UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D C 20460
PRO*-
OFRCE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD
SUBJECT: Transmittal of Science Advisory Board Report
FROM:
TO:
Vanessa T. Vu
Director, Science Advisory Board Staff Office (1400R)
Karen Sheffer
EPA Headquarters Library Repository (3404T)
This is to advise you that the Science Advisory Board Clean Air Scientific Advisory
Committee, Oxides of Nitrogen (NOx) and Sulfur Oxides (SOx) Secondary Review Panel,
issued a report numbered EPA-CASAC-11-003, Review of the Policy Assessment for the
Review of the Secondary National Ambient Air Quality Standard for NOx and SOx: Second
Draft, dated December 9,2010.
Two copies of the report are attached and a third copy has been sent electronically to
the attention of Ms. Jeannie Turner at turner.]eannie@epa.gov. The report is available in
electronic format on the Science Advisory Board's Web site at http://www.epa.gov/sab.
If you have any questions regarding this report, please contact the Designated Federal
Officer, Dr. Angela Nugent directly at (202) 564-2218.
Attachments (2)
Internet Address (URL) • httpJAwww epa gov
Recycled/Recyclable • Printed with Vegetable Oil Based Inks on 100% Postconsumer. Process Chlorine Free Recycled Paper
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
\ WASHINGTON D.C. 20460
^i PRO*
OFFICE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD
December 9, 2010
EPA-CASAC-11-003
The Honorable Lisa P. Jackson
Administrator
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, NW
Washington, DC 20460
Subject: Review of the Policy Assessment for the Review of the Secondary National
A mbient A ir Quality Standard for NOx and SOx Second Draft
Dear Administrator Jackson:
The Clean Air Scientific Advisory Committee (CASAC or Committee) Oxides of
Nitrogen (NOX) and Sulfur Oxides (SOX) Secondary National Ambient Air Quality Standards
(NAAQS) Review Panel met on October 6-7, 2010 and held public teleconferences on
November 9, 2010 and November 10, 2010 to review EPA's Policy Assessment for the Review of
the Secondary National Ambient Air Quality Standards for NOx and SOx Second Draft. This
letter provides CASAC's overall comments and evaluation. The CASAC and Panel membership
is listed in Enclosure A. The Panel's responses to EPA's charge questions are presented in
Enclosure B. Finally, Enclosure C is a compilation of individual panel member comments. This
letter provides our views on: 1) the need for retaining the current and developing new secondary
standards; 2) limitations in our review of the second draft Policy Assessment and priority needs
to be addressed in making revisions; and 3) CASAC recommendations regarding the secondary
standard.
Need for retaining current and for designing new secondary standards
The current public-welfare-based (secondary) NAAQS standards for oxides of nitrogen
(NOX) and sulfur oxides (SOX) were designed to protect vegetation from exposures to injurious
concentrations of gaseous NOX and SOX. This protection is a desirable goal and for that reason
the CASAC Panel recommends that the current secondary NOX and SOX NAAQS standards
should be retained. The current standard forNOx is an annual arithmetic standard of 53 ppb,
using NO2 as the indicator species, identical to the primary annual health-based standard, with no
short-term secondary standard. The current secondary standard for SOX is 0.5 ppm 3-hour
average, not to be exceeded more than once per year, using SOj as the indicator species, and it is
separate from the primary standard and there is no long-term secondary standard.
-------
EPA staff has demonstrated through the findings of the Integrated Science Assessment
(ISA), the Risk and Exposure Assessment (REA). and the draft Policy Assessment that ambient
NOX and SOX can have, and are having, adverse environmental impacts on some ecosystems
across the United States due to deposition of NOX and SOX, even at current concentrations under
existing standards. Those impacts include ecosystem acidification and undesirable levels of
nutrient enrichment in some ecosystems. For this reason, CASAC has concluded that different
and more protective standards are needed for NOX and SOX. A NAAQS is needed that would be
formulated in a way that would provide requisite protection in sensitive areas, while not
providing protection that is more stringent than necessary.
We compliment EPA Staff on the progress that has been made since CASAC reviewed
the first draft Policy Assessment and appreciate Staffs responsiveness to CASAC's initial
comments. There are significant scientific challenges in developing multipoMutant, ecologically
relevant secondary standards. These challenges are made more difficult by the regulatory
constraints under which EPA staff operates: the standard must use ambient air pollutant
concentrations as the indicator, rather than pollutant depositional flux and only oxidized
nitrogen, not chemically-reduced forms (both inorganic and organic) of reactive nitrogen, is
currently included as a criteria pollutant. To meet these constraints, EPA has developed a new
index, the Atmospheric Acidification Protection Index (AAPI). This innovative index integrates
the combined effects of atmospherically deposited oxides of nitrogen and sulfur (NOX and SOX),
as well as chemically reduced forms of reactive nitrogen (NHX). The AAPI also takes into
account a series of ecosystem characteristics that determine sensitivity to total acidifying
deposition in various regions of the United States. The AAPI approach is responsive to recent
recommendations by the National Research Council for multi-pollutant air quality management
(Air Quality Management in the United States, 2004).
Despite the regulatory constraints described above, the AAPI approach would
appropriately integrate the combined effects of NOX and SOX deposition on aquatic acidification,
and could provide protection for sensitive aquatic ecosystems at an appropriate spatial scale.
Use of the AAPI, however, introduces a number of technical complexities because it considers
depositional effects of multiple pollutants within diverse and complex ecological systems.
Because of these complexities, it is not apparent how to construct an equally appropriate, and
significantly simpler, approach to capture the many important processes that influence the
relationship between observable atmospheric concentrations and aquatic acidification.
CASAC notes that EPA did not have sufficient time in preparing the second draft Policy
Assessment to formulate a complementary approach that would be the basis for standards to
protect against nutrient enrichment effects on aquatic ecosystems and nutrient enrichment and/or
acidification effects on terrestrial ecosystems. Although a standard that focuses on aquatic
acidification would provide some co-benefits in addressing these other adverse effects, in the
future, EPA should consider developing approaches for protecting against nutrient enrichment
effects on aquatic ecosystems and nutrient enrichment and/or acidification effects on terrestrial
ecosystems.
-------
Challenges to the CASAC review of the second draft Policy Assessment and priority needs
to be addressed in final revisions
CASAC's review of this document has been constrained by the limited time available for
review and the incomplete status of the submitted document. Even though the second draft
Policy Assessment was novel and complex, the CASAC Panel received the document only three
weeks before the review meeting. As described below and in our responses to the Charge
Questions, there are critical sections of the Policy Assessment that are unclear and/or that require
further analyses. In addition, and in contrast to policy assessments for other pollutant reviews,
EPA did not provide staff recommendations for key elements of the secondary NAAQS for NOX
and SOX along with supporting rationales. As a result, CASAC is not able to provide specific
comments on the EPA Staff recommendations, nor is CASAC able to use Staff recommendations
to help frame CASAC discussions or recommendations about the four key elements of the
NAAQS.
The final Policy Assessment needs to more clearly and fully set out the basis for the
recommended ranges of each of the four elements (indicator, averaging time, level and form) of
the proposed NAAQS. The implications of choosing specific combinations from within the
ranges of elements should be thoroughly discussed, with justifications provided for the specific
options or range of options that staff recommends. It would also be useful to see a map and/or
tabular estimates of the spatial extent and degree of severity of NAAQS exceedances expected to
result from the recommended combinations of the elements of the standard.
The final Policy Assessment needs to more fully describe the proposed alternative
approaches for landscape categorization of inherent sensitivities to acidification, and how these
alternative approaches relate to the different target fractions of water bodies that would be
protected at different acid neutralizing capacity thresholds. The implications of choosing
specific landscape categorization approaches combined with specific fractions of water bodies to
be protected should be thoroughly discussed, with justification provided for the specific
combinations or range of combinations that staff recommends.
The final Policy Assessment needs to provide a more detailed analysis of the uncertainties
associated with the entire AAP1 calculation, and of the relative sensitivities of the allowable
ambient concentrations to uncertainties in its individual components, including uncertainties
introduced by use of models (e.g., CMAQ and the ecological model, MAGIC). The AAPI
formulation is unavoidably complex and dependent on critical assumptions and model
calculations, which are characterized by various levels of uncertainty. This more complete
uncertainty analysis should focus on the overall, cumulative uncertainty estimation including the
possible application of Monte Carlo techniques.
CASAC recommendations regarding the secondary standard
While the CASAC Panel is supportive of the AAPI approach developed, CASAC is not
able to provide consensus recommendations on all elements of the standard because of the
constraints to its review identified above. Further, each element of the standard should be
considered in the context of the choices for the other elements. The Policy Assessment does not
yet provide adequate analyses on all the specifics of a new (and novel) NOX-SOX secondary
NAAQS. However, CASAC will offer initial comments on some elements of the standard. We
-------
agree that a 3-5 year averaging time appears appropriate. CASAC supports the general structure
of the A API equation and agrees that acid neutralizing capacity is an appropriate ecological
measure for reflecting the effects of aquatic acidification. Acid neutralizing capacity targets in
the range of 20 to 100 u,eq/L appear appropriate to consider at this time.
CASAC recognizes the very tight time lines associated with revising the NOX and SOX
secondary NAAQS. However, CASAC should have the opportunity to review a more complete
Policy Assessment, one that provides staff recommendations, the basis for the choices made, the
direct supporting analyses for those choices, and the ramifications of alternative choices within
the ranges of the alternatives. To this end, we are working through the SAB Staff Office to plan
a meeting to review the completed Policy Assessment in February 2011. Without this information,
we are unable to provide the level of advice that you need and to fill our role under the Clean Air
Act. EPA staff must provide complete documents with sufficient time to permit an in-depth
review.
Summary
While we have identified various needs for additional analyses and enhanced clarity
before the final Policy Assessment is published, CASAC remains supportive of the novel
approach described in the Policy Assessment for the Review of the Secondary National Ambient
Air Quality Standard for NOX and SOX. Second Draft. We support EPA staffs continuing work
on revising the Policy Assessment to establish a foundation for a revised NOX-SOX secondary
NAAQS. This work is groundbreaking and significant for several reasons:
1) The current NAAQS review for welfare-based effects was conducted separately from the
review of the health-based standard and has allowed focus on ecological impacts.
2) The review was designed to consider two criteria pollutants at the same time, and set the
stage for a "multi-pollutant/multi-media/multi-effect" approach as recommended in the
2004 National Research Council report, and
3) The AAPI takes into account another chemical form of biologically reactive nitrogen,
NHX that is important to aquatic acidification, but is not a criteria pollutant.
In closing, CASAC trusts that our comments will be useful in revising the Policy
Assessment.
Sincerely,
/Signed/ /Signed/
Dr. Armistead (Ted) Russell, Chair Dr. Jonathan M Samet, Chair
CASAC Oxides of Nitrogen and Sulfur Oxides Clean Air Scientific Advisory Committee
Secondary NAAQS Review Panel
-------
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 - Rosters
U.S. Environmental Protection Agency
Clean Air Scientific Advisory Committee
Oxides of Nitrogen (NOx) and Sulfur Oxides (SOx) Secondary Review Panel
CHAIR
Dr. Armistead (Ted) Russell, Professor, Department of Civil and Environmental Engineering,
Georgia Institute of Technology, Atlanta, GA
MEMBERS
Dr. Praveen Amar, Director, Science And Policy, NESCAUM, Boston, MA
Dr. Andrzej Bytnerowicz, Senior Scientist, Pacific Southwest Research Station, USDA Forest
Service, Riverside, CA
Ms. Lauraine Chestnut, Managing Economist, Stratus Consulting Inc., Boulder , CO
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. Charles T. Driscoll, Jr., Professor, Department of Civil and Environmental Engineering,
College of Engineering and Computer Science, Syracuse University, Syracuse, NY
Dr. H. Christopher Frey. Professor, Department of Civil, Construction and Environmental
Engineering, College of Engineering, North Carolina State University, Raleigh, NC
Dr. Paul J. Hanson, Distinguished R&D Staff Member, Environmental Sciences Division, Oak
Ridge National Laboratory, Oak Ridge, TN
Dr. Rudolf Husar, Professor, Mechanical Engineering, Engineering and Applied Science,
Washington University, St. Louis, MO
Dr Dale Johnson, Professor, Nat. Res. Env. Sci., College of Agriculture, Biotechnology, and
Natural Resources, University of Nevada, Reno, Reno, NV
Dr. Narcsh Kumar, Senior Program Manager, Environment Division, Electric Power Research
Institute, Palo Alto, CA
-------
Dr. Myron Mitchell, Distinguished Professor and Director of Council on Hydrologic Systems
Science, College of Environmental Science and Forestry, State University of New York,
Syracuse, NY
Mr. Richard L. Poirot, Environmental Analyst. Air Pollution Control Division, Department of
Environmental Conservation, Vermont Agency of Natural Resources, Waterbury, VT
Mr. David J. Shaw*, Director, Division of Air Resources, Department of Environmental
Conservation - New York State, New York State, Albany, NY, United States of America
Dr. Kathleen Weathers, Senior Scientist, Gary Institute of Ecosystem Studies, Millbrook, NY
SCIENCE ADVISORY BOARD STAFF
Dr. Angela Nugent, Designated Federal Officer, U.S. Environmental Protection Agency,
Science Advisory Board (1400R), 1200 Pennsylvania Avenue, NW, Washington, DC, Phone:
202-564-2218, Fax: 202-565-2098, (nugent.angela@epa.gov)
* Provided individual comments only. Did not participate in the Panel's deliberations.
-------
U.S. Environmental Protection Agency
Clean Air Scientific Advisory Committee
CASAC
CHAIR
Dr. Jonathan M. Samet, Professor and Flora L. Thornton Chair, Department of Preventive
Medicine, University of Southern California, Los Angeles, CA
SAB MEMBERS
Mr. George A. Allen, Senior Scientist, Northeast States for Coordinated Air Use Management
(NESCAUM), Boston, MA
Dr. Joseph D. Brain, Cecil K. and Philip Drinker Professor of Environmental Physiology,
Department of Environmental Health, Harvard School of Public Health, Harvard University,
Boston, MA
Dr. H. Christopher Frey, Professor, Department of Civil, Construction and Environmental
Engineering, College of Engineering, North Carolina State University, Raleigh, NC
Dr. Armistead (Ted) Russell, Professor, Department of Civil and Environmental Engineering,
Georgia Institute of Technology, Atlanta, GA
Dr. Helen Suh, Senior Lecturer in Environmental Chemistry and Exposure Assessment,
Department of Environmental Health, School of Public Health, Harvard University, Boston, MA
Dr. Kathleen Weathers, Senior Scientist. Gary Institute of Ecosystem Studies, Millbrook, NY
SCIENCE ADVISORY BOARD STAFF
Dr. Holly Stallworth, Designated Federal Officer, U.S. Environmental Protection Agency,
Science Advisory Board (1400R), 1200 Pennsylvania Avenue, NW, Washington, DC, Phone:
202-564-2073, Fax: 202-565-2098, (stallworth.holly@epa.gov)
-------
Enclosure B
CASAC Oxides of Nitrogen (NOx) and Sulfur Oxides (SOx) Secondary
Review Panel Consensus Responses to Charge Questions
Chapter 3: Considerations of Adversity to Public Welfare
1 What are the Panel's views on the definitions of adversity that are appropriate to
consider in determining what constitutes adversity to public welfare relative to the NOx
and SOx secondary standards7
Ecosystem services provide a framework to characterize and describe how changes in
ecosystem structure and function affect public welfare, even if they cannot be specifically
quantified. The link is well-documented between the selected ecosystem effects indicator, acid
neutralizing capacity (ANC), and the public welfare effects of lost value of recreational fishing,
biodiversity, and habitat. Fish populations (and in some cases whole species) become
unsustainable in lakes and streams with decreases in ANC levels caused by elevated inputs of
acidic deposition. The text mentions non-use values several times, but it would be helpful to
make explicit that this includes values for the preservation of habitat and biodiversity that are
independent of human use value. These services generally fall into the category of cultural
services. More could be done to explain and characterize the qualitative links between acidic
deposition and lost ecosystem services that are known and documented but cannot be specifically
quantified for a specific amount of acidic deposition. While it is clear that the total value of these
services is large; what is important to convey is the degree to which they are diminished at
current acidic deposition levels.
Evidence of community, local and state actions to decrease acidification is informative
regarding adversity even though such evidence doesn't provide specific estimates of welfare
changes. Also including federal actions, such as the Title IV (Clean Air Act Amendments of
1990) program, to address acidification would be appropriate here.
Chapter 4: Addressing the Adequacy of the Current Standards
2 What are the Panel's views on staff's approach to translating the available evidence and
risk information and other relevant information into the basis for reaching conclusions
on the adequacy of the current standards and on alternative standards for consideration?
a) In light of the Panel's views on the appropriate definitions of adversity to public
welfare (see Chapter 3), do you agree that the current levels ofNOy and SOx deposition
are adverse to public welfare?
Yes, the Panel agrees that current amounts of NOy and SOX deposition are adverse to
public welfare especially with regard to effects on aquatic ecosystems in acid-sensitive regions in
various parts of the United States. The Panel also agrees with EPA's historical interpretation that
-------
air-pollution-induced effects on ecosystems should be considered "adverse to public welfare"
whenever these effects include "disruptions in ecosystem structure and function" that are considered
important to the people of this country.
3 Has staff appropriately applied this approach in reviewing the adequacy of the current
standards and potential alternative standards?
Yes, the panel finds that EPA staff has appropriately reviewed the adequacy of the
current standards and potential alternative standards. The current NOX and SOX standards were
designed to protect vegetation against exposures to SOX and NOX Thus, the current standards
address only a fraction of the total nitrogen and sulfur compounds that are causing adverse
effects on aquatic ecosystems, and these standards are not designed to protect ecosystems from
acidic deposition. None of the elements of the current NAAQS standards - indicator, form,
averaging time, and level - are suitable for addressing the long-term (multi-annual) cumulative
acidification effects of total atmospheric loads of total reactive nitrogen and sulfur on aquatic
ecosystems.
The ISA and REA for the current review (as summarized in Chapters 2 and 3) make it
clear that current ambient concentrations of airborne nitrogen and sulfur compounds (including
not only NOy and SOX. as asked in Charge Question 2 but also ambient NHX as well as organic
forms of nitrogen) are now causing significant "disruptions in the structure and function of aquatic
ecosystems" in various acid-sensitive regions of the United States.
4. Has staff appropriately acknowledged the potential beneficial effects of nitrogen inputs
info nutrient limited ecosystems, while maintaining the focus of the review on preventing
adverse effects in nitrogen sensitive ecosystems?
While the potential beneficial effects of nitrogen inputs into nutrient-limited ecosystems
have been acknowledged, the tone and emphasis given has not been appropriately balanced. As
an example, the last few lines of page 4-45 in Chapter 4 and especially the first four words, may
suggest the potential benefits to be very limited: "In certain limited situations, additions of
nitrogen can (word inserted) increase rates of growth, and these increases can have short-term
benefits in certain managed ecosystems...."
A better balanced presentation of these same ideas could read as follows:
"Most ecosystems in the United States are nitrogen-limited, so regional decreases
in emissions and deposition of airborne nitrogen compounds will lead to some
decrease in growth of the vegetation that surrounds the targeted aquatic system.
Whether these changes in plant growth are seen as beneficial or adverse will
depend on the circumstances, fncreased carbon sequestration due to increased
growth in N-limited ecosystems may be the most significant category of potential
beneficial effects of N deposition."
10
-------
Carbon sequestration is not addressed in the Policy Assessment. Carbon sequestration is
implied, however, by the inclusion of climate-related issues in Table 3-1 on page 3-11. As
indicated above, increased carbon sequestration due to increased growth in nitrogen-limited
ecosystems may be the most significant category of potential beneficial effects of nitrogen
deposition.
While the Policy Assessment and supporting documents acknowledge the possibilities of
beneficial effects, they tend to minimize them. The panel believes that while such unintended
effects by no means justify continuing current levels of air pollution, a balanced document
should discuss these unintended effects more thoroughly.
Chapter 5: Conceptual Design of an Ecologically Relevant Multi-pollutant
Standard
5. What are the Panel's views on staff's revised conceptual framework for the structure of a
mullipollulant, ecologically relevant standard for NOx and SOx? To what extent does the
Panel agree that this suggested structure adequately represents the scientific linkages
between ecological responses, water chemistry, atmospheric deposition, and ambient
NOxandSOx?
With some exceptions noted below, the revised conceptual framework and structure of
the proposed standard(s) are well-thought out for addressing various components and
connections between these components (ecological effects, atmospheric wet and dry deposition,
atmospheric concentrations of NOy and SOX, and surface-water chemistry).
For example, the framework and the structure "take into account" the reduced ambient
NHX and its deposition in designing the AAPI (atmospheric acidification potential index). The
revised treatment of ammonia and deposition of reduced nitrogen is an improvement over the
first draft in that AAPI will reflect periodic changes in NHX concentrations. Emissions of
ammonia (which is currently an unregulated air pollutant) and resulting ammonia and
ammonium concentrations and deposition amounts are expected to increase over the next few
decades because of increased food production and increased activity in CAFO sources (confined
animal feeding operations) in the United States.
The conceptual framework for the proposed multipollutant ecologically relevant standard
forNOx and SOX is sound with considerable support from the scientific literature on how the
generation of strong mobile acids results in the acidification of soils and water. Some of the
information, however, is not correct or is incomplete. For example, the discussion of sources of
nitrate during snowmelt is incorrect in that it suggests that most of the nitrate released is of
atmospheric origin. However, most nitrate mobilized during snowmelt is derived from
nitrification in the soil itself. Also the assumptions associated with atmospheric sulfur input
being equal to drainage water losses are not correct. For example, the soil can serve as a
substantial source or sink of sulfur depending upon soil properties and history of atmospheric
sulfur sources.
II
-------
Even though the AAPI conceptual framework is sound and useful in principle, further
evaluation of robustness is required. One way to evaluate robustness of the AAPI framework is
by using sensitivity and/or uncertainty analysis, as discussed in our responses to Charge
Questions 14 and 21. The AAPI can also be tested by the use of time series chemistry data.
Where data are available, one could use the AAPI to estimate a functional relationship between
AAPI and changes in SOX and NOy concentrations. The values of other components of the AAPI
(Q5 Neco, [BC]0, LNHx, TNOy and TSOX) have already been estimated by EPA or can be
determined from measured values. The recommended analysis of historical data should be
carried out at more than one location. The changes in predicted AAPI should more or less match
the changes in ANC (perhaps with some lag).
Notwithstanding these concerns, the proposed structure adequately represents the
scientific linkages between ecological effects, surface water chemistry, atmospheric deposition,
and ambient levels of NOy, SOX, and NHX.
6. What are the Panel's views on the appropriateness of considering a single national
population ofwaterbodies in establishing standards to protect against aquatic
acidification? What are the Panel's views on consideration of alternative subdivisions of
the U.S. to identify/ the spatial boundaries of populations ofwalerbodies and acid-
sensitivity categories, specifically
a) the use ofEcoregion HI areas to aggregate waterbodies ?
b) the use of ANC to further aggregate Ecoregion 111 areas into different categories of
sensitivity?
c) the relative appropriateness of the suggested methods for categorizing spatial
boundaries of sensitivity, e.g one nation, binary sensitive/less-sensitive classes, cluster-
analysis based sensitivity classes, and individual ecoregions?
The justification, logic, and necessity of the spatial grouping classifications were not
clear to the panel. The ecoregions approach has conceptual appeal, but the rationale and
limitations for classification and aggregation methods must be better articulated for all options
described in the Policy Assessment before the CASAC Panel can provide meaningful advice.
The first approach (option 1). which considers the whole United States as one unit, has
the advantage that it provides for a single deposition metric and is simple and easy to use.
However, the single-region approach also has many weaknesses (e.g., over protection for the
least sensitive areas and under protection for areas that are most sensitive necessitates having a
system with higher spatial resolution) and is probably not a desirable approach. Nonetheless, the
panel finds it useful to include discussion of this option for the overall context. On the other
hand, the option 2d, which includes 85 ecoregions, may provide an unnecessary amount of
complexity, but future analyses could provide support for such a choice. The use of clustering is
also conceptually appealing, although the optimal number of sensitivity categories and the
degrees of protection that would be provided under the different sensitivity categories are not
clear. It does appear to strike a more reasonable balance between oversimplification and
12
-------
unnecessary complexity. The use of ANC appears to be a reasonable basis for grouping
ecoregions into a relatively small number of categories, each containing surface waters with
similar inherent sensitivities to acidification. This approach is consistent with the overall goal of
developing an ecologically relevant secondary standard to protect sensitive surface waters from
further acidification and decrease acidifying deposition to levels that will allow those water
bodies (that have been deleteriously impacted by acidic deposition) to recover as indicated by
increasing ANC values.
The Panel recommends that the final Policy Assessment include a more detailed
description of the clustering approach and other options, along with clear illustrations of the
advantages and disadvantages of the recommended options.
7. What are the Panel's views on the appropriateness of the critical loads that form the
basis for the population assessment to determine deposition metrics7
Using the concept of critical loads is logical and appropriate for development of a
secondary standard for the biological effects of NOy SOX and NHX. This approach links
concentrations of the atmospheric oxidized forms of nitrogen and sulfur with N & S deposition
and their acidifying effects on aquatic ecosystems and includes consideration of chemically
reduced forms of atmospheric N.
a) What are the views of the Panel on the appropriateness of generalizing thef-factor
approach to apply to lakes and streams in the Western US and other portions of the
Eastern U.S.
The f-factor approach is a reasonable initial approach to evaluate the response of aquatic
ecosystems to changes in atmospheric deposition. However the f-factor approach is based on
steady-state calculation but ecosystems are simply not at steady state. Ultimately, it would be
useful to apply dynamic models as management tools to evaluate effects of atmospheric
deposition on non-steady state ecosystems.
Differences between the use of MAGIC and the SSWC methods to determine background
concentrations of base cations are not adequately described in the Policy Assessment. The
proposed procedures and differences between the two approaches need to be described more
clearly
b) What are the views of the Panel on the Jittering criteria used to remove lakes and
streams that are naturally acidic or not sensitive to atmospheric deposition7
It is justifiable to exclude in advance water bodies impacted by mine drainage. It is not
clear, however, why water bodies with low background ANC and high concentrations of
naturally occurring organic acid are, likewise, excluded from further consideration since these
are often very sensitive water bodies. The rationale for this approach needs to be better explained
with examples given, with some discussion of the implications for eliminating these water
13
-------
bodies. The panel needs this information before it can fully and meaningfully respond to this
charge question.
8 What are the Panel's views on the suggested methods for determining appropriate values
of reduced nitrogen deposition in establishing NOx/SOx tradeoff curves7
The proposed approach is reasonable and utilizes the available knowledge on levels and
distribution of reduced N based on the CMAQ outputs. Potentially the NADP chemistry and
PRISM precipitation results could also be utilized. Due to the high deposition velocity of NHs,
steep concentration gradients near the NHs source areas can be expected. Therefore averaging
N^ concentrations over larger areas may lead to missing smaller areas where NHs
concentrations may be elevated with potentially high ecological effects. Consequently, option
"2'" is preferable since it allows for additional spatial refinement of sensitive areas to reflect the
spatial and temporal heterogeneity of NHX deposition. A better understanding of spatial and
temporal distribution of reduced N5 especially NH3> in the United States is critical. Realizing that
estimates of chemically reduced N deposition are viewed as highly uncertain, efforts should be
continued to assure the nationwide monitoring of Nred, especially in remote areas.
9 What are the Panel's views on the revised characterization of the deposition transference
ratios (TNOy and TSOx)?
A major concern with Twoy and Tsox is that although they are the critical links between
NOy and SOX ambient concentrations and their deposition, they are derived using a model that
has not been thoroughly evaluated for its ability to accurately simulate N and S deposition
because of lack of measurements of the required concentration and deposition components. It is
recommended that EPA evaluate the stability of these ratios using different models, emissions
and meteorological conditions. It is recommended to calculate these ratios for the following
model simulations (in addition to what has already been done):
• CMAQ and CAMx models (it is acceptable, in fact preferable, to use different emissions
and meteorological conditions)
• Different model grid resolutions (36-km v/s 12-km or even 4-km, if available)
The CMAQ TSOX calculation could also be evaluated using a combination of measured
wet deposition data from NADP and the measured concentrations and estimated dry deposition
of SO2 and pS04 from CASTNET.
The draft Policy Assessment notes the possibility of significant amounts of sulfur and
nitrogen deposition in the forest ecosystems in the coarser particle mode and further notes that
CMAQ may not adequately account for coarse particle sulfate deposition. At the same time, most
currently available measurement programs do not specifically quantify coarse mode sulfate or
nitrate concentrations or deposition, so there are only very scant measurement data with which to
evaluate related CMAQ estimates. It is not clear how important coarse particle depositoni is and
how it should/would be addressed. The Panel requests more clarification on this issue.
14
-------
On a related topic, the Panel suggests that the Agency consider the feasibility of
calculating an alternative deposition transfer ratio for oxidized nitrogen, using a combination of
(or perhaps the sum of) nitric acid and particulate nitrate, as an alternative to using NOy. A
possible advantage of this approach is that nitric acid is the component of NOy that deposits most
efficiently, and correlates best with total oxidized N deposition. Consequently, the resulting total
deposition estimate would be less dependent on CMAQ model performance. A second possible
advantage is that this calculation (as well as the TSOx calculation) could be made using currently
available and relatively low-cost CASTNET filter pack measurements, without a need to
establish a large new network of continuous NOy and SC>2. A disadvantage of this approach is
that while CASTNET measurements of total (gas + particle) nitrate are considered reliable, the
CASTNET measurements of the separate HNOs and p-NOs components are subject to large
sampling artifacts.
Estimates of total oxidized N deposition calculated using the original TNOY method and
the suggested alternative approach could be evaluated against both CMAQ estimates of total
deposition as well as wet deposition measurements from the NADP plus dry deposition estimates
from the CASTNETnetwork. It would also be important to consider whether the alternative
approach would perform as well as the original TNOY when calculated over broad spatial scales,
and over long time periods when NOX emissions and NOy species compositions may change.
As an alternative approach, EPA should attempt to further evaluate the stability of
the TSOX and TSOX ratios over time and space recognizing that these ratios are a function of both
air concentrations and deposition velocities. One possibility would be to use information from
other sources (e.g., CASTNET) to make some comparisons among air concentrations for these
chemical species with respect to their modeled deposition velocities and resultant estimated
deposition where such data are available. The Panel recognizes that the suite of chemical species
that can be used in this analysis is less extensive than that modeled in CMAQ. If these ratios
obtained from other data sources show substantial variation over time or space, it would be
useful to evaluate the relationship between meteorological and/or emissions sources.
10 What are the Panel's views on staff's conclusion that an averaging time of 3 to 5years is
appropriate given the AAPIform of the standard7
The EPA staff makes a good case for using the averaging time of three to five years and
the panel agrees with that recommendation.
7 7 What are the Panel's views on the preliminary staff conclusions regarding alternative
target ANC levels that are appropriate for consideration and the rationale upon which
those conclusions are based7
Based on the available scientific data, the range of target ANC values considered in the
Policy Assessment is appropriate, i.e., 0, 20, 50 and 100 ueq/L as target levels. These values
encompass the range of sensitive ANC classes for surface waters in the literature, and there is a
range of biological responses corresponding to this range of ANC levels. There will likely be
15
-------
biological effects of acidification at higher ANC values within this range, and there are relatively
insensitive organisms that are not impacted at ANC values at the low end of this range. Adverse
effects of acidification on aquatic biota are fairly certain at the low end of this range of ANC and
incremental benefits of shifting waters to higher ANC become more uncertain at higher ANC
levels. There is substantial confidence that there are adverse effects at ANC levels below 20
u.eq/L. and reasonable confidence that there are adverse effects below 50 ueq/L. Levels of 50
u.eq/L and higher would provide additional protection, but the Panel has less confidence in the
significance of the incremental benefits as the level increases above 50 jxeq/L. As indicated in
the draft Policy Assessment, there are clear and marked biological effects at ANC values near 0
ueq/L, so this is probably not an appropriate target value for the AAPI. At a target value of 20
ueq/L, aquatic biota experience acidification effects. Moreover, at this level of ANC many
surface waters experience episodic acidification and associated biological effects. As a result,
target ANC values of 20 to 100 ueq/L are in the range of appropriate values, while recognizing
that there is additional protection at 50 to 100 ueq/L.
a) In light of the Panel's views on the appropriate definitions of adversity to public
welfare (see Chapter 3), what are the Panel's views on the appropriateness of the
information related to adversity considered by staff in evaluating alternative target
ANC levels?
The information on adversity to public welfare associated with the effects of aquatic
organisms and ecosystems at different levels of ANC is appropriate given the available literature.
The reduction in or loss of sensitive species that would otherwise have been present in that
ecosystem is an appropriate pointer to adversity to public welfare. There is relatively little
information on the temporal biological response of acid-impacted aquatic ecosystems to marked
decreases in acidic deposition. Most of the information on biological response to acidification is
developed from spatial data. It may be useful to emphasize that it is unclear if the biological
patterns observed for spatial data of varying ANC will similarly occur temporally in surface
waters following increases in ANC due to any future decreases in nitrogen and sulfur deposition.
12 What are the Panel's views on the approaches considered by staff for assessing
alternative target percentages of water bodies for protection at alternative ANC levels?
This question is difficult to address without specifying the filtering criteria for the
watersheds at specific ANC thresholds. As noted in our response to Charge Question 7b, the
rationale for the filtering criteria should be better explained. It would be helpful to see an
analysis of the implications of different choices of the filtering criteria for the target percentages.
It is difficult to suggest target percentages without more information on subdivisions of the
United Slates to be used and the distribution of ANC values in these subdivisions. Since effects
at current deposition levels are adverse, the target should be a higher percentage than is currently
adversely affected in sensitive areas.
16
-------
The DL factors, which clearly are numerical indices of some kind, should either be
formally defined in the form of equations or it should be made clear how the numerical values
for them presented in Tables 5-12 and 5-13 were derived.
Chapter 6: Co-protection for Other Effects Using Standards to Protect Against
Aquatic Acidification
13. What are the Panel's views on the utility of the additional analyses of co-protection
benefits to inform consideration of alternative levels of the standard7
The analyses and conclusions in Chapter 6 are important because the decision to focus on
the effects of acidification on aquatic ecosystems means that in the current standard setting
process, other important effects on ecosystems (documented in the ISA), are not being explicitly
taken into account. To the extent that standards set to protect against effects of acidification on
aquatic ecosystems result in decreased amounts of nitrogen and sulfur deposition there may be
additional beneficial and detrimental effects to other ecosystems. It is important to acknowledge
these even if they are not quantified.
The analyses reported in Chapter 6 are adequate for this purpose, but the interpretation of
the conclusions could be broadened. One analysis suggests that sensitive terrestrial systems
located in the same watersheds with sensitive aquatic systems would be protected by the
deposition levels that would be needed to protect the aquatic systems. A relevant question then is
the extent to which sensitive terrestrial ecosystems are co-located with sensitive aquatic systems
throughout the country.
Similarly, even though the standard would not decrease N deposition to the extent
required to meet the target share of the TMDL in the Chesapeake watershed, the discussion could
say more about what percentage of the target TMDL might be achieved.
The discussion in this chapter should acknowledge that the level of protection from
undesirable effects of N deposition in terrestrial ecosystems is not addressed in this analysis and
remains uncertain, especially in the arid and semi-arid ecosystems of the Southwest. Negative
effects of N deposition on lichen communities are observed in some locations at very low
amounts of N deposition.
Introduction of mobile sulfate or nitrate anions into acidic soils (whether naturally acid or
acidified by pollution) can result in near instantaneous acidification of waters, whereas
acidification of soils is a long-term process occurring over decades or longer. Similarly,
recovery of surface waters from acidification could happen relatively quickly if mobile sulfate
and nitrate are removed, but recovery of acidic soils is highly questionable as soils in humid
systems naturally acidify but do not spontaneously become less acid. The rate of acidification of
soils should decrease with decreased atmospheric inputs of sulfur and nitrogen, however.
17
-------
Chapter 7: Evaluation of Uncertainty and Variability in the Context of an AAPI
standard, including Model Evaluation, Sensitivity Analyses, and Assessment of
Information Gaps
14 What are the Panel's views on the follow ing:
a. The degree to which the chapter appropriately characterizes the potential role of
information on uncertainty, sensitivity, and variability in informing the standards?
The discussion of uncertainty and sensitivity analysis is much improved compared to the
first draft in consolidating the uncertainty, sensitivity and variability in the proposed indices;
however, a more complete quantitative analysis of uncertainty is needed for the AAPI. The
reason it is particularly important to conduct a comprehensive uncertainty and sensitivity
analysis for the AAPI form of the standard is that many of the components of the AAPI cannot
be evaluated because of lack of measurements. One way to gain confidence in using the AAPI is
to examine quantitatively how sensitive the SOX and NOy response surfaces are to different
components of the AAPI. The Panel's major comments on this chapter include:
1. Summarize the general framework applied for uncertainty analysis, e.g. the World Health
Organization framework used in other NAAQS assessments.
2. Extend the uncertainty analysis beyond the components and examine the propagation of
the uncertainties through the entire AAPI, with particular focus on how uncertainties
impact the levels of NO2 and SO2. Include constraints from available observations.
3. Aggressively pursue the identification and reduction of biases in the CMAQ model that
are relevant to the AAPI.
It is recognized that it is difficult to quantify the uncertainties associated with the AAPI.
Nevertheless, the Panel recommends that EPA pursue more complete uncertainty analysis
focusing on the overall, end-to-end, cumulative estimation, including the possible application of
Monte Carlo techniques. EPA may want to assess the report presented by the Electrical Power
Research Institute at the October 6-7, 2010 meeting.
While there is significant uncertainty associated with model calculations both in CMAQ
and the MAGIC/SSWC, there is a considerable amount of empirical observations that provide
constraints on the magnitude of these uncertainties. The combined use of uncertainty propagation
and the observational constraints should be pursued.
b The appropriateness and completeness of the evaluation of CMAQ model performance
and sensitivity to critical inputs7
The inclusion of comparisons of CMAQ and CASTNET results and the related
discussion on the CMAQ limitations in Chapter 7 is very helpful. It should be useful for future
improvements of CMAQ. As indicated, the "sensitivity of CMAQ derived deposition
transformation ratios to changes in emissions and treatment of chemistry" is not yet completed.
This should be a high priority for EPA.
18
-------
The performance evaluation of the CMAQ model is incomplete. A more complete
evaluation with measurements is key to improving confidence in the calculations of the AAPI.
The overestimation of SOi is a significant systematic error that may lead to a bias and may have
a major impact on the estimated deposition and the AAPI overall. A figure similar to Fig 7-5 for
CMAQ-CASTNET comparison for SC>2 could be very revealing because the model cannot be
directly evaluated for dry deposition due to the lack of measurements, and evaluation of the
ability of the model to represent ambient levels can serve as a proxy for its ability to represent
dry deposition. It is recommended that following evaluations (using daily or weekly averaged
quantities, and showing mean normalized statistics as well as normalized mean statistics) be
performed to assess the model performance:
1. Model performance for simulating nitric oxide, nitrogen dioxide, sulfur dioxide, nitrate,
ammonium and aerosol nitrate, ammonium, and sulfate for different networks for which
the data are routinely available,
2. Model performance for capturing observed levels of wet deposition of sulfate, nitrate, and
ammonium using the National Atmospheric Deposition Program (NADP) network
3. Model evaluation using the continuous measurements of nitric oxide, nitrogen dioxide,
nitric acid and NOy from the SEARCH network in the southeastern United States.
EPA should use the results from the model performance evaluation to describe the uncertainties
and limitations of CMAQ simulation of total deposition of reactive nitrogen and sulfur oxides
more completely,
c The utility of the analyses of temporal and spatial variability in the deposition
transference ratios (TNOy and TSOx)?
The figures in the Policy Assessment showing the spatial pattern of TNOyand Tsox are
insufficient to provide the reader with adequate understanding of the spatial variation of the
transfer ratios and how they are linked to acid-sensitive ecosystems. The meaning and
implications of the box-and-whisker plots are not obvious. The terms "stiff5 and "stiffness" are
not explained. The TNQy and TSOX are critically important to the APPI calculations, are entirely
dependent on CMAQ simulations, and are impossible to fully evaluate with currently available
measurements. It is therefore important to demonstrate that their spatial patterns appear
reasonable, that the resultant deposition calculations are consistent with (limited) available
measurements, and that these ratios remain stable as emissions, concentrations and deposition
rates are changed over time.
15 What are the Panel's views on the insights provided by the AAPI sensitivity analysis
including.
a The evaluation of elasticities of response7
b The multivariable ANOVA analysis7
Evaluation of elasticity of response is a good way to get an initial estimate of the AAPI
sensitivity to its components. However, the Panel recommends doing this analysis also for the
SOx and NOx response surfaces to meet a particular standard, as those are the quantities for
which compliance with the standard would be determined. The sensitivity analyses should
19
-------
include a larger range of scenarios, such as the sensitivity to the 40% SO2 over-estimation in
CMAQ.
A summary is needed for the relative sensitivities of the various parameters that make up
the AAPI to show the parameters of the AAPI that have the most and least impact and their
influence on confidence levels, e.g.. the role of non-atmospheric inputs, including base cation
weathering and runoff rates. Such information should be used in driving research and monitoring
efforts by EPA.
16 What are the Panel's views on the discussion of uncertainty in the critical loads models
including MAGIC and SSWC?
There is clearly uncertainty associated with model calculations. However, the Policy
Assessment does not acknowledge that there is a considerable amount of empirical field data to
support application of this secondary standard. Through monitoring studies there are about 30
years of observations providing a quantitative understanding of the nature and extent of soil and
surface water responses to decreases in atmospheric deposition. Through these observations and
some field-based experiments, there is also a good understanding of the compensatory response
of ANC to changes in concentrations of sulfate and nitrate. These empirical data should be used
to evaluate the quality of the AAPI calculations and to support the justification and target
parameter values for the AAPI.
There have been limited uncertainty analyses of the ecological models MAGIC and the
SSWC. Some uncertainty analysis for MAGIC is presented in the REA. The panel encourages
EPA to continue uncertainty analysis of these ecosystem models in the future
Beyond uncertainty analysis, efforts should also be made evaluate model structure and
compare this to the structure of other models available for use. Efforts should be made to
• test models, although it is difficult to test steady state models.
• improve and test the Neco calculation by using observed data and improved model
simulations of N-deposition.
• compare results from steady-state with dynamic models to obtain a sense for the time
scale to achieve target ANC values.
• evaluate the effects of variation and changes in climate on model calculations.
Some of these evaluations may be feasible within the current NAAQS review cycle,
while others will help to refine the standards in future reviews.
17 What are the Panel's views on the areas for future research and data collection outlined
in this chapter, on relative priorities for research in these areas, and on any other areas
that ought to be identified?
The future research areas outlined in Chapter 7 are appropriate However, there arc other
areas that should be considered for future research and data collection.
20
-------
• Key uncertainties identified in the qualitative uncertainty analysis of Chapter 7 including
pre-industrial sulfate, nitrate, base cation, and ANC levels, dry deposition, and ecological
indicators, with particular emphasis on parameters that were found to be most influential
from the analyses conducted as part of this NAAQS review.
• There is a need to improve understanding of the sources, atmospheric dynamics ambient
concentrations, bi-directional transport and deposition of both chemically reduced and
organic forms of nitrogen.
• Efforts should be made to develop dynamic models to simulate effects of acidic
deposition on soil, drainage waters and biota, to test these models and to apply these as
tools in determining critical loads. Research should be conducted comparing results from
steady-state and dynamic models.
• There is a need for research to improve the linkages between atmospheric chemistry and
transport models with watershed models. Atmospheric models typically have relatively
large spatial scales and simulate over relatively short temporal scales. Watershed models
simulating acidification of soil and surface waters, in contrast, have small spatial scales
and simulate processes over long temporal scales. It is important to quantify the subgrid
scale variability in atmospheric deposition and how this variability can be addressed in
simulations of watershed response to changes in atmospheric deposition.
• It is essential that surface water monitoring programs be maintained and soil and
biological monitoring programs be strengthened.
• There needs to be improvements of tools and models to predict nitrogen retention in
watersheds.
• There is a need to better understand the compensatory response of naturally occurring
organic acids to decreases in acidic deposition.
• Since the current assessment was unable to address endpoints other than aquatic
acidification, there is a need for research regarding endpoints such as terrestrial
acidification and aquatic system nutrient enrichment.
• EPA should organize a future workshop to further enumerate and prioritize research
identified in the Integrated Science Assessment, Risk and Exposure Assessment, and the
Policy Assessment for the secondary MAAQS review for NOX and SOX-
The CASAC Panel will discuss priorities among these research and data collection needs at
its planned meeting in February 2011.
Chapter 8: Monitoring
18 What are the Panel's views on using an open inlet to capture all paniculate size fractions
for the purpose of analyzing for sulfate7 What is your opinion on using existing
CASTNETfilter packs as a future Federal reference method for sulfate?
21
-------
As a prefacing comment on these monitoring questions (18-20). this Panel is pleased to
learn that the Agency plans to consult with CASAC to identify the most appropriate monitoring
approaches for this NAAQS, and we expect that more informed responses to these and other
monitoring questions will be provided in that process. In conducting that monitoring review, we
encourage the Agency to emphasize not just compliance determination, but the multiple
monitoring objectives outlined in chapter 8 of the Policy Assessment, and to consider whether
some of those objectives might be most effectively addressed by enhancement of and
coordination among existing monitoring programs. In addition, we recommend that the
membership of the CASAC monitoring and methods subcommittee be enhanced for that review
by adding individuals with expertise in conducting deposition measurements, as well as in
assessing the effects of deposited S and N pollutants on aquatic and/or terrestrial ecosystems.
The Panel is not opposed to considering the use of open-faced samplers, and possibly the
CASTNET sampler in particular, as a possible federal reference method (FRM) for paniculate
sulfate, as a component of the multiple pollutant measurements needed to determine compliance
with this secondary standard. It should be recognized, however, that the inclusion of coarse
particle sulfate (excluded in sulfate measurements by more commonly deployed fine particle
samplers) will not by itself provide any information on how much of the sulfate is present in
coarse mode particles and which would contribute proportionately more to deposition than their
fine particle counterparts. It should also be noted that inclusion of coarse particles, which tend to
be alkaline, could lead to formation of positive sampling artifacts from reactions with acidic S
and N gasses.
Since the open-faced CASTNET samplers also measure particulate nitrate, and since
coarse particle nitrate can contribute to nitrogen deposition, especially in areas influenced by
marine aerosols, consideration should be given to evaluating the quality of CASTNET filter pack
methods for particulate nitrate as well. CASTNET samplers also measure sulfur dioxide and
nitric acid, and so an alternative (to the TNOY) nitrogen deposition transfer ratio could be
developed (see response to question 9) based on combined measurements of HNOs and pNOs.
As such, all the measurements needed to determine compliance with this standard could be made
by the existing CASTNET methods, which could be enhanced by adding new sites in acid and
nitrogen sensitive regions, and by adding more detailed measurements like continuous NOy,
SO2, etc. at a subset of those sites to better address important objectives other than compliance. It
would be unprecedented to have a compliance network operated by an EPA contractor (as
CASTNET currently is) rather than by state agencies. And as indicated above, there are also
serious concerns with the quality of CASTNET HNOs and particulate nitrate (p-NOs) data. For
these reasons, it would be helpful if the proposed AAMMS review of monitoring methods for
implementing this standard includes consideration of both continuous and filter pack
measurements of all the relevant S and N species, as well as other approaches like passive
samplers and diffusion denuders.
22
-------
19 What are the Panel's views on requiring measurements of ammonia and ammonium to
assist implementation of the standard7
Although NHX deposition estimates could be supplied by CMAQ model output,
additional NH4 and especially NHs measurements would be extremely valuable for supporting
and implementing the standard both directly - to quantify an unregulated but varying element of
the compliance metric - and indirectly, to help evaluate and improve emissions inventories and
CMAQ model performance. NH4 measurements are currently available from the CASTNET and
(urban) Chemical Speciation (CSN) networks, and could conceivably be added to the
Interagency Monitoring of Protected Visual Environments. NHs measurements are currently
much sparser and are more critically needed, not only for assessing a key parameter in the AAPI
used in the proposed secondary standard but also for better understanding sources and trends of
PM25, regional haze, and sources and effects of N deposition on nutrient enrichment. The
passive NHs sampling approach currently being deployed in the AMoN Network (at a subset of
NADP sites) appears promising and would benefit from more dedicated EPA funding support.
Given the current state of development of instruments and methods and their costs, the Panel
does not believe that comprehensive ammonia monitoring should be required. The panel
strongly encourages, however, that monitoring be conducted at a minimum using the passive
NHs sampling approach currently being deployed in the AMoN Network.
20. What are the Panel's views on having a subset (e g, 3-5 sites) of monitoring stations in
different airsheds that measure for the major NOy species; nitric acid, true NOz, NO,
PANandp-NO3?
Appropriate design of a network required to determine attainment with the proposed
standard and to inform future reviews will require a major effort. An appropriate design will be
impacted by the choices made in formulating the standard, including the form, indicator,
ecoregion approach and fraction of lakes protected. Insufficient information is available at this
time, and we commend EPA staffs desire to involve the CASAC monitoring and methods
subcommittee in addressing the monitoring related issues. Some initial thoughts are provided
below.
As suggested in the response to question 18, a slight modification to the proposed
calculation of the deposition transfer ratio (currently expressed as TNOY) for oxidized nitrogen
deposition, might allow the use of a modestly expanded version of the existing CASTNET
network to determine compliance with the proposed secondary SOX/NOX NAAQS.
Disadvantages of this approach include the loss of valuable temporal resolution in the weekly
aggregated CASTNET filter pack data, uncertainties in the portioning between nitric acid and
particulate nitrate, and the exclusion of important NOy components like NO, NC>2 and PAN,
which better reflect the sources of oxidized nitrogen emissions, which eventually contribute to N
deposition downwind, and/or which may represent important components of total deposition at
some locations.
23
-------
For these reasons, it is important for implementing this secondary standard that the
existing monitoring network be expanded by adding sites in different kinds of sensitive areas,
and refined by adding more detailed supplemental measurements at a small subset of these sites.
Valuable supplemental measurements would include continuous NOy and trace SOi, PAN, true
NC»2, and possibly nitric acid, as well as p-NOs, and NHs that would be measured continuously.
if practicable, or by integrated methods. Possibly there will be opportunities to add CASTNET
filter packs, passive samplers, denuder analyses, and/or other supplemental measurements to
several of the existing or planned rural National Core (NCore) Multipollutant Monitoring
Network sites. Such measurements would not only help respond to the multiple objectives for
this secondary standard outlined on page 8-1 of the PA, they would also be of great value for
improving data analysis and modeling assessments of sources, atmospheric chemistry, transport
of and the effectiveness of control strategies for ozone, PM25 and regional haze.
Chapter 9: Conclusions
21 What are the Panel's views on the overall characterization of uncertainty as it relates to
the determination of an ecologically-relevant multi-pollutant standard for NOx andSOx?
EPA has done a good job of qualitatively discussing uncertainties in Chapter 7 and
reviewing them in Chapter 9. As noted in response to charge questions 14-16, CASAC believes
it is important that there is further progress on quantitative analysis of sensitivity and uncertainty,
for key components of the AAPI, for the combined effect of multiple uncertainties on the AAPI,
and implications for specification of the trade-off between NOy and SOx allowable
concentrations. In Chapter 9, EPA should provide a concise summary of those key uncertainties
that are most likely to lead to bias, imprecision, or both, in the AAPI, and the implications of
such uncertainties when translating an ANC target into an associated AAPI level. For example,
given biases, should the selected AAPI be higher or lower than implied by a specific target
ANC? Given imprecision, what range of AAPI might be consistent with a particular target
ANC? EPA should conduct a more complete evaluation of the CMAQ simulations used to
calculate the deposition transfer coefficients and consider additional processes, such as internal
sulfur sources, in the AAPI.
The choice of averaging times needs to inform the variability and uncertainty analyses of
Chapter 7. This is because the range of variability and uncertainty depend on averaging time.
Similarly, the geographic scope needs to be taken into account in the analysis of variability and
uncertainty.
22 What are the Panel's views on the following.
a The insights that can be gained into potential alternative additional secondary
standards (using the AAPI form) by considering.
i Informal ion from studies on the relationship between mortality in aquatic
organisms and pH and ANC?
ii Information from studies on the relationship between fish health and/or
biodiversity metrics and pH and ANC?
24
-------
///' Information on the relationship between pH, At, andANC?
iv Information on target ANC levels identified by stales and regions, as well
as other nations'''
Each of the sources of information mentioned in the charge question both separately and
taken together provide a compelling case on the relationships between ANC and other water
quality metrics that are associated with biotic health of waters, and provide insight regarding
target ANC values. Text should be provided on the validity of spatial survey data when applied
to infer temporal relationships. Different states and nations have identified different target
levels. Some use pH, others use ANC. It will be helpful to explain and compare how these
values were developed. Chapter 9 could clearly and briefly summarize possible co-benefits and
unintended consequences of various alternatives for the standard. For example, to what extent
might a standard focused on aquatic acidification also be protective of terrestrial acidification or
aquatic nutrient enrichment? Would higher levels of target ANC provide more protection for
these other effects? These points can be made while still placing emphasis on the sufficiency of
the scientific evidence supporting the need for a revised standard to protect from aquatic
acidification.
b The appropriate role of qualitative and quantitative characterizations of uncertainty in
developing standards using the AAPI form7
Conceptually, the AAPI approach is compelling and appropriate. There are uncertainties
associated with the practical use of AAPI that should be more fully evaluated. The sensitivity and
uncertainty characterization of the AAPI needs to include not only statistical analyses associated with
specific model parameters individually, but also consider their joint effect (taking into account
covariance and dependencies) and an evaluation of possible omissions (e.g. reduced nitrogen inputs,
contribution of sulfate sources and sinks in soil). To the extent possible, biases and imprecision in
values of AAPI associated with a target level of ANC should be quantified, and these uncertainties
should be used to inform specification of ranges of AAPI associated with a target ANC that may be
more or less protective within the range of scientific uncertainty. This would lead to a family of
NOy-SOx trade-off curves associated with each target ANC for a given geographic location. A
specific standard would be set by choosing an AAPI within the range of scientific uncertainty, which
would then be associated with just one NOy-SOx trade-off curve per region. EPA staff is encouraged
to offer reasonable judgments about the range of uncertainty in AAPI for a given ANC target based
on factors difficult to quantify within the time period of the assessment, such as the preindustrial
cation weathering, the deposition transfer ratios, unmodeled factors, ancillary benefits, and
unintended consequences
c The role of considerations regarding the relationship of the standard to.
i. the time trajectory of response, e g. when specific ANC levels are likely to be
realized given a specific level of the AAPI?
Based on recent observations and dynamic model calculations, the time response to recovery
in ecosystems from decreases in acidic deposition is very slow. Because of accumulation of sulfur
in soils, it is likely that the timescale for recovery of watersheds, especially in the Southeast, would
25
-------
likely be even longer. Factors such as changes in climate and carbon dioxide concentration in the
atmosphere could affect the time trajectory, and the effects may be substantially different between
aquatic and terrestrial ecosystems.
ii the likelihood of damages to aquatic ecosystems due to episodic acidification
events given a specific target for chronic ANC?
Based on surface waters studied in the Northeast, decreases in ANC associated with
snowmelt is approximately 50 ueq/L. Thus, based on these studies, a long term ANC target level of
75 ueq/L would generally guard against effects from episodic acidification down to a level of about
25 ueq/L.
//I the levels of co-protection for terrestrial ecosystems against acidification effects
and the for aquatic and terrestrial ecosystems against effects of excess nutrient
enrichment?
There may be co-benefits for terrestrial and coastal ecosystems with respect to decreases in
methylization of mercury associated with decreases in sulfur dioxide emissions and decreases in
bioaccumulation of mercury associated with increases in pH and ANC. Aquatic ecosystems may not
be more sensitive to acidic deposition than terrestrial ecosystems. Many soil time series studies
suggest ongoing soil acidification while surface waters are recovering from acidic deposition. This
may also suggest that soil is more "sensitive" to inputs of acidic deposition than surface waters.
Levels of protection provided by the proposed standard against nutritional nitrogen effects in
terrestrial ecosystems are uncertain, especially in arid and semi-arid zones, and should be evaluated
in the future.
23 What are the Panel's views on Staff's conclusion that the existing secondary standards for
NOX and SOX should be retained to provide protection against direct adverse effects to
vegetation due to gas phase exposures7
Based on the information presented in the PA, the scientific understanding of effects from
direct foliar exposures to gaseous sulfur and nitrogen oxides has not changed appreciably, and the
existing secondary standards for SC>2 and NC<2 should be retained. The indicators, averaging times,
levels and forms of the current standards are not appropriate for addressing the (indirect) effects of
SOX and NC\ deposition to acid-sensitive ecosystems. Therefore, additional secondary standards to
protect against adverse effects from acidic deposition should be added to the existing secondary
standard.
24 In light of the Panel's views on what constitutes adverse effects to public welfare (see
Chapter 3), what are the Panel's views on
a) the degree to which current levels ofNOv and SOX deposition are adverse to public
welfare based on evidence and risk information, and information on adversity
provided in Chapters 2,3, and 4?
26
-------
Current and cumulative levels of NOy and SOX deposition have been shown to result in
environmental damage to an extent that is adverse to public welfare. The effects include acidification
of aquatic and terrestrial ecosystems and nutrient enrichment. However, the panel believes that
descriptive information about that adversity and its significance could be better and more
comprehensively articulated, and that additional discussion of the possible benefits of S and N
deposition would be helpful.
b) target values for ANC that protect against adversity to public welfare in light of the
information presented in Chapter 5 concerning levels of ANC and the ecosystem
effects associated with those target ANC levels?
There is substantial confidence that there are adverse effects at ANC levels below 20 u.eq/L,
and reasonable confidence that there are adverse effects below 50 u.eq/L. Levels of 50 u,eq/L and
higher would provide additional protection, but the Panel has less confidence in the significance of
the incremental benefits as the level increases above 50 u.eq/L.
c) factors relevant in selecting target percentages ofwaterbodies to protect at
alternative target ANC levels to protect against adverse effects to public welfare, and
weights to place on those factors?
The justification and implications of alternative spatial grouping classifications were not
clear to the panel It is not clear what is gained by the added complexities of going beyond the two
groups of sensitive and not sensitive, although there is inherent appeal to taking into account
available information about variations across eco-regions. Because there is large variability in
inherent sensitivities of water bodies to acidification effects among different regions and even within
regions, protecting a target percentage of lakes from the populations which are potentially susceptible
to acidification seems logical, however. It seems that the target should be higher than the current
percentages of sensitive water bodies that are below the target ANC.
d) alternative standards for NOx and SOx that would protect against adverse effects to
public welfare based on the AA PI form, and taking into account
• consideration of target levels of ANC (chapter 5),
The panel concurred that ANC levels from 20 to 100 were appropriate to consider.
• target percentage of water bodies to protect (chapter 5),
The panel believes that this choice was a value judgment and somewhat arbitrary. Insufficient
analysis was provided to adequately support a choice at this time The panel stresses that the target
percentage will also be influenced by whether and how the filters used to exclude lakes (e.g ,
naturally acidified lakes, for example) are applied.
• consideration of relevant uncertainties in AAPI components (chapter 7),
The Panel spent considerable time discussing how and what is necessary to characterize
relevant uncertainties in AAPI components in order to answer this and other questions about the PA.
27
-------
The current sensitivity and uncertainty analysis should be strengthened. For suggestions on specific
AAPI components see the responses to charge questions 5, 9, 14, 15, 16 and 22. The panel would
also particularly like to see some assessment of the cumulative uncertainties associated with the
complete AAPI calculation. One approach to this might be to employ available measurement data
and model calculations to compare levels and changes in AAPI estimates over the past 20 years with
concurrent ANC levels in surface waters. In some cases, individual components of the APPI could
also be compared with their measured counterparts over the same recent time period. The goal of
these syntheses and analyses would be to lend defensibility to the approach, provide broad bounds on
uncertainties, or, in some cases to provide reality checks on the components of the AAPI.
• any other potentially relevant factors, such as levels of co-protection against
terrestrial acidification and nutrient enrichment (chapter 6)?
It seems likely that a standard that decreases acidifying deposition to acid-sensitive
ecosystems would provide some co-protection benefits to acid-sensitive terrestrial ecosystems. If
attaining such a standard results in regional-scale reductions in nitrogen deposition, there may also be
reductions in plant growth rates in aquatic or terrestrial ecosystems or components of those
ecosystems. These growth rate changes might be viewed as either benefits or dis-benefits, depending
on the specific ecosystem and management objective. It is not currently possible to provide
quantitative estimates of co-protection benefits or detrimental effects, but it would be useful to
qualitatively discuss these associated effects in the final Policy Assessment document.
28
-------
Enclosure C
Compilation of Comments from Individual Members of the CASAC Oxides of
Nitrogen (NOx) and Sulfur Oxides (SOx) Secondary Review Panel
Comments from Dr. Praveen Amar 30
Comments from Dr. Andrzej Bytnerowicz 33
Comments from Ms. Lauraine Chestnut 36
Comments from Dr. Ellis Cowling 41
Comments from Dr. Charles Driscoll 53
Comments from Dr. Christopher Frey 61
Comments from Dr. Paul Hanson 67
Comments from Dr. Rudolph Husar 69
Comments from Dr. Dale Johnson 71
Comments from Dr. Naresh Kumar 76
Comments from Dr. Myron Mitchell 81
Comments from Mr. Richard Poirot 95
Comments from Dr. Armistead Russell 104
Comments from Dr. David Shaw 106
Comments from Dr. Kathleen Weathers 108
29
-------
Comments from Dr. Praveen Amar
Chapter 3: Considerations of Adversity to Public Welfare
Charge Question 1. What are the Panel's views on the definitions of adversity that are
appropriate to consider in determining what constitutes adversity to public welfare relative to the
NOy and SOx secondary standards
Chapter Three covers three areas of :a) adversity to public welfare, b) application of ecosystem
services framework (provisioning, regulating, cultural, and supporting) as a way to address
adversity to public welfare, and , c) usefulness of economic valuation approaches to
"value'Vmonetize ecosystem services, when possible. The second draft of the Policy Assessment
is a great improvement over the first draft. The definition of adversity used in this document is
derived from, and based on, recent applications of the concept by EPA in other recent
environmental policy contexts and is quite applicable to ecosystem effects from exposure to
ambient levels of SOx and NOy.
The Chapter is much improved in describing the current level of ecosystem services as well as
scale of adversity to public welfare driven by changes to ecosystem services as a function of
changes in atmospheric deposition of SOx, NOy, and potentially no changes (potentially
increases) in atmospheric deposition of reduced NHx. The Chapter presents many quantitative
estimates in dollars when economic valuation/monetizing were possible. Also, monetized
benefits of current status of ecosystem services are clearly presented in many Tables.
Specific Comments on Chapter 3:
It would be useful if public welfare/adversity was more clearly discussed for, and separately
allocated to, NOy and NHx (both through atmospheric deposition and through water runoffs)
instead of just atmospheric NOy alone, (see page 3-8; TMDL discussion for Chesapeake Bay;
this discussion should be more explicit in describing the role of NHx through water discharge
and air deposition; please also see Page 3-25, Line 15, nutrient enrichment refers there "only to
that due to NOy deposition" ).
For Figure 3-6, the range for high end of N and S deposition (300- 1,337 eq/ha-yr) for the
Western U.S. is too large and needs to be sub divided (say, in two or three parts) for finer
representation of high-end deposition levels in the West.
Sections 3.3.2 and 3.3.3 on economics framework and its role in defining adversity are very well
written.
The Section 3.3.4 on "collective action as an indicator of public preferences" correctly notes the
actions and efforts on the part of communities, NGOs, and States to reduce acidity of lakes and
streams. This Section overlooks what I believe is the most important action/effort taken so far in
the U.S. by the federal government: Title IV of the 1990 CAAA to lower SO2 emissions by 10
30
-------
million tons per year as well as NOx emissions by 2 mm tpy to address ecosystem acidification.
The value of the this national "revealed preference" should be valued/monetized at about $5
billion/year, based on $500/ton of SOi controlled and should be noted in this section.
Section 3.3.1.1 needs to be written more clearly to make the points it is "trying" to make with
references to table 3-2 and 3-3.1 found it hard to understand.
Finally, Table 3-6 (page 3-33) needs major improvement in format including column headings.
Chapter 5: Options for Elements of the Standard
Charge question 5: What are the Panel's views on staffs revised conceptual framework for the
structure of a multipollutant, ecologically relevant standard for NOx and SOx? To what extent
does the Panel agree that this suggested structure adequately represents the scientific linkages
between ecological responses, water chemistry, atmospheric deposition, and ambient NOx and
SOx?
The revised conceptual framework and structure of the proposed standard (s) is very well-
thought out for addressing various components and connections between these components
(ecological effects, atmospheric wet and dry deposition, atmospheric concentrations of NOy and
SOx, and surface water chemistry), with one major exception noted below. I had made this same
point for the first draft of the policy assessment document.
Even though the framework and the structure "takes into account" the reduced ambient NHx and
its deposition in designing AAP1 (atmospheric acidification potential index), it does so in a
manner such that future control strategies and policy options most probably will not allow EPA
to address and require reductions in U.S. ammonia emissions under proposed standard setting
structure. Ammonia emissions are currently at about 4 to 5 million tons per year. Emissions of
ammonia (which is an unregulated air pollutant) and resulting ammonia and ammonium
concentrations and reduced nitrogen deposition levels are only expected to increase by as much
as ten percent over the next few decades because of increased food production and increased
activity in CAFO sources (confined animal feeding operations) in the U.S.
Notwithstanding my concern about not addressing reduced nitrogen/NHx directly, the proposed
structure more than adequately represents the scientific linkages between ecological effects,
surface water chemistry, atmospheric deposition, and ambient levels of NOy and SOx.
Charge Question 6: What are the Panel's views on the appropriateness of considering a single
national population of waterbodies in establishing standards to protect against aquatic
acidification? What are the Panel's views on consideration of alternative subdivisions of the U.S.
to identify the spatial boundaries of populations of waterbodies and acid-sensitivity categories,
specifically:
a) the use of Ecoregion III areas to aggregate watrebodies?
31
-------
b) the use of ANC to further aggregate Ecoregion III areas into different categories of
sensitivity?
c) The relative appropriateness of the suggested methods for categorizing spatial boundaries
of sensitivity, e.g., one nation, binary sensitive/less-sensitive classes, cluster-analysis
based on sensitivity classes, and individual ecoregions?
The first approach (option 1) that considers the whole U.S. as one unit and provides for a single
deposition metric is simple and easy to calculate, but its weaknesses are too many to consider
this as the preferred approach (e.g., over protection for the least sensitive areas and under
protection for areas that are most sensitive).
The three sub-options under second option seem to have merits. However, they are based on the
concept of "Level 3 Ecoregions," which is rather poorly described in the document and 1 was
not sure how this approach divides US into 120+ acid-sensitive categories. A reference is made
to Omernik's (1987, 1995) and other works about "the analysis of the patterns and the
composition of biotic and abiotic phenomena that affect or reflect differences in ecosystem
quality and integrity " What is not explained is how "hierarchical levels are developed" at
various levels (Level I, II, 111, and IV (future)).
Between options 2a, 2b, and 2c, the approach based on cluster analysis (option 2b) seems to
provide the right balance when compared to approach that is not detailed enough (option 2a) or
detailed too much (option 2c).
Charge Question 9: What are the Panel's views on the revised characterization of the deposition
transference ratios (TNOy and T SOx)?
The policy assessment document proposes to use the output of CMAQ model to calculate
deposition transference ratios for both NOy and SOx. The CMAQ hourly predictions at the scale
of 12-km grid will be averaged to provide annual transference ratios so as to be consistent with
depositional loads derived from ecosystem models. It is not clear how to account for wet and dry
deposition of those nitrogen and sulfur species (which ones?) that are not explicitly modeled in
the CAMQ. The PAD does note the possibility of large amount of sulfur and nitrogen deposition
in the forest ecosystems in the coarser particle mode and that CMAQ's simulations do not
account for deposition in the coarse particle mode. It is not clear how big this issue is and how it
should/would be addressed.
Charge Question 10: What are the Panel's views on staffs conclusion that an averaging time of 3
to 5 years is appropriate given the AAP1 form of the standard?
The PAD makes a good case for using the averaging time of three years (Figure 5-22 on the
magnitude of coefficient of variation (CV) shows that it is less than 25%, based on CAMQ
simulations for the years 2002-2005).
32
-------
Comments from Dr. Andrzej Bytnerowicz
7. What are the Panel's views on the appropriateness of the critical loads that form the
basis for the population assessment to determined deposition metrics?
Using a concept of Critical Loads is logical and appropriate for development of a
secondary (welfare) standard for biological effects of NOX and SO x. This approach links
concentrations of the atmospheric oxidized forms of nitrogen and sulfur with N & S
deposition and their acidifying effects on aquatic ecosystems. What is important is also a
fact that the proposed approach includes reduced forms of atmospheric N as a contributor
to acidification of lakes and streams.
a) What are the views of the Panel on the appropriateness of generalizing the f-factor
approach to apply to lakes and streams in the Western U.S. and other portions of the
Eastern U.S.
The purpose of the F-factor is to obtain estimates of the pre-industrial surface water base
cation concentrations needed for calculation of critical loads. These values can be
obtained from the SSWC and MAGIC models.
At this point I am not able to adequately answer the posed question. Explanation of the
problem and graphs illustrating differences between the two approaches are not sufficient
for understanding the proposed concept.
I believe this question could be modified to: "Is the proposed methodology for obtaining
BC0 values adequately described and what is the Panel's opinion on extrapolating the
knowledge gained for the Adirondacks lakes and the Southern Appalachian streams to the
rest of the US water bodies?"
b) What are the views of the Panel on the filtering criteria used to remove lakes and
streams that are naturally acidic or not sensitive to atmospheric deposition?
It's logical to eliminate in advance water bodies impacted by mine drainage, however,
advance elimination of water bodies impacted should be considered and the rationale
better explained with examples given, preferably in a distribution form. The panel needs
this information before it can fully respond to the question. The panel is concerned about
removing lakes and streams with high concentrations of organic acids and with low
historical ANC from the analysis since these are often highly sensitive water bodies.
8 What are the Panel's views on the suggested methods for determining appropriate
values of reduced nitrogen deposition in establishing NO^/SOxx tradeoff curves?
The proposed approach makes sense and utilizes the best available knowledge on levels
and distribution of reduced N deposition, however, reliance on the CMAQ-derived values
provides high level of uncertainty (see my comments below).
33
-------
Due to a high NHs deposition velocity, steep concentration gradients near the NHa source
areas exist. Therefore averaging Nred concentrations over larger areas may lead to missing
smaller areas where NHs concentrations may be seriously elevated and with potentially
high biological and ecological effects. Therefore option 2 "allow for additional spatial
refinement of sensitive areas to reflect the heterogeneity of NHX deposition" seems to be
preferable.
As stated in previous CASAC reviews, a better understanding of spatial and temporal
distribution of reduced N, especially NHs, in the US is critical. Efforts should be
continued to assure the nation-wide monitoring of NHs in remote areas.
Additional remark: It would be good to develop similar methodologies that account for
atmospheric organic N.
1 1 . What are the Panel's views on the preliminary staff conclusions regarding alternative
target ANC levels that are appropriate for consideration and the rationale upon which
those conclusions are based?
Focusing on a range of ANC values between -50 and 50 Deq/L makes sense from a
perspective of the expected pH changes, Al toxicity and related biological effects. At
values <-50 Deq/L no further damage should occur, while at values > 50 Deq/L no more
improvement is expected.
Improved biodiversity offish populations may continue up to!60 Deq/L and therefore the
best protection would be achieved at ANC values >100 Deq/L. However, considering
that such recommendation could be impractical, the proposed ANC 50 Deq/L as a target
value seems to be reasonable and should be supported.
a) In light of the Panel's views on the appropriate definitions of adversity to public
welfare (see Chapter 3), what are the Panel's views on the appropriateness of the
information related to adversity considered by staff in evaluating alternative target ANC
levels?
Appropriate information has been provided for the aquatic ecosystems. However, I would
like to see a better discussion of what the main ANC values considered (20, 50 and 100
Deq/L) would mean to the surrounding terrestrial ecosystems in various eco-zones. That
could be discussed for such sensitive indicators and sugar maple and red spruce in the
eastern part of the country, and for lichen communities in such areas as Sierra Nevada
Mountains in the West.
12. What are the Panel's views on the approaches considered by staff for assessing
alternative target percentages of.water bodies for protection at alternative ANC levels?
This question comes to an issue of toxicity (damage to individual species) versus the
biodiversity changes. What should be more important is where or if there is a common
denominator for these two approaches? An approach that would provide various levels of
34
-------
protection against toxic effects and biodiversity changes would be most desirable for
scientists, managers and decision makers.
Additional general comments
In general I support the proposed CL-based approach for the newly developed secondary
standard. However, at the same time I have to emphasize that such an approach could
only work if NOy (NOx plus HMOs, HONO, PAN, particulate NO3 and other oxidized
reactive N species) replaces NOx (mainly NO and NO2) as the secondary pollution
standard. Correlation between ambient concentrations of NOx and dry deposition of
oxidized N to watersheds is poor. This is not surprising because while NO2 and NO
dominate atmospheric budget of the oxidized N, their contribution to that deposition is
low due to their low deposition velocities to vegetation and soils. It is well established
that N dry deposition of oxidized N is dominated by HNOs (which is a component of
NOy, but not of NOx) characterized by very high deposition velocity. These problems
are well covered on page 4-21 and shown on Fig. 4-21 of the reviewed PA.
Another major problem of the proposed approach is its reliance on the modeled
concentrations of NOy, NHx and deposition of oxidized & reduced N from the CMAQ
model runs. These modeled outputs are characterized by high level of uncertainty and
temporal limitation (model runs are just for certain years). Therefore verification of the
model results, both for ambient concentrations and estimates of deposition, is critically
needed. Additionally, deriving deposition transference ratios (T values) by dividing
modeled deposition of oxidized N by a sum of concentrations of dry and wet N oxidized
species will very likely result in highly unreliable values. These values will also differ in
time and space considering huge regional and year-to-year differences in ratios of
gaseous /particulate/water-dissolved oxidized N and ratios of dry/wet deposition. Better
understanding of such issues is needed while developing this approach.
Another problem, mentioned many times in the CASAC deliberations, is that while
concentrations of NOx and NOy slowly diminish, concentrations of NH3 from
agricultural emissions and from 3-way catalytic converters are increasing as well as the
importance of Nred in N deposition. In my opinion there a clear need for regulating all
those NHx emissions and development of a new federal air pollution standard
35
-------
Comments from Ms. Lauraine Chestnut
Charge Question 1:
What are the panel's views on the definitions of adversity that are appropriate to consider
in determining what constitutes adversity to public welfare relative to the NOx and SOx
secondary standards?
Overall, the presentation and explanation of available information on losses in ecosystem
services and associated economic valuation as a result of NOx/SOx deposition is much
improved in clarity and context from the first draft PA.
The link is clear and well-documented between the selected ecosystem effects indicator,
ANC, and the welfare effects of lost value of recreational fishing as fish populations (and
in some cases whole species) are not sustained in lakes and streams with lower ANC
levels. The available quantitative information is well presented and explained (except for
a few specific questions noted below). However, more could be done to explain the
qualitative links between deposition and lost ecosystem services that are known and
documented, but cannot be specifically quantified for a specific amount of deposition.
For example, on page 3-13, changes in biodiversity, which are listed as an ecosystem
effect of deposition, are associated with changes in cultural ecosystem services related to
the preservation of natural areas (nonuse values) in addition to productivity, recreational
viewing and aesthetics services that are listed currently in the text. It is well-established
that there is public welfare value to protection and preservation of natural ecosystems in
condition that supports the flora and fauna species that are native to the system, even
when there is no direct use value. This is evidenced in the state and federal statutes that
set aside parks and wilderness areas (noted in the first sections of chapter 3), and in
willingness-to-pay study results such as the Banzhaf et al. (2006) study discussed on page
3-29. The text mentions nonuse value several times, but it would be helpful to make
explicit that this includes value for the preservation of habitat and biodiversity
independent of human use value.
Specific comments/questions in Chapter 3
page 3-9: What is the pollutant referenced in the critical loads shown for Europe?
page 3-11: Add nonuse to ecosystem services listed for water.
pages 3-14 and 3-15: Figure 3-5 includes federal and state public lands according to the
legend, but the text on page 3-14 just references Class I areas, which 1 think are just
federal. Please make clear what areas are included in the maps, and what oher natural
areas may not be included that the public may also care about protecting.
36
-------
page 3-17, line 20: Ecosystem services provide a framework to characterize and describe
how changes in ecosystem function affect public welfare, even if they cannot be
specifically quantified.
page 3-18: It is important to recognize and include language that preferences are not just
about people's own use and enjoyment of an ecosystem, but also include preservation and
bequest value.
page 3-18, second paragraph: Good discussion and explanation of how preferences
depend on information.
page 3-25: figure 3-7. It would be helpful for more general audiences to include words
such as biodiversity and habitat preservation under cultural services. This is part of
"nonuse" services, but I'm not sure most people are aware of this.
page 3-26: It might be useful to reference the language on page 3-22 about the goal of
keeping the Adirondack Forest Preserve as "wild forest lands" and "kept in natural
conditions." This is a significant motivation behind the public's willingness to pay to
prevent the effects of deposition in these areas and is part of what makes the effects
adverse to public welfare.
page 3-28: I'm a little confused by Table 3-2. Are these threshold categories mutually
exclusive?
page 3-29: Are the results in table 3-3 additive? For example, if a threshold of 100 is met,
is the annual value of additional recreational fishing services for NY residents the amount
in the bottom row only, or the sum of the 3 rows? If they are all based on a comparison
to background, then why are the numbers smaller for the 100 threshold than for the 50
threshold? These numbers reflect just a portion of benefits, as noted in the text, so it is
important to include more information in the table title and headings about what they are:
recreational fishing services for NY residents.
page 3:30: Same question for table 3-4.
page 3-34: Is there some descriptive information from the REA or the ISA to give a sense
of the overall magnitude of the red spruce and maple decline attributable to deposition?
The estimates of lost commercial forestry value in the second paragraph are interesting,
but are these forests significant timber resources? What can be said to describe the
implications of the health of these tree species on the natural habitat and health of the
natural ecosystems where these species are prominent? It seems like a more
comprehensive story could be summed up here about the loss in services that is linked to
deposition, even though specific quantitative valuation is not possible. Perhaps more
could be said about the Jenkins et al. (2002) results for avoiding a "significant decline in
health" of high elevation spruce in the Southern Appalachians. How does the description
of decline in this study compare to what has been linked to current deposition levels? The
37
-------
results indicate substantially greater value than was estimated as commercial forest
losses.
page 3-36: Need to say something about how these services are hurt or impaired by
eutrophication. The total value of these services is only relevant if something can be said
about how they are diminished by the effects of deposition. It is okay if this is only
descriptive, but the link needs to be made.
pages 3-40 and 3-41: There is better clarity than in the first draft PA when total values of
ecosystem services are presented to give context for the potential effects of deposition.
The discussion on page 3-41 is helpful in describing why the CSS and MCF ecosystems
are important and how the effects of deposition are likely to diminish the services that
these systems provide. Anything that could be added about the extent of the current
degradation of these ecosystems due to deposition would be helpful for understanding
whether the effects of current deposition are adverse to public welfare.
Charge question 4
Has staff appropriately acknowledged the potential beneficial effects of nitrogen inputs
into nutrient limited ecosystems, while maintaining the focus of the review on preventing
the adverse effects in nitrogen sensitive ecosystems?
It seems to me that the PA is careful now to acknowledge the potential beneficial nutrient
effects of N deposition in some systems. This will come up again when it is time for
regulatory assessment, because there may be some loss in benefits when N deposition is
reduced.
Charge question 13
What are the panel's views on the utility of the additional analyses of co-protection
benefits to inform the consideration of alternative levels of the standards?
The analysis and conclusion in Chapter 6 are important because the decision to focus on
the effects of acidification on aquatic ecosystems means that in this current standard
setting process, other important effects on ecosystems (documented in the ISA), are not
being explicitly taken into account. To the extent that standards set to protect against
effects of acidification on aquatic ecosystems also provide some amount of protection
against the other effects of deposition, then this provides support that the proposed
standards are justified and beneficial.
The analyses reported in Chapter 6 seem adequate for this purpose, but the interpretation
of the conclusions could perhaps be broadened. It is clear that standards set to protect
aquatic resources from adverse effects of acidification would not fully protect against the
effects of deposition on acidification of terrestrial resources and nutrient effects on
terrestrial and aquatic resources. However, some partial protection that would be
provided could be characterized more fully. For example, the analysis suggests that
38
-------
terrestrial systems located in the same watersheds with acid sensitive aquatic systems
would be protected by the deposition levels that would be needed to protect the aquatic
resources. So, the question that comes to mind is what do we know about where sensitive
terrestrial systems are located relative to sensitive aquatic resources throughout the
country. Are they mostly located near one another, or do they occur in completely
separate locations in significant amounts? Given the regional nature of ambient NOx and
SOx concentrations, how close together would sensitive aquatic and terrestrial resources
have to be for protections for one to extend to the other?
Related to this is whether there is benefit to reductions in deposition that are short of the
targets for full protection. This depends on whether the dose-response relationships are
continuous or substantially nonlinear.
Similar questions come up for the analysis of reductions in N deposition relative to the
TMDLs for the Chesapeake watershed. The discussion on page 6-6 shows that N
deposition could be higher under an ANC target of SO than would be allowed given the
TMDL target. However, this is the maximum that the N deposition could be if SOx
deposition were zero. There is a good chance it would be lower than this. Also, how does
this N deposition compare to current levels? How much of the reduction to the target
TMDL would be achieved?
Charge question 24
In light of the panel's views on what constitutes adverse effects to public welfare, what
are the panel's views on:
a) the degree to which current levels of NOy and SOx deposition are adverse to
public welfare?
The case is well made in the PA, based on the information from the REA and the
ISA and information added in the PA, that current levels of NOy and SOx
deposition are harming sensitive ecosystems to an extent that is adverse to public
welfare. A bit more can be done to carry forward the descriptive information
about the significance of the current effects that cannot be fully quantified so that
the implications for adversity to public welfare are more comprehensive.
b) target levels of ANC that protect against adversity to public welfare?
The case seems well supported for a target ANC of at least 50. The wording used
to describe the benefits of a target higher than 50 seems unnecessarily cautious.
What I understand is that at 50, most sensitive species would survive, but not
necessarily thrive. It is certainly clear that loss of an entire species of fish that
would otherwise be expected to live in such waters is an adverse effect, so a target
of 50 to prevent loss of species is justified. To the extent that the size and
robustness of the populations matter to public welfare (and I think there is
evidence that they do) then it seems there would be further benefits of an ANC
target higher than 50. It may be difficult to quantify the value of this additional
39
-------
benefit, but is it really all that uncertain that there would be some additional
benefit?
c) factors relevant in selecting target percentages of waterbodies to protect?
This is a tough question. The choice seems a bit arbitrary. It is key that those
bodies that are naturally acidic and would not benefit from reductions in
deposition have already been excluded. Protecting only half the sensitive water
bodies seems clearly like not enough. What percentage of water bodies in the
Adirondacks are currently affected? It is already established that current effects
are adverse?
d) alternative standards for NOx and SOx.. .taking into account target ANC, target
percentages of water bodies protected, relevant uncertainties, other factors such as
co-protection?
The question of how to group resources seems an important one that needs to be
resolved. At a minimum the split into two categories seems necessary. It is not
clear that the benefits of going to the ecoregion level are worth the extra effort. A
key question is whether further disaggregation would put less restriction on
locations that are not sensitive—which is the whole reason why something other
than a uniform national standard is being developed.
40
-------
Comments from Dr. Ellis Cowling
Charge Question 24. In light of the Panel's views on what constitutes adverse effects to
public welfare (see Chapter 3), what are the Panel's views on:
a) the degree to which current levels of NOy and SOx deposition are adverse to public
welfare based on evidence and risk information, and information on adversity provided in
Chapters 2, 3, and 4?
The ISA and REA for the current review of the NAAQS for oxides oY nitrogen and sulfur
(as summarized in Chapters 2 and 3) make very clear that current levels (ambient
concentrations) of air-borne nitrogen and sulfur compounds (that include not only NOy
and SOx, as asked about in this Charge Question, but also include NHx and some as yet
poorly characterized organic forms of nitrogen — which I would abbreviate RHx) - see
page 7-35) are now causing significant "disruptions in the structure and function of
aquatic ecosystems" in various acid-sensitive regions of the US.
In this connection please note especially the following paragraphs in Chapter 2 page 2-3:
"The scientific evidence is sufficient to infer a causal relationship between
acidifying deposition and effects on biogeochemistry and biota in aquatic
ecosystems (ISA 4.2.2). The strongest evidence comes from studies of surface
water chemistry in which acidic deposition is observed to alter sulfate and nitrate
concentrations in surface waters, the sum of base cations, ANC, dissolved
inorganic aluminum and pH. (ISA 3.2.3.2). Consistent and coherent
documentation from multiple studies on various species from all major trophic
levels of aquatic systems shows that geochemical alteration caused by
acidification can result in the loss of acid sensitive biological species (ISA
3.2.3.3). For example, in the Adirondacks, of the 53 fish species recorded in
Adirondack lakes about half (26 species) were absent from lakes with pH below
6.0 (Baker et al., 1990b). Biological effects are linked to changes in water
chemistry including decreases in ANC and pH and increases in inorganic Al
concentration."
Chapter 3 also makes clear that although the Clean Air Act provides a very broad
definition of different kinds of air-pollution-induced "effects" on public welfare, the Act
in fact does not define "public welfare" as such, and also does not define "adversity to
public welfare." Nevertheless EPA has historically interpreted air-pollution-induced
"adversity" to include "disruptions in ecosystem structure and function" that are regarded
as important to the people of this country. This working definition of "adversity" seems
very sensible to me.
Chapter 3 also includes a brief introduction to the concept of "Ecosystem Services" and
describes various economic valuation and "Willingness to Pay" (WTP) studies that show
very clearly that many citizens of our country are willing to pay the administrative and
operational costs of both private-sector and public-sector efforts to decrease the presently
41
-------
ongoing acidification of freshwater lakes and streams in such places as the Adirondack
Mountains of New York and New England and the Shenandoah National Park in the
eastern US and in acid-sensitive landscapes such as the grasslands of Minnesota and
Coastal Sage Scrub (CSS) areas of California.
Chapter 4 makes very clear that the current NAAQS standards for oxides of nitrogen and
sulfur are not adequate to protect sensitive aquatic and terrestrial ecosystems from
acidification- and nutrient-enrichment effects induced by atmospheric deposition of total
reactive nitrogen and sulfur compounds. See especially Chapter 4 page 4-2:
"... the current standards are not directed toward depositional effects, and none
of the elements of the current NAAQS - indicator, form, averaging time, and
level - are suited for addressing the effects of [total reactive] N and S deposition.
Thus, by using atmospheric NO2 and SO2 concentrations as indicators, the
current standards address only a fraction of total atmospheric NOX and SOX, and
do not take into account the effects from deposition of total atmospheric NOX and
SOX. By addressing short-term concentrations, the current S02 standards, while
protective against direct foliar effects from gaseous SOX, do not take into account
the findings of effects in the ISA, which notes the relationship between annual
deposition of S and acidification effects which are likely to be more severe and
widespread than phytotoxic effects under current ambient conditions, and include
effects from long term deposition as well as short term."
Thus my response to Charge Question 24a is:
Based on the evidence and risk information as well as the information on
adversity provided in Chapters 2. 3, and 4, and in light of my professional views
about what constitutes adverse effects to public welfare, 1 conclude that current
atmospheric deposition loads of total reactive nitrogen and sulfur (including NOy,
SOx, NHx, and probably RHx as well) are causing very substantial and publicly
unacceptable adverse effects on public welfare in various parts of the US.
I also believe that the AAPI approach currently being developed through the
currently proposed and well-integrated "two criteria-pollutant" approach (with
acidifying NHx emissions and deposition also being taken "as given") is well
grounded in the present state of scientific understanding about acidification
effects on aquatic ecosystems.
In addition I believe that the present focus on adverse effects in aquatic
ecosystems will very likely provide some important co-benefits with regard to
decreased adverse acidification effects and decreased nutrient-enrichment effects
in sensitive terrestrial and estuarine ecosystems as well as decreased air
concentrations of methyl mercury.
42
-------
b) target values for ANC that protect against adversity to public welfare in light of the
information presented in Chapter 5 concerning levels of ANC and the ecosystem effects
associated with those target ANC levels?
Thus, my response to Charge Question 24b is:
I regard the three suggested target values outlined in Chapter 5 for use of ANC as
the ecological indicator of choice - 20 u,eq/L, 50 fieq/L, and 100 u,eq/L - to be
very reasonable alternative levels for the Administrator of EPA to use in making
her final decisions about the target value of ANC that would be appropriate for
various acid-sensitive regions of our country.
c) factors relevant in selecting target percentages of water bodies to protect at alternative
target ANC levels to protect against adverse effects to public welfare, and weights to
place on those factors?
Thus, my response to Charge Question 24c is
The addendum we received on September23 provided some clarification of
"factors relevant in selecting target percentages of water bodies to protect at
alternative ANC levels" and some information about "weights that could be
placed on these factors" but after careful and repeated rereading of this addendum
and other parts of Chapter 5,1 still am not able to figure out how to formulate an
appropriate response to Charge Question 24c, other than the obvious idea that
protecting 90% of the water bodies would be more stringent than protecting 75%
of the water bodies, and that protecting only 50% of the water bodies would be
even less stringent.
d) alternative standards for NOx and SOx that would protect against adverse effects to
public welfare based on the AAPI form, and taking into account
(i) consideration of target levels of ANC (chapter 5),
Chapter 5 describes the range of ANC values that are necessary to both understand and
then make decisions about protection of freshwater lakes and streams from acidification
caused by atmospheric deposition of total reactive nitrogen and sulfur:
• Water bodies with ANC values near or above 100 u,eq/L have little or no
risk of acidification.
• Water bodies with ANC values between 100 and 50 u,eq/L are at
progressively increasing risk of acidification,
• Water bodies with ANC values between 50 and 20 u,eq/L are at even
greater risk of acidification, and
• Water bodies with ANC values of 20 u,eq/L or lower already are so acidic
that most of them will not support viable populations offish and many
other aquatic biota.
43
-------
Thus my response to Charge Question 24d(i) is:
The NOx and SOx standards that will be necessary to protect water bodies in an
acid-sensitive region of the US will be an inverse function of the ANC values
already existing in the population of water bodies that are to be protected - i.e.,
the lower the existing ANC values in the water bodies to be protected, the more
stringent must be the NOx and SOx standards that must be met.
In addition, the final SOx standards that are established for each acid-sensitive
region that is to be protected will need to be adjusted in part by determinations of
the percentage of sensitive water bodies in the region that are to be protected and
also by calculations or measurements of the nitrogen-assimilative capacity of the
ecosystems that are to be protected.
(ii) target percentage of water bodies to protect (chapter 5),
My response to Charge Question 24d(ii) is essentially the same as my response to Charge
Question 24b (above)
The addendum we received on September23 provided some clarification of
"factors relevant in selecting target percentages of water bodies to protect at
alternative ANC levels" and some information about "weights that could be
placed on these factors" but after careful and repeated rereading of this addendum
and other parts of Chapter 5,1 still not figure out how to formulate an appropriate
response to Charge Question 24c, other than the obvious idea that protecting 90%
of the water bodies would be more stringent than protecting 75% of the water
bodies, and that protecting only 50% of the water bodies would be even less
stringent.
(iii) consideration of relevant uncertainties in AAPI components (chapter 7).
Chapter 7 provides a very succinct and thorough introduction to many of the still existing
uncertainties that are inherent in the AAPI approach to setting welfare-based NAAQS
standards for NOx and SOx. As Chapter 7 and both the earlier ISA and REA documents
make clear, however, major advances have been made in'recent years in decreasing many
of the scientific uncertainties that were considered in previous NAAQS reviews for NOx
and SOx. Thus, a much more robust scientific foundation has been developed for
establishing NOx and SOx NAAQS standards that will diminish the frequency and
intensity of nitrogen and sulfur induced adverse effects on the structure and function of
ecosystems and on ecosystem services important to public welfare in this country.
These important decreases in scientific uncertainty have resulted from the following
developments in recent years:
l)-The decision to take a two-criteria pollutant (nitrogen and sulfur) integrated approach
rather than to continue to consider NOx and SOx separately,
44
-------
2) Separating the development of public-welfare-based NAAQS standards from the
formerly always dominating public-health-based NAAQS review processes,
3) Including in the AAPI approach to management of acidifying nitrogen and sulfur
deposition, both chemically oxidized and chemically reduced inorganic forms of nitrogen
(and even recognizing that organic as well as inorganic forms of nitrogen also must be
considered) in the current ecosystem-focused secondary NAAQS review process,
4) Considering both acidification effects and nutrient-enrichment effects on whole
ecosystems (including interactive effects among all types of plants, animals, insects, and
microorganisms) rather than just direct effects on individual species of plants and/or
animals,
5) Focusing on nitrogen and sulfur effects on naturally occurring and unmanaged
terrestrial and aquatic ecosystems (that include natural grasslands; open range lands;
unmanaged coniferous, hardwood, and mixed-species forests; and riverine, estuarine, and
coastal ecosystems ~ rather than trying also to consider at the same time, air-borne
nitrogen and sulfur effects on commercially important plant and animal agricultural
production systems and intensively managed commercial forests,
6) Greatly improved mathematical models (especially CMAQ) of spatial and temporal
relationships among air emissions of pollutants, meteorological transport phenomena,
chemical and physical transformations of airborne nitrogen and sulfur compounds, and
wet, dry, and occult (cloud and fog) deposition processes at both high and low elevations,
7) Greatly improved concepts and descriptions of the diversity array of eco-regions that
exist across this great continent of ours,
8) Much improved understanding of linkages among bed rock geology, soils, vegetative
cover, temperature and moisture-supply gradients, episodic phenomena such as droughts,
floods, snow melt processes, physical climate process, and chemical-climate-induced
changes in the physical climate,
9) Recognition that our present scientific understanding of nitrogen- and sulfur-induced
acidification and nutrient-enrichment processes in aquatic ecosystems is much more
thoroughly developed than acidification and nutrient-enrichment phenomena in terrestrial
and estuarine ecosystems.
Thus, my response to Charge Question 24d(iii) is:
Yes, there are still some important uncertainties about how many different
categories of sensitivity to aquatic ecosystems should be recognized, how
adequately the estimates of chemically reduced forms of nitrogen from the
CMAQ air quality model can be trusted, how large the co-benefits for terrestrial
ecosystems will be from use of the present AAPI approach with its primary focus
45
-------
on protection of aquatic ecosystems, and in the several kinds of Willingness to
Pay (WTP) and other kinds of benefit estimates, but 1 am confident that these and
other sources of uncertainty will continue to decrease during the next few years
and that the present scientific foundation is adequate to implement the AAPI
approach as soon as final decisions about the indicator, level, statistical form, and
averaging time of the proposed NAAQS standard can be resolved.
(iv) any other potentially relevant factors, such as levels of co-protection against
terrestrial acidification and nutrient enrichment (chapter 6)?
Chapter 6 contains a very short but persuasive description of the likelihood of significant
co-benefits in protection of terrestrial ecosystems from acidification and nutrient
enrichment effects from implementation of an AAPI approach aimed primarily at
protection for aquatic ecosystems.
Thus, my response to Charge Question 24d(iv) is:
Although it is difficult to develop quantitative estimates of the co-benefits that are
likely to accrue in terrestrial ecosystems from NAAQS standards designed
specifically to diminish adverse effects on aquatic ecosystems in acid-sensitive
regions, 1 believe there is no uncertainty at all that such co-benefits will occur and
will not be surprised if these co-benefits turn out to be significant in magnitude.
Chapter 4: Addressing the Adequacy of the Current Standards
Charge Question 2. What are the Panel's views on staffs approach to translating the
available evidence and risk information and other relevant information into the basis for
reaching conclusions on the adequacy of the current standards and on alternative
standards for consideration?
My response to Charge Question 2 is essentially the same as my response to Question
24a:
"Chapter 4 makes very clear that the current NAAQS standards for oxides of
nitrogen and sulfur are not adequate to protect sensitive aquatic and terrestrial
ecosystems against acidification and nutrient enrichment induced by atmospheric
deposition of total reactive nitrogen and sulfur compounds.
In this connection, please note especially Chapter 4 page 4-2:
"... the current standards are not directed toward dcpositional effects, and none
of the elements of the current NAAQS - indicator, form, averaging time, and
level - are suited for addressing the effects of [total reactive] N and S deposition.
Thus, by using atmospheric NO2 and SO2 concentrations as indicators, the
current standards address only a fraction of total atmospheric NOX and SOX, and
do not take into account the effects from deposition of total atmospheric NOX and
46
-------
SOX. By addressing short-term concentrations, the current S02 standard, while
protective against direct foliar effects from gaseous SOX, does not take into
account the findings of effects in the ISA, which notes the relationship between
annual deposition of S and acidification effects which are likely to be more severe
and widespread than phytotoxic effects under current ambient conditions, and
include effects from long term deposition as well as short term."
a) In light of the Panel's views on the appropriate definitions of adversity to public
welfare (see Chapter 3), do you agree that the current levels of NOy and SOx deposition
are adverse to public welfare?
Yes, I do agree that the current levels of NOy and SOx deposition are adverse to
public welfare. Once again let me explain my response by repeating parts of my
response to Charge Question 24a:
Chapter 3 makes clear that although the Clean Air Act provides a very broad
definition of different kinds of air-pollution-induced "effects" on public welfare,
the Act in fact does not define "public welfare" as such, and also does not define
"adversity to public welfare." Nevertheless EPA has historically interpreted air-
pollution-induced "adversity" to include "disruptions in ecosystem structure and
function" that are regarded as important to the people of this country. Thus
EPA's working definition of "adversity" seems very sensible to me.
The ISA and R.EA for the current review of the NAAQS for oxides of nitrogen
and sulfur (as summarized in Chapters 2 and 3) make very clear that current levels
(ambient concentrations) of air-borne nitrogen and sulfur compounds (including
not only NOy and SOx, as asked about in this Charge Question (but also include
ambient NHx and some as yet poorly characterized organic forms of nitrogen (see
Chapter 7 page 7-35) are now causing significant "disruptions in the structure and
function of aquatic ecosystems" in various acid-sensitive regions of the US.
In this regard, please note especially Chapter 2 page 2-3:
"The scientific evidence is sufficient to infer a causal relationship between
acidifying deposition and effects on biogeochemistry and biota in aquatic
ecosystems (ISA 4.2.2). The strongest evidence comes from studies of surface
water chemistry in which acidic deposition is observed to alter sulfate and nitrate
concentrations in surface waters, the sum of base cations, ANC, dissolved
inorganic aluminum and pH. (ISA 3.2.3.2). Consistent and coherent
documentation from multiple studies on various species from all major trophic
levels of aquatic systems shows that geochemical alteration caused by
acidification can result in the loss of acid sensitive biological species (ISA
3.2.3.3). For example, in the Adirondacks. of the 53 fish species recorded in
Adirondack lakes about half (26 species) were absent from lakes with pH below
6.0 (Baker et al., 1990b) Biological effects are linked to changes in water
47
-------
chemistry including decreases in ANC and pH and increases in inorganic Al
concentration."
3. Has staff appropriately applied this approach in reviewing the adequacy of the current
standards and potential alternative standards?
Yes, I believe that EPA staff has very appropriately noted that:
1) the very short-term present secondary NOx and SOx standards (calculated as
the arithmetic mean of 1-hour concentrations of NC>2 and as the arithmetic mean
of 3-hour concentrations of SCh) are wholly inadequate to protect aquatic or
terrestrial ecosystems from the long-term cumulative acidifying loads of total
reactive nitrogen and sulfur compounds;
2) the "indicators" used in the present NAAQS standards do not include all of the
acidifying and nutrient-enriching forms total reactive nitrogen and sulfur that are
now causing significant adverse impacts on the structure and function of aquatic
and terrestrial ecosystems in various acid-sensitive regions of the US.
4. Has staff appropriately acknowledged the potential beneficial effects of nitrogen inputs
into nutrient limited ecosystems, while maintaining the focus of the review on preventing
adverse effects in nitrogen sensitive ecosystems?
Yes, in this connection please note the following discussion in Chapter 4 pages 4-
44 and 4-45:
"In certain limited situations, additions of nitrogen can increase rates of growth,
and these increases can have short term benefits in certain managed ecosystems.
As noted earlier, this review of the standards is focused on unmanaged
ecosystems. As a result, in assessing adequacy of the current standards, we are
focusing on the adverse effects of nutrient enrichment in unmanaged ecosystems.
However, the following discussion provides a brief assessment of effects in
managed ecosystems.
Impacts of nutrient enrichment in managed ecosystems may be positive or
negative depending on the levels of nutrients from other sources in those areas.
Positive effects can occur when crops or commercial forests are not receiving
enough nitrogen nutrients. Nutrients deposited on crops from atmospheric sources
are often referred to as passive fertilization. Nitrogen is a fundamental nutrient for
primary production in both managed and unmanaged ecosystems. Most
productive agricultural systems require external sources of nitrogen in order to
satisfy nutrient requirements. Nitrogen uptake by crops varies, but typical
requirements for wheat and corn arc approximately 150 kg/ha-yr and 300 kg/ha-
yr, respectively (NAPAP, 1990). These rates compare to estimated rates of
passive nitrogen fertilization in the range of 0 to 5.5 kg/ha-yr (NAPAP, 1991).
48
-------
Chapter 6: Co-protection for Other Effects Using Standards to Protect Against Aquatic
Acidification
Charge Question 13. What are the Panel's views on the utility of the additional analyses
of co-protection benefits to inform consideration of alternative levels of the standard?
My view is that the additional analyses of co-protection benefits contained in
Chapter 6 is a well-reasoned and valuable addition to this Policy Assessment
Document. I was especially well-pleased with the following summary paragraph
in Chapter 6 page 6-2 and the additional detailed information contained in Tables
6-1 and 6-2 on page 6-3:
"Results of the comparison between the aquatic critical acid load (ANC = 50
ueq/L) and
the terrestrial critical acid loads (Bc:Al 1.2 and 10.0) for the 32 watersheds are
presented in Tables 6.1 and 6.2. In the 16 Adirondack watersheds, 13 of the 29
lakes had aquatic critical acid loads that were lower (more protective) than the
terrestrial critical acid loads when a Bc:Al ratio of 10.0 was used. Based on
terrestrial critical acid loads determined with a Bc:Al ratio of 1.2, 21 of the 29
lakes in the Adirondacks had aquatic critical acid loads lower than the terrestrial
critical acid loads. More importantly, for the terrestrial critical acid loads
determined with a Bc:AI ratio of 10.0, 13 of the 16 lakes in the Adirondacks
classified as "highly" and "moderately" sensitive to acidification had aquatic
critical acid loads lower than the terrestrial critical acid loads, and all 16 lakes in
these two sensitivity classes had critical acid loads lower than the terrestrial loads
determined with a Bc:Al of 1.2 The watersheds within the Shenandoah region
showed similar results (Table 6.1)."
Let me turn now to a few general remarks deriving from my experience as the designated
"liaison person" serving as a member of both this NOx/SOx Secondary NAAQS Review
Panel and the Integrated Nitrogen Committee (FNC) developed within EPA's Science
Advisory Board.
The first and perhaps most important linkage between the INC and the NOx/SOx Panel
was the following "Resolution" developed by the INC and communicated to the
NOx/SOx Panel on October 31, 2007:
Resolution: The current air pollution indicator for oxides of nitrogen, NOx, is an
inadequate measure of reactive nitrogen in the atmospheric environment. The SAB's
Integrated Nitrogen Committee recommends that inorganic reduced nitrogen (ammonia
plus ammonium) and total oxidized nitrogen, NOy, be monitored as indicators of total
chemically reactive nitrogen.
49
-------
The NOx/SOx Panel has accepted this resolution and incorporated NOy as the
recommended "indicator" of choice for implementation of the proposed revision of the
NOx and SOx Public Welfare based NAAQS standards using the AAPI approach.
The second important linkage between the INC and the NOx/SOx Panel was a
presentation in September 2008 by Chairman Russell of the then emerging AAPI
approach, with its incorporation of chemically reduced forms as well as chemically
oxidized forms of reactive nitrogen as an "as given" feature of regions of the US to which
the AAPI approach could be applied. This novel approach was useful in developing an
integrated way of recognizing that chemically reduced inorganic forms of nitrogen
(gaseous NHs and ammonium ion (NHx+) as well as chemically oxidized forms of
reactive nitrogen and sulfur (NOy+SO2+ SO*) are all very important parts of the total
acidifying deposition that leads to adverse ecosystem impacts in acid-sensitive regions of
the US.
This choice to include NHx "as given" in the AAPI index for ecosystem effects of oxides
of nitrogen and sulfur was is an artful means of avoiding the large administrative and
probably nearly prohibitive political challenges of trying to designate ammonia and
ammonium ion as a seventh Criteria Pollutant and thus including three rather than two
Criteria Pollutants in this initial step that EPA is now taking in exploring options for
multi-pollutant approaches in air quality management in the US as recommended in the
National Research Council's 2004 report on "Air Quality Management in the United
States."
The third important linkage between the NOx/SOx Panel and the INC came about during
EPA's renegotiation of the original court-ordered deadline for completion of the
NOx/SOx NAAQS review process. This change in the court-ordered deadline provided
approximately 18 moths of additional time that EPA Staff sorely needed to complete the
additional analyses and assessments that we presently have available in this Second
External Review Draft Policy Assessment.
The fourth and last important linkage between scientific findings and recommendations
from the INC and the findings and recommendation of the NOx/SOx Panel has to do with
the magnitude of air emissions from various US sources of the reactive nitrogen and
sulfur. As indicated in the attached Table 2 from the June 2010 draft report of the INC,
in 2002 the total air emission of reactive nitrogen from industrial and transportation
sources totaled about 6.2 Tg of NOx-N compared to agricultural sources that totaled
about 3.1 Tg/yr of NHx-N - roughly a two-fold difference in air emissions of total
reactive nitrogen from these three major sources.
Table 1: Mr fluxes for the United States, Tg N in 2002."
fir inputs to the Atmo\plieric environmental system
N^O-N emissions '
Agriculture - livestock (manure) N2O-N
Tg N/vr
08
003
%.
8
50
-------
Agriculture - soil management N2O-N
Agriculture - field burning agricultural residues
Fossil fuel combustion - transportation*
Miscellaneous
NH.-N emissions 2
Agriculture livestock NHj-N
Agriculture fertilizer NH3-N
Agriculture other NH3-N
Fossil fuel combustion - transportation *
Fossil fuel combustion - utility & industry *
Other combustion
Miscellaneous
NO«-N emissions 2
Biogenic from soils
Fossil fuel combustion - transportation *
Fossil fuel combustion - utility & industry *
Other combustion
Miscellaneous
Total Atmospheric inputs
Nr inputs to the Terrestrial environmental system
Atmospheric N deposition11
Organic N '
Inorganic NO,-N '
Inorgamc-NH,-N *
*N fixation in cultivated croplands 5
Soybeans*
Alfalfa*
Other leguminous hay *
Pasture*
05
0001
01
01
31
16
09
01
02
003
02
01
62
03
35
1 9
04
02
100
69
2 1
27
2 1
77
33
2 1
1 8
05
31
61
100
19
21
51
-------
Dry beans, peas, lentils *
N fixation in non-cultivated vegetation * '
N import in commodities *7
Synthetic N *'
Fertilizer use on farms & non-farms
Non- fertilizer uses
Manure N production '
Human waste N '"
Total Terrestrial inputs
Nr inputs to the Aquatic environmental system
Surface water N flux "
01
64
02
15 1
109
42
60
1 3
435
48
IS
03
41
16
3
100
Also attached please find the following Concluding Statement from the June 2010 draft
report from the INC:
Concluding Statement
Fossil fuel combustion and food production have significantly increased the introduction
of Nr (reactive nitrogen) into the US environment and, while there have been tremendous
benefits, there are also tremendous damages to the health of both ecosystems and people.
Optimizing the benefits of Nr while minimizing its problems will require an integrated
nitrogen management strategy that not only involves EPA, but also other federal agencies
(e.g., USDA, DOE, NOAA), state agency managers, the private sector, and a strong
public outreach [educationally focused] program.
52
-------
Comments from Dr. Charles Driscoll
Executive Summary
ES-3, paragraph 4
ES-3, paragraph 5, line 2
ES-4, paragraph 2, line 4
ES-4, paragraph 4, line 3
ES-4, paragraph 6
ES-7, paragraph 2
ES-10-11, 1st paragraph ES-11
ES-13, last bullet
What is meant by balance of base cations? Change
to "dissolved inorganic aluminum."
Change to "depth of soil and surficial deposits"
ANCof50ueq/L
Change to "may result in nutrient imbalance"
Change to "as trout are eliminated due to
acidification"
Why would watersheds with low base-cation
weathering be eliminated from consideration?
These should be the most acid-sensitive watersheds.
1 don't believe this statement on naturally acidic
ecosystems is true. There are surface waters that
are naturally acidic due to low rates of base cation
supply and/or high inputs of naturally occurring
organic acids. However, these systems can also be
impacted by elevated inputs of acidic deposition.
This is a widespread occurrence in the Adirondack
region of New York. These naturally acidic surface
waters will exhibit loss of ANC and elevated
aluminum concentrations from acidic deposition.
This statement is problematic. Brook trout is not a
sensitive species. Maybe the sentence should be
changed to state "...protection against declines in
fitness of less sensitive species (e.g., brook trout,
zooplankton) ..."
Also, what is meant by "the overall health of
aquatic communities may not be impacted." If
species are lost, isn't this an impact on the health of
aquatic communities? I think this bullet needs to be
re-phrased.
Chapter 1
53
-------
PI-9, line 4-6
Chapter 2
P 2-2, line 8
P 2-9 in PnET box
P 2-21, line 1
P 2-21, paragraph 2
Chapter 3
P 3-6, line 13-15
P 3-9, line 11
P3-15, Figures 3-5 and 3-6
Page 3-30, line 8
Is this sentence correct? Aren't both direct and
indirect effects considered in this PA? Also, total
deposition is not just paniculate forms. This
sentence should be re-written.
This sentence needs to be changed to something like
"in some instances unless strongly retained by soil
or biota, leach out..."
Change last line to "The model can be set to operate
on any time set, but is generally run on a monthly
time-step. It is applied at the stand to small-
watershed scale."
It might be good to reference Goodale et al. (2010)
here.
The article by Thomas et al. (2010) on nitrogen
deposition on northern tree species should be
mentioned in this paragraph.
This statement about alkalinity is incorrect. For all
intents and purposes, alkalinity and ANC are the
same. Often alkalinity involves titration to a fixed
pH endpoint (-4.2), while ANC generally involves
Gran Plot determination of the equivalence point.
The difference between the two is subtle at best.
For this document, the two should be used
interchangeably.
This sentence on the units conversion does not
make sense. Sulfur and nitrogen have different
molecular weights. Therefore one cannot have a
single mass conversion for a nitrogen map (left) and
a sulfur map (right).
Docs the N deposition include NH4+, is the map
total N deposition or NOs deposition? Please
clarify.
100 ueq/L
54
-------
Chapter 4
P 4-1, line 6 Change to "associated with elevated deposition of
NOX..."
P 4-1, line 22 Change to "nutrients and acid neutralizing capacity
P 4-2, line 25 Change to "as the ability of the watershed to
counteract acidic inputs is decreased as the supply
of acid neutralizing capacity is used more rapidly
than can be replaced through geological
weathering."
P 4-20, line 2 The text refers to sulfur fields but the figures
referenced (Figure 4-5, 4-6) depict NOX and NHX.
Is there a mistake here?
P 4-20, line 4 The text refers to concentration patterns, but Figure
4-4 shows deposition.
P 4-20, lines 24, 29 The text refers to correlation between NOX and N
deposition. Is this NOX concentrations? If so, this
should be clarified.
P 4-49, line 27 Change to "and methyl mercury can be taken up..."
P 4-50, line 14-15 Methylation of mercury occurs in watersheds all
over the U.S. (and the world) where conditions are
appropriate. Please change this sentence, it is
incorrect.
P 4-50 Note there are other linkages between acidification
and fish mercury accumulation. Mercury is
accumulated to a greater degree in aquatic biota as
pH and ANC decreases. Dittman and Driscoll
(2009) noted that as fish condition increased
associated with decreases in acidic deposition, fish
mercury concentrations decreased.
Chapter 5
P 5-8, line 21-22 Change to "some fraction of the acid neutralizing
capacity. ."
55
-------
P5-13(Figure5-4b)
P 5-15, line 6
P 5-15, line 16
P 5-17, entire page
P 5-17, line 16
P 5-17, line 24
P 5-19, line 7
P 5-19 (around here)
P 5-19, Figure 5-5
I am not familiar with this paper, but the figure does
not make sense, why would the ANC curve have
different lines for a wet US average year? This
figure should be explained or deleted.
Change to "to neutralize the deposition."
This definite of steady-state models is horrible. A
steady state model is one with time invariant inputs,
outputs and pools. This section should be re-
written.
The authors are using the term equilibrium
incorrectly. Equilibrium is a thermodynamic term.
Throughout this page, the word equilibrium needs
to be replaced with the word "steady-state" (lines 1,
4,6,8,14,29,30).
There is a problem here with the term critical load.
Critical load is a steady-state phenomenon. For a
value of critical load that is not at steady-state, the
term dynamic critical load or target load should be
used.
Change to "implying that watersheds with greater
inherent supply of acid neutralizing capacity
respond ..."
Change to "in-lake retention of SC>42" and N.
A critical issue needs to be addressed if steady-state
models are going to be used. Steady-state models
will give relative high values of the level of
atmosphere deposition needed to protect ecosystem
(critical loads) because they assume steady-state
conditions. At any reasonable time frame, the
dynamic critical load will be much lower. If steady-
state models are going to be used a ''safety factor"
should be applied to account for this discrepancy.
Does the trade-off figure consider background
deposition? In other words, is the zero deposition
value really 0? There is background deposition of S
and N that will not be changed by controls of
emissions.
-------
P 5-21, equation 3 This equation does not make sense and needs to be
explained better. What is the difference between
nitrogen uptake and nitrogen immobilization?
P 5-23, lines 10, 16 Change to "as the acid neutralizing capacity of the
watersheds increase..."
P 5-26, 5-27, line 18-19, Figure 5-7 The text refers to a map of critical loads, but Figure
5-7 is a map of sites where critical loads have been
calculated.
P 5-27 This approach of eliminating low ANC sites seems
foolish. These are the most sensitive sites. Why
would you throw them out? If you don't want to
include these sites, the percentage of the lakes
targeted should be relaxed. This would be a more
honest, transparent approach rather than throwing
out the most sensitive watersheds. Also I would
recommend against throwing out the high DOC
lakes. High DOC waters can be impacted by acidic
deposition. A better approach would be to include
these waters and check the DOC concentrations of
the waters that would not be protected.
Undoubtedly these would include many high DOC
waters.
P 5-51, Figure 5-18 I would like some additional explanation of this
figure. It appears that the NHX deposition shifts the
"dog-leg" to lower values of N deposition for
graphs a and c, but not b. Why?
P 5-52, Figure 5-19 This figure is also difficult to follow. I think I have
the sense of it, but a more detailed explanation
would be helpful.
P 5-54, line 14 What is meant by pure nitrogen and sulfur? Do you
mean total nitrogen and total sulfur? Please clarify.
P 5-54, line 26 Would better wording be "N and S atoms of NOX
and SOX removed from the atmosphere, which ..."
P 5-56, line 8 Would better wording be "species that affect the
health of ecosystems would..."
P 5-56, line 25 Should you specify which lake in the Adirondacks?
57
-------
P 5-58, Figure 5-20
P 5-65, line 6
P 5-66, line 23
P 5-66, line 24
•>nd
P 5-66, 2na paragraph
P 5-67, line 3
P 5-70, paragraph ????
P 5-75, Table 5-10
P 5-81, line 13
P 5-82, line 24
P 5-82, line 25
This figure suggests that T values are relatively
invariant for the eastern U.S. Is this correct? If so,
this would be important information to clarify. It
would also be helpful to explain why this is the case
(also discussed in Chapter 7).
It is important to define what is meant by uptake
and immobilization. It also would be helpful to
indicate how uptake, denitrification and
immobilization are calculated.
Rather than nitrogen buffering capacity, do you
mean nitrogen retention capacity?
Again this is confusing. Do you mean "when
reduced nitrogen deposition exceeds the ability of
the ecosystem to retain nitrogen?
This paragraph is confusing and needs to be re-
worded. The term buffering capacity is not being
used properly. There is confusion on distinguishing
between nitrogen retention and loss of acid
neutralizing capacity.
Clarify units 50 ueq/L.
This paragraph is horrible. For example, line 7
indicates that below pH 4.5 ANC appears to be
uncorrelated with pH. As at pH values below the
equivalence point
ANC = - [H+] this shows what an incorrect
statement this is. This paragraph is filled with mis-
statements and needs to be completely re-written.
Aren't these species listed from most sensitive to
least sensitive? See table title.
How about "as ANC decreases, the probability of
very low pH values occurring increases."
Brook trout is a relatively insensitive fish species.
How about "When ANC values are <50 ueq/L, the
probability of acidic episodes increases
substantially.'5
58
-------
P 5-82, line 30
How about "At these levels during acidic episodes
brook trout populations..."
Chapter 5 General Comments (the devil is in the details)
• How will the probability of lakes to be protected be determined?
• How many sites per region/category will be evaluated?
• How will the time-dependence of recovery be addressed? Critical loads vs. target
loads (dynamic critical loads).
Chapter 6
P6-1
P 6-4, line 17
P6-6
Chapter 7
P 7-13, 7.4.3.2
P 7-17-7-20, Figures 7-1,7-4
P 7-32, line 1
References:
This analysis is nice but I am skeptical. There are
limited field observations on this. Many soil time
series studies over the past 15 years show ongoing
depletion of soil exchangeable calcium and
magnesium which many waters, particularly in the
Northeast, are showing recovery of ANC. This
pattern, if true, suggests ongoing soil acidification
while surface waters are recovering from acidic
deposition. This may also suggest that soil is more
"sensitive" to inputs of acidic deposition than
surface waters.
50 ueq/L
How about a short blurb about co-benefits
associated with decreases in fish mercury and
wildlife mercury concentrations associated with
decreases in sulfate loading and/or increases in
surface water pH?
It would be helpful and important to discuss why T
values are relatively homogenous.
Indicate what the lines on the figures represent.
Change to "as the supply of acid neutralizing
capacity of watersheds increases..."
59
-------
Dittman. J.A., Driscoll, C.T., 2009. Factors influencing changes in mercury
concentrations in yellow perch (Percaflavescens) in Adirondack lakes.
Biogeochemistry 93 (3), 179-196.
Goodale, C.L., Thomas, R.Q., Dentener, F., Adams, M.B.. Baron. J.S., Emmett, B.A.,
Evans, C.D., Fernandez, I.J., Gundersen, P., Hagedorn, F., Lovett, G.M..
Kulmatiski, A., McNulty, S.G., Melvin, A.M., Moldan, F., Ollinger, S.V.,
Schleppi, P., Weiss, M.S., In press. Nitrogen deposition and forest carbon
sequestration: A quantitative synthesis from plot to global scales. Global Change
Biology.
Thomas, R.Q., Canham, C.D., Weathers, K.C., Goodale, C.L., 2010. Increased tree
carbon storage in response to nitrogen deposition in the US. Nature Geosciences
3,13-17.
60
-------
Comments from Dr. Christopher Frey
Chapter 7: General Comments
Chapter 7 should have an introduction regarding the spatial scope and temporal averaging
that is the basis for each of the inputs or factors discussed in the chapter. Without a
context regarding spatial scope and temporal averaging, it is not possible to characterize
either variability or uncertainty. It needs to be clear, for example, as to over what
geographic domain and for what averaging time quantities such as deposition
transformation ratios are to be estimated.
Charge Question 16: What are the Panel's views on the discussion of uncertainty in the
critical loads models including MAGIC and SSWC?
The chapter discusses the MAGIC critical load simulations on pages 7-3 to 7-4, page 7-6,
and in three paragraphs on pages 7-30 to 7-31. The discussion is with respect to an
uncertainty analysis conducted in the REA. Figures 7-14 and 7-15 illustrate that the
MAGIC model estimates for surface water chemistry at two locations compared very
well with observed values. The simulated values fall very close to the parity line,
demonstrating that the model is both precise and accurate in estimating concentrations,
ANC, and pH.
On lines 8-9 on page 7-31, the text is "The estimated confidence bounds on predicted
ANC suggest that the 95 percent upper confidence bound is on average 10 percent higher
in lakes, and 5 percent higher in streams." It is not clear as to what is being compared -
higher than what? Is the intended meaning that the 95 percent upper confidence bound is
10 percent higher than the mean value? Or 10 percent higher than the observed value?
For MAGIC, given that there are comparisons of model predictions to observed values
that demonstrate the model performance, the discussion appears to be adequate.
The discussion of uncertainty in the SSWC makes several points:
• The F-factor approach is widely used in Europe and Canada, but not in the U.S.
• Critical loads estimated by steady-state MAGIC and SSWC F-factor approaches
had similar trends and results converged for low critical loads
• In the REA, a Monte Carlo-based uncertainty analysis was conducted by varying
runoff rates, water chemistry variables, and acid deposition. The coefficients of
variations were 5 to 9 percent for critical load limits of 20 to 50 ueq/L.
The statement that the uncertainties introduced by the SSWC F-Factor model are likely to
be moderate seems to be supported by the discussion in the text.
Charge Question 17: What are the Panel's views on the areas for future research and data
collection outlined in this chapter, on relative priorities for research in these areas, and on
any other areas that ought to be identified?
61
-------
The chapter does not clearly outline areas for future research or data collection, nor does
it offer priorities for research.
Perhaps this charge question refers to the section on modeling and data gaps. For
consistency with the PA's for other criteria pollutants, consistent header titles should be
used. If the purpose of this section is to identify areas for future research and data
collection, there should be an introductory paragraph that states this.
From the current draft, one infers that the key needs for better models and data are:
• Occult deposition from cloud and fog processes, especially in high elevation
watersheds
• Lightning generated NOx, to be incorporated into CMAQ in 2012
• The role of organic bound nitrogen in wet deposition, and re-entrainment from the
surface
• Increased spatial coverage of ambient measurements, and specific measurement
"needs" (better instruments? Could be more clear) related to NOy, speciated NOy,
ammonia, and ammonium.
• Better emissions estimates for soil and agricultural emissions of [NOy? Or just
ammonia and NOx?].
• More data for sensitive ecosystem areas, to have more even coverage
geographically.
There is no discussion, based on Table 7-1, of the key sources of uncertainty other than
gaps that should be areas for future research or data collection. Examples include:
• Pre-industrial base cation concentration is listed as having a high magniture of
uncertainty and a high knowledge base uncertainty. Should this be a priority for
future research or data collection?
• Dry deposition (generically - N and S species) is listed as having a medium
magnitude of uncertainty and a medium-high knowledge base uncertainty.
• Ecological indicator to changes in the value of ecosystem services is listed as
having "medium-high" magnitude of uncertainty, and a "negative" bias. Is there a
need for better information here?
Not mentioned is what research is needed in order to improve the knowledge base for
ecosystem effects other than aquatic acidification, such as terrestrial acidification and
aquatic system nutrient enrichment.
In the recent CASAC PM Panel review of the PM policy assessment second draft, a fairly
substantial list of research needs was identified with a recommendation that EPA hold a
follow-up workshop to further refine the research agenda and consider how to implement
research in order to improve the state of knowledge before the next review cycle.
Charge Question 21. What arc the Panel's views on the overall characterization of
uncertainty as it relates to the determination of an ecologically-relevant multi-pollutant
standard for NOx and SOx?
62
-------
This charge question is rather broad and it is not clear as to the specific focus. Chapter 7
appears to do a thorough job of listing many sources of uncertainty and providing
qualitative characterization of each.
Charge Question 22. What are the Panel's views on the following:
a. The insights that can be gained into potential alternative additional secondary standards
(using the AAPI form) by considering:
i. Information from studies on the relationship between mortality in aquatic organisms
and pH and ANC?
ii. Information from studies on the relationship between fish health and/or
biodiversity metrics and pH and ANC?
iii. Information on the relationship between pH, Al, and ANC?
iv. Information on target ANC levels identified by states and regions, as well as other
nations?
Chapter 9 could more clearly, and very briefly, summarize what are the key hazards
associated with aquatic acidification, and list or briefly describe the key adverse
effects, while stating conclusions regarding the relationship between mortality in
aquatic organisms, pH, and ANC. The chapter presumes that the reader is already
familiar with the endpoint effects, and tends to focus only on the indicator.
Likewise, Chapter 9 could more clearly summarize the relationship between fish
health and biodiversity metrics with respect to pH and ANC. This should be done
qualitatively, for a lay reader, leaving the technical details to the earlier chapters.
Although Chapter 9 implies that the use of the indicator for acidic deposition to
aquatic systems may afford some protection with respect to acidic deposition to
terrestrial systems, there could be a paragraph that more clearly but succinctly
addresses the significance of. and linkages between, pH, Al, and ANC.
There is no discussion in chapter 9 that I could find regarding comparison of target
ANC levels among states, regions, or nations. Perhaps a paragraph could be provided
that summarizes this, leaving the details to previous chapters.
b. The appropriate role of qualitative and quantitative characterizations of uncertainty in
developing standards using the AAPI form?
EPA has appropriately taken a weight of evidence approach to hazard identification.
EPA has done a reasaonable job in identifying and characterizing individual sources
63
-------
of uncertainty, as summarized in Table 7-1. Other committee members may have
specific comments on those sources of uncertainty, or others that might be included.
The discussion of model and data gaps in Chapter 7 was useful, but needs to be
expanded (see earlier charge question).
A key conclusion from Chapter 7 may be worth repeating in Chapter 9 (page 7-37,
lines 19-29):
"there is no apparent directional bias in the uncertainty regarding the biological,
chemical and physical processes incorporated in the AAPI. From the perspective of
valuation of ecosystem services, the estimates generally are believed to be biased low,
meaning the values of reaching a target level of protection are underestimated.
However, quantification of these values is perhaps the most uncertain of all aspects
considered. Consequently, the level of the AAPI should be relatively high in a
buffering context to account for the existence of uncertainties in several components.
In addition to, but related to these uncertainties discussions, are considerations of time
lag to reach a target level ANC due to ecosystem response dynamics, as well the
uncertainties in the severity and prevalence of episodic events. Both of these
considerations suggest support for an AAPI that is somewhat higher than the target
ANC supported by the specific evidence and risk information."
It is not clear, for example, that the conclusion that the AAPI should be somewhat
higher than the target ANC (nor is this statement itself particularly clear, because it
implies that AAPI and ANC are the same quantity) has been taken into account in the
discussion of Target ANC limits in Chapter 9.
c. The role of considerations regarding the relationship of the standard to:
i. the time trajectory of response, e.g. when specific ANC levels are likely to be
realized given a specific level of the AAPI?
ii. the likelihood of damages to aquatic ecosystems due to episodic acidification
events given a specific target for chronic ANC?
iii. the levels of co-protection for terrestrial ecosystems against acidification effects
and the for aquatic and terrestrial ecosystems against effects of excess nutrient
enrichment?
These questions go beyond my expertise, but they imply a question regarding short-
term versus long term responses of ecosystems, and the dynamics of eco-system
response, that do not appear to be discussed in Chapter 9. For example, if a 3 to 5
year averaging time is used, should the level be set so as to also be protective against
short term (e.g., seasonal, or shorter episode) events? A related question is whether
the modeling tools, such as steady-state MAGIC simulations, adequately account for
the adverse effects associate with dynamic responses. As another reviewer suggests
64
-------
(Dr. Driscoll), perhaps a safety factor is needed to account for differences in dynamic
versus steady-state response.
There is discussion of the implications of the aquatic system acidification-based
standard with regard to its effect on terrestrial acidification and nutrient enrichment,
which seems adequate but domain experts may have more comment on this issue than
do I.
General comments on Chapter 9
The summary of suggested options given in table form is very useful and should be
incorporated into chapter 9.
The choice of averaging times and whether only such averages or used, or whether
shorter term episodes will be considered, needs to inform the variability and uncertainty
analyses of Chapter 7. This is because the range of variability and uncertainty depend on
averaging time. Similarly, the geographic scope needs to be taken into account in the
analysis of variability and uncertainty. The spatial options for components of the AAPI
equation need to be further discussed or refer the reader to specific text earlier in the
document.
This standard may be challenging to communicate to the public and the regulated
community. Finding a way to state the basic ideas in qualitative language that a lay
reader can understand would be very helpful.
Although Chapter 9 alludes to attributes of a standard, it seems to require referring back
to previous chapters in order to have a complete picture. In communicating this to the
Administrator and other stakeholders, the EPA staff have a significant challenge of how
to present the indicator, averaging time, form, and level in a manner that is relatively easy
to explain and complete without overwhelming the audience with details.
• For example, the summary table is helpful in defining the "indicator" but seems
incomplete. Isn't AAPI actually being used as an indicator? Not just NOy and
SOx?
• The rationale for the 3 year, 5, year, and other averaging times should be
explained.
• The form of the standard seems to be expressed in a complex manner. The cogent
information is that NOy and SOx are related through AAPI. This could be stated
more succinctly in the summary table. Explanations of other points can be in
footnotes or the text.
• While the AAPI equation is important, perhaps it can be a footnote if it is not
actually considered to be an indicator.
• The spatial options should be illustrated graphically in Chapter 9. This chapter
should be thought of as the synthesis material that any reader will go to for the
bottom line of this policy assessment. Thus, it should be complete in terms of
information needed to completely define a standard.
65
-------
If the spatial categories are only regarding non-air quality inputs to the AAPI
equation, then perhaps this material is not critical to Chapter 9 as long as it is
clearly laid out earlier in the policy assessment. Or, it can be in a different table.
On the other hand, there is a need for clarity on the geographic scale and
averaging time for each of the AAPI inputs. Possibly this material needs to be
moved to Chapter 7.
To be consistent with how NAAQS are structured, the summary table should
include a section on "Level."
66
-------
Comments from Dr. Paul Hanson
General Comments:
1 approve of the suggested approach taken by EPA staff to retain the existing
standards for phytotoxic direct effects of NOX and SOX, and the development of an
additional standard based on the deposition of combined forms of N and S. The focus on
aquatic acidification effects for that additional standard is appropriate to the available
data. My expertise lies in the area of terrestrial ecosystems and I will not comment on the
specifics of the proposed aquatic-acidification-based standard.
In most instances, the document does an appropriate job of clarifying that
demonstrated effects associated with acidification and nutrient enrichment from N and S
deposition are limited to 'sensitive' ecosystems. Nevertheless, in a number of places the
document slips back into language that could be interpreted as inferring that all
ecosystems will benefit from such a new standard. Such language should be corrected to
ensure that the reader understands that protection afforded by the proposed new standards
would be limited to sensitive ecosystems located in specific regions of the United States.
The executive summary and conclusion (Chapter 9) fail to provide a clear picture
about the extent of total US lands subject to the benefits of the proposed standard.
Comments on selected Charge Questions:
What are the Panel's views on the definitions of adversity that are appropriate to consider
in determining what constitutes adversity to public welfare relative to the NOx and SOx
secondary standards?
Without a defined understanding of how ecosystem changes in species
composition would proceed through natural succession processes in the absence of
anthropogenic N and S inputs, it is difficult to understand when changes in species or
biodiversity should be considered adverse. The text tends to suggest that any change from
the status quo should be considered adverse. I am not in agreement with such a
conclusion. Unfortunately, I am also not providing a definition of the boundary of
acceptable vs. unacceptable rates of change. The document should deal with this issue. If
species changes in lichens or grassland composition are to be proposed as metrics of
adverse effects, an attempt needs to be made to characterize levels of change to be
expected in non-polluted ecosystems.
Comments on specific sections of the document:
Executive Summary:
1. Line numbers should be added to this section to facilitate comments.
2. On Page ES-3 Paragraph 3 the word "often" might be removed. It is insufficiently
quantitative to be informative. Its inclusion also tends to lead the reader to
conclude that deposition of NOx and SOX routinely leads to leaching from
watersheds, which then result in the acidification of aquatic systems. This may be
the case in some, but not all watersheds throughout the US where deposition of N
and S forms are largely retained.
67
-------
3. Page ES-4 paragraph 3: The claim at the end of this paragraph should be
referenced to the ISA or REA.
4. Page ES-5 fourth paragraph: I don't agree with the first sentence of this
paragraph, but rather think mostN-limited ecosystems would have the capacity to
absorb or buffer incoming N and S forms to limit or avoid perturbation. The
examples of species change driven by deposition are provided (I guess) as
definitive examples of an adverse result without adequate justification of the rates
of species change that could or should be considered adverse.
5. ES-7 last paragraph: Deposition velocity is not the rate of pollutant deposition.
Deposition velocity times the concentration gradient is the rate of deposition.
68
-------
Comments from Dr. Rudolph Husar
Chapter 5: Conceptual Design of the Standard
5.3.1 Conceptual Design of the form.
1 like the revised Figure 5.2. Still, instead of bidirectional arrows, two rows of arrows,
one heading left the other to the right would more clearly communicate the intent.
9. What are the Panel's views on the revised characterization of the deposition
transference ratios (TNOyand TSOx)?
In principle, I do agree with the approach to derive the ambient concentration -
deposition relationship for individual S and N species using CMAQ. The aggregation
approach also makes sense.
A repeat of my past comment on nomenclature: The Deposition/Concentration IS an
effective total wet+dry deposition velocity. So why the addition of the confusing
'Transference" term? The concepts and formulae are already messy enough, so why not
keep it simple: Effective Total Deposition Velocity. On p 7-3 its 'transformation ratio' .
On p 7-12 its "Ambient Concentration to Deposition Transformation Ratio"....etc.
Figure 5.23 is the same as Fig 5.4
10. What are the Panel's views on staffs conclusion that an averaging time of 3 to 5
years is appropriate given the A API form of the standard?
Agree that 3-5 years is adequate for the characterization of inter-annual variability of key
measures and indicators.
However, I am not clear as to why and how the episodic events and damage (and the
springtime 'acid shock') are dismissed as is done in Section 5.2. Where is the evidence
that an annual average standard will be more protective than a companion short-term
standard?
Chapter 7 Uncertainty
14. What are the Panel's views on the following?
a. The degree to which the chapter appropriately characterizes the potential role of
information on uncertainty, sensitivity, and variability in informing the standards?
This version of the Policy Assessment came a long-long way in characterizing the
uncertainty, sensitivity and variability. My main concern remains to be known systematic
errors in CMAQ simulations and their inadequate inclusion in the uncertainty analysis.
69
-------
b. The appropriateness and completeness of the evaluation of CMAQ model performance
and sensitivity to critical inputs?
The overestimation of the SO2 is a significant error that is not being 'corrected for'. Is
that correct? High SO2 will probably result in high (unverifiable) dry deposition. So,
focusing and showing the propagation of this known error would make the uncertainty
analysis more convincing. For instance a figure like Fig 7-5 for CMAQ-CASTNET
comparison for SO2 could be very revealing.
c. The utility of the analyses of temporal and spatial variability in the deposition
transference ratios (TNOyand TSOx)?
The spatio-temporal variability of the Effective Total Deposition Velocity makes sense,
since it indeed varies considerably in space and time. However, the CMAQ calculations
of the Dep/Conc ratio for S and N compounds in Fig 5-20 do not appear to have the right
spatial texture. Near sources where dry deposition of gases dominate, Dep/Conc( dep
velocity) should be significantly higher than in the far field where wet deposition
dominates. 1 don't see that texture. Inverting T in the plots FIG. 5.2 makes the
interpretation even more difficult.
15. What are the Panel's views on the insights provided by the AAPI sensitivity analysis
including: a. The evaluation of elasticities of response? b. The multi variable ANOVA
analysis?
The value of the sensitivity/elasticity analysis heavily depends on what magnitude of
perturbations one assumes. Are the assumptions reasonable?? I don't know. The CMAQ
SO2 concentration bias is a big, known systematic error. Is it properly incorporated in the
uncertainty.
70
-------
Comments from Dr. Dale Johnson
Charge Question 2: What are the Panel's views on staffs approach to translating the
available evidence and risk information and other relevant information into the basis for
reaching conclusions on the adequacy of the current standards and on alternative
standards for consideration?
Response: The Second Review Draft focuses almost entirely on aquatic effects with the
rationale that such effects are better known and better documented than terrestrial effects.
This is certainly true. But this raises a question in my mind: is it the purpose of this
document to provide evidence to support changing standards, more or less in a lawyerly
fashion? Should this be the purpose? Or should the purpose of the document be to
examine all the potential pluses, minuses, and potential unintended consequences of
changing standards? In short, is this a mission? If so, is the review panel expected to sign
on to the mission? These questions came to mind as I considered my response to Charge
Question 4: many of the points that I have raised in the past as potential benefits of
increased N deposition to forest ecosystems and C sequestration are now moot with the
change in focus toward aquatic effects. I reiterate that I am in no way against changing
standards to protect aquatic ecosystems, I am only trying to see that the approach to it
includes a balanced assessment of the effects of such changes. If EPA does not do it, I am
quite sure that someone else will.
Charge Question 2a) In light of the Panel's views on the appropriate definitions of
adversity to public welfare (see Chapter 3), do you agree that the current levels of NOy
and SOx deposition are adverse to public welfare?
Response: This almost becomes a philosophical issue. It is hard to conceive of an effect
of some perturbation that does not have some adverse as well as some beneficial effect to
public welfare, with the probable exception of Hg deposition. The example that comes to
mind is agricultural fertilization, which is adverse to public welfare when it is done in
excess and leads to groundwater nitrate pollution, yet on the other hand, it is certainly
adverse to public welfare to preclude fertilization with the resultant substantial decline in
crop and food production! There is little doubt that current levels of NOy (combined with
NH4) and SOx deposition are having adverse effects on some sensitive ecosystems; how
many of such ecosystems can be protected at what cost, and what are the magnitudes and
importance of unintended consequences (such as forest ecosystem C balance or crop S
fertilization) that might result from such protection, and how does this compare to the
benefits of protecting these sensitive ecosystems? The question becomes one of assessing
the balance between these two effects, and, while recognizing the considerable
uncertainties in some unintended consequences, I do feel that further discussions along
such lines will add considerable credibility to this document.
71
-------
Charge Question 4: Has staff appropriately acknowledged the potential beneficial effects
of nitrogen inputs into nutrient limited ecosystems, while maintaining the focus of the
review on preventing adverse effects in nitrogen sensitive ecosystems?
Response: The staff acknowledged at various places in the document that some benefits
of N deposition might occur in very limited circumstances in commercial forests. They
do not mention the C balance issue, which could occur in any forest, although the
inclusion of Climate in Table 3-1 implies this, and is certainly of more practical relevance
than" Climate Control" or "Regulating Climate" as is now shown in the table. While
dismissing the potentially positive effects of N fertilization in non-commercial forests,
the staff does, however, spend a considerable amount of time considering exactly the
same phenomenon in non-commercial forests of the southwestern US where increased
growth probably provides unwanted fuel for the next wildfire. I find this to be
unbalanced. In general, however, any benefits of N (or S) deposition to terrestrial
ecosystems is far less relevant in this document than in previous ones in that the staff has
limited their scope largely to aquatic effects, none of which (to my knowledge) are
beneficial. The wording in the Executive Summary regarding the ISA is correct (p. 1-9,
lines 7-9) in that "The ISA highlights the ecological effects to sensitive ecosystems
other than commercially managed forests and agricultural lands..." Thus, the
consideration of any potential benefits to ecosystems which by definition are "sensitive"
(implying sensitive to negative effects) becomes moot, irrelevant, and dismissable.
However, I would note that the first part of the quote from the Clean Air Act in page 2-1,
lines 7-9, does not necessarily dismiss any benefits, but simply addresses effects. It does
indeed mention damage to property, etc in the middle part of the quote, but the beginning
and end do not specify that only negative effects be considered (although this may well
have been what was intended). In a nutshell, if the scope of this effort is now limited to
aquatic effects, then the issue of potential benefits becomes nearly irrelevant. It is not
irrelevant, however, in the larger scheme of things where C balance effects of N
deposition are now being hotly debated in the literature and sure to come up at some later
time if new standards are proposed.
Charge Question 11: What are the Panel's views on the preliminary staff conclusions
regarding alternative target ANC levels that are appropriate for consideration and the
rationale upon which those conclusions are based?
The ANC levels at which negative effects on aquatic systems occur appear to be well
established and I see no problem with considering alternate target ANC levels in this
context. The premise of the section on alternate levels of ANC does not directly assess
whether a given AAPI standard will or will not achieve a certain target ANC because of
uncertainties, but presumes that the target ANC levels are reached and discusses them in
that context (p. 5-69, lines 9-16). 1 confess to some degree of confusion as to the logic in
some of this section, for example, the discussion of alternative target ANC's and timing
on p. 5-85, lines 2-8. Clarification of this logic would help this reader. I would also point
out that the capacity (change in soil) and intensity (change in soil solution) considerations
raised before should enter into this discussion, as changes in the intensity factor could be
very rapid in response to changes in deposition whereas changes in the capacity factor
72
-------
could be as slow as envisioned here, if indeed they occur at all (for example, it seems
very unlikely that soil exchangeable acidity will decline and base saturation will increase
in response to a decrease in acid deposition inputs - soils in humid environments would
not likely become more basic with time, but only continue to acidify at a slower rate). I
also confess to some confusion as to the derivation of the DL factors discussed on page 5-
88 and 5-89 - if they were explicitly defined by other than the general terms used here, 1
missed it, and these terms are not in the list of abbreviations and acronyms up front.
Charge Question 11 a) Jn light of the Panel's views on appropriate definitions of
adversity to public welfare (see Chapter 3), what are the Panel's views on the
appropriateness of the information related to adversity considered by staff in evaluating
alternative target ANC levels?
Here I have nothing further to add in addition to what was stated above.
Charge Question 12: What are the Panel's views on the approaches considered by staff
for assessing alternative target percentages of water bodies for protection at alternative
ANC levels?
Again, it is not clear how this was done - the DL factors, which clearly are numerical
indices of some kind, are not defined in pages 5-88 and 5-89, and thus I am unable
adequately address this question. If they were defined (preferably in the form of
equations) elsewhere, that should at least be clearly referenced here, and preferably
formally defined again. It is unclear to me how the DLo/oeco terms in Tables 5-12 and 5-13
were derived, and yet it seems to be a critical element of the assessment. This needs to be
more clearly described.
Other Specific Comments
Table 1-1: It would probably be better to use SI units here, as nearly all journals demand
these days.
p. 1-9, lines 7-8: I do not understand why ag systems and commercial forest systems
should be left out.
p. 2-1, line 24: need a space between "9" and "of
p. 2-2, lines 13-28: Again, the intensity effects need to be included here - that is,
introduction of strong acid anions such as sulfate and nitrate to an already acid soil (and
acid soils do occur in nature, without any air pollution effects), then acidification of
waters can occur instantly without any change at all in base saturation.
p. 2-5: Some discussion of natural acidification processes by natural carbonic and organic
acids and by plant uptake should be discussed here. The uninitiated may erroneously
conclude that acid soils only occur in the presence of air pollution.
73
-------
p. 2-19, lines 20-21: This statement seems to fly in the face of other statements later on
which say that there it too much uncertainty in terrestrial effects and therefore the
document will concentrate on aquatic effects.
pp. 2-21 through 2-23: As I have said in many previous reviews, I believe some mention
of the C balance issues as related to N deposition deserves a mention here.
p. 3-7: What about DOE? They have funded a considerable amount of ecological
research, including air pollution research.
p. 3-11: Table 3-1 is incomplete. Soils not only provide the service of nutrient cycling,
they also provide filters for providing clean water (or, in some cases, provide pollution to
clean water). And where are the timber values for forests in here? It is NOT implied only
in crops, because there are still forest products removed from National Forests these days,
even though their primary purpose is no longer for timbering. The vaguest of all is
Climate and especially Climate Control? "Regulating" climate? I did not realize we had
that technology yet. What should really be shown here is C balance considerations, but I
am aware that the authors are very loathe to do this.
p. 3-25, lines 14-19: Again, potential beneficial effects, for example for C sequestration,
are not necessarily limited to managed ecosystems.
p. 3-89 through 3-41: So while potential beneficial effects of N enrichment are summarily
dismissed as irrelevant, three full pages are spent on the negative, fire related effects of
N enrichment. Certainly the fire effects are very valid ones, that is not the point - it is a
matter of selective emphasis on only the negative.
p. 4-2, line 25: Again, acidification of waters can take place in minutes as mobile strong
acid anions enter an acidic soil.
p. 4-39, lines 10-16: and D) C sequestration, which can benefit national C balance if
permanent and be of benefit, or cause enhanced fire danger in drier systems and thus be
extremely negative.
p. 4-45, lines 28-30: This is parsed out very specifically, but is not completely true, as
potential benefits can accrue even in unmanaged systems, as stated above.
p. 5-16, lines 4-5 and page 5-7, lines 1-6: These sections clearly point out the problems
with steady-state models which assume that base cation flux is equal to soil weathering
rate. This simply cannot be true because if it were, soils would never acidify. Acid soils
are found all over the world, including in pristine, unpolluted areas.
p. 5-23: CL is not in the list of acronyms
p. 7-7, lines 3-15: This section has me lost. I am unclear as to what is being said here.
74
-------
p. 7-32, lines 8-13: The real source of uncertainty here is in the assumptions and premises
upon which these calculations are based. This source of uncertainty has not been
quantified (and perhaps cannot be quantified).
p. A-5, lines 6-11: Nearly all of these assumptions is false. 1. Steady-state conditions
never exist, as is well recognized by nearly all ecologists these days. 2. Nutrient cycling
effects on soils are profound, often far more important than inputs by deposition and
outputs by leaching. 3. N inputs by N-fixation are still greater on a global scale (last time
I looked) than those of air pollution (although N fixation is more spotty). 4.1 will not
contest. 5. Some sesquioxide rich soils can absorb sulfate for a very very long time.
p. A-5, equation 7: This equation and the premises upon which it is based clearly point
out the problems with steady-state models which assume that base cation flux is equal to
soil weathering rate. This simply cannot be true because if it were, soils would never
acidify. Acid soils are found all over the world, including in pristine, unpolluted areas.
p. A-5, line 26: Where is equation 5?
75
-------
Comments from Dr. Naresh Kumar
Charge question 5: What are the Panel's views on staffs revised conceptual framework
for the structure of a multipollutant, ecologically relevant standard for NOx and SOx? To
what extent does the Panel agree that this suggested structure adequately represents the
scientific linkages between ecological responses, water chemistry, atmospheric
deposition, and ambient NOx and SOx?
From a theoretical standpoint the conceptual framework looks fine, but the practical
usefulness of the framework depends on its robustness. One way to determine the
robustness is by a comprehensive uncertainty analysis, as discussed in my response to
Charge Question 14. Another way the AAPI can be tested is by use of historical data.
Where data are available, one could use the AAPI (Equation 18 on Page 5-63) to get a
trajectory of changes in AAPI in response to changes in SOx and NOy concentrations (I
don't know whether concentrations of all forms of SOx and NOy would be available, but
some assumptions may have to be made for non-measured components of NOy). The
values of other components of AAPI (Q, Neco, [BC]o, LNHx, TNOy and TSOx) have
already been estimated by EPA or can be determined using models or measurements. It is
critical to do this "hindcasting" at more than one location. The changes in AAPI
predicted using Equation 18 should more or less match the changes in ANC (may be with
some lag).
Charge Question 6: What are the Panel's views on the appropriateness of considering a
single national population of waterbodies in establishing standards to protect against
aquatic acidification? What are the Panel's views on consideration of alternative
subdivisions of the U.S. to identify the spatial boundaries of populations of waterbodies
and acid-sensitivity categories, specifically:
d) the use of Ecoregion III areas to aggregate watrebodies?
e) the use of ANC to further aggregate Ecoregion III areas into different categories
of sensitivity?
f) The relative appropriateness of the suggested methods for categorizing spatial
boundaries of sensitivity, e.g., one nation, binary sensitive/less-sensitive classes,
cluster-analysis based on sensitivity classes, and individual ecoregions?
Using a single national population of water bodies in establishing standards to protect
against aquatic acidification may seem attractive for its simplicity, but it has many
shortcomings as noted in the PAD and by other members of the panel. Other approaches
discussed by the staff are not too complicated either, especially given the overall
complexity of the framework, so I see no reason to settle for a single population of water
bodies to represent the whole nation. I would recommend using Option 2 (d) unless EPA
finds it too complicated in which case 2 (b) or 2 (c) could be used.
Overall, the whole concept of how ecoregions will be used in establishing standards to
protect against aquatic acidification was very confusing and EPA needs to do a better job
of clarifying this concept.
76
-------
Charge Question 9: What are the Panel's views on the revised characterization of the
deposition transference ratios (TNOy and T SOx)?
A major concern with TNOy and TSOx is that although they are the critical links
between NOy and SOx ambient concentrations and their deposition, they are derived
using a model that can not be evaluated because of lack of measurements of dry
deposition. EPA has shown that inter-annual variability of these ratios can be large even
when using the same model. This is most likely due to variability in wet deposition
because of changing rainfall patterns from year-to-year. May be EPA should evaluate the
stability of the dry and wet deposition ratios separately. It is also critical to show how
these ratios vary when using different models and different chemical mechanisms.
It is recommended that EPA evaluate the stability of these ratios by examining these
ratios for the following model simulations (in addition to what has already been done):
• CMAQ and CAMx models (it is okay, in fact preferable, to use different
emissions and meteorological conditions)
• Different chemical mechanisms
• Different model grid resolutions (36-km v/s 12-km or even 4-km, if available)
Charge Question 10: What are the Panel's views on staffs conclusion that an averaging
time of 3 to 5 years is appropriate given the A API form of the standard?
The Agency makes a good case for using the averaging time of three years and I agree
with that recommendation.
Charge Question 14: What are the Panel's views on the following?
a. The degree to which the chapter appropriately characterizes the potential role of
information on uncertainty, sensitivity, and variability in informing the standards?
The discussion of uncertainty and sensitivity analysis is much improved compared to the
first draft; however a complete quantitative analysis of uncertainty is needed for the
AAPI. Staff has mentioned some good reasons for conducting this analysis, which
includes gaining confidence in the data and the models used in defining the form of the
standard. The reason it is particularly important to conduct a comprehensive uncertainty
and sensitivity analysis for the standard in review is that the many of the components of
the AAPI can not be evaluated because of lack of measurements. So, the only way to gain
confidence in using AAPI is to examine how sensitive the SOX and NOY response
surfaces are to different components of the AAPI. Although. EPA has evaluated
uncertainty of some components - some quantitative, but mostly qualitative - the
analysis falls short of what the CASAC panel had requested as part of the review of the
first draft of the PAD.
77
-------
The panel had asked the EPA conduct a sensitivity study to characterize uncertainty with
different components of the conceptual framework and propagate the resulting
uncertainty at every step to arrive at an ensemble of SOX and NOY response surfaces to
meet a given level of the AAP1. The range of the SOX and NOY response surfaces would
have given a confidence level in the use of the A API. By not conducting that level of
analysis, EPA has failed to really show how robust an AAP1 standard is. It is mentioned
in the text that a Monte-Carlo analysis was not performed, but no reason is mentioned for
that. I believe it is imperative that EPA conduct a full quantitative analysis of uncertainty
before moving forward on using the AAPI construct for setting a standard. One way to
conduct the quantitative uncertainty analysis is to use the Monte Carlo approach to arrive
at an ensemble of SOX and NOy response surfaces to meet a given AAPI standard. As an
example, the Monte Carlo analysis for Adirondacks could be conducted by using the
following approach:
• Use a probable range (or a mean value with some assumed or derived standard
deviation) of estimates for Neco,
• Use a probable range (or a mean value with some assumed or derived standard
deviation) of estimates for [Bc]o,
• Use an estimate of Q based on different years (dry vs. wet year),
• Calculate L(NHX), TNOV and TSOX using different air quality model simulations
that may already be available and use the distributions in Monte Carlo
simulations,
o Use CMAQ, CAMx, or other model simulations for different years (with
different meteorological conditions and different emissions) to get a range
of these variables,
o Use different chemical mechanisms, if available
• Use covariant constraints for quantities that may be correlated when running the
Monte Carlo simulations
b. The appropriateness and completeness of the evaluation of CMAQ model performance
and sensitivity to critical inputs?
The performance of the CMAQ model is still incomplete after repeated requests by
the CASAC panel. The performance statistics for the major species shown in Table 7-
1 can be misleading. One should use mean normalized bias (and not normalized mean
bias) when conducting model performance using concentrations that are averaged
over long periods (e.g. one week or more). A complete evaluation of CMAQ derived
values with measurements is needed before any confidence can be placed on the use
of the model to generate the desired parameters. Although the model cannot be
evaluated for dry deposition because of lack of measurements, evaluation of the
ability of the model to represent ambient levels can serve as a proxy for its ability to
represent dry deposition. It is recommended that following evaluations (using daily or
weekly averaged quantities, not annual) be performed to assess the uncertainty in the
model:
78
-------
4. Model performance for nitric oxide, nitrogen dioxide, sulfur dioxide, nitrate,-
ammonium and aerosol nitrate, ammonium, and sulfate for different networks for
which the data are routinely available,
5. Model performance for wet deposition of sulfate, nitrate, and ammonium using
the National Atmospheric Deposition Program (NADP) network,
6. A regional model evaluation using the continuous measurements of nitric oxide,
nitrogen dioxide, nitric acid and NOvfrom the SEARCH network in the
southeastern U.S.
c. The utility of the analyses of temporal and spatial variability in the deposition
transference ratios (TNOyand TSOx)?
Showing year-to-year change using the same model is not sufficient. A complete
analysis should also show how these ratios vary with use of different models and
different chemical mechanisms. EPA has the modeling data from the CAMx model
used for the transport rule, so it should be able to compare these ratios easily from
model-to-model.
Charge Question 15: What are the Panel's views on the insights provided by the AAPI
sensitivity analysis including: a. The evaluation of elasticities of response? b. The
multivariable ANOVA analysis?
Evaluation of elasticity of response is a good way to get a first hand picture of the
sensitivity of the AAPI wrt to different components of the AAPI. However, I would
suggest doing this analysis for the SOX and NOX concentrations needed to meet a
standard, as those are the quantities for which the standard is being set.
Charge Question 24. In light of the Panel's views on what constitutes adverse effects to
public welfare (see Chapter 3), what are the Panel's views on:
b) target values for ANC that protect against adversity to public welfare in light of the
information presented in Chapter 5 concerning levels of ANC and the ecosystem effects
associated with those target ANC levels?
d) alternative standards for NOx and SOx that would protect against adverse effects to
public welfare based on the AAPI form, and taking into account
(i) consideration of target levels of ANC (chapter 5),
Given the lack of complete quantitative uncertainty analysis, I don't think the
administrator would have a high degree of confidence in whether a particular level of
AAPI would indeed provide requisite level of protection or be overly protective for that
matter. Qualitative way the uncertainty has been discussed has the potential to give a
false sense of confidence about the overall uncertainty On the bottom of Page 7-2 and
top of Page 7-3, there is a discussion on what a low level of confidence in the
components of the standard mean. It goes on to say that if the confidence is low then
79
-------
AAPI could be adjusted upwards or downwards depending on whether you want to put
more emphasis on providing requisite level of protection or on not making the standard
being overly protective. I would add one more thing there is that if the level of
confidence is too low, then it makes the AAPI form of the standard questionable whether
it is a robust standard or not. The reason it is particularly important to conduct a
comprehensive uncertainty and sensitivity analysis for the standard in review is that many
of the components of the AAPI can not be evaluated because of lack of measurements.
So, the only way to gain confidence in using AAP1 is to examine quantitatively how
sensitive the SOX and NOY response surfaces are to different components of the AAPI.
Given the lack of a quantitative sensitivity analysis we have no idea of the confidence
interval and in my view it is premature to talk of the levels or ranges of standards unless
that shortcoming is overcome.
80
-------
Comments from Dr. Myron Mitchell
Three important areas that need to be addressed are:
1) Evaluation of AAPI using historical data will be very helpful in developing this
combined NOy and SOx standard. There are a number of historical data sets
including the Adirondacks, Catskills, Hubbard Brook, etc. that have long-term
water chemistry data. There may be some difficulties in obtaining all of the
needed AAPI inputs including atmospheric deposition estimates, but at least some
approximation will be very useful. This use of historical data will provide
additional confidence in the AAPI approach.
2) The reliance on the CMAQ model with respect to providing estimates of
deposition input is important to clearly link this effort by EPA with effects. The
importance of the CMAQ output for developing this secondary standard clearly
suggests that more effort is needed by EPA in the evaluation of the CMAQ
output. This should evaluation should be a high priority for EPA in monitoring
and research efforts.
3) Treatment of sulfur in the AAPI. There is no consideration in the AAPI
formulation of internal (soil) sulfur sinks (e.g., soil sulfate adsorption) or sulfur
sources (organic S mineralization, S mineral weathering, sulfate desorption). It is
assumed that watershed sulfur outputs equal sulfur inputs in deposition.
Mitchell1 et al. (2010, Biogeochemistry) found that watersheds that had
previously had substantial portions of atmospheric S input that from 1985 through
2002 that internal sources contribute 1-6 kg S ha'1 year'1. This would equal 6 to
37 meq/m2/year. This contribution is substantial when compared to various
analyses provided in the document (e.g., figures 5-15, 5-18, 5-19, etc.). Not
including this internal sulfur source will result in an underestimate in the amount
of reduction for nitrogen and sulfur deposition needed to meet target loads. Other
studies in North America and Europe have also emphasized the importance of
internal sulfur sources.
My other comments are given below.
Executive Summary
'Mitchell, M J, G Lovett, S Bailey, F BeaH, D. Burns, D Buso, T A Clair, F Courchesne, L
Ouchesne, C Eimers, D Jeffries, S Kahl,, G Likens, M D Moran, C Rogers, D Schwede, J. Shanley, K
Weathers and R. Vet. 2010 Comparisons of Watershed Sulfur Budgets in Southeast Canada and Northeast
US New Approaches and Implications Biogeochemistry (In Press and Available on Line)
81
-------
Page(s)
ES-8 The term Neco used in the figure has not been defined prior to its use.
ES-9 The sentence "Snowmelt can release stored N deposited throughout the
winter" is conceptually not correct. The vast majority of N released is
nitrate that has been generated microbially within the soil, not
atmospherically deposited N in the snowpack.
Chapter 1: Introduction
[no questions]
Chapter 2: Known or Anticipated Ecological Effects
[no questions]
Chapter 3: Considerations of Adversity to Public Welfare
3-15 to 16 It would be useful to indicate in figure legends 3-5 and 3-6 that it is
assumed that all the N and S deposited are converted to nitrate and sulfate
respectively for calculation eq/ha/yr.
3-28 Table 3-2 needs further clarification. It is not completely clear why the
lake count is 0 for years 2005 for all ANC thresholds.
3-29 For Table 3-3 include within the table legend a replacement of "present"
with "Year 2007".
3-37 In Table 3-7, a delineation of the arrows used in the value column needs to
be provided.
I. What are the Panel's views on the definitions of adversity that are appropriate to
consider in determining what constitutes adversity to public welfare relative to the NOx
and SOx secondary standards?
The Chapter does a good job of describing the various attributes of
diversity with particular emphasis on those areas expected to be most
sensitive to NOx and SOx effects in the USA. The impact of the Chapter
could be improved by a summary section that clearly indicates which of
the adversity components will be the primary focus of the proposed
standards.
Chapter 4: Addressing the Adequacy of the Current Standards
4-5 line 4 This sentence needs to be changed from "oxidized nitrogen" to
82
-------
"reactive nitrogen"-as it currently is written it excludes reduced forms of
N including ammonium.
4-15 In describing the issues related to the differences between the rural (e.g.,
CASTNET) and urban deposition monitoring sites, it is clear that there is a
disconnect. Would it be appropriate to recommend that a unified network
is needed that includes both rural and urban sites?
4-17 to 18 Certainly there is justification for using CMAQ as a predictor of
deposition. It is somewhat curious, however, that NADP is used for wet
deposition and CMAQ is for dry deposition. Certainly, there are more
problems associated with the estimates of dry deposition than those for
wet deposition. However, to gain more confidence in the CMAQ
predictions it would be very important to compare the NADP
(measurements) and CMAQ (predictions) for wet deposition. This type of
comparison is needed to confirm that "CMAQ promotes analytical
consistency and efficiency across analyses of multiple pollutants" and
"CMAQ provides a consistent platform incorporating the atmospheric and
deposition species of interest over the entire United States".
4-18 The issues related to scaling up in time the CMAQ estimates of hourly
estimates needs to be discussed.
4-21 to 36 It would be very helpful to use the same color ranges for each gases
pollutants for comparing estimates from CMAQ, CASTNET and SLAMS.
Also, there is very limited discussion on the differences in the results
associated with these different monitoring networks. For example, there
appear to be major differences in CMAQ (Figure 4-11) and CASTNET
(Figure 4-13) sulfate concentrations with respect to the absolute values
and spatial distribution.
2. What are the Panel's views on staffs approach to translating the available evidence
and risk information and other relevant information into the basis for reaching
conclusions on the adequacy of the current standards and on alternative standards for
consideration?
The general information is certainly contained within this document and other
supporting information such as within the ISA and REA, but the actual linkages
of evidence and the translation to the generation of the standards could be
improved.
a) In light of the Panel's views on the appropriate definitions of adversity to public
welfare (see Chapter 3), do you agree that the current levels of NOy and SOx
deposition are adverse to public welfare?
83
-------
b)
Yes, the evidence is sufficient that the current levels of NOy and SOx deposition
are adverse to public welfare in some systems which are particular sensitive to
acidification and orN addition causing nutrient enrichment.
3. Has staff appropriately applied this approach in reviewing the adequacy of the current
standards and potential alternative standards?
Yes, the approach is valid, but more attention to the linkages between evidence
and the generation of the standards would be helpful.
4. Has staff appropriately acknowledged the potential beneficial effects of nitrogen inputs
into nutrient limited ecosystems, while maintaining the focus of the review on preventing
adverse effects in nitrogen sensitive ecosystems?
The current balance is appropriate in the context of the standard and the
protection of sensitive systems.
Chapter 5: Conceptual Design of an Ecologically Relevant Multi-pollutant Standard
5-5 lines 5-6 This statement is not correct. The vast majority (>95%) of
the nitrate released during episodic snowmelt is derived from the forest
floor and mineral soil and not from the snow. A possible rewording could
be as follows: Snowmelt results in the mobilization to drainage waters of
nitrate most of which has been generated within the forest floor and
mineral soil. This release of this nitrate can result in episodic
acidification. Literature citations would include (Kendall, 1998, Tracing
Nitrogen Sources and Cycling in Catchments, Book Chapter; Piatek et al.,
2005, WASP; Campbell et al. 2006, J. Geophys. Res.).
5-5 lines 7-8 The statement that "inputs of nitrogen and sulfur from
snowpack and atmospheric deposition" suggests that snowpack N and S is
not derived from atmospheric deposition. Change to "inputs of nitrogen
and sulfur from atmospheric deposition".
5-12 18 Lien et al 1992 not in References.
5-28 lines 10-22 This section is difficult to follow. Inclusion of a figure
illustrating the issue associated with the skcwness of the distribution of
critical loads would be helpful so that the reader does not need to go ahead
to section 5.3.2.7 to understand the issue
5-33 Figure 5-9 is very difficult to read. The numerical designations of
ecoregions especially in the dark blue areas are not readable.
84
-------
5-45 Figure 5-13. In its present form it is difficult to distinguish differences
between the one nation versus binary categorization.
5-55 lines 26-29 The statement "Due to lack of direct measurements, no
performance evaluations of CMAQ's dry deposition calculation can be
found; however, the current state of MCI P is the product of research that
has been based on peer-reviewed literature from the past two decades
(EPA, 1999) and is considered to be EPA's best estimate of dry deposition
values" is rather weak and suggests that effort is needed to further evaluate
CMAQ using available information. This issue comes up a number of
times in the document (e.g., page 5-64, lines 23-25).
5-56 lines 22-28 The time unit for these depositions and ratios needs to be
provided. Is this a yearly interval?
5-57 In showing these coefficient of variation values in Figure 5.22, it is
difficult to see the actual values and respective ranges in the Adirondack
and Shenandoah case study areas. Instead of stating that "values are
relatively small", it would be better to provide the means and standard
error of the means of these ratios.
5-58 In Figure 5-20, for sulfur deposition/concentration, how is marine sulfur
accounted for? For sulfur it seems somewhat curious that there is a
change in the isopleths a substantial distance into the Atlantic Ocean. I
would expect the difference if it includes marine components would be
more related to the coastal outline. For this figure, the deposition
component needs a time unit as previously stated.
5-61 to 70 I am concerned with the treatment of sulfur in the AAPI. There is no
consideration in the formulation of sulfur sinks (e.g., soil sulfate
adsorption) or sulfur sources (organic S mineralization, S mineral
weathering, sulfate desorption). Mitchell et al. (2010, Biogeochemistry)
found that watersheds that had previously had substantial portions of
atmospheric S input that from 1985 through 2002 that internal sources
contribute 1-6 kg S ha"1 year"1. This would equal 6 to 37 meq/m2/year.
This contribution is substantial when compared to various analyses
provided in the document (e.g., figures 5-15, 5-18, 5-19, etc.)
5-84 lines 12-23 In considering issues related to recovery there is a need to
not only consider the issues related to weathering of base cations, but also
to internal generation of the mobile nitrate and sulfate anions.
Particularly for sulfate this sulfate will likely result in substantial delays in
recovery in those systems with net losses of soil sulfur and low levels of
base cation weathering.
85
-------
5. What are the Panel's views on staffs revised conceptual framework for the structure of
a multipollutant, ecologically relevant standard for NOx and SOx? To what extent does
the Panel agree that this suggested structure adequately represents the scientific linkages
between ecological responses, water chemistry, atmospheric deposition, and ambient
NOx and SOx?
The conceptual framework for a multipollutant, ecologically relevant standard for
NOx and SOx is sound with considerable support from the scientific literature on
how the generation of strong mobile acids result in the acidification of soils and
water. Some of information, however, is not correct or incomplete. Note for
example the discussion of nitrate sources during snow melt as discussed above for
page 5-5, lines 5-6. Also the assumptions associated with atmospheric sulfur
input being equal to drainage water losses are also not correct (see page 5-61 to
70).
6. What are the Panel's views on the appropriateness of considering a single national
population of waterbodies in establishing standards to protect against aquatic
acidification?
Although having a single national population of waterbodies makes is more facile
to explain the standard, the problems associated with under protecting sensitive
systems and overprotecting insensitive systems necessitates having a system with
more spatial resolution.
What are the Panel's views on consideration of alternative subdivisions of the U.S. to
identify the spatial boundaries of populations of waterbodies and acid-sensitivity
categories, specifically:
a) the use of Ecoregion III areas to aggregate waterbodies ?
This seems to be a reasonable approach that takes advantage of the extensive
information on various ecosystem components including both abiotic and biotic
components.
b) the use of ANC to further aggregate Ecoregion III areas into different categories of
sensitivity?
The use of ANC is consistent with the overall emphasis on the standard to protect
sensitive surface waters from further acidification and have deposition that will
allow those water bodies that have been deleterious impacted by acidic deposition
to recover as indicated by increasing ANC values.
c) the relative appropriateness of the suggested methods for categorizing spatial
boundaries of sensitivity, e.g. on nation, binary sensitive/less-sensitive classes,
cluster analysis based sensitivity classes, and individual ecoregions?
86
-------
The analysis is interesting showing the different distributions of sensitivity and
that categorization that captures substantial variation should be used. This wili
be a compromise between one nation versus using individual ecoregions.
7. What are the Panel's views on the appropriateness of the critical loads that form the
basis for the population assessment to determined deposition metrics?
The use of critical loads has been found to be a useful approach for looking at
spatial and temporal aspects of acidification. This concept was originally applied
to Europe and more recently has been extended to other regions including North
America.
a) What are the views of the Panel on the appropriateness of generalizing the f-factor
approach to apply to lakes and streams in the Western U.S. and other portions of the
Eastern U.S.
The application of the f-factor is a useful approach for evaluating the potential for
mineral weathering to contribute to the generation of base cations and enhance
acid neutralization.
b) What are the views of the Panel on the filtering criteria used to remove lakes and
streams that are naturally acidic or not sensitive to atmospheric deposition?
Yes, it is reasonable to exclude lakes in two of these classes: 1) CL < 10
meq/m2/yr and for which pre-industrial ANC values could not be calculated; and
2) waters affected by acid mine drainage (>400 Deq/L SO42" twice or more than
expected by atmospheric deposition. Lakes with low ANC values (e.g. < 50
Deq/L) and waters with >IO mg C/L DOC those dominated by organic acids
should be included in the analyses and possibly flagged with respect to these
characteristics. Alternatively, the PA should show the consequences related to
excluding or including these with respect to the development of the A API and
critical loads.
8. What are the Panel's views on the suggested methods for determining appropriate
values of reduced nitrogen deposition in establishing NOx/SOx tradeoff curves?
The presentation of the NOx/SOx tradeoff curves is difficult to follow since the
linkages between the various components in the various figures and tables are not
always consistent (e.g. Table 5-7 versus Figure 5-12). Also it would be most
helpful to keep the axes lengths the same in plots within the same figures for
comparisons (e.g. Figure 5-15). I am somewhat concerned with respect to the
sulfate portion of the curve on how systems are handled in which sulfate losses in
drainage waters is not in balance with sulfur deposition. We know that for a
number of sites in the United States that there can either be net retention or net
loss of sulfur. These imbalances can be substantial especially under conditions of
decreasing atmospheric sulfur deposition.
87
-------
9. What are the Panel's views on the revised characterization of the deposition
transference ratios (TNOy and TSOx)?
Implicit in the use of such an aggregated deposition ratio is that the relative
portion of the chemical species in deposition remain constant both with space and
time. This analysis was done for only from 2002 - 2005 and hence may have
covered a range of meteorological conditions, but little difference in nitrogen and
sulfur sources. I don't believe that there is strong evidence suggesting that this is
the case. At a minimum some error analyses associated with this assumption is
needed. This analysis should be expanded beyond the results of CMAQ and to
include other estimates of the proportion of gaseous species and their respective
deposition velocities. It is recognized that this other analysis will include fewer
chemical species than that provided in CMAQ.
10. What are the Panel's views on staffs conclusion that an averaging time of 3 to 5
years is
appropriate given the AAPI form of the standard?
There should be consideration of not only looking at the averages, but also the
minima and maxima for the period of examination. Care will need to be taken on
issues related to the water regimes and other climatic factors among these years.
Droughts or other extreme events such as freezing rain can have a substantial
impact on N and S drainage losses and resultant effects on ANC.
11. What are the Panel's views on the preliminary staff conclusions regarding alternative
target ANC levels that are appropriate for consideration and the rationale upon which
those conclusions are based?
The use of alternative ANC levels is appropriate and based upon sound science
that has shown different levels of sensitivity of various biotic taxa with respect to
sensitivity to low ANC.
a) In light of the Panel's views on the appropriate definitions of adversity to public
welfare (see Chapter 3), what are the Panel's views on the appropriateness of the
information related to adversity considered by staff in evaluating alternative target ANC
levels?
The information provided is adequate for showing at least some of the major
concerns that are documented with respect to public welfare.
12. What are the Panel's views on the approaches considered by staff for assessing
alternative target percentages of water bodies for protection at alternative ANC levels?
This approach is useful in providing a range of water bodies to be covered with
respect to these alternative ANC levels. This also provides flexibility with
88
-------
respect to the administrator regarding choices for protection, overall protection of
public welfare and costs for implementation of the standard.
Chapter 6: Co-protection for Other Effects Using Standards to Protect Against Aquatic
Acidification
13. What are the Panel's views on the utility of the additional analyses of co-protection
benefits to inform consideration of alternative levels of the standard?
This discussion is helpful in showing the linkage of protection between terrestrial
and aquatic components of watersheds and emphasizes as indicated elsewhere in
the document that in general the projection of sensitive aquatic resources results
in terrestrial protection with the aquatic resources being more sensitive to
deposition. One issue, however, that needs some consideration is that the time
frames for the recovery are substantially different from aquatic and terrestrial
components with a greater time lag expected for terrestrial systems.
Chapter 7: Evaluation of Uncertainty and Variability in the Context of an AAPI standard,
including Model Evaluation, Sensitivity Analyses, and Assessment of Information Gaps
14. What are the Panel's views on the following:
a. The degree to which the chapter appropriately characterizes the potential role of
information on uncertainty, sensitivity, and variability in informing the standards?
7-4 lines 13-15 The document states that "Confidence regarding the
fundamental science supporting causal determination about the effects of
acid deposition, and the translation of those efforts into ecosystem services
and values is less amenable to quantification". Even though it is difficult,
providing even some approximate evaluations would be helpful. Some of
this uncertainty may be substantial and having uncertainly analysis focus
on those components for which the calculations are more facile may result
in a misunderstanding of the impact of the proposed standard on human
welfare.
7-6 lines 14-17 In addition to the uncertainties associated with the estimate
of catchment supply of base cations via weathering the exclusion of
sulfate dynamics (or possibly considering a range of internal S supply)
will have a major impact on uncertainty especially associated with future
recovery.
7-8 lines 13-23 Some further elaboration of the Banzhaf survey would be
helpful.
b. The appropriateness and completeness of the evaluation of CMAQ model performance
and sensitivity to critical inputs?
89
-------
7-10 to 24 The inclusion of comparisons of CMAQ and CASTNET results (i.e.,
Figures 7-1 to 7-7) is very helpful. The discussion related to the
limitations of CMAQ (page 7-10, lines 15-26) are insightful and should be
useful in providing future modifications of CMAQ.
7-12 line 1 Is it appropriate to utilize a manuscript in preparation (e.g.,
Dennis and Foley, 2010) for this document?
7-3 lines 12-27 It is indicated that "sensitivity of CMAQ derived deposition
transformation ratios to changes in emissions and treatment of chemistry"
is not yet completed. This should be a high priority for EPA.
7-12 lines 1-7 Do these results suggest that CMAQ needs to be changed
such that precipitation estimates are derived from actual measurements
versus modeled estimates. Isn't this approach more similar to that
employed by the Canadian AURAMS model?
c. The utility of the analyses of temporal and spatial variability in the deposition
transference ratios (TNOy and TSOx)?
7-13 The terms "stiff' and "stiffness" are introduced. Is the use of these terms
identical to "invariate"? In indicating that the absolute values remain
"stable", it is difficult to ascertain how these relative large ranges of ratios
will affect the overall results in using mean Ts and Tn values.
7-14 This comparison to emission change overtime is for only two years (2005
and (2030) and is highly dependent on assumptions of changes in emission
sources. What were the underlying assumptions of these changes? Do
the range in values in Figure 7-12 show the differences based upon these
assumptions?
7-16 lines 14-26 With the continual evolution of CMAQ and likely changes
in the predictions of AAPI, will there be problems in the standard itself
being affected by changes in CMAQ?
7-27 to 28 For Figures 7-11 and 7-12, the figure legend needs to include a description
of the statistical values (mean, ranges, confidence intervals, ??) associated
with these box-and-whisker plots.
15. What are the Panel's views on the insights provided by the AAPI sensitivity analysis
including: a. The evaluation of elasticities of response?
b. The multivariablc ANOVA analysis?
7-30 It is challenging to use the results provided in Appendix A and see how
90
-------
the various analyses are used in this evaluation. A summarization is
needed on the relative sensitivities of the various parameters that make up
the AAPI. An important result is provided by the statement that
emphasizes the need to focus on the uncertainties of the non-atmospheric
inputs, including base cation weathering and runoff rates. As indicated
previously some inclusion of internal generation of sulfate is also needed.
An important outcome of this analysis should to show the parameters of
the AAPI have the most and least confidence. Such information should
be used in driving research and monitoring efforts by EPA.
16. What are the Panel's views on the discussion of uncertainty in the critical loads
models including MAGIC and SSWC?
These descriptions of uncertainty for the model calculations for MAGIC
and SSWC are adequate. A more quantitative term than "moderate5
should be used in describing the uncertainty in SSWC (Page 7-32, line
15). The development of the information in Table 7-2 is helpful in
summarizing the uncertainty associated with the AAPI.
7-33 to 34 For Figures 7-14 and 7-15 within the figure legend it needs to be indicated
that these MAGIC model simulations.
17. What are the Panel's views on the areas for future research and data collection
outlined in this chapter, on relative priorities for research in these areas, and on any other
areas that ought to be identified?
The AAPI needs to include some estimates of the role of internal sulfur sources in
contributing to sulfate in drainage waters. The absence of including this factor
will result in an underestimate of the deposition required to achieve a desired
level of ANC.
Chapter 8: Monitoring
8-1 to 18 The most critical aspect of monitoring is that there needs to be a more
explicit linkage between the monitoring networks and the evaluation and
further refinement of the CMAQ model. This interplay between the
monitoring and modeling efforts will help ensure that both the monitoring
and modeling are most relevant to the environmental issues being
addressed.
18. What are the Panel's views on using an open inlet to capture all particulate size
fractions for the purpose of analyzing for sulfate?
This should be the focus of a research question versus an overall modeling
component.
91
-------
What is your opinion on using existing CASTNET filter packs as a future Federal
reference method for sulfate?
This has considerable advantages with respect to spatial and temporal patterns
since the CASTNET network has been in place for a number of years and
includes a generally good representation of sites across the US.
19. What are the Panel's views on requiring measurements of ammonia and ammonium
to assist implementation of the standard?
There is a clear need to expand monitoring to include measurements of ambient
ammonium and ammonium concentrations. This reduced form of nitrogen is a
major component of nitrogen deposition for many sites including those within
areas with intensive agricultural activities.
20. What are the Panel's views on having a subset (e.g., 3-5 sites) of monitoring stations
in different airsheds that measure for the major NOy species; nitric acid, true NO2, NO,
PAN and p-NO3?
This could be an important research question.
Chapter 9: Conclusions
21. What are the Panel's views on the overall characterization of uncertainty as it relates
to the determination of an ecologically-relevant multi-pollutant standard for NOx and
SOx?
The current document does a commendable job in showing were some of the
major uncertainly lies with respect to the development of a multi-pollutant
standard. Areas that should be targeted for improvement include a more
complete evaluation of the CMAQ predictions and the consideration of additional
processes, especially internal sulfur sources in the AAPI.
22. What are the Panel's views on the following:
a. The insights that can be gained into potential alternative additional secondary
standards (using the AAPI form) by considering:
I. Information from studies on the relationship between mortality in
aquatic organisms and pH and ANC?
ii. Information from studies on the relationship between fish health and/or
biodiversity metrics and pH and ANC?
iii. Information on the relationship between pH, Al. and ANC?
iv. Information on target ANC levels identified by states and regions, as
well as other nations?
Each of these sources of information both separately and taken together
92
-------
provide a compelling case on the relationships between ANC and other
water quality metrics that are associated with biotic health of waters. The
findings from these different studies all provide a rather unified picture
suggesting appropriate ANC values to be the target for the standard.
b. The appropriate role of qualitative and quantitative characterizations of
uncertainty in developing standards using the AAPI form?
As mentioned previously the uncertainty of the AAPI needs to include not only
statistical analyses associated with specific model parameters, but also include an
evaluation of possible omissions in the AAPI (e.g. reduced nitrogen inputs,
contribution of sulfate sources and sinks in soil, etc).
c. The role of considerations regarding the relationship of the standard to:
i. the time trajectory of response, e.g. when specific ANC levels are likely
to be realized given a specific level of the AAPI?
ii. the likelihood of damages to aquatic ecosystems due to episodic
acidification events given a specific target for chronic ANC?
iii. the levels of co-protection for terrestrial ecosystems against
acidification effects and the for aquatic and terrestrial ecosystems against
effects of excess nutrient enrichment?
There may be some problems associated with the time trajectory of the
response due to the understanding and ability to model the relative
contribution of net N uptake and net S loss from the terrestrial portion of
the system. Any factor (e.g. changes in climate, CC«2 concentration in the
atmosphere) could have important effects on the time trajectory. Also, the
effect of factors may be substantially different between aquatic and
terrestrial ecosystems. For example, although for aquatic systems the sum
of base cations may be adequate, for terrestrial systems the availability of
specific base cation, calcium, may be a critical factor in affecting tree
health. Important tree species such as sugar maple have a high demand
for calcium.
23. What are the Panel's views on Staffs conclusion that the existing secondary
standards for NOx and SOx should be retained to provide protection against direct
adverse effects to vegetation due to gas phase exposures?
There is no reason not to retain these existing standards since these concentration
levels will likely be substantially greater than those associated with join NOx and
SOx standards. More importantly the scientific justification is still valid for
protecting against deleterious impacts to vegetation.
24. In light of the Panel's views on what constitutes adverse effects to public welfare (see
Chapter 3), what are the Panel's views on:
93
-------
a) the degree to which current levels of NOy and SOx deposition are adverse to
public welfare based on evidence and risk information, and information on
adversity provided in Chapters 2,3. and 4?
b) target values for ANC that protect against adversity to public welfare in light of
the information presented in Chapter 5 concerning levels of ANC and the
ecosystem effects associated with those target ANC levels?
c) factors relevant in selecting target percentages of waterbodies to protect at
alternative target ANC levels to protect against adverse effects to public welfare,
and weights to place on those factors?
The information provided substantiates that the current levels of NOy and SOx
deposition are producing adverse effects to the public welfare. The target values
selected for ANC are congruent with current scientific understanding with respect
to which ANC values and any resultant change are most sensitive to biotic
components. Selecting a target subset of waterbodies to be protected by
alternative target ANC values is a useful approach.
d) alternative standards for NOx and SOx that would protect against adverse effects to
public welfare based on the AAPI form, and taking into account
• consideration of target levels of ANC (chapter 5),
• target percentage of water bodies to protect (chapter 5),
• consideration of relevant uncertainties in AAPI components (chapter 7), and
• any other potentially relevant factors, such as levels of co-protection against
terrestrial acidification and nutrient enrichment (chapter 6)?
It may be important to consider alternate standards especially for protecting those
systems where nutrient enrichment (e.g. western U.S.) is a substantial effect
associated with N deposition.
94
-------
Comments from Mr. Richard Poirot
5. What are the Panel's views on staffs revised conceptual framework for the structure of
a multipollutant, ecologically relevant standard for NOx and SOx?
The revised conceptual framework for the structure of the multi-pollutant secondary
standard has been substantially improved from the first draft policy assessment. The
inherently complex framework is more clearly presented and more carefully justified,
with revisions that are directly responsive to previous review comments.
To what extent does the Panel agree that this suggested structure adequately represents
the scientific linkages between ecological responses, water chemistry, atmospheric
deposition, and ambient NOx and SOx?
The proposed structure adequately reflects the current state of scientific understanding of
the complex linkages between ambient concentrations of SOx and NOx, wet and dry
deposition of these and other acidifying pollutants (i.e. NHx), environmental processing
of these deposited S and N compounds, resultant changes in surface water chemistry, and
subsequent ecological effects.
6. What are the Panel's views on the appropriateness of considering a single national
population of waterbodies in establishing standards to protect against aquatic
acidification?
Use of a single national population of water bodies as the basis for selecting (a percent of
water bodies to be protected from reaching or maintaining a specific minimal ANC
component of) a national standard has the "advantage" of "simplicity". But a large
fraction of national surface waters are located in areas where underlying soils, bedrock
and other local environmental factors effectively preclude adverse acidification effects
from past, current, and expected future deposition rates of S and N, while other water
bodies are extremely sensitive to effects from relatively low rates of S and N deposition.
Use of a single national population and associated percentage level of protection
unnecessarily disregards the large regional variations in inherent sensitivity to
acidification, and is likely to lead to under-protection in some areas and over-protection
in others (or both).
Since there are various methods and data available to allow refined estimates of inherent
sensitivity to be calculated on a regional basis, and since many other location-specific
environmental variables are included in the calculation of compliance with the proposed
standard, it makes sense (I think - but need more info) to consider protection in the more
refined context of the populations of water bodies at risk from acidification effects.
What are the Panel's views on consideration of alternative subdivisions of the U.S. to
identify the spatial boundaries of populations of waterbodies and acid-sensitivity
categories, specifically:
a) the use of Ecorcgion III areas to aggregate waterbodies ?
95
-------
Not my area of expertise, but this seems like a reasonable approach, and possibly one that
could be considered for refining secondary NAAQS for these and other criteria pollutants
in the future. Offhand, it seems like using 120+ different Ecoregion III categories for
aggregating water bodies is unnecessarily complex. However, it also appears that there
are reasonably ways to simplify, group or sort these many categories into a much smaller
number of Ecoregion subsets which are inherently sensitive (or most sensitive) to
acidification, and which would make for a more efficient standard better focused on
protecting those systems at greatest risk.
b) the use of ANC to further aggregate Ecoregion III areas into different categories of
sensitivity?
This seems like a logical (almost obvious) metric for sorting/grouping the Ecoregion
categories. If only we could just use the readily measured direct ANC indicator of effects
as the NAAQS indicator...
c) the relative appropriateness of the suggested methods for categorizing spatial
boundaries of sensitivity, e.g. on nation, binary sensitive/less-sensitive classes, cluster
analysis based sensitivity classes, and individual ecoregions?
I don't have a strong opinion on the relative appropriateness of these alternative
approaches. None of them seems inappropriate. Offhand, I think 1 like the cluster
analysis approach, for its inherent scientific merit, its direct focus on sensitivity, and the
relative simplicity of a 5-class grouping scheme (especially for the initial roll-out of an
extremely complex NAAQS).
However, I also don't think that the advantages/disadvantages/ consequences of the
various options (2a, b, c, d) are presented here with sufficient clarity to allow an informed
choice by the Administrator (or by me anyway). Hopefully, these options can be
presented, discussed and illustrated more clearly in the final PA document, and staff
might propose and defend a preferred option. For example, the page 5-50 statement (and
associated figures) that "In option I, the Adirondack air quality is slightly out of
attainment for a 75%-tile deposition metric based on a CL at ANC=50. In option 2a, the
Adirondack air quality is out of attainment for the curve for the sensitive areas, but in
attainment for the less sensitive areas." helps convince me that option 2a is preferable to
option 1. But I don't have a similar feel for the relative strengths or weaknesses of the
other options.
8. What are the Panel's views on the suggested methods for determining appropriate
values of reduced nitrogen deposition in establishing NOx/SOx tradeoff curves?
Since reduced nitrogen in the air or in (dry) deposition is not currently measured, and not
currently considered as a regulated component of the NAAQS, but does contribute to the
acidifying (and N enrichment) effects of SOx and NOx deposition, I think it's reasonable
to estimate its location-specific deposition with CMAQ. At the same time, there is also a
need to verify and refine the CMAQ estimates with direct measurements, especially for
NH3.
96
-------
I also strongly support the proposed approach to consider NHx as a temporally varying,
location-specific component of the AAPI calculation. This is a scientifically preferable
approach to the previous proposal which would have considered NHx deposition as a
fixed constant. If NHx increases, larger reductions in SOx and NOx would be required,
and conversely, if NHx decreases, SOx and NOx reductions would be lower.
9. What are the Panel's views on the revised characterization of the deposition
transference ratios (TNOy and TSOx)?
So far as I can tell, the (CMAQ) methods for calculating these deposition transference
ratios are the same as they were in the last draft PA, but are described, illustrated and
evaluated more clearly in the current document. These transfer functions are logically
conceived, but seem like such critical elements of the proposed standards, which are
uncomfortably dependent entirely on CMAQ model performance. The illustrations
(Figures 5-20 - 5-22) showing the spatial characteristics are helpful, and the illustrations
(Figures 7-1 1 and 7-12) showing that the transfer functions remain stable over time with
large changes in emissions add some confidence. Although since the model chemistry is
fixed, S and N species totals are conserved, and meteorology held constant, highly
variable modeled results would not be expected. Some additional confidence might be
provided by comparing CMAQ estimates of dry and total deposition (wet is already
shown) at selected CASTNET sites in recent years, and perhaps breaking out the Figure
7-3 performance for TNOs into separate figures for particulate nitrate (which deposits
inefficiently) and
One additional analysis that might be informative would be to calculate and evaluate a
modified TNOY function (call it TNOY*) that would be based on CMAQ modeled total N
deposition as a joint function of CMAQ HNOs and pNOs (separate coefficients could be
derived for each species). This empirically derived relationship would be no more of a
"black box fudge factor" than the current TNOY calculation (ratio of CMAQ estimate of
total N deposition to CMAQ estimate of NOy). Potential benefits of this approach are
that it would be less dependent on CMAQ's ability to accurately predict and apportion all
the separate NOy components (with their widely different deposition velocities); it can be
applied (as can the sulfur TSOX function) to currently available and relatively low cost
CASTNET measurement data; and the measured species would directly represent major
components of dry N deposition, compared to NOy, which has no relationship (R = 0.067
in Figure 4-21) without benefit of the black box CMAQ conversion. In evaluating
whether this alternative approach is "close enough" to the original TNOY, both
calculations could be compared to both the CMAQ estimates and CASTNET (+ NADP)
measurements of total N deposition.
12. What are the Panel's views on the approaches considered by staff for assessing
alternative target percentages of water bodies for protection at alternative ANC levels?
As indicated previously, I think the alternative approaches seem reasonable, and that the
objective should be to focus as tightly as possible on protecting water bodies that are
inherently sensitive to acidification, without adding too much complexity to the
97
-------
regulatory metric. It might also be recognized in this case that a metric that provided
some "over-protection" in areas less sensitive to aquatic acidification might provide
added protection for terrestrial ecosystems or against nitrification and would unavoidably
improve visibility and reduce mortality and morbidity associated with PMis
concentrations.
13. What are the Panel's views on the utility of the additional analyses of co-protection
benefits to inform consideration of alternative levels of the standard?
This seems like a reasonable concept to explore in more detail, although I don't really see
any discussion of this in Chapter 6. It seems clear from the analysis that there are areas -
without surface waters or with relatively insensitive surface waters where adverse affects
on terrestrial ecosystems are expected - and for which adding a "co-protection" element
to the standard would provide added benefits. If other welfare effects of SOx and NOx -
such as on materials damage and visibility had been considered in this review, the co-
protection benefits would have been substantial.
14. What are the Panel's views on the following:
a. The degree to which the chapter appropriately characterizes the potential role of
information on uncertainty, sensitivity, and variability in informing the standards?
The additional information and discussion uncertainty, sensitivity, and variability in
Chapter 7 is extremely helpful, and represents a major improvement to the previous draft
PA.
b. The appropriateness and completeness of the evaluation of CMAQ model
performance and sensitivity to critical inputs?
While various CMAQ model performance evaluations have been presented elsewhere,
the model performance evaluations and sensitivity analyses presented here are most
helpful. Since pNO3 and KNO3 have such different deposition velocities and are
measured separately in CASTNET, it might be informative to show comparisons of the
separate modeled species and CASTNET measurements, and perhaps also for the CMAQ
and CASTNET estimates of dry deposition of the separate pNOs and HNOs species, as
well as for the CMAQ and CASTNET estimates of TNO3 dry deposition.
c. The utility of the analyses of temporal and spatial variability in the deposition
transference ratios (TNOy and TSOx)?
I have a hard time understanding what the spatial variability in these transfer ratios
actually means, though it is comforting to see that the patterns seem relatively "smooth"
rather than abrupt. Is there is seasonal or diurnal variability in these ratios that might
give us a better feeling for what's really going on inside the model (and in the
atmosphere)? I wonder if it would be informative to see maps analogous to Figure 5-20
which separately showed the ratios of S cone to S wet dcp and S dry dcp, and of N cone
to N wet dep and N dry dep. Maps showing the ratios of S and N deposition to S and N
emissions (perhaps aggregated on a state by state basis) could also be interesting...
98
-------
The illustrations of the relative absence of temporal variability are comforting, as it is key
to have stable regulatory metric which is linearly responsive to emissions changes over
time. Some (any) discussion which helped explain the causes and implications (if any) of
some of the spatial or temporal variations would be helpful.
17. What are the Panel's views on the areas for future research and data collection
outlined in this chapter, on relative priorities for research in these areas, and on any other
areas that ought to be identified?
I thought this section of Chapter 7 was especially well done, and well supported by the
preceding discussions. A chapter like this should become standard practice in future
NAAQS reviews!
18. What are the Panel's views on using an open inlet to capture all particulate size
fractions for the purpose of analyzing for sulfate? What is your opinion on using existing
CASTNET filter packs as a future Federal reference method for sulfate?
I don't oppose these proposals, although I think the case is somewhat overstated,
especially in relation to aquatic acidification effects. A major concern is that this would
require exclusive use of CASTNET methods or network and preclude use of fine fraction
sulfate measurements which are more abundant, and not demonstrably grossly inferior.
Conversely, there's no reason not to include a similar open inlet approach for pNOs, for
which coarse particle deposition may be especially important forN deposition
contributions to nutrient enrichment of coastal estuaries. I also think an argument could
be made to consider CASTNET HNOs and pNOs as an (interim) alternative to continuous
NOy measurements (more detail on this below and in #9 above).
Some counterpoints to the open faced sulfate proposal:
• Away from coastal areas with coarse sea salt or arid or agricultural areas with
windblown dust, (and especially in the remote humid, high elevation areas where
acidification occurs) there is relatively little coarse sulfate (or coarse nitrate) period.
• The particle cut size characteristics of the open faced collectors have not been well
characterized, nor is any information provided on what (small) fraction of the open
faced S or N sample is composed of more rapidly depositing coarse mode particles.
You need to add a fudge factor, which you could do just as well using fine fraction
data.
• Open-faced collectors may take in fog or cloud water. In addition, since coarse
particles tend to be alkaline, additional artifacts may occur as gaseous SC>2 or HMOs
reacts with the alkaline coarse material collected on the sample filter.
• Sulfate and nitrate in coarse mode particles which are formed in the air (and are not
sampling artifacts) typically result from reactions of acidic S and N gases and alkaline
crustal material or sea salt. Consequently, these particles carry their own cations and
represent uniquely well buffered forms of S and N deposition.
• Lastly, the total fractions of S and N depositions from particulate matter- especially
in the higher elevations where acidification is an issue - is not very large. Below are
the 2006-08 estimates of total S and N deposition for the Huntington Forest
CASTNET site (relevant to deposition in the Adirondack case study area). Total
99
-------
participate sulfate and nitrate were estimated to account for 2.1% and 0.2% of the
total S and N deposition respectively.
As indicated above, I'm not opposed to the proposal to specify an open faced FRM or
even the CASTNET filter pack for pSO4, but think it could also be specified for pNOj
(assuming problems with loss of volatile NOs during summer sampling can be corrected),
but also think accommodations could be made (FEM) to accommodate use of fine
fraction 864 and NOa data (with adjustments) to avoid being too prescriptive at this early
stage of the NAAQS process, to mandate a compliance network (CASTNET) which is
not operated by states but by EPA contractors (funded by $ taken from state monitoring
pots), or - in combination with the proposed continuous NOy indicator - to require
deployment of a costly new network which may not be currently feasible, or which might
indefinitely delay implementation of the NAAQS.
Total Wet and Dry Deposition of S and N at Huntington Wildlife Forest. 2006-2008
Particulate Matter Deposition Accounts for 2.11% S and 0.23% N Respectively
Composition of S deposition tor 2006-2008 ComposHion of N deposition for 2006-2008
HWFH7
19. What are the Panel's views on requiring measurements of ammonia and ammonium
to assist implementation of the standard?
and NH4 measurements would be useful for implementing the sample both directly,
to quantify an unregulated but varying element of the compliance metric, and indirectly,
to help evaluate and improve emissions inventories and CMAQ model performance.
NH4 measurements are currently available from CASTNET and (urban) CSN networks,
and could conceivably be added to IMPROVE. NH^ measurements are currently very
sparse but would be useful - and have added relevance to better understanding sources
and trends of PM25, regional haze, and sources and effects of N deposition on nutrient
enrichment. However, I'm not sure current methods have been sufficiently well
developed and evaluated for use in routine network operations.
100
-------
20. What are the Panel's views on having a subset (e.g., 3-5 sites) of monitoring stations
in different airsheds that measure for the major NOy species; nitric acid, true NO2, NO,
PAN and p-NO3?
Conceptually this is a good idea and can be (and needs to be) justified for reasons beyond
just compliance with the proposed NAAQS (for acidification effects). Possibly some of
these measurements could be added to existing (or planned) rural NCore sites. NOy, NO
and pNOs (fine fraction), SO2 (continuous) and SO4 (fine fraction) are currently
measured at these sites. Add a CASTNET sampler and you've got HNOs, Open faced
(vs fine) SO4 and NOs (^ you Wiii know the coarse fractions), and comparative SO2 by filter
pack and continuous analyzer. Adding true NO2 would be an excellent addition at some
sites (1 understand there's a photolytic unit currently applying for FEM status), and this
would allow calculating NOy minus (NO, NOi, pNOs, HNOs)...
However, while 1 support the need for these kinds of more detailed measurements at a
few sites in a "clustered network" approach, I'm not sure they can be or should be
justified just to determine compliance with this secondary NOx/SOx standard or for
evaluation and improvement of model performance just for this standard alone. Along
similar lines, I'm not sure a large new network of continuous NOy analyzers (at new,
remote rural sites) can be justified (or can be afforded, or could be maintained by
shrinking numbers of state personnel). I'm uncomfortable with the relatively vague
picture of how these new measurements would be conducted. Will this be a state-
operated network (like NCore), an enhanced CASTNET network (operated by EPA
contractors), an enhanced IMPROVE network - or some combination of the above?
I think a reasonably good argument could be made to specify CASTNET filter pack
methods (possibly with some tweaks such as adding Nhh passive sampler) as the basic
monitoring approach, as it does capture the key species - albeit over longer averaging
times but which are plenty short enough for the long-term 3-5 year standard. As
indicated in the response to question 9 above, 1 think an alternative TNOS* N deposition
transfer function could be developed that would calculate total (CMAQ modeled; or
CASTNET + NADP measured) NOx deposition as a function of HNO3 and pNO3. If
these calculations performed reasonably well, it would allow use of existing and
relatively low cost data, use measurements which actually relate to deposition (NOy does
not, without a huge assist from the model), and minimize the reliance on complex internal
CMAQ calculations. Using this approach, a slightly expanded CASTNET filter pack
network might become the initial compliance network, with new CASTNET samplers at
a limited number of rural NCore or IMPROVE sites. Continuous NOy could be added to
a few of these sites (NCore sites have continuous NOy and SO2, as well as fine particle
SO« and NOs), but should not be required in a large new network. More exotic
measurements such as true NO2, PAN, continuous nitric acid, etc should be considered at
only a very few "Level 1" type sites.
23. What are the Panel's views on Staffs conclusion that the existing secondary
standards for NOx and SOx should be retained to provide protection against direct
adverse effects to vegetation due to gas phase exposures?
101
-------
I agree that existing single-pollutant secondary standards forNO2 and SC>2 should be
retained to protect against direct effects to vegetation due to gas phase exposures.
24. In light of the Panel's views on what constitutes adverse effects to public welfare (see
Chapter 3), what are the Panel's views on:
a) the degree to which current levels of NOy and SOx deposition are adverse to public
welfare based on evidence and risk information, and information on adversity provided in
Chapters 2,3, and 4?
I believe the evidence and risk information provided in previous chapters indicates that
environmental damage has occurred and continues to occur as a result of cumulative and
continuing SOx and NOx deposition, and that these effects, including acidification of
aquatic and terrestrial ecosystems and nutrient enrichment, represent adverse effects on
public welfare.
b) target values for ANC that protect against adversity to public welfare in light of the
information presented in Chapter 5 concerning levels of ANC and the ecosystem effects
associated with those target ANC levels?
ANC is an appropriate environmental indicator of effects from acidification on aquatic
ecosystems, and the target levels of ANC being considered - about 50 ^icq/L - would
represent an appropriate target level that - if attained - could be expected to substantially
reduce the adverse welfare effects due to aquatic acidification. A somewhat higher ANC
target of say 75 or 100 u.eq/L would provide a greater degree of protection for both
aquatic and terrestrial ecosystems, although the degree of protection is co-dependent on
the target ANC and the target percentage of water bodies (in what sensitivity class or
classes) for which the target ANC is to be attained.
c) factors relevant in selecting target percentages of waterbodies to protect at alternative
target ANC levels to protect against adverse effects to public welfare, and weights to
place on those factors?
Since there can be large variability in the inherent sensitivities of water bodies to
acidification effects among different regions and within individual regions, it seems
logical to consider protecting a target percentage of lakes from the populations which are
potentially susceptible to acidification. The proposed use of Ecoregion 111 classifications
clustered into 5 groups on the basis of ANC seems like an appropriate accommodation of
scientific detail without adding unnecessary complexity. As indicated previously, I had a
hard lime understanding the implications of the different proposed options for selecting
target percentages of water bodies for protection. Comparison of the various metrics
applied to case study regions where there are sensitive and insensitive lakes which are
chemically well characterized would be a useful way to judge the appropriate
combinations.
d) alternative standards for NOx and SOx that would protect against adverse effects to
public welfare based on the AAPI form, and taking into account
102
-------
• consideration of target levels of ANC (chapter 5),
• target percentage of water bodies to protect (chapter 5),
• consideration of relevant uncertainties in A API components (chapter 7), and
• any other potentially relevant factors, such as levels of co-protection against
terrestrial acidification and nutrient enrichment (chapter 6)?
As indicated above, I suggest considering a modification ( TNQY*) to the N deposition
transfer function such that it would not require extensive new measurements of
continuous NOy. I'm not sure it would work well enough, but think it could be
considered.
Considering the relatively long time frames associated with acidification and recovery,
and also considering that there is no pre-specified time frame for attaining secondary
standards, some consideration might be given to standards which require reasonable rates
of progress over time toward increasing ANC levels (APPI levels) in water bodies
(watersheds) with low ANC levels.
103
-------
Comments from Dr. Armistead Russell
EPA staff is to be complemented for this work towards developing a Policy Assessment
document (PAd or PA) that can be used to support the promulgation of an ecologically-
relevant, combined NOx-SOx standard. The PA has evolved considerably since our last
review, and the 2nd draft shows that a significant amount of work and additional thought
have gone in to its further development. This undertaking demonstrates just how
complex it may be to develop multipollutant standards. They are also to be
complemented for addressing CASAC's prior comments. They have gotten very far
addressing a very complex issue. However, the current document is noticeably not as
informative as desired. It is difficult to get the "big picture" of the impacts of choosing
different elements of the potential standard. There is little doubt in my mind that this is
the most difficult PA (or equivalent document) that I have reviewed.
In the current document, having a clear and comprehensive description of the AAPI is
key, as well as the associated components of the AAPI, how the AAPI would be applied,
and the consequences of various decisions about the AAPI level, ecoregions and percent
of lakes protected, and this makes Chapter 5 a key chapter. At present, however, it does
not provide the material needed for someone to read the document and get a clear
understanding of all of the concepts, and the tremendous complexities, without a
significant amount of work. In part, more data (or distillations of the data) are needed to
give the reader an idea of just how big and varied are the various quantities that are being
used (e.g., distributions of key variables used to compute the AAPI). Second, sample
calculations could be shown. A particular weakness is a demonstration of how the
fraction of lakes protected interacts with the choice of ANC. The skeleton is there, but
not enough. For example, a more complete demonstration of what went in to making
Table 5-12 would be very useful. (Also, as is true in a number of places, some of the
variables are not defined or ambiguous, e.g., DLo/0£co: is it for NOy+S (I think) or N+S).
How do the calculated DL%ECO'S compare with the current estimate of its level? How,
specifically, do you find the sites protected from ANC<20 in this case? Further, it would
be of interest to see a distribution of DL%ECo current vs. distributions of DLo/oECo at one of
the candidate control levels (e.g.. ANC 50. %Prot.=75), along with a description of the
decrease in DL%eco (again, potentially a frequency distribution). (Figs. A11,16 & 20 are
informative, but not as so). It might be even more insightful if a spatial distribution of
the control levels were able to be shown (again as % reduction in DL%ECO)- It may be
necessary to have a set of assumptions for developing that spatial map, and those should
be clearly described. It is difficult to see how the administrator and CASAC can provide
their best guidance on the level of the standard without further understanding of the level
of control that would likely be required to meet various combinations of the
level/form/ecosystem choices made to specify the standard. In terms of advising the
Administrator, one of the sections that should be strengthened is the one on the approach
to defining/choosing ecosystems. While this section is substantial, it was difficult to
distill.
I appreciate the chapter on uncertainties. In this analysis, more than those conducted for
other standards, the uncertainties in the modeling have to be put forth in a very
transparent fashion because those uncertainties impact directly on the translation of the
104
-------
estimated depositional flux to the monitored quantity. They are not small. The
magnitude of such uncertainties (e.g., in the transfer ratios) should be quantitative or
semi-quantitative. Having an "unknown" for that uncertainty is a weakness of the current
document, and indicates an area needing intense assessment. Air quality modeling
uncertainty is important to the overall viability of the approach, impacting not only the
transfer ratio, but also the estimated NHs flux.
The Executive Summary does provide a reasonable overview of the rest of the PA, and is
a valuable component of the PA. However, it does suffer from the same elements from
which the rest of the PA suffers (e.g., see above for discussions of what is needed to
bolster Chapter 5). It also suffers from trying to overly simplify the complexities of the
proposed approach. In particular, the AAP1 equation really should be included, with
explanations of the origin and importance (including magnitude) of each term. Symbols
should be defined and figures should have explanatory captions. I would add more
headings to show the steps in the conceptual design of an ecologically-relevant standard.
I would also note that at this time the secondary standard is most strongly supported by
the demonstrated relationship between ambient NOx and SOx and aquatic acidification.
Like Chapter Five, there is a need to provide more information as to the consequences of
making various decisions about the choice of components of the standard.
It is interesting that this PA does not include staff recommendations as to a range of
AAPI, % protected or choice of ecosystem approach for the standard (i.e., the elements of
the standard upon which we are supposed to give guidance). Similar information has
been provided in prior reviews, and that information has been very useful to assess the
reasoning behind the choices. As part of our review, it would have been very helpful to
see similar information, particularly from people who have been intimately immersed in
doing the analyses. This is particularly true given the complexity and novelty of the
approach. Further, having only had a rather limited time to review the PA magnifies the
problem.
105
-------
Comments from Dr. David Shaw
This draft demonstrates a marked improvement over the first draft and 1 feel it is
responsive to many of the CASAC member comments. The addition of details with
models and uncertainties has resulted in a more informative document.
While this assessment seems to touch on the need for meaningful data from other regions,
and it is more specific about what parameters need to be measured to guage the standard,
it does not address the resources needed to expand air monitoring into these other
regions. I must emphasize the need for data. While this proposed NAAQS is innovative
and I appreciate the efforts being made to identify an appropriate NAAQS, it is also
model dependent and because of that, it calls for commitments to get better data for
future analysis. I am concerned that the USEPA is taking steps towards ranking model
data higher than monitored ambient data, and I want to ensure that this is not the direction
which NAAQS will take. I believe that real ambient data should be considered in higher
regards than model data.
J feel that the PA is a report that should be more readable and user friendly than the
highly technical ISA and REA, and in that vein I suggest that it be clear what each
indicator represents. For example, the ANC for lakes is relatively easy to measure and
therefore represents a large amount of available data, but it doesn't represent streams. In
addition, ANC data is typically a summer target which leaves us dependent on models to
estimate or adapt for year round use. As a result, 1 suggest that when ANC is discussed
in the broad context, it should be prefaced as lake surface water ANC, not to be confused
with stream ANC levels which were not evaluated. On that same note, I suggest that the
base cation to aluminum ratio (Bc/AI) be specified as soil water.
Flexibility
There is limited data in regions that currently do not have sufficient monitoring data or
modeling efforts to characterize their own sensitive ecosystems at this point. Therefore, I
suggest that sufficient flexibility be build into the policy to allow for future
monitoring/modeling efforts and characterization.
Models & Data
There are several issues with a heavy reliance modeling. There are differing levels of
uncertainties associated with each of the models (thank you for including the updated
document on 9/23/2010). One example of these uncertainties is that it is very difficult to
assess the dry deposition estimates in CMAQ. As a result, how much confidence do we
have in the NOX/SOV transference ratios which are based on modeled deposition?
The PA states that the current monitoring networks (IMPROVE, CASTnet, and
NADP/NTN) are not adequate to cover all sensitive areas while Chapter 5 suggests that
CMAQ will be used to help develop the spatial patterns needed to create the NAAQS.
Without sufficient measurements of ambient NOy and 862 in sensitive areas is a serious
106
-------
limitation, again leading to the conclusion that a clear commitment be developed to
provide adequate data.
Specific Comments
Brook trout is listed as a sensitive species, it is generally not. 1 suggest using a more
general term like fish which is more accurate. This is also comparable to zooplankton
which is not specific either.
ES 12 bullet 1. The statement "at least as protective" does not seem to be appropriate.
The secondary standard should is held separate to be protective of human welfare. We
most likely will identify different areas of the nation which have different sensitivities.
Perhaps something could be added at the end of the sentence to say in the nation, and
more protective than the current standard in several parts of the US.
ES- 10 figure caption, delete stream since this model is based on lakes only.
ES 12 bullet 2. Specify lake surface water ANC.
ES 12 bullet 3. Specify soil water Bc/AI values
ES-13 bullet 4. Change to "Less protective against species mortality during acidification
episodes"
ES-13 bullet 6. Change may to will. Substitute "fish" for brook trout, because brook
trout are not generally considered a sensitive species.
9-18 line 13 change 'may' to 'would'; change 'brook trout' to 'fish'.
A-31-32 Alkalinity section. Text and table need more explanation.
A-34 Line 8. Simplify the explanation, something to the effect that in glaciated areas, the
parent material over the bedrock (e.g. glacial till) has been deposited miles from its
origin. The soils develop from these parent materials and can be very different from the
bedrock.
B-4 Table 1. Suggest identifying the lakes by name, instead of or in addition to their ID
number. This would improve the readability and connect the reader to the landscape
feature. Likewise with naming the Shenandoah streams (Table 2). Same for Tables 6, 7, 8
&9.
B-l 1. Line 29. Change 'lakes" to 'streams'.
107
-------
11/17/10 Draft text for review at the chartered CASAC Public Teleconference 12/06/2010 Teleconference-- Please
Do not Cite or Quote -
This draft is a work in progress, does not reflect consensus advice or recommendations, has not been reviewed or
approved by the chartered CASAC and does not represent EPA policy
Comments from Dr. Kathleen Weathers
Clearly EPA staff has put considerable effort into the conceptual design, analyses, associated
uncertainties, and various policy options outlined in this 2nd version of the PA. The stated goal
of developing a conceptual design for a standard has been significantly advanced. Many kudos.
It is also true that the timeline was tight for generating this (substantial) revision, and the
resultant text is rough and unclear in many places, and lacks well-placed, succinct descriptions
of, for example, the AAPI and its component parts (the Introduction notwithstanding; Chapter 5
was a challenge in places).
A few general comments:
That the importance of NHx as a source of pollution/nutrient addition to ecosystems is explicitly
considered is an important improvement, even if it must be taken back out of the calculations
because of the current regulatory focus on NOx and SOx only.
Perhaps best summarized by the statement on page 5-69: "A critical issue for the Administrator
to address in setting a NAAQS based on the AAPI is to consider and weigh the varying degrees
of uncertainty in establishing the elements of the AAPI. These uncertainties impact the
likelihood that a specific AAPI standard would in fact achieve a target ANC level for a specified
percentage of a population of water bodies." While the addition of Chapter 7 is most welcome, I
still have some concerns with the large and important influence of CMAQ deposition estimates
and as [BC]o, in particular, on the AAPI. Some additional analyses, comparisons, and
explanatory text would be helpful in at least bounding these concerns.
For example, some comparisons and analyses that would be likely to either increase or bound
confidence in, or understanding of— CMAQ deposition estimates are still lacking in this
document. 1 do not object to using CMAQ here, in fact, as I said in my comments with the first
draft, 1 think it is appropriate, but since this model is so fundamental to the standard setting, it is
critical that the PA is clear about model details, caveats and how it compares to other estimates
of deposition (e.g., NADP wet, CASTNET modeled deposition, etc). Despite the fact that
CASTNET also models (vs measures) dry deposition, it is instructive to compare the two
modeled dry deposition estimates. Similarly CMAQ (or CMAQ/PRISM, as the case may be) vs
NADP wet deposition should also be compared. Also, for the case study areas, what are the
current deposition estimates and how do they compare to NADP + CASTNET for these
locations? 1 note as well that some of the discussions about the use of CMAQ and its
uncertainties often appear defensive, or unnecessarily offensive. And, please, wherever the
CMAQ model appears in the text, include the version used.
In regard to the importance of BC (7-30/31) ("largest uncertainty and most influence on the
CL"), is there any confidence that can be gained by doing spatial aggregation based on BC? It
was dismissed along with bedrock geology by the staff as a way to subdivide or aggregate
waterbodies?
108
-------
11/17/10 Draft text for review at the chartered CASAC Public Teleconference 12/06/2010 Teleconference- Please
Do not Cite or Quote -
This draft is a work in progress, does not reflect consensus advice or recommendations, has not been reviewed or
approved by the chartered CASAC and does not represent EPA policy
In regard to spatially aggregating or dividing waterbodies, The Ecoregions include vegetation
and vegetation influences both deposition and the processes involved in Neco, for example.
Therefore, I favor them over the one nation approach. That said, 70+ regions seems rather
unwieldy.
I think that a few more categories should be added to the monitoring section if a goal is to be
able to evaluate the efficacy of the standard. Surface water ANC, and soil monitoring should be
included along with field deposition measurements that can extend networks, such as NADP and
CASTNET.
A few more specific items:
I'm confused by the first paragraph on page 5-85 in regard to the timeframe consideration. While
it is of course important, I'm not exactly sure what, if I were the Administrator, I would do with
this statement.
5-53: For specified regions (high elevation watersheds along the Appalachians, for example),
occult precipitation can represent a large fraction of N and S deposition, larger than precipitation
or dry deposition. I suggest modifying the statement and referencing, for example, MADPro
results. Perhaps it is the case that occult precipitation is insignificant at the scale of CMAQ
model output, which would be important to note.
7-13: Shenandoah and the Adirondacks are regions of reasonably high heterogeneity with respect
to variables that are likely to influence not only the deposition of N and S but all the other
parameters in the AAPI index, thus while the statement may be true for large grid cell sizes, it's
unlikely to be true over the relevant ecological scales.
9-4 and 9-5—here is an example of a Faulkner-like paragraph whose meaning is completely
opaque to me (but I think that there's important information there)
Charge Questions:
Chapter 5 (needs work to make it clearer).
7. What are the Panel's views on the appropriateness of the critical loads that form the basis for
the population assessment to determined deposition metrics?
The critical loads concept has been used by EU countries for quite some time; I think it a good
and tested framework for considering ecosystem effects of air pollution and around which to
build a secondary standard. I find the aggregation method proposed (ES-7) compelling.
a) What are the views of the Panel on the appropriateness of generalizing the f-factor
approach to apply to lakes and streams in the Western U.S. and other portions of the Eastern U.S.
The f-factor did not seem well described to me, despite the fact that there were a few places
where it showed up in the text. I wished for a "so what" statement on 5-23 with regard to
estimation of [BC]o. and what it might mean within the AAPI. So. I'm not sure exactly what the
109
-------
11/17/10 Draft text for review at the chartered CASAC Public Teleconference 12/06/2010 Teleconference-- Please
Do not Cite or Quote -
This draft is a work in progress, does not reflect consensus advice or recommendations, has not been reviewed or
approved by the chartered CASAC and does not represent EPA policy
question is here—is the f-factor likely to hold for other locations, or does it matter whether lakes
or streams are the target freshwater (given the divergences in Fig. 5-6b)? As noted above, the
[BC] is a critically important variable.
b) What are the views of the Panel on the filtering criteria used to remove lakes and
streams that are naturally acidic or not sensitive to atmospheric deposition?
These are important and appropriate criteria on which to filter freshwater systems from the
database. This filtering is responsive to the Panel's previous recommendation.
8. What are the Panel's views on the suggested methods for determining appropriate values of
reduced nitrogen deposition in establishing NOx/SOx tradeoff curves?
As with all of the deposition estimates, CMAQ is the source, and there aren't really any other
great options. NADP chemistry and PRJSM precipitation might be used to get at wet inputs (or
compare them with CMAQ). The inclusion of NHx is important.
CHAPTER 9
23. What are the Panel's views on Staffs conclusion that the existing secondary standards for
NOx and SOx should be retained to provide protection against direct adverse effects to
vegetation due to gas phase exposures?
That there are adverse effects to vegetation as a result of exposures to SOx and NOx has been
established and is reiterated in this document. However, isn't it likely that the proposed new
standard would result in concentrations at, or lower than the current concentrations, and therefore
be protective? Or, is the question about whether these existing secondary standards should
remain in place in addition to a deposition standard? (I do think that protection against adverse
effects to vegetation due to gas phase exposures should continue.)
24. In light of the Panel's views on what constitutes adverse effects to public welfare (see
Chapter 3), what are the Panel's views on:
a) the degree to which current levels of NOy and SOx deposition are adverse to public
welfare based on evidence and risk information, and information on adversity provided in
Chapters 2,3, and 4?
Research has shown that acidifying atmospheric deposition at its current levels results in
ecosystem functional changes, and aquatic acidification, in sensitive ecosystems; the information
in Chapters 2-4 reinforces this.
b) target values for ANC that protect against adversity to public welfare in light of the
information presented in Chapter 5 concerning levels of ANC and the ecosystem effects
associated with those target ANC levels? These levels are supported in the literature; they are
defensible and appropriate.
110
-------
11/17/10 Draft text for review at the chartered CASAC Public Teleconference 12/06/2010 Teleconference- Please
Do not Cite or Quote —
This draft is a work in progress, does not reflect consensus advice or recommendations, has not been reviewed or
approved by the chartered CASAC and does not represent EPA policy
c) factors relevant in selecting target percentages of waterbodies to protect at alternative target
ANC levels to protect against adverse effects to public welfare, and weights to place on those
factors?
It seems to me that the challenges of creating and enforcing a spatially explicit standard at fine
spatial scales will be large. The graphical representation of percentages of waterbodies that
would, or would not be protected is very informative and suggests that an aim of 50% is too low.
But this is clearly a value judgment. What are the socio-ecological characteristics of the
waterbodies that would not be protected? Are they public water supplies? High elevation
waterbodies?
d) alternative standards for NOx and SOx that would protect against adverse effects to
public welfare based on the AAPI form, and taking into account • consideration of target levels
of ANC (chapter 5),
These appear to be consistent with what's in the literature (see especially Driscoll et al. 2001),
and what others have used for policy targets.
• target percentage of water bodies to protect (chapter 5),
The ranges presented are wide; I cannot come up with rationale to suggest different targets,
however.
• consideration of relevant uncertainties in AAPI components (chapter 7), and
Thank you for chapter 7. As noted, almost all AAPI components have significant uncertainties,
Transference and BC, especially. I continue to wonder whether uncertainties are exacerbated, or
could be reduced, by choosing appropriate spatial scales for aggregation (see comments from last
review). Also, I still find not wholly adequate the discussion and analysis of the CMAQ model
(see above). While I agree that it is an appropriate model to be used for examining deposition
across the US, the uncertainties that have to do with estimating deposition to heterogeneous
terrain were not addressed, for example.
The elasticity analysis was an interesting way to examine uncertainties. It was not clear to me
why 1% was chosen for each of the variables, however.
• any other potentially relevant factors, such as levels of co-protection against
terrestrial acidification and nutrient enrichment (chapter 6)?
I agree that using ANC as a target for chemical protection is a good one, and a defensible place
to focus. Further, by definition and as pointed out, there will be some level of protection afforded
to the adjacent terrestrial systems that influence downstream freshwater systems. The analysis of
protection that would be afforded for terrestrial systems using Bc:Al was very useful. However.
since data are limited, and the linked biogeochemical reactions within terrestrial ecosystems as a
result of N and S deposition are complex, and may have opposite effects (nutrient enrichment vs
111
-------
11/17/10 Draft text for review at the chartered CASAC Public Teleconference 12/06/2010 Teleconference- Please
Do not Cite or Quote -
This draft is a work in progress, does not reflect consensus advice or recommendations, has not been reviewed or
approved by the chartered CASAC and does not represent EPA policy
acidification). I do not think that addressing co-protection further than what has been
accomplished in this document is warranted. It will be crucial, however, to monitor terrestrial
ecosystem responses to changing deposition scenarios so that sufficient data are available for the
next review.
112
------- |