlike health, there is much less reason to address this
issue by relying on statistics/finding associations. The underlying physics is well known. This
should be the major focus. For one, it will focus the uncertainty assessment, and it will shift the
discussion more to how certain are the effects. After laying out that the physics are well known,
and that this can be addressed by first principle analyses if desired, one can then say that the
correlations follow the physics. Indeed, after going through the rest of the SM-Welfare, I get the
feeling that the problem is being made too big, and that the analyses are going to be overly
complex, in part because we know so much more about the physics. I would reconsider what
needs to be done here to provide the type of analysis that would lead to a different standard than
for the primary NAAQS, and recognize that there is going to be much more known about the
21
-------
physics here, so a certain amount of less detailed analysis can be tolerated because the
uncertainties will still be quite reasonable.
On page 1-19, line 27, there is the comment that visibility effects are less well understood at
night. This is true, but it misses something larger. We still understand the optics, and the
visibility effects are probably understood quite well enough. What is less well understood is the
importance/value. Most folks, I suspect, would say that visibility at night is less important, and
that is what drives how to proceed, not our lack of understanding.
On page 1-20, it would be best, if it is not too much trouble, to incorporate both composition and
RH, and this should be relatively straight forward using CMAQ results. On the other hand, if
this is a very burdensome analysis, (though I would think it actually might be easier), then the
former approach is fine.
I am positively disposed to having a PM light-extinction-based standard and using a
nepholometer/aetholometer pair to provide PM light extinction. This would minimize some
issues in computing light extinction. However, it also adds an issue about how to simulate
attainment as some additional uncertainties are added, though I think those are quite reasonable.
Chapter 2:
The plan to assess urban visibility conditions is reasonable in most cases, though it does get a bit
unclear, and it reads as though what level of uncertainty is allowable has not been decided,
leading to some areas where the levels of analysis do not match.
In regards to the PRB, they suggest using CMAQ results, but if that is too time consuming, they
might use the prior review's analyses. Using CMAQ results should not be overly complex, and
is preferable. They plan to use those results in many other ways, so they should make it so doing
the PRB-light extinction calculation just part of what is routinely found. There is one problem
with relying on CMAQ results, and this is found (I think) in the ISA. CMAQ will have
considerable uncertainty in simulating PMcoarse from soil, and this can complicate finding the
PRB. Table 3-26 in the ISA does not provide an evaluation of CMAQ PMcoarse results. Note,
like the SM-Health, I would still be cautious about laying the poor performance of CMAQ in the
west to grid resolution problems.
On page 2-4, they start dealing with how to address PM10-2.5. I am not convinced that much
ado about nothing is being made. How important is the anthropogenic component of PM10-2.5
to visibility? I did not see this in the ISA (I saw course contribution, but not anthropogenic
coarse contribution). Next, I think the approach being proffered in lines 28-29 is going to add
unnecessarily to the uncertainty as it integrates two relatively large uncertainties. The first is the
uncertainty in the source apportionment. The second is in the coarse-to-fine ratio. Both can be
large. I would like to see how much anthropogenic PMcoarse adds to VAQ issues in urban
22
-------
areas. If anthro-PMcoarse is high on days when dust is high, one is much less concerned given
the optical relationship.
Section 2.1.3 assumes that the standard will be based on PM-mass, possibly with compositional
information as well. A deficiency is that it glosses over how compositional information will be
used as few places will have sub-daily compositional information. Unlike the SM-health
analysis, compositional information plays a much larger role here. They should also add a
section on the approach if they use aetholometer/nepholometer pairs to measure light extinction
directly. In that case, meeting an alternative secondary PM standard will be much less uncertain:
a whole set of calculations need not be done.
In response to the specific charge questions associated with Chapter 2:
la (or c): Role of PMcoarse: As noted above, you should consider the importance of this
component to visibility, noting that we are interested in the anthropogenic component, and that
the anthropogenic component may be large when the natural component is large, so that has
much less importance.
Ib. As noted above, I am positively disposed to this approach. It does shift some uncertainties,
but I think it reduces them and simplifies the process. In general, given our knowledge of the
physics and chemistry of the system, we have a pretty good way of relating emission changes in
the more important species to visibility changes.
Ic. Revising the IMPROVE algorithm may not be necessary. A first assessment can be to see
what might be gained, and see if much effort is required. Is a reanalysis going to reduce an
uncertainty or bias to the degree it will influence the process. I suspect it is relatively
straightforward, so it may be deemed appropriate even if it only slightly modifies the results.
2. For the most part, following the same method to estimate PRB is fine, though more
attention to compositional information and PMcoarse must be included, including model
evaluation and how high levels of PRB-light extinction correlate to high levels of
anthropogenic light extinction.
3a. I applaud the use of more high resolution compositional information. You have identified
SEARCH as providing such data. There are other high resolution data (e.g., from the Supersites
and other special studies) that can provide information on how composition changes with time,
and CMAQ can as well (though this will add other uncertainties).
3b. The uncertainties section is rather short at present, though correctly notes that there will be
uncertainties in the light extinction calculation. However, I suspect those will be small. How
CMAQ results are used, and uncertainties in those results, will be significant and should be
assessed. How (and if) source apportionment is done on ambient data is also an issue. They
23
-------
need to address the uncertainty in how well visibility improvement can be calculated and how
well they can predict how PM composition will change with controls.
Chapter 3.
The central theme in Chapter 3 is the development of how to quantitatively value the impact of
the perception of urban visual air quality on individuals, citing the lack of information about
public preferences.
Appendix A.
Again, climate may be the single biggest welfare (and it is also linked to health) concern from
PM. In the REAI think that much more emphasis on climatic impacts, positive and negative,
and compositionally dependencies, should be given than is indicated here. It is actually quite
possible to do a quantitative analysis, though such an analysis may be fraught with uncertainties.
However, if the conclusion is that certain species that increase urban light extinction also
enhance global warming, one might be very tempted to develop a standard that addresses those
components. For example, if an aethalometer/nephelometer pair is used, one might be tempted
to weight the light absorption component, or if it is done by component based on speciated PM
mass, one might provide additional control on light absorbing species. Might one have a
standard that sets a limit on total extinction, however, doubling (or more) the absorbing
component?
24
------- |