£ JPlk \
ISEz)
Review of the Secondary Standards for
Ecological Effects of Oxides of Nitrogen,
Oxides of Sulfur, and Particulate Matter: Risk
and Exposure Assessment Planning Document
-------
-------
EPA-452/D-18-001
August 2018
Review of the Secondary Standards for Ecological Effects of Oxides of Nitrogen, Oxides of
Sulfur, and Particulate Matter: Risk and Exposure Assessment Planning Document
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Health and Environmental Impacts Division
Research Triangle Park, NC
-------
DISCLAIMER
This document has been prepared by staff in the Health and Environmental Impacts
Division, Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency
(EPA). Any findings and conclusions are those of the authors and do not necessarily reflect the
views of the Agency. This document is being circulated to facilitate discussion with the Clean
Air Scientific Advisory Committee and for public comment to inform the EPA's consideration of
the primary national ambient air quality standard for oxides of nitrogen, oxides of sulfur and
particulate matter. This information is distributed for the purposes of pre-dissemination peer
review under applicable information quality guidelines. It does not represent and should not be
construed to represent any Agency determination or policy.
Questions or comments related to this document should be addressed to Dr. Travis Smith,
U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, C539-07,
Research Triangle Park, North Carolina 27711 (email: smith.itravis@epa.gov) and Ms. Ginger
Tennant, U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards,
C504-06, Research Triangle Park, North Carolina 27711 (email: tennant.ginger@epa.gov).
i
-------
TABLE OF CONTENTS
DISCLAIMER i
TABLE OF CONTENTS ii
LIST OF FIGURES v
LIST OF TABLES v
LIST 01 ACRONYMS AND ABBREVIATIONS vi
EXECUTIVE SUMMARY 1
1 INTRODUCTION 1-1
1.1 Background 1-3
1.2 Exposure and Risk Analyses for Oxides of Nitrogen, Oxides of Sulfur and PM 1-5
1.2.1 Context for Analyses 1-5
1.2.2 Key Technical Issues 1-6
1.2.3 Conceptual Model 1-8
1.3 Organization of this Document 1-9
2 OVERVIEW OF PREVIOUS ASSESSMENTS 2-1
2.1 Air Quality 2-1
2.2 Direct Effects 2-4
2.3 Non-N and S Effects of PM 2-4
2.4 Deposition-Related Effects of N and S 2-5
2.4.1 Aquatic Acidification 2-5
2.4.2 Aquatic Nitrogen Enrichment 2-10
2.4.3 Terrestrial Acidification 2-12
2.4.4 Terrestrial N Enrichment 2-14
2.4.5 Mercury Methylation 2-15
3 CONSIDERATION OF NEWLY AVAILABLE INFORMATION 3-1
3.1 Key Considerations 3-1
3.2 Air Quality Information 3-2
3.2.1 Emissions, Atmospheric Concentrations and Deposition 3-2
3.2.2 Linking Atmospheric Concentration Changes to Changes in Deposition 3-9
3.3 Direct Effects 3-10
3.4 Non-Nitrogen and Non-Sulfur Effects of PM 3-11
3.5 Freshwater Acidification and Nitrogen Enrichment 3-11
ii
-------
3.5.1 Freshwater Acidification 3-11
3.5.2 F reshwater Nitrogen Enri chment 3-14
3.6 Terrestrial Acidification and Nitrogen Enrichment 3-16
3.6.1 Forests 3-17
3.6.2 Lichens and Mycorrhizal Fungi 3-18
3.6.3 Herbs and Shrubs 3-19
3.6.4 Summary 3-20
3.7 Other Nitrogen and Sulfur Effects 3-20
3.7.1 Estuarine Nitrogen Enrichment 3-20
3.7.2 Wetlands Nitrogen Enrichment 3-21
3.7.3 Coastal Acidification 3-22
3.7.4 S Enrichment in Freshwater and Wetland Ecosystems 3-23
3.8 Ecosystem Services 3-25
3.9 Conclusions 3-27
4 PLAN FOR QUANTITATIVE ASSESSMENT 4-1
4.1 Analytical Framework 4-1
4.1.1 Ambient Air Quality and Atmospheric Deposition 4-2
4.1.2 Critical Loads and Exceedances 4-3
4.1.3 Changes in Ecological Responses 4-4
4.2 Air Quality Analyses 4-4
4.2.1 Consideration of Current Conditions for National-scale Air Quality Concentrations
and Deposition 4-4
4.2.2 Consideration of Air Quality Scenarios 4-5
4.3 Ecological Effects Assessment 4-13
4.3.1 Aquatic Acidification and Nitrogen Enrichment 4-13
4.3.2 Terrestrial Effects 4-16
4.4 Case Study Area Selection 4-20
4.5 Consideration of Variability and Uncertainty 4-22
4.5.1 Air Quality 4-23
4.5.2 Critical Loads Data 4-24
4.5.3 Exposure-Response Functions 4-29
5 REFERENCES 5-1
6 APPENDIX 6-1
111
-------
iv
-------
LIST OF FIGURES
Figure 1-1. General NAAQS Review Process 1-5
Figure 1-2. Conceptual Model 1-9
Figure 3-1. Key considerations for updated or new quantitative analyses 3-2
Figure 3-2. Annual total deposition of sulfur, kg S ha"1 in 2002 3-6
Figure 3-3. Annual total deposition of sulfur, kg S ha"1 in 2015 3-7
Figure 3-4. Change in annual total sulfur deposition, 2015 - 2002 (kg S ha"1) 3-7
Figure 3-5. Annual total deposition of nitrogen, kg N ha"1 in 2002 3-8
Figure 3-6. Annual total deposition of nitrogen, kg N ha"1 in 2015 3-8
Figure 3-7. Change in annual total nitrogen deposition, 2015 - 2002 (kg N ha"1) 3-9
LIST OF TABLES
Table 2-1. Noted Uncertainties from the REA and PA in the Prior Review 2-3
Table 4-1. Current National Ambient Air Quality Standards for PM, NO2, and SO2 4-6
Table 4-2. Initial Assessment of Available Deposition Response Factors 4-11
Table 4-3. Major Sources of Uncertainty for the CLs Considered for the REA 4-25
v
-------
LIST OF ACRONYMS AND ABBREVIATIONS
AAI Aquatic Acidification Index
A1 Aluminum
AMoN Ammonia Monitoring Network
ANC Acid Neutralizing Capacity
AQ Air quality
ASSETS EI Assessment of Estuarine Trophic Status Eutrophication Index
Bc/Al Base cation/Aluminum
BCS Base Cation Surplus
BCw Base Cation weathering
C Carbon
CAA Clean Air Act
CALFIRE State of California fire occurrence, prevention and fire-fighting expenditures
CAMx Comprehensive Air Quality Model with extensions
CASAC Clean Air Scientific Advisory Committee
CASTNet Clean Air Status and Trends Network
CH4 Methane
CL Critical Load
CLAD Critical Loads of Atmospheric Deposition Science Committee
CMAQ Community Multiscale Air Quality Model
CONUS Contiguous United States
CO2 Carbon dioxide
CSN Chemical Speciation Networks
CSS Coastal Sage Scrub
CV Coefficient of Variation
DOC Dissolved Organic Carbon
EMAP Environmental Monitoring and Assessment Program
EPA Environmental Protection Agency
F AB First-order Aci d B al ance
FASOM Forest and Agriculture Sector Optimization Model
FIA Forest Inventory and Analysis
FR Federal Register
FRM Federal Reference Method
GIS Geographic Information Systems
GLNC Georeferenced Lake Nutrient Chemistry database
HNO3 nitric acid
vi
-------
IMPROVE Interagency Monitoring of Protected Visual Environments
IRIS Internet Research Information Series
IRP Integrated Review Plan
ISA Integrated Science Assessment
LTM Long-Term Monitoring
MAGIC Model of Acidification of Groundwater In Catchments
MeHg Methylmercury
MCF Mixed Conifer Forest
|ig/m3 micrograms per cubic meter
N Nitrogen
NAAQS National Ambient Air Quality Standard
NADP National Atmospheric Deposition Program
NARS National Aquatic Resource Surveys
NCCA National Coastal Condition Assessment
NCLD National Critical Loads Database
NCore National Core network
NEI National Emissions Inventory
NH3 Ammonia
NH4+ Ammonium
NH4NO3 Ammonium nitrate
NLA National Lakes Assessment
NOAA National Oceanic and Atmospheric Administration
NO Nitric Oxide
NO2 Nitrogen Dioxide
NO3" Nitrate
NOx Oxides of Nitrogen
NOy Sum of oxidized N in the gas phase and particle nitrate
N2O Nitrous Oxide
NPS National Park Service
NRSA National River and Stream Assessment
NTN National Trends Network
NWCA National Wetland Condition Assessment
OAQPS Office of Air Quality Planning and Standards
ppb parts per billion
ppm parts per million
PA Policy Assessment
Vll
-------
PAN Peroxyacetyl Nitrate
PM Particulate Matter
PM2.5 particles with a diameter less than or equal to 2.5 |ig
PM10 particles with a diameter less than or equal to 10 |ig
PRISM Parameter-elevation Regression on Independent Slopes Model
REA Risk and Exposure Assessment
S Sulfur
SAV Submerged Aquatic Vegetation
SLT State and Local agencies and Tribes
SMB Simple Mass Balance
502 Sulfur dioxide
503 Sulfur trioxide
SOx Oxides of Sulfur
S042" Sulfate
SPARROW SPAtially Referenced Regressions on Watershed
SRB Sulfur-reducing bacteria
SRP Sulfur-reducing prokaryotes
SSWC Steady State Water Chemistry
TDEP Total Deposition
TIME Temporally Integrated Monitoring of Ecosystems
USFS U.S. Forest Service
USGS U.S. Geological Survey
VOC Volatile organic compounds
WHO World Health Organization
Vlll
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
EXECUTIVE SUMMARY
The U.S. Environmental Protection Agency (EPA) is conducting a review of the air
quality criteria and the associated secondary (welfare-based) national ambient air quality
standards (NAAQS) for oxides of nitrogen, oxides of sulfur and particulate matter (PM). This
review is focused on the contribution of oxides of nitrogen, oxides of sulfur, and PM to
ecological effects and particularly the contribution of these pollutants to such effects through
atmospheric deposition. Based on analysis of the information available in this review regarding
support for a quantitative risk and exposure assessment (REA) to inform the review, this
document outlines a plan, including scope and methods, for conducting a REA, and is intended
to facilitate consultation with the Clean Air Scientific Advisory Committee (CASAC), as well as
an opportunity for public participation. The information considered includes newly available
scientific evidence, new and/or improved data, methods, and tools and other information or data
supporting a quantitative REA particularly those assessed in the second draft Integrated Science
Assessment for Oxides of Nitrogen, Oxides of Sulfur and Particulate Matter - Ecological
Criteria (ISA).
The current review of the secondary standards for oxides of nitrogen, oxides of sulfur and
PM considers secondary standards for these three pollutants together with regard to protection
against adverse ecological effects on public welfare. This review differs from the review
completed in 2012 in that the current review includes consideration of the secondary PM
standards, in addition to the secondary standards for oxides of nitrogen and sulfur. Given the
contribution of nitrogen compounds to PM, including but not limited to those related to oxides of
nitrogen, the current review provides for an expanded and more integrated consideration of N
deposition and the current related air quality information.
Since the last review, the scientific evidence for all effect categories has been expanded,
especially for terrestrial nitrogen enrichment effects. New critical loads data and exposure-
response functions for nitrogen enrichment and acidification effects on terrestrial and freshwater
ecosystems are now available that together provide a basis for analysis of ecological effects of
atmospheric deposition across the nation. This new information contributes to the stronger
weight of evidence and expanded causality determinations in the second draft ISA and enables
an evaluation of growth and mortality effects, as well as species or community composition
changes, that was not available at the time of the last review. In addition, new air quality data
suggest that the spatial variability and distribution of atmospheric deposition has changed in
recent years. New techniques are also available for combining these measurements and modeling
outputs to estimate total deposition with lower uncertainty. Thus, given the new scientific
ES-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
information and data available since the last review, we believe there is support for a new and
updated assessment of risk and exposure to inform the current review.
EPA's proposed plans for the risk and exposure assessment include quantitative analyses
of terrestrial and aquatic effects using critical loads and exposure-response functions. These
analyses will be conducted at a national-scale and within select case study areas. To estimate
deposition for this assessment, we intend to develop new estimates of total N and S deposition
that will utilize the methodologies developed by the National Atmospheric Deposition Program
(NADP) Total Deposition (TDEP) Science Committee and combine the most recently available
measured ambient air concentration and wet deposition data with modeled deposition velocity
and dry deposition data. To assess ecological risk under current conditions at a national scale, we
intend to compare these TDEP estimates to national critical loads and exposure-response datasets
for aquatic and terrestrial acidification and N enrichment. This will inform our understanding of
sensitivity and risk for acidification and nitrogen enrichment on a national scale for multiple
species.
In addition, to assess ecological risk when just meeting the current standards and any
potential alternative standards, as appropriate, we propose to evaluate the impacts of changing
ambient concentrations and atmospheric deposition on aquatic and terrestrial acidification and N
enrichment in selected study area locations. To do this, we intend to evaluate the relationship
between emissions, ambient concentrations and atmospheric deposition of N and S based on data
from air quality models and ambient measurements and to use this information to adjust air
quality in the study area locations to reflect just meeting the appropriate air quality scenarios.
These atmospheric deposition estimates would then be compared to critical loads and exposure-
response datasets for aquatic and terrestrial acidification and N enrichment that are most
applicable for the individual study area locations. Together, the national-scale and study area
assessment information would be used to inform decisions regarding potential adverse effects to
public welfare.
ES-2
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
1 INTRODUCTION
The U.S. Environmental Protection Agency (EPA) is conducting a review of the air
quality criteria and the associated secondary (welfare-based) national ambient air quality
standards (NAAQS) for oxides of nitrogen, oxides of sulfur, and particulate matter (PM). The
purpose of this planning document (titled Review of the Secondary National Ambient Air Quality
Standards for Ecological Effects of Oxides of Nitrogen, Oxides of Sulfur and Particulate Matter:
Risk and Exposure Assessment Planning Document - hereafter referred to as the REA Planning
Document) is to describe EPA's consideration of the extent to which newly available scientific
evidence, tools or methodologies, and/or information warrant the conduct of a quantitative risk
and exposure assessment (REA) that might inform this review, and may expand our
characterization of exposure and risk estimates provided by the assessments conducted for the
last review. Based on these considerations, and as described below, we plan to develop a new
REA to inform the current review of the secondary NAAQS for oxides of nitrogen, oxides of
sulfur, and PM. Accordingly, this document's additional purpose is to describe the general plan,
including scope and methods for conducting the REA.
This review of the secondary (welfare-based) NAAQS for oxides of nitrogen,1 oxides of
sulfur2 and PM,3 is focused on the contribution of these pollutants to ecological effects and
particularly the contribution of these pollutants to such effects through atmospheric deposition.4
In so doing, we recognize that oxides of nitrogen, oxides of sulfur, and PM contribute to
1 In this document, the term oxides of nitrogen refers to all forms of oxidized nitrogen (N) compounds, including
nitric oxide (NO), nitrogen dioxide (NO2), and all other oxidized N-containing compounds formed from NO and
NO2. This follows usages in the Clean Air Act Section 108(c): "Such criteria [for oxides of nitrogen] shall include
a discussion of nitric and nitrous acids, nitrites, nitrates, nitrosamines, and other carcinogenic and potentially
carcinogenic derivatives of oxides of nitrogen." By contrast, within air pollution research and control
communities, the terms "oxides of nitrogen" and "nitrogen oxides" are restricted to refer only to the sum of NO
and NO2, and this sum is commonly abbreviated as NOx. The category label used by this community for the sum
of all forms of oxidized nitrogen compounds including those listed in Section 108(c) is total oxidized nitrogen
(NOy).
2 Oxides of sulfur are defined here to include sulfur monoxide (SO), sulfur dioxide (SO2), sulfur trioxide (SO3),
disulfur monoxide (S2O), and sulfate (SO12 ). however, SO, SO3, and S2O are present at much lower ambient
levels than SO2 and SO/- .
3 PM is the generic term for a broad class of chemically and physically diverse substances that exist as discrete
liquid and/or solid particles over a wide range of sizes. Particles may be emitted directly from anthropogenic and
natural sources, or formed in the atmosphere by transformations of gaseous emissions such as SO2, nitrogen
oxides (NOx), ammonia (NH3) and volatile organic compounds (VOC). The chemical and physical properties of
PM vary greatly with time, region, meteorology, and source category.
4 In addition, these air pollutants contribute to effects on vegetation, soils, and biota both through direct exposure to
the pollutant in air, and indirect exposure after deposition.
1-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
ecological effects together, through deposition of N and/or S to the surface of vegetation, soils or
water bodies. In recognition of these linkages, addressing the pollutants together to assess
ecological effects enables a comprehensive look at the nature and interactions of the pollutants.
The current secondary NAAQS for oxides of nitrogen and oxides of sulfur were
established to protect against direct damage to vegetation by exposure to gas-phase oxides of
nitrogen and oxides of sulfur. The secondary standard for oxides of nitrogen is an annual average
not to exceed 0.053 parts per million (ppm) NO2. The secondary standard for oxides of sulfur is a
3-hour average of 0.5 ppm SO2, not to be exceeded more than once per year. The secondary PM
standards were established to provide protection against a variety of PM-associated welfare
effects, including effects on vegetation as well as visibility impairment and materials damage
(e.g., soiling, corrosion).5 The annual secondary PM2.5 standard is set at a level of 15 |ig/m3, with
an annual arithmetic mean averaged over three years, and a secondary 24-hour PM2.5 standard is
set at a level of 35 |ig/m3, as the 98th percentile of the 24-hour average, averaged over 3 years.
The annual secondary PM10 standard is an annual arithmetic mean, averaged over three years
with a level of 50 |ig/m3, and the secondary 24-hour PM10 standard is a 24-hour average of 150
|ig/m3, not to be exceeded more than once per year on average over a three-year period.
Welfare effects associated with PM that are not ecologically related (and therefore not
included in this review), such as visibility impairment, climate effects and materials damage, and
the health effects of PM (including particulate transformation products of oxides of nitrogen and
oxides of sulfur) are being considered as part of a separate review of the NAAQS for PM (U.S.
EPA, 2016). The health effects of oxides of nitrogen were considered in a separate assessment
that was completed recently as part of the review of the primary (health-based) NAAQS for
oxides of nitrogen (83 FR 17226). Similarly, the health effects of oxides of sulfur are currently
being considered in a separate assessment as part of the review of the primary NAAQS for
oxides of sulfur (83 FR 26752). In addition, NH3 is not a criteria pollutant but is a precursor to
PM (ammonium sulfate (NH4SO4) and ammonium nitrate (NH4NO3)) and is considered in this
review to the extent that it contributes to atmospheric transformations and loading to
ecosystems.6
This REA Planning Document presents an evaluation of information related to ecological
effects that is newly available in the second draft of the Integrated Science Assessment for
Oxides of Nitrogen, Oxides of Sulfur, and Particulate Matter—Ecological Criteria (U.S. EPA,
5 The secondary PM standards were most recently reviewed in 2012 (78 FR 3086, January 15, 2018).
6 The scientific and technical information and analyses in this review are expected to inform our understanding of
the contribution of NH3 to total N deposition and N-related ecological effects, as well as on the role of NH3 in the
formation of PM (which is a criteria pollutant).
1-2
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
2017a; hereafter referred to as the second draft ISA). Advances in modeling tools and techniques
and air quality data that have become available since the last review are also considered. This
document is intended to facilitate consultation with the Clean Air Scientific Advisory Committee
(CASAC), as well as an opportunity for public participation, on the evaluation of the potential
support in the newly available scientific evidence (including existing and historical air quality
patterns and trends), new and/or improved data, methods, and tools, for conducting updated
quantitative assessments, and on the plan for such analyses, as warranted. Additionally, this
document evaluates the extent to which new information, tools or methodologies, will address or
improve our consideration of important limitations or uncertainties associated with the analyses
from the last review (summarized in chapter 2). Based on these considerations and our
preliminary conclusions on the extent to which updated quantitative analyses of ecological risks
and/or exposures are warranted in the current review, this document presents general plans for
such analyses.
1.1 BACKGROUND
Sections 108 and 109 of the CAA govern the establishment and periodic review of the
NAAQS. Section 108 [42 U.S.C. 7408] directs the Administrator to identify and list certain air
pollutants and then to issue air quality criteria for those pollutants. The Administrator is to list
those air pollutants that in his/her "judgment, cause or contribute to air pollution which may
reasonably be anticipated to endanger public health or welfare," "the presence of which in the
ambient air results from numerous or diverse mobile or stationary sources;" and "for which...
[the Administrator] plans to issue air quality criteria..CAA section 108(a)(1). The NAAQS are
established for these pollutants. The CAA requires that NAAQS are to be based on air quality
criteria, which are intended to "accurately reflect the latest scientific knowledge useful in
indicating the kind and extent of all identifiable effects on public health or welfare that may be
expected from the presence of [the] pollutant in the ambient air..CAA section 108(a)(2).
Under CAA section 109 [42 U.S.C. 7409], the EPA Administrator is to propose, promulgate, and
periodically review, at five-year intervals, "primary" (health-based) and "secondary" (welfare-
based)7 NAAQS for such pollutants for which air quality criteria are issued. Section 109(b)(2) of
the CAA directs that a secondary standard is to "specify a level of air quality the attainment and
maintenance of which, in the judgment of the Administrator, based on such criteria, is requisite
7 Section 302(h) of the CAA provides that all language referring to effects on welfare includes but is not limited to,
"...effects on soils, water, crops, vegetation, man-made materials, animals, wildlife, weather, visibility and
climate, damage to and deterioration of property, and hazards to transportation, as well as effects on economic
values and on personal comfort and well-being.
1-3
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
to protect the public welfare from any known or anticipated adverse effects associated with the
presence of [the] pollutant in the ambient air."8 Based on periodic reviews of the air quality
criteria and standards, the Administrator is to make revisions in the criteria and standards, and
promulgate any new standards, as may be appropriate. The CAA also requires that an
independent scientific review committee review the air quality criteria and standards and
recommend to the Administrator any new standards and revisions of existing air quality criteria
and standards as may be appropriate, a function now performed by the CASAC.
The overall plan for this review was presented in the Integrated Review Plan (IRP) (U.S.
EPA, 2017b), which discusses the preparation of key documents in the NAAQS review process
including an ISA, an REA, as warranted, and a Policy Assessment (PA). In general terms, the
ISA provides a critical assessment of the latest available scientific information upon which the
NAAQS are to be based. The purpose of the REA in a secondary standards review is to estimate
risk and exposure to public welfare associated with the current standards and, if appropriate,
evaluate potential improvements in public welfare that could be achieved from meeting potential
alternate standard(s). The PA evaluates the policy implications of the information contained in
the ISA and of any policy-relevant quantitative analyses, such as a quantitative REA, that were
performed for the review. Based on that evaluation, the PA presents conclusions regarding
standard-setting options for the Administrator to consider in reaching decisions on the NAAQS.9
The general NAAQS review process is illustrated in Figure 1-1, below.
8 Section 109(b)(1) [42 U.S.C. 7409] of the CAA defines a primary standard as one "the attainment and maintenance
of which in the judgment of the Administrator, based on such criteria and allowing an adequate margin of safety,
are requisite to protect the public health."
9 Review of the NAAQS involves consideration of the four basic elements of a standard: indicator, averaging time,
form, and level. The indicator defines the pollutant to be measured in the ambient air in order to determine
compliance with the standard. The averaging time defines the time-period over which air quality measurements
are to be obtained and averaged or cumulated. The form of a standard defines the air quality statistic that is to be
compared to the level of the standard in determining whether an area attains the standard. The level of a standard
defines the air quality concentration used (i.e., an ambient air concentration of the indicator pollutant).
1-4
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Planning
Workshop on
science-policy issues
Integrated Review Plan (IRP): timeline and key
policy-relevant issues and scientific questions
Peer-reviewed
scientific studies
Integrated Science Assessment (ISA): evaluation and
synthesis of most policy-relevant studies
REA Planning Document
Clean Air Scientific
Advisory Committee
(CASAC) review
Public comment
Assessment
Risk/Exposure Assessment (REA):
quantitative assessment, as warranted, focused
on key results, observations, and uncertainties
t I
Policy Assessment (PA): staff analysis of
policy options based on integration and
interpretation of information in the ISA and REA
Interagency
review
Agency decision
making and draft
proposal notice
Rulemaking
Public hearings
and comments
on proposal
Agency decision
making and draft
final notice
Interagency
review
Figure 1-1. General NAAQS Review Process
1.2 EXPOSURE AND RISK ANALYSES FOR OXIDES OF NITROGEN,
OXIDES OF SULFUR AND PM
This REA Planning Document serves to provide support for conducting any new or
updated quantitative assessments in the current review. Conclusions regarding such support are
based on our consideration of the available scientific evidence; the available technical
information, tools, and methods; and judgments as to the likelihood that quantitative assessments
will add substantially to our understanding of risk or exposure related to ecological effects,
beyond the insights gained from the assessments conducted in the last review. Specifically, this
consideration also includes an evaluation of the newly available data, tools and methods and
whether they would be expected to reduce previously identified uncertainties or limitations from
the last review.
1.2.1 Context for Analyses
The most recent review of the secondary NO2 and SO2 NAAQS was completed in 2012.
Technical analyses for that review focused on two general types of effects (1) direct effects on
vegetation of exposure to gaseous oxi des of nitrogen and sulfur, which are the type of effects that
the current secondary NO2 and SO2 standards were developed to protect against, and (2) effects
associated with nitrogen and sulfur deposition to sensitive aquatic and terrestrial ecosystems (77
1-5
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
FR 20218, April 3, 2012). The REA for the 2012 review presented assessments of aquatic and
terrestrial acidification-related and nutrient enrichment-related effects. Based on the REA
findings, the PA gave primary attention to aquatic acidification effects from deposition of
nitrogen and sulfur, and developed an aquatic acidification index (AAI)— an equation that relied
upon ecosystem and air quality modeling to estimate the ecosystem's natural ability to buffer
acidic deposition from ambient nitrogen and sulfur deposition.
In the Agency's consideration of the scientific and technical information and analyses
with regard to the potential use in establishing a new standard based on aquatic acidification,
several important uncertainties and data limitations were recognized, both those related to
characterizing the relationships between N and S deposition and ecological effects generally, and
those specifically relating to the development of an AAI. Thus, while recognizing the
scientifically-supported conceptual basis for an AAI-based standard,10 the EPA also recognized
current limitations in relevant data and uncertainties that, together, were concluded to be too
great to support a standard that would meet the requirements of the Act (77 FR 20218, April 3,
2012).
The current review of the secondary NO2 and SO2 standards differs from the review
completed in 2012 in that the current review also includes consideration of the secondary PM
standards. Specifically, the Agency is considering secondary standards for these three pollutants
together with regard to protection against adverse ecological effects on public welfare and
particularly such effects related to atmospheric deposition. Given the contribution of nitrogen
compounds to PM, including but not limited to those related to oxides of nitrogen, the current
review provides for an expanded and more integrated consideration of N deposition and the
current related air quality information.
1.2.2 Key Technical Issues
Because NOx, SOx and PM are deposited from the ambient air into ecosystems where
they have the potential to affect individual organisms and communities, considerations in this
review will include potential impacts on the public welfare from alterations in structure and
function of ecosystems. Important considerations for the review, which the REA will inform,
include:
• Source contributions and loading. It is important to consider the emissions sources,
chemical species and contributions to overall loading to ecosystems. The ability to
evaluate modeled data as well as the ability to characterize components of ambient air
10 The Administrator recognized that while an AAI-based standard was innovative and unique, the structure of the
proposed standard was well-grounded in the science underlying the relationships between ambient concentrations
of oxides of nitrogen and sulfur and the aquatic acidification related to deposition of nitrogen and sulfur
associated with such ambient concentrations (77 FR 20251, April 3, 2012).
1-6
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
and atmospheric deposition are important to understanding the potential effects of the
pollutants on ecosystems. In some systems atmospheric deposition is not the only source
of loadings and N and S can be introduced into the system from other sources (e.g.,
surface water runoff). Additionally, the composition of emissions in a given area will
have an effect on what types of chemical species are contributing to atmospheric
deposition. Using analytical tools and data to help understand what pollutants are
contributing to the overall effect will be important, particularly in the case of reduced
nitrogen impacts11.
• Historic air quality impacts on deposition-related effects. Historic air quality and
associated deposition can have an appreciable impact on ecosystems in many areas and
will be an important consideration in understanding the role of current air quality versus
historic air quality on deposition-related effects. For example, in areas where deposition
levels have been historically high there may be terrestrial systems where the soil is
saturated or aquatic systems where improvements to water quality are impeded by the
area's geology (e.g., underlying bedrock). In such cases, the continued role of this
historic deposition in the area's deposition-related impacts may mean that we would not
expect the deposition-related risk and exposures to be as responsive to changes in
atmospheric deposition, and that there would be a substantial delay prior to any such
response.
• Timescale of effects and potential for recovery. Another consideration for this review is
to understand the timescale for the effects being considered. For example, the response
time for trees to stressors is very different than for fish populations or herbaceous plants.
This affects both the exposure period necessary for effects to become adverse, as well as
the uncertainty associated with the ability of the REA to assess these impacts on public
welfare. In addition, while information may be limited, understanding the scientific
evidence indicating the potential for recovery or improvement is also important to
consider.
• Spatial scale of effects. It is important to consider the spatial scale, magnitude, and
associated variability, of some ecological effects. The effects of N, S and PM deposition
on aquatic and terrestrial ecosystems occur over time and are a result of atmospheric
reactions and transport. The evaluation of environmental responses to these pollutants,
accordingly, will need to consider the variability of environmental characteristics of
ecosystems across the nation, including those related to ecosystem susceptibility and to
the relative importance of individual effects (such as acidification orN enrichment).
• Impacts to public welfare. Importance to the public welfare is an important
consideration to the review. For example, potential impacts in areas with special federal
protections, and lands set aside by states, tribes, and public interest groups to provide
similar benefits to the public welfare, may be of particular importance. The relevance of
such areas to consideration of effects on the public welfare has been recognized in past
NAAQS decisions (e.g., 80 FR 65292, October 26, 2015). Such areas include Class I
11 NH3 is not a criteria pollutant but is a precursor to PM and is considered in this review to the extent that it
contributes to atmospheric transformations and loading to ecosystems.
1-7
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
areas12 which are federally mandated to preserve certain air quality related values. Class
I areas and other parks have been afforded special federal protection to preserve services
that provide for the enjoyment of these resources for current and future generations.
1.2.3 Conceptual Model
In each NAAQS review, selection of the approach most appropriate for the
characterization of risks is influenced by the nature and strength of the evidence for the subject
pollutants. Depending on the type of evidence available, analyses may include quantitative risk
assessments based on exposure-response, or ambient air concentration/deposition-response
relationships. Figure 1-2 below illustrates our conceptual model for evaluating ecological risk
and exposure of oxides of nitrogen, oxides of sulfur and PM in this review, including exposure
from direct pathways and through atmospheric deposition. Included in Figure 1-2, below is a
basic conceptual diagram highlighting the relevant pathways of exposure for the pollutants in
this review.
As described in chapter 4, the REA for this review will include an evaluation of the
relationship between emissions of the three criteria pollutants (and their precursors) and ambient
air concentrations, as well as their contribution to direct as well as deposition-related exposures
to biota, soils, sediments and surface waters. While the review focuses on atmospheric inputs,
non-atmospheric inputs of the pollutants that may occur through runoff of the pollutants from
other human activities, such as agriculture and industrial processes, will also be considered as
part of the total loading of the pollutant to the ecosystem. The ecological effects of these inputs
(e.g., nutrient enrichment) will then be evaluated based on the biogeochemical and/or ecological
responses of "receptors" 13 (e.g., plants, wildlife) within ecosystems.
12 Areas designated as Class I include all national parks, national wilderness areas which exceed 5,000 acres in size,
national memorial parks which exceed 5,000 acres in size, and national parks which exceed six thousand acres in
size, provided the park or wilderness area was in existence on August 7, 1977. Other areas may also be Class I if
designated as Class I consistent with the Act.
13 A receptor is defined as a biological element that is impacted by the conditions created by atmospheric deposition.
1-8
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Emissions to Ambient Air
NOyand SOx, PM
Direct
Exposures
Deposition
Exposures
Non-
Atmospheric
Loadings
Soils
Plants
Plants, Wildlife
Plants, Wildlife
Lichen
Surface Waters
and Sediments
Ambient Air
Figure 1-2. Conceptual Model.
1.3 ORGANIZATION OF THIS DOCUMENT
The remainder of this document presents our evaluations and preliminary conclusions
regarding the degree to which available evidence and information address important
uncertainties and the support for updated or new quantitative analyses in the current review.
Chapter 2 provides background information regarding the assessments conducted in the prior
reviews, evaluating deposition-related effects of oxides of nitrogen, oxides of sulfur, and PM as
well as direct effects of oxides of nitrogen and oxides of sulfur. Chapter 2 also includes
information regarding any uncertainties that were identified in the last reviews. Chapter 3
includes consideration of information newly available in this review, including scientific
evidence and data for various ecological effects. This chapter also includes consideration and
conclusions regarding whether this new information is expected to reduce uncertainties or
limitations from the prior review and inform updated or new quantitative assessments for this
review. Chapter 4 presents an overview of the REA analysis approach, summarizes approaches
for characterizing air quality, and describes the plan for quantitatively evaluating ecological
effects in various ecosystems.
1-9
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
2 OVERVIEW OF PREVIOUS ASSESSMENTS
This chapter summarizes the analyses and limitations and/or uncertainties related to
ecological risk and exposure as assessed in the two previous reviews: the secondary NAAQS
review of oxides of nitrogen and sulfur; and the review of particulate matter. More detailed
information on the body of evidence and causal determinations for these reviews can be found in
the Integrated Science Assessment for Oxides of Nitrogen and Sulfur - Ecological Criteria (U.S.
EPA, 2008; hereafter referred to as the 2008 NOx/SOx ISA) as well as the Integrated Science
Assessment for Particulate Matter (U.S. EPA, 2009; hereafter referred to as the 2009 PM ISA).
More detailed descriptions of the assessment approaches in the previous oxides of nitrogen and
sulfur joint review can be found in the Risk and Exposure Assessment for Review of the
Secondary National Ambient Air Quality Standards for Oxides of Nitrogen and Oxides of Sulfur:
Final (U.S. EPA, 2009; hereafter referred to as 2009 NOx/SOx REA) as well as the Policy
Assessment for the Review of the Secondary National Ambient Air Quality Standards for Oxides
of Nitrogen and Oxides of Sulfur: Final (U.S. EPA, 201 la; hereafter referred to as the 2011
NOx/SOx PA). More details about the previous PM review can be found in the Particulate
Matter Urban-Focused Visibility Assessment - Final Document (U.S. EPA, 2010a)14 and in the
Policy Assessment for the Review of the Particulate Matter National Ambient Air Quality
Standards (U.S. EPA, 201 lb).
In each NAAQS review, selection of the approach most appropriate for the
characterization of risk is influenced by the nature and strength of the evidence for the subject
pollutants. This chapter summarizes the assessment approach used in the last review of NOx and
SOx, which focused on evaluating the protection provided by secondary standards for oxides of
nitrogen and sulfur for two general types of effects: (1) direct effects on vegetation associated
with exposure to gaseous oxides of nitrogen and sulfur in the ambient air, which are the effects
that the current NO2 and SO2 secondary standards were set to protect against; and, (2) effects
associated with the deposition of nitrogen (N) and sulfur (S) to aquatic and terrestrial
ecosystems. This chapter also summarizes the assessment approach used in the last PM review to
evaluate direct and indirect effects of non-N and non-S deposited particles (including metals and
organics).
2.1 AIR QUALITY
The 2009 NOx/SOx REA examined the path from emissions to air concentrations,
providing characterizations of (1) major emissions sources of NOx, NH3 and SO2, (2)
14 Note that this review does not cover visibility.
2-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
atmospheric concentrations of NOy and SO2, and (3) estimates of deposition of total N and S
across the U.S. and for specific case study areas (see 2009 NOx/SOx REA, sections 3.2 and 3.3).
Data from the 2002 National Emission Inventory (U.S. EPA, 2006) were assessed in the 2009
NOx/SOx REA to understand the major emissions sources of N and S across the U.S., and used
as inputs to the Community Multiscale Air Quality (CMAQ) model to provide estimates of
atmospheric concentrations and deposition of N and S in the U.S. Spatial fields representing
2002 annual average N and S deposition were also created by extracting the dry deposition
predictions from the CMAQ model simulations and combining them with the wet deposition
measurements from the National Atmospheric Deposition Program (NADP) National Trends
Network (NTN).
Based on these analyses, the REA concluded that there was significant variation both
spatially and seasonally across the country with the eastern U.S. receiving much greater
deposition of both N and S than the states west of the Mississippi river. CMAQ simulations of
scenarios with emission reductions of NOx, SOx, and NH3 were used to examine the impact of
these emissions to specific ecosystems. At the time of the REA analysis, oxidized N was the
largest contributor to deposition, and the 2009 NOx/SOx REA found that a 50% reduction in
NOx emissions led to a 30% - 40% reduction in N deposition across most of the Eastern US. In
locations with more NH3 emissions, such as portions of the Potomac River Estuary and the
Neuse River Estuary, the 50% NH3 emission reduction led to a 40% - 50% reduction in N
deposition. For SO2 emissions, a 50% reduction resulted in a nearly 50% reduction in S
deposition. The 2009 NOx/SOx REA concluded that most of the N and S deposition can be
attributed to emissions of NOx, NH3, and SO2. However, the REA also recognized that not all
loadings of N and S compounds to freshwater, estuarine, and wetland ecosystems are due to
atmospheric deposition. Other inputs, such as runoff from agricultural soils and point-source
discharges, also contribute to acidification and N enrichment. Thus, some ecosystems may be
solely impacted by atmospheric deposition (e.g., high elevation lakes), while ecological effects
attributed to N and S in other systems might be largely due to non-atmospheric sources (e.g.,
high order streams). Sources to total loading of N and S was discussed more in the 2011
NOx/SOx PA in consideration of the standard-setting process.
Building upon these 2009 NOx/SOx REA analyses, the 2011 NOx/SOx PA further
explored the relationship between atmospheric concentrations and deposition of N and S. In
doing so, the 2011 NOx/SOx PA introduced the concept of a "transference ratio" and defined it
as the ratio of deposition to ambient atmospheric concentrations. The transference ratios were
calculated in the PA by dividing the 2002 CMAQ model outputs of the annual average of
deposition of NOy or SOx by the annual average of ambient air concentrations of NOy or SOx,
respectively. These annual average deposition and ambient air concentration values were
2-2
-------
1 estimated for every CMAQ model grid cell (e.g., 12 km by 12 km) and then averaged across
2 areas of ecological similarity (i.e., spanning 100 km or more, as defined by level III ecoregions).
3 The transference ratios were then used to characterize how changes in air quality concentrations
4 of NOy or SOx related to changes in deposition of NOy or SOx in each of these ecological areas.
5 Limitations and uncertainties related to linking atmospheric concentrations to deposition
6 are summarized in the 2011 NOx/SOx PA in Table 7-3, and the 2009 NOx/SOx REA in Section
7 3.6. Frequently identified issues include lack of spatial coverage and reference methods for
8 concentration measurements, CMAQ modeling uncertainties, especially with respect to emission
9 estimates, and the inherent difficulty in measuring dry deposition.
10 Table 2-1. Noted Uncertainties from the REA and PA in the Prior Review
Source of uncertainty
Description of uncertainty from the previous NOx/SOx review
Measurements of atmospheric
concentrations of N0y
A Federal Reference Method (FRM) for NOy instruments was not
available and existing techniques were known to have a negative
bias. The spatial coverage of existing NOy measurements was not
adequate for assessing effects outside of urban areas.
Measurements of atmospheric
concentrations of S0X
A lack of adequate spatial coverage was the primary concern for
SO2 + SO42" observations.
Atmospheric deposition of N0y, derived
from NTN measurements of wet
deposition and CMAQ estimates of dry
deposition
Quantifying uncertainty in CMAQ estimates of dry deposition was
hampered by a lack of dry deposition measurements.
Atmospheric deposition of S0X, derived
from NTN measurements of wet
deposition and CMAQ estimates of dry
deposition
There was general consensus that the overall mass balance of S
was treated well in CMAQ. As with all dry deposition estimates,
technologies for direct measurements were not available.
Atmospheric deposition of NHX, derived
from NTN measurements of wet
deposition and CMAQ estimates of dry
deposition
A lack of both Nbhand ammonium (NH4+) ambient observations
made it difficult to characterize uncertainty in NHX deposition. As
with all dry deposition estimates, technologies for direct
measurements were not available routinely.
Deposition T ransference Ratios,
calculated using CMAQ
Transference ratios were not evaluated in a traditional model to
observation context. Uncertainty was attributed to the information
(e.g., NH3 emissions) driving these calculations and availability of
observations to evaluate model behavior.
Emissions of N0X, S0X, and NH3
Uncertainty depends on the emission source. Emissions from
electricity generating facilities are directly measured and have low
uncertainty; however, emission estimates of livestock and fertilized
fields do not capture all of the variability in these sources.
2-3
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
2.2 DIRECT EFFECTS
The 2008 NOx/SOx ISA determined that there was a causal relationship between
exposure to nitric acid (HNO3) and changes in vegetation, and a causal relationship between
exposure to NO, NO2, peroxyacetyl nitrate (PAN), and SO2 and injury to vegetation. It was also
noted that oxides of N and S gases have different degrees of phytotoxic effects on vegetation. In
addition, while there was relatively little information about the direct effects of HNO3 vapor on
vegetation in previous reviews, research on the decline of sensitive lichen species was
highlighted in the 2008 NOx/SOx ISA. In the mixed conifer forest of the San Bernardino
Mountains, HNO3 has been estimated to provide 60% of all dry deposited nitrogen, and it has
been suspected as the cause of a dramatic decline in lichen species (see 2009 NOx/SOx REA,
Section 6.4.3). At high concentrations over the short-term, HNO3 can damage vascular plants
such as seedlings of ponderosa pine (Pinus ponderosa) and California black oak (Quercus
kelloggii) (see 2009 NOx/SOx REA, Section 6.4.3). The 2009 NOx/SOx REA noted that more
research was needed to determine long-term exposure effects at lower concentrations.
Additionally, at the time of the last review, there were no tools available for use in quantitative
assessments of direct gaseous effects. For these reasons, the 2009 NOx/SOx REA included a
qualitative discussion of direct gaseous effects but no quantitative analyses were conducted.
2.3 NON-N AND S EFFECTS OF PM
Direct and indirect ecological effects associated with deposited PM components
(including metals and organics) were evaluated in the 2009 PM ISA. Direct deposition-related
effects include alteration of leaf processes from deposition of PM ("dust") to vegetative surfaces.
Indirect deposition-related effects encompass physiological responses associated with uptake of
PM components and alterations to ecosystem structure and function. Evidence reviewed in the
2009 PM ISA was sufficient to determine that there is a likely causal relationship between PM
deposition and a variety of effects on individual organisms and ecosystems.
In the REA for the prior PM review, issued in 2009,15 the EPA determined that despite
this likely causal relationship data was insufficient to support quantitative assessments for non-N
and S deposition-related PM effects. After a careful evaluation of the evidence, the EPA
determined that data needed to conduct quantitative assessments for ecological welfare effects in
the last review were not available. A qualitative assessment was included in an Appendix of the
REA (see 2009 PM REA, Appendix A).
15 U.S. EPA, Risk and Exposure Assessment for Review of the Secondary National Ambient Air Quality Standards
for Oxides of Nitrogen and Oxides of Sulfur, September 2009, EPA-452/R-09-008a.
2-4
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
2.4 DEPOSITION-RELATED EFFECTS OF N AND S
Deposition-related ecological effects are those related to the deposition of mixtures of isl-
and S-containing compounds onto aquatic or terrestrial surfaces (e.g., vegetation, soil, or
waterbodies). The ways in which ambient air pollution gets assimilated into ecosystems is a
complex process, with many geochemical and biological factors contributing to the resulting
deposition loading. The 2009 NOx/SOx REA assessed the ecological effects associated with
deposition of N and S, focusing on four main targeted ecosystem effects on terrestrial and
aquatic systems: (1) aquatic acidification16; (2) terrestrial acidification; (3) aquatic nitrogen
enrichment17; and (4) terrestrial nitrogen enrichment18'19. In conducting these analyses, the 2009
NOx/SOx REA evaluated the relationships between atmospheric concentrations, deposition (wet
and dry), biologically relevant exposures, targeted ecosystem effects, and, to the extent possible,
associated ecosystem services. In doing so, the 2009 NOx/SOx REA recognized a lack of broad-
scale data and that deposition-related effects were not evenly distributed across the U.S.
Accordingly, the 2009 NOx/SOx REA used a case study approach for the quantitative
assessments, building from the scientific information presented in the 2008 NOx/SOx ISA as
well as the identification of ecosystems that are sensitive to N and/or S deposition.
To show the impacts of ecological effects on public welfare, the 2009 NOx/SOx REA
and 2011 NOx/SOx PA qualitatively and quantitatively associated changes in ecological effects
with changes in their ecological benefits and ecosystem services or welfare effects.
The sections that follow summarize the analysis of deposition-related effects as presented
in the prior reviews, including (1) the relationship between emissions, air concentrations, and
deposition; (2) the qualitative and quantitative approaches used to assess deposition-related
ecological effects and related ecosystem services, where relevant; and (3) key limitations or
uncertainties.
2.4.1 Aquatic Acidification
With respect to aquatic acidification, the 2008 NOx/SOx ISA for the prior review of
oxides of nitrogen and sulfur determined that the evidence was sufficient to infer a causal
16 These analyses were then expanded upon in the 2011 NOx/SOx PA which focused on outlining an approach for a
standard which focused on the ecological effects associated with acidifying deposition of oxides of nitrogen and
sulfur in aquatic ecosystems, (see 2011 NOx/SOx PA, Chapter 7).
17 Referred to as "nutrient enrichment" in the 2009 NOx/SOx REA. This effect includes eutrophication.
18 Referred to as "nutrient enrichment" in the 2009 NOx/SOx REA.
19 The REA also qualitatively addressed the influence of sulfur oxide deposition on MeHg production; nitrous oxide
effects on climate; nitrogen effects on primary productivity and biogenic greenhouse gas fluxes; and phytotoxic
effects on plants.
2-5
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
relationship between N and S deposition and changes in biogeochemistry and changes in biota in
freshwater ecosystems.
The 2009 NOx/SOx REA conducted a case study evaluating acidifying effects on streams
in the Shenandoah region of the southern Appalachian Mountains of Virginia (hereafter referred
to as the Shenandoah), and lakes in the Adirondack Mountains of New York (hereafter referred
to as the Adirondack). These areas were chosen because they were among the most severely
acid-impacted regions in North America at the time the case study was conducted and had
substantially high levels of N and S deposition known to have acidified a large number of
waterbodies in both regions.
2.4.1.1 Assessment Approach
For the case studies, the impacts of acidification on waterbodies were assessed by (1)
examining the amount and trends of N and S deposition using NADP data, (2) relating water
quality condition to known biological impacts from acidification, (3) examining past, present,
and future water quality conditions using long-term monitoring data and the Model of
Acidification of Groundwaters in Catchments (MAGIC) biogeochemical model,20 and (4)
calculating critical loads (CLs) that relate deposition levels with water quality conditions. N and
S deposition from all NADP monitors within the case study areas showed acidifying deposition
had declined substantially since 1990, but still remained high relative to other regions in the
United States and historical levels.
Both case studies used sulfate (SO42 ) concentrations, nitrate (NO?) concentrations, and
acid neutralizing capacity (ANC)21 as the chemical indicators of deposition-driven acidification.
ANC was particularly of interest, given its well understood relationship between deposition,
water quality, and biological conditions, as well as its common use in modeling platforms (see
2009 NOx/SOx REA, section 4.2.1). Status of current conditions and trends in S042", NO3",
ANC concentrations measured in surface water were used to characterize links to the effects of
acidifying deposition on the acid-base chemistry of the studied waterbodies, and to determine if
conditions of the waterbodies were improving and recovering or were still degrading. These
trends were analyzed along with monitoring data from the EPA-administered Temporally
Integrated Monitoring of Ecosystems (TIME)/Long-Term Monitoring (LTM) programs along
with survey data from EPA's Environmental Monitoring and Assessment Program (EMAP) for
20 MAGIC is a lumped-parameter model of intermediate complexity, developed to predict the long-term effects of
acidic deposition on surface water chemistry (Cosby et al. 1985a,b,c, 2001). The model simulates soil solution
and surface water chemistry to predict average concentrations of the major ions.
21 ANC was defined in the 2009 NOx/SOx REA as "a key indicator of the ability of water to neutralize the acid or
acidifying inputs it receives. This ability depends largely on associated biogeophysical characteristics, such as
underlying geology, base cation concentrations, and weathering rates." (see "Key Terms" in U.S. EPA 2009)
2-6
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
the years 1990 to 2006. At the time the case studies were completed, the trends analysis found
that the SO42" concentrations were strongly declining in nearly all lakes in the Adirondack while
ANC in many lakes were improving. The results of NO3" concentrations were mixed. For the
Shenandoah case study, SO42", NO3", and ANC concentrations were all variable with some sites
improving while many were still declining.
MAGIC was used to produce estimates of past, present, and future water chemistry in 60
streams in the Shenandoah case study area and 44 lakes in the Adirondack case study area.
Furthermore, MAGIC was used to evaluate the associated risk and uncertainty of the current
levels of acidification given the pre-acidification water quality and the levels of uncertainty in
the input parameters to the model. The MAGIC model output for each waterbody was
summarized into five ANC levels that relate to biological impacts of acidification. The five
ANC levels correspond to the aquatic status categories: Acute Concern, Severe Concern,
Elevated Concern, Moderate Concern, and Low Concern. For each of these levels, the expected
ecological effects were identified (see 2009 NOx/SOx REA, Table 4.2-1).
Additionally, after considering the five ANC levels that correspond to aquatic status
categories, ANC "limits" of biological protection were selected: 0 [j.eq/L (acidic), 20 [j,eq/L
(minimal protection), 50 [j,eq/L (moderate protection), and 100 [j,eq/L (full protection). The
MAGIC modeling results showed that SO42" and NO3" concentrations in the waterbodies at the
time of the case studies were still well above pre-historical conditions. In addition, ANC levels
were much lower, as compared to pre-historical levels, fostering ecological impacts. Lastly, the
MAGIC model indicated that recovery of ANC was unlikely, assuming no changes in emissions
between 2002 and 2050.
The 2009 NOx/SOx REA also calculated CLs22 for waterbodies in each of the case study
areas in order to assess whether current deposition of N and S was high enough to cause
ecological effects. For these case studies, the calculated CLs used ANC limits of 0, 20, 50, and
100 [j,eq/L to define the biological risk to biota. From the 169 modeled lakes and 60 streams in
the Shenandoah and Adirondack case study areas, respectively, the number and percentage of
waterbodies that receive acidifying deposition above their CLs for a given ANC limit of 0, 20,
50, and 100 [j,eq/L were determined. These case studies indicated that large number of
waterbodies were still being impacted by N and S deposition. Between 18 to 58 percent of
modeled lakes in the Adirondacks and 52 to 93 percent of modeled streams in Shenandoah had
exceeded their critical load in 2002 for a given ANC level of 0 to 100. (see 2009 NOx/SOx REA,
section 4.2.4.2).
22 A critical load is formally defined as a quantitative estimate of exposure to one or more pollutants below which
significant harmful effects on specified sensitive elements of the environment do not occur according to present
knowledge (Nilsson and Grennfelt 1988, UNECE 2004).
2-7
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Building from the 2009 NOx/SOx REA, the 2011 NOx/SOx PA focused on a potential
new standard that could protect against ecological effects associated with acidifying deposition
of N and S in aquatic ecosystems. As mentioned in Chapter 1 of this document, the 2011
NOx/SOx PA developed an approach to account for the geographical differences in sensitivity to
acidification across the United States, called the Aquatic Acidification Index (AAI). In general,
the AAI was meant to allow for consistent protection of ecosystems across the U.S. by selecting
a single ANC value for each ecoregions in the U.S., and calculating how much acidifying
deposition from N and S an ecosystem could accept before crossing the selected ANC limit,
accounting for differences in ecosystem sensitivity.
The PA recognized that both dynamic (e.g., MAGIC) and steady state [e.g., Steady state
water chemistry (SSWC)] models calculate ANC and can be used to calculate CLs, and
discussed these models in the context of answering the question as to which modelling approach
would be appropriate for development of a nationally applicable standard (see 2001 PA, section
7.2.2). The 2011 NOx/SOx PA additionally noted that information provided by steady state
modeling of aquatic acidification would be sufficient to develop and analyze alternative NAAQS
and the kind of protection they would afford. In addition, the 2011 NOx/SOx PA noted that the
First-order Acid Balance (FAB) model includes more explicit modeling of N processes including
soil immobilization, denitrification, in-lake retention of N and S, as well as lake size. Hence, the
PA used a combination of the SSWC and the FAB model to inform development of the form of
the standard (the AAI).
Recognizing the spatial variability across the U.S. of the factors in the AAI equation, the
PA suggested that AAI values were meant to be calculated specifically for each ecologically
relevant region (i.e., Ecoregion III (Omernik, 1987)). With regard to a level for the AAI, the PA
concluded consideration should be given to a ANC threshold within the range of 20 to 75 [j,eq/L
noting that a target ANC value of 20 [j,eq/L would be a reasonable lower end of this range, so as
to protect against chronic acidification-related adverse impacts which have been characterized as
severe on fish populations at ANC values below this level. Further, a target ANC value of 75
[j,eq/L would be a reasonable upper end of this range in recognition that the potential for
additional protection at higher ANC values is substantially more uncertain in light of evidence
that acidification-related effects are far less responsive to increases in ANC above this value (see
2011 NOx/SOx PA, section 7.7).
Ecosystem services related to aquatic acidification were discussed in the 2009 NOx/SOx
REA and the 2011 NOx/SOx PA. The 2009 NOx/SOx REA included qualitative discussions of
recreational fishing, which was identified as the service most relevant to aquatic acidification by
freshwater lakes and streams (see 2009 NOx/SOx REA, section 4.2.1). The 2011 NOx/SOx PA
included a quantitative assessment of the recreational fishing benefits to New York residents of
2-8
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
reducing deposition of anthropogenic N and S for the Adirondacks case study area, but also
recognized that there are numerous other ecosystem services that may be related to the ecological
effects of acidification (see 2011 NOx/SOx PA, section 4.4).
2.4.1.2 Key Uncertainties/Limitations
An evaluation of the uncertainty in the parameters for the steady state critical load model
was performed using a Monte Carlo approach in the 2009 NOx/SOx REA.23 This probabilistic
framework specifically assessed the degree of confidence in the exceedance values and the
coefficient of variation (CV) of the critical load and exceedance values. The results suggested
that the estimates of CLs and exceedances for the case study areas were robust. It was noted in
the 2009 NOx/SOx REA, however, that this analysis may understate the actual uncertainty
because some of the range and distribution types of critical load model parameters were not well
understood for aquatic systems in the U.S. at the time.
An evaluation of the uncertainty for the water quality estimates and parameters used in
the MAGIC model was also performed in the 2009 NOx/SOx REA. These uncertainty estimates
were derived by running multiple calibrations of each site using the "optimization" tool and
procedure as part of the MAGIC model. Direct comparison of simulated versus observed water
chemistry values were compared to determine the uncertainty and variability in the MAGIC
model output. Average water chemistry (SO42", NO3", and ANC) simulated versus observed
values during the calibration period (i.e., reference year) were compared for all modeled sites.
The 2009 NOx/SOx REA found that the simulated and observed water quality values were in
close agreement, (see 2009 NOx/SOx REA, section 4.2.8).
No formal analysis of the uncertainty in the AAI was performed in the 2011 NOx/SOx
PA. However, the 2011 NOx/SOx PA noted that uncertainty and natural variability existed in all
of the components of the AAI (for uncertainty analyses conducted for the 2011 NOx/SOx PA,
see Appendix G). In addition, the 2011 NOx/SOx PA noted that there was no apparent
directional bias in the uncertainty regarding the biological, chemical and physical processes
incorporated in the AAI. Lastly, the 2011 NOx/SOx PA noted that the estimates for ecosystem
services generally were believed to be biased low, meaning the monetary and non-monetary
value of reaching a target level of protection is underestimated. However, quantification of these
values was recognized as perhaps the most uncertain of all aspects considered.
Additionally, no formal analysis of uncertainty was performed for ecosystem services in
the 2011 NOx/SOx/PA.
23 For a summary of the Monte Carlo approach used in the 2009 NOx/SOx REA, see section 4.2.8.1.
2-9
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
2.4.2 Aquatic Nitrogen Enrichment
With respect to aquatic N enrichment, the 2008 NOx/SOx ISA determined that the
evidence was sufficient to infer a causal relationship between N deposition and the following:
• Biogeochemical cycling of N in freshwater and estuarine ecosystems
• Alteration to the biogeochemical cycling of carbon (C) in freshwater, estuarine, and near
coastal marine ecosystems; and
• Alteration of species richness, species composition and biodiversity in freshwater and
estuarine ecosystems
The 2009 NOx/SOx REA conducted a case study for estuaries in the Chesapeake Bay and
the Pamlico Sound areas. Both areas were selected primarily based on the availability of data.
Because the Chesapeake Bay and Pamlico Sound are fed by multiple river systems, the case
study was scaled to one main stem river for each system: the Potomac River/Potomac Estuary (in
the Washington, DC metropolitan area) and the Neuse River/Neuse River Estuary (in North
Carolina).
2.4.2.1 Assessment Approach
The 2009 NOx/SOx REA focused the case study on aquatic nitrogen enrichment of
estuaries and used the National Oceanic and Atmospheric Administration (NOAA)'s Assessment
of Estuarine Trophic Status Eutrophication Index, commonly referred to as ASSETS EI (Bricker
et al., 2007), as the ecological indicator for the case study. ASSETS EI is an estimation of the
likelihood that an estuary is experiencing eutrophication or will experience eutrophi cation in the
future based on five chemical and/or biological indicators: chlorophyll a, macroalgae, dissolved
oxygen, nuisance/toxic algal blooms, and submerged aquatic vegetation (SAV) (Bricker et al.,
2007).
Specifically, the analysis in this case study sought to determine the change in N loading
required to improve the ASSETS EI from its current level set in the 2002 current condition
analysis.24 To create response curves for the ASSETS EI based on changes in N loads to an
estuary, the case study used the U.S. Geological Survey's (USGS's) SPAtially Referenced
Regressions on Watershed attributes (SPARROW) model to calculate the N loads.25
The case study found that a decrease of 78 percent of atmospheric N deposition would be
required to improve the eutrophication index category ASSETS EI Score for the Potomac
24 Current conditions were evaluated using 2002 CMAQ model year and NADP monitoring data.
25 SPARROW relies on a nonlinear regression formulation to relate water quality measurements throughout the
watershed of interest to attributes of the watershed. SPARROW can be used to predict total N loads at the outlet
of the watershed that result from changes in the total N atmospheric deposition loads.
2-10
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
River/Potomac Estuary from Bad to Poor (see 2009 NOx/SOx REA, section 5.2.7.1), and that
the Neuse River/Neuse River Estuary ASSETS EI score could not be improved from Bad to
Poor with only decreases in the 2002 atmospheric N deposition load to the watershed (see 2009
NOx/SOx REA, section 5.2.7.2). The 2009 NOx/SOx REA found that even if all atmospheric N
deposition inputs were eliminated (100% decrease), the total annual N load to the Neuse River
Estuary would only decrease by 4%. This small effect is because atmospheric deposition of
nitrogen to the Neuse River watershed are small relative to other sources, such as run-off and
agricultural sources.
The 2009 NOx/SOx REA included qualitative and quantitative assessments of ecosystem
services related to aquatic nutrient enrichment in the Potomac River/Potomac Estuary and Neuse
River/Neuse River Estuary (see 2009 NOx/SOx REA, section 5.2.1). The 2009 NOx/SOx REA
evaluated several cultural ecosystem services, including recreational fishing, boating, and beach
use. In addition, aesthetic and nonuse values were evaluated; the impacts on recreational fishing
(e.g., closings, decreased species richness) to eutrophication symptoms through monitoring data
were quantitatively linked; other recreational activities and aesthetic and non-use services to
eutrophi cation symptoms were quantitatively related through user surveys and valuation
literature; and the current commercial fishing markets were described. The 2011 NOx/SOx PA
projected the quantitative change in the provision of these services based on the changes in water
quality related to a policy scenario that eliminated the deposition of anthropogenic N to the
Chesapeake Bay. The PA also included a hedonic study for aesthetic benefits related to
improved water quality for near-shore residents, (see 2011 NOx/SOx PA, section 4.4.5)
2.4.2.2 Key Uncertainties/Limitations
The 2009 NOx/SOx REA noted potential uncertainties in the inputs and outputs of the
SPARROW model, as well as sensitivity of the SPARROW model. The 2009 NOx/SOx REA
also noted uncertainties in the inputs to the ASSETS EI, given the numerous data requirements
and sources required to conduct a full ASSETS EI analysis.
The 2009 NOx/SOx REA determined that the small effect of decreasing atmospheric
deposition in the Neuse River watershed is because the other N sources within the watershed are
more influential than atmospheric deposition to the total nitrogen loadings to the Neuse River
Estuary as estimated with the SPARROW model. The 2009 NOx/SOx REA noted that future
application of the methods to case study areas where atmospheric deposition plays a larger role
in the N loading to an estuary will likely provide more tangible results.
The 2011 NOx/SOx PA noted that the relative lack of empirical models and valuation
studies imposed obstacles to the estimation of ecosystem services affected by N deposition
2-11
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
resulting in underestimation of the monetary and non-monetary value of changes in ecosystem
service provision, (see 2011 NOx/SOx PA, section 4.4.5)
2.4.3 Terrestrial Acidification
With respect to terrestrial acidification, the 2008 NOx/SOx ISA determined that the
evidence was sufficient to infer a causal relationship between N and S deposition and changes in
biogeochemistry and biota.
At the time of the 2009 NOx/SOx REA, Picea rubens (red spruce) and Acer saccharum
(sugar maple) were the best studied species in North America with regard to impacts of
acidification from N and S deposition to terrestrial systems. Hence, the 2009 NOx/SOx REA
conducted a case study to evaluate the effects of acidifying deposition on sugar maple and red
spruce physiological condition leading to impacts on tree growth and/or mortality in the Kane
Experimental Forest and Hubbard Brook Experimental Forest, located in the Allegheny Plateau
region in Pennsylvania and New Hampshire's White Mountains, respectively.
2.4.3.1 Assessment Approach
In the case study areas, critical load calculations were applied to multiple areas within 24
states for sugar maple and in 8 states for red spruce—site locations within each state were
determined by the U.S. Forest Service's Forest Inventory and Analysis (FIA) database permanent
sampling plots. The 2009 NOx/SOx REA noted that the Simple Mass Balance (SMB) model
was a commonly used tool for evaluating soil acidification. The SMB model specifically
examines a long-term, steady-state balance of base cation, chloride, and nutrient inputs, "sinks,"
and outputs within an ecosystem, and base cation equilibrium is assumed to equal the system's
critical load for ecological effects. The 2009 NOx/SOx REA also noted that the base
cation/aluminum (Bc/Al) ratio is a good indicator for soil acidification and relates well to the
calcium/aluminum (Ca/Al) ratio in the soil solution, (see 2009 NOx/SOx REA, section 4.3.1.1)
Hence, the 2009 NOx/SOx REA quantitative assessment included the use of the SMB model to
calculate CLs for soil acidification, with the Bc/Al ratio as the indicator.
It should be noted that in order to properly use the SMB model, which is expressed as an
equation, the user must first choose a critical level or levels for the Bc/Al ratio. For purposes of
the 2009 NOx/SOx REA, data published by Sverdrup and Warfvringe (1993) was used to make
decisions regarding the Bc/Al ratio as well as percent estimates for negative tree responses.
Bc/Al values of 0.6 and 1.2 were selected based on a 50% and 75% chance of negative tree
response (i.e., >20% reduced growth) for sugar maple and red spruce. A Bc/Al level of 10 was
also chosen to represent the lowest impact (greatest level of protection) to tree growth. At the
time of the 2009 NOx/SOx REA, it was considered the most conservative value used in studies
that had calculated CLs in the U.S. and Canada, (see 2009 NOx/SOx REA, section 4.3.4).
2-12
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
One important input into the SMB model is base cation weathering (BCw) rates, which
the 2009 NOx/SOx REA estimated using a clay-substrate method. At the time of the REA, this
method was one of the most-commonly used methods for estimating this parameter.
Case study results suggested that the health of at least a portion of the sugar maple and
red spruce growing in the U.S. may have been compromised with acidifying total N and S
deposition in 2002. We concluded that the pattern of case study results suggests that N and S
acidifying deposition in the sugar maple and red spruce forest areas studied were very close to, if
not greater than, the CLs for those areas and both ecosystems are likely to be sensitive to any
future changes in the levels of deposition.
The 2009 NOx/SOx REA also identified several ecosystem services that a deterioration
of sugar maple and red spruce tree health could negatively impact. These include provisioning
(use of trees for timber and maple syrup), cultural (endangered and threatened species habitat),
regulating (soil stabilization and erosion control, water regulation, and climate regulation) and
recreational ecosystem services (fall color viewing). The 2009 NOx/SOx REA used preference
studies for the southern Appalachians to capture willingness to pay (WTP) for forest
improvements that specifically addressed non-use values. Most services were qualitatively
evaluated due to lack of data. However, the 2009 NOx/SOx REA did include a pilot study using
the Forest and Agriculture Sector Optimization Model (FASOM) to estimate changes in timber
harvest for red spruce and sugar maple. Uncertainties in model parameterization, however, were
deemed too high to allow utilization of this model (see 2009 NOx/SOx REA, section 4.3.1).
2.4.3.2 Key Uncertainties/Limitations
As mentioned earlier, the 2009 NOx/SOx REA used the SMB model to estimate CLs for
soil acidification, and used BCw as an input into the model. The 2009 NOx/SOx REA noted that
one limitation of the SMB model is that it is a steady-state model and therefore does not capture
the cumulative changes in ecosystem conditions. Additionally, the 2009 NOx/SOx REA also
noted that the estimates for BCw rates and forest soil ANC input parameters were the main
sources of uncertainty since these parameters are rarely measured and require researchers to use
default values. The 2009 NOx/SOx REA also stated that the BCw value is strongly influenced by
the classified acidity of the soil parent material, and is poorly measured in non-glaciated soils,
(see 2009 NOx/SOx REA section 4.3.9). The 2009 NOx/SOx REA determined that the BCw
rates contributed 49% to the total variability in the critical load estimates, and forest soil ANC
contributed 46% to the total variability, (see 2009 NOx/SOx REA, section 4.4)
2-13
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
2.4.4 Terrestrial N Enrichment
With respect to terrestrial N enrichment, the 2008 NOx/SOx ISA determined that the
evidence was sufficient to infer a causal relationship between N deposition and the alteration of
the following:
1. Biogeochemical cycling of N and C;
2. Biogeochemical flux of methane (CH4) and nitrous oxide (N20); and
3. Species richness, species composition, and biodiversity
At the time of the last review, there was a large body of evidence showing that some of
the highest N deposition has occurred in Southern California, where researchers have
documented measurable ecological changes related to atmospheric deposition. These changes
include increases in nonnative grasses and fire susceptibility for Coastal Sage Scrub (CSS), as
well as tree mortality, increased fire intensity, and a change in nutrient cycling in mixed conifer
forests (U.S. EPA, 2008).
2.4.4.1 Assessment Approach
The 2009 NOx/SOx REA included a case study for CSS and Mixed Conifer Forest
(MCF) communities in the Sierra Nevada Range and San Bernardino Mountains of California.
No quantitative assessment was conducted for this case study. Rather, geographic information
systems (GIS) analysis supported a qualitative review of past field research to identify ecological
benchmarks associated with CSS and mycorrhizal communities, as well as MCF's nutrient-
sensitive acidophyte lichen communities, fine-root biomass in Ponderosa pine, and leached NO3"
in receiving waters. The benchmarks were identified from empirical studies, including CLs, in
the southern California region. These benchmarks, ranging from 3.1 to 17 kg N/ha/yr for CSS
and MCF, were compared to 2002 CMAQ/NADP data to discern any associations between
atmospheric deposition and changing communities. Evidence supported the finding that N alters
CSS and MCF communities.
The 2009 NOx/SOx REA also used data from Rocky Mountain National Park to examine
the sensitivity and effects of nutrient enrichment on aquatic and terrestrial ecosystems, and found
that exposure levels at which negative effects were observed appeared to be generally
comparable to levels identified in other sensitive U.S. ecosystems (benchmarks range from 1.5 to
30.5 kg N/ha/yr). In addition, it included a qualitative assessment of ecosystem services,
focusing on cultural services, including: habitat for endangered and threatened species,
recreational (e.g., hiking, fishing, and hunting), aesthetic (view of the landscape), non-use
(existence value), and regulation (e.g., water, regulation of fire intensity).
To assess the impact of fire risk, the 2009 NOx/SOx REA used GIS to overlay N
deposition with the locations of MCF and CSS, and CALFIRE data (State of California fire
2-14
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
occurrence, prevention and fire-fighting expenditures) to describe the potential benefit of
reducing N enrichment to these fire-prone areas. This was accompanied by a discussion of the
hedonic benefits to homeowners and aesthetic benefits to general public for reductions in fire
damages, (see 2009 NOx/SOx REA, section 5.3.1)
The 2009 NOx/SOx REA provided a qualitative discussion of the services offered by
CSS and MCF and a sense of the scale of benefits associated with these services. Specifically,
the 2009 NOx/SOx REA stated the following: CSS and MCF are an integral part of the
California landscape, and together the ranges of these habitats include the densely populated and
valuable coastline and mountain areas of the state. Through recreation and scenic value, these
habitats affect the lives of millions of California residents and tourists. Numerous threatened and
endangered species at both the state and federal levels reside in CSS and MCF. Both habitats
may play an important role in wildfire frequency and intensity, an extremely important problem
for California. The potentially high value of the ecosystem services provided by CSS and MCF
justify careful attention to the long-term viability of these habitats, (see 2009 NOx/SOx REA, p.
5-60).
2.4.4.2 Key Uncertainties/Limitations
The 2009 NOx/SOx REA noted that the exact relationship between atmospheric N
loadings, fire frequency and intensity, and nonnative plants, particularly in the CSS ecosystem,
had not been quantified in the scientific literature. The 2009 NOx/SOx REA noted that although
various conceptual models linking these factors had been developed, an understanding of cause
and effect, seasonal influences, and thresholds remained undeveloped.
Overall, the REA concluded that although the available data used for the targeted effect
of terrestrial N enrichment were considered high quality, there was a limited ability to
extrapolate these data to a larger regional area.
2.4.5 Mercury Methylation
With respect to mercury methylation, the 2008 NOx/SOx ISA determined that the
evidence was sufficient to infer a causal relationship between S deposition and increased
methylation of mercury in aquatic environments, where the value of other factors is within
adequate range for methylation.
Information available at the time of the last review demonstrated that methylmercury
(MeHg) production is mediated primarily by sulfur-reducing bacteria (SRB), and changes in
SO42" deposition result in changes in both Hg methylation rates and Hg concentration in fish. It
was also shown that watersheds with conditions known to be conducive to Hg methylation could
be found in the northeastern U.S. and southeastern Canada, though biotic Hg accumulation had
2-15
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
been widely observed in other regions that had not been studied as extensively, and where a
different set of conditions may exist.
The 2009 NOx/SOx REA acknowledged that a number of factors influence or modify the
relationship between SO42" and the methylation of Hg. It noted that while there appears to be a
relationship between SO42" deposition and mercury methylation, the rate of mercury methylation
varies according to several spatial and biogeochemical factors whose influence has not been fully
quantified (2009 NOx/SOx REA, section 6.2.1).
2.4.5.1 Assessment Approach
Given the factors considered, the 2009 NOx/SOx REA included a qualitative assessment
of mercury methylation effects, focusing on Little Rock Lake in Wisconsin. It noted that
decreases in SO42" deposition were linked to observed decreases in MeHg fish tissue
concentrations in the lake.
The 2009 NOx/SOx REA also discussed qualitatively the provisioning and cultural
services potentially impacted by mercury methylation. The REA referenced commercial and
sport fishing, shell fishing, fishing for subsistence, and the cultural and spiritual significance
derived from fishing and consuming local fish or shellfish (particularly for Native Americans
and Alaska Native villagers).
2.4.5.2 Key Uncertainties/Limitations
Overall, the REA noted that decreases in SO42" deposition will likely result in decreases
in MeHg concentrations, but that the rate of methylation varies spatially and with
biogeochemical factors so that the correlation of SO42" deposition and MeHg could not be
quantified for the purpose of interpolating the association across waterbodies or regions (see
2009 NOx/SOx REA, section 6.2.2.2). This limitation hindered the ability to establish large scale
dose-response relationships.
2-16
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
3 CONSIDERATION OF NEWLY AVAILABLE
INFORMATION
In this chapter, we consider the extent to which our characterization of risk and exposure
in the last review remains appropriately informative to the key questions in the current review, as
summarized in the IRP. As noted in that document, the ISA, REA (if warranted) and PA
developed in this review will provide the basis for addressing the key policy-relevant questions
and will inform the Administrator's judgment as to the adequacy of the secondary NAAQS for
oxides of nitrogen, oxides of sulfur, and PM. Accordingly, this chapter considers the extent to
which any new assessment of risk and exposure is warranted to inform decisions regarding
potential adverse effects to public welfare. Based on these considerations, as described below,
we plan to develop a new REA to inform the current review.
3.1 KEY CONSIDERATIONS
In any NAAQS review, considerations contributing to a decision to conduct a new risk
and exposure assessment include the role played by the risk and exposure information in the
EPA's previous decisions on the existing standards and the role of risk and exposure information
expected in the current review. Another important consideration is the robustness of the risk and
exposure estimates for the existing standards that are available from the last review. In reaching a
decision on conducting a new REA, we also consider the extent to which results of a new
quantitative risk and exposure assessment are expected to appreciably change our understanding
of risk and exposures beyond the insights gained from the assessments conducted for the last
review. More specifically, we consider questions such as the following, which are summarized in
Figure 3-1:
• Is appropriate scientific and technical information available to support quantitative
assessments?
• Is the scientific and/or technical information that could inform updated quantitative
assessments substantially different from that available in the previous reviews?
• Would the new information appreciably reduce the uncertainties and limitations identified
in previous reviews?
• Would updated quantitative assessments likely inform decision making in the current
review by adding substantially to our understanding of pollutant exposures or pollutant-
attributable risks, beyond the insights gained from previous reviews and assessments?
Regarding these questions, section 3.2 below considers the newly available air quality
information with a focus on current air quality concentrations and deposition and whether those
levels across the U.S. have changed significantly since the last review. Additional consideration
3-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
is also given to new tools, data and information that could inform improvements in the
quantification of the relationship between ambient air quality concentrations and deposition, a
key uncertainty in the previous review. Sections 3.3 through 3.7 then consider the scientific
evidence and data on ecological effects available in the second draft ISA that play critical roles
in our characterization of risk and exposure. Particular attention is given to consideration of new
information pertaining to those aspects, and the extent to which it might be expected to address
key areas of uncertainty identified in the last review. Based on these considerations, section 3.9
then provides preliminary conclusions on the extent to which updated quantitative analyses of
ecological risk and exposure are warranted in the current review. This document then presents
our proposed plan for performing such analyses in Chapter 4.
Is appropriate
scientific and
technical
information
available to
support
quantitative
assessments?
No
Yes
Is the scientific and/or technical
information that could inform
updated quantitative assessments
substantially different from that
available in previous reviews?
-and-
Would the new information
appreciably reduce the
uncertainties and limitations
identified in previous reviews?
No
Yes
Would updated quantitative
assessments likely inform
decision making in the current
review by adding substantially
to our understanding of
pollutant exposures or
pollutant-attributable
risks, beyond the insights
gained from previous reviews
and assessments?
No
Yes
I
Quantitative assessments are not supported in the current review
Quantitative assessments
are supported in the
current review
Figure 3-1. Key considerations for updated or new quantitative analyses
3.2 AIR QUALITY INFORMATION
3.2.1 Emissions, Atmospheric Concentrations and Deposition
Sulfur oxides are emitted into the air from specific sources (e.g., fuel combustion
processes) and are also formed in the atmosphere from other atmospheric compounds (e.g., as an
oxidation product of reduced sulfur compounds, such as sulfides). Sulfur oxides are also
transformed in the atmosphere to particulate sulfur compounds, such as SO42". The most
prevalent sulfur oxide in the atmosphere is SO2. Emissions of SChin the U.S. are largely due to
coal-fired power plants and diesel fuel combustion (second draft ISA, Appendix 2, section 2.2.1).
3-2
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Oxidized nitrogen species considered here range from NO and NO2, collectively referred
to as NOx, to higher order organic and inorganic oxidation products, collectively referred to as
NOz (e.g., pN03, HNO3, HONO, PAN, other organic nitrates). The largest sources of NOx
emissions are related to combustion sources, which includes power plants, industrial facilities,
vehicles, and wood burning stoves. Non-controllable sources of NOx include wildfires, biological
soil processes, and lightning (second draft ISA, Appendix 2, section 2.2.1).
Reduced nitrogen species (i.e., NH3 + NH4+ = NHx) are considered in this review. While,
NH3 is not a criteria pollutant, it is considered to the extent that it contributes to atmospheric
transformations and loading to ecosystems. NH3 is a precursor for atmospheric PM, reacting with
gas phase nitric acid (HNO3) to form ammonium nitrate (NH4NO3), a major contributor to N
deposition. Livestock and fertilized fields are the largest sources of NH3, but there are
combustion related sources as well. Particles containing SO42", NO3", and NH4+ are also directly
emitted into the atmosphere from sources such as wind-blown dust or sea-salt spray, but in total,
direct particulate emissions are a small contribution to total emissions of N and S containing
compounds (second draft ISA, Appendix 2, section 2.2.1).
The geographic distributions of NOx and SOx emissions reported in the National
Emissions Inventory (NEI) reflect the fact that transportation and power generation source
sectors dominate NOx and SOx emissions, respectively. In the last review, the density of
emissions sources of NOx and SOx was highest in the eastern U.S., and around population centers
and transportation corridors in both the east and west. This remains true for this review, though
these areas of the country have also experienced some of the largest reductions in emissions of
NOx and SOx (second draft ISA, Appendix 2, section 2.2.1). National average SO2 emissions are
estimated to have declined by 81% and NOx emissions have declined by 53% over the period
from 2002 to 2016. Such declines in emissions are likely related to the implementation of
national control programs developed under the Clean Air Act Amendments of 1990, including
Phase I and II of the Acid Rain Program, the Clean Air Interstate Rule, the Cross-State Air
Pollution Rule, and the Mercury Air Toxic Standards, as well as changes in market conditions,
e.g., reduction in energy generation by coal (U.S. EIA, 2018).
The largest sources of NH3 emissions are agricultural: livestock, including confined
animal feeding operations, and soils after addition of N containing fertilizers. Motor vehicles can
also be a substantial contributor to total NH3 emissions in urban areas. In contrast to SOx and
NOx, Xing et al. (2013) estimated that the national emissions of NH3 increased between 1990 to
3-3
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
2010. Because of this, in many places in the U.S., reduced forms of N are now the largest
contributor to N deposition26 (second draft ISA, Appendix 2, section 2.6.2).
In the atmosphere, NOx, SO2, and NH3 undergo chemical transformations and the
resulting compounds are removed from the atmosphere by dry deposition and wet deposition at
different rates. For example, NO2 can oxidize to form nitric acid (HNO3), which dry deposits
rapidly. However, in the presence of NH3, particulate NH4NO3, which deposits via dry
deposition slower than both HNO3 and NH3. Wet deposition, the scavenging of gases and
particles by cloud droplets and precipitation, is the most important removal process for NH4NO3.
The deposition rates of oxidized sulfur also depend on the chemical form; particulate sulfate does
not dry deposit as rapidly as freshly emitted SO2. Since the chemical form is important to
determining the rate of dry and wet deposition, as well as the relationship between air
concentrations and deposition, we use process-based models and quality-assured ambient
measurements to understand the transformation from emissions to concentrations to deposition.
Both measurements, models, and techniques to develop merged model-measurement datasets
have been advanced since the previous review (see second draft ISA, Appendix 2, sections 2.5
and 2.6).
Several monitoring networks measure the atmospheric concentrations and wet deposition
of NOx, SOx, and PM. The most relevant routinely operating networks measuring ambient air
concentrations include the Interagency Monitoring of Protected Visual Environments
(IMPROVE), the National Core network (NCore) and EPA Chemical Speciation Networks
(CSN), Clean Air Status and Trends Network (CASTNet), and the ammonia monitoring network
(AMoN). The National Trends Network (NTN) is part of the National Atmospheric Deposition
Program and measures wet deposition of NO3", SO42", NH4+, and other ions. Several new
monitoring efforts are underway since the previous review. Gas phase NH3 has expanded and as
of May 2018, is measured at 98 sites as part of the AMoN network. The NCore network includes
measurements of NO, NOy, SO2 and PM chemical composition at 63 urban and 17 rural sites.
The near-road network includes NO2 measurements at 80 monitoring sites located near
roadways. Of these monitoring networks, the co-located CASTNet, AMoN and NTN sites are
of particular interest because they measure both air concentrations and deposition of N and S
containing compounds.
However, some limitations remain. Particulate NH4+ measurements from CASTNet and
IMPROVE networks are thought to have biases that depend on meteorological conditions and
require careful consideration when used for analysis. While methods for directly measuring dry
26 A portion of the reduced N deposition is from NH4+ bound in PM. The NH4+ fraction changes according to the
relative levels of sulfate and NO3" available to form particles.
3-4
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
deposition in routine networks have been deployed in Europe (Skiba et al., 2009), direct
measurements of dry deposition in the U.S. are sparse and infrequent. The spatial coverage of the
monitoring networks is representative of some, but not all, of the variability in ecosystems and
concentrations nationally. To help address the limitations in the available observations of air
concentration and deposition, a combination of observations and computational models is often
employed (see second draft ISA, section 2.6).
The National Atmospheric Deposition Program's total deposition science committee has
been advancing the state of the science to create datasets that estimate annual average deposition.
The Total deposition (TDEP) dataset combines measurements and models to estimate annual
average deposition for years 2002 - 2016 (Schwede and Lear, 2014). Spatial coverage includes
the continental U.S. at 4-km horizontal resolution. Wet deposition is estimated by spatial
interpolation of NTN measurements and dry deposition is calculated by the Community
Multiscale Air Quality (CMAQ) model. CMAQ is a chemical transport model that simulates the
fate and transport of gases and particles by using physically-based, numerical models and
explicitly including the processes of advection, dispersion, chemistry, aerosol physics, cloud
processes, dry deposition, and wet deposition. The use of modeled estimates is necessary to
account for dry deposition, which is not measured on a broad enough scale to inform national
applications. Using the TDEP dataset to analyze ecosystem effects could potentially reduce
uncertainties associated with the use of deposition values estimated by CMAQ alone, as was
done in the last review. In addition, we note the representation of chemical and physical
processes in CMAQ has been further improved since the last review, with updates that include
gas-phase chemistry relevant to the oxidation of NOx, dry deposition of particles by gravitational
settling, interactions between meteorology and particles, and the influence of temperature, wind,
and precipitation on NH3 emissions (see second draft ISA, Appendix 2, section 2.5).
To further assess how deposition has been estimated to have changed since the last
review, Figures 3-2 and 3-3 are included to illustrate TDEP estimates of S deposition across the
U.S. for the years of 2002 and 2015, respectively, with Figure 3-4 showing the estimated
changes in S deposition between these two years. Similarly, the TDEP estimates of N deposition
for 2002 and 2015 are shown in Figures 3-5 and 3-6, with changes between these two years
shown in Figure 3-7. These figures mimic the same general patterns, discussed above, for the
changes in emissions since the last review. Generally, substantial changes in S deposition has
occurred across most of the U.S., with the largest decreases occurring in the eastern U.S.
Similarly, N deposition has also generally decreased across the entire U.S., with increases in
areas of the country generally dominated by agricultural sources, due to an increase in reduced
nitrogen deposition in these areas.
3-5
-------
1
2
3
4
5
6
7
8
9
10
11
12
Given that emissions, concentrations and deposition levels of N and S species have
changed substantially across the U.S. since the last review and given the availability of new and
improved tools and methods for estimating this deposition (e.g., TDEP), we judge that new
analyses are supported for this review. Assessment of the most recently available emissions,
concentration and deposition datasets are likely to influence important differences in any new or
updated quantitative assessments and inform our understanding of changes in pollutant
exposures and pollutant-attributable risks since the last review. In addition, application of the
improved tools and methods will likely reduce uncertainties and limitations in the
characterization of air concentrations and deposition from the last review.
*
Total Deposition of Sulfur (kg S ha 1) 2002
2002
kg/ha
I | 0(o2
2.1 >0 4 ,
H4
IB ei»3
Figure 3-2. Annual total deposition of sulfur, kg S ha"1 in 2002.
3-6
-------
Total Deposition of Sulfur (kg S ha1) 2015
2015
kg/ha
I 10to2
H 21104
B 4 1 K>6
¦I 6 110 8
1
2 Figure 3-3. Annual total deposition of sulfur, kg S ha"1 in 2015.
Change in Total Deposition of Sulfur
(kg S ha1) 2015 - 2002
Difference (2015-2002)
kgJha
H -76 to -26
I 2* 9 to -10
¦ -99to-5
| -4.9 to -1
I | -9to0
Hoi
ito6
3
4 Figure 3-4. Change in annual total sulfur deposition, 2015 - 2002 (kg S ha"1)
5
3-7
-------
Total Deposition of Nitrogen (kg N ha1) 2002
j jfc* Q
1
2 Figure 3-5. Annual total deposition of nitrogen, kg N ha"1 in 2002.
201S
kali.
LU
0»1
»1*>»
»'to*
MM It
12
Total Deposition of Nitrogen (kg N ha1) 2015
j
H ...
3
4 Figure 3-6. Annual total deposition of nitrogen, kg N ha"1 in 2015.
5
3-8
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Change in Total Deposition of Nitrogen (kg N ha x) 2015 - 2002
Figure 3-7. Change in annual total nitrogen deposition, 2015 - 2002 (kg N ha"1)
3.2.2 Linking Atmospheric Concentration Changes to Changes in Deposition
The new modeling approaches, additional years of measurements, and combined model-
measurement TDEP estimates described in the previous section are also relevant for linking
atmospheric concentration changes to changes in deposition. In the last review, the "transference
ratio" approach was used to estimate the relationship between air quality concentrations and
deposition of N and S. The uncertainties relevant to this transference ratio approach are
summarized in Chapter 2. Also, since the last review, several studies have examined the
transference ratio using measurements and modeling, as described in more detail in the second
draft ISA, Appendix section 2.5.2.4. Specific areas of concern are the drivers of the spatial
variability of the computed transference ratio, differences in the transference ratio when
calculated by different chemical-transport models, and errors associated with using the air
concentration of NOy as a predictor of nitrogen deposition. These recent findings and
uncertainties, listed in Chapter 2, will be used to inform the updated approach to relating changes
in ambient concentrations to changes in atmospheric deposition, which is described in Chapter 4.
Briefly, the updated approach analyzes 20-years of concentration and deposi tion measurements
and CMAQ model results, includes consideration of uncertainty and variability using multiple
chemical-transport models, and carefully evaluates how the concentration/deposition relationship
may be different near sources versus in remote areas.
3-9
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
3.3 DIRECT EFFECTS
New scientific evidence supports the causality determinations made in the 2008
NOx/SOx ISA regarding the gas-phase effects of NO, NO2, PAN, SO2. and HNO3 on vegetation;
there are no new causal determinations.27
At the time of the 2008 NOx/SOx ISA, it was known that NO, NO2, and PAN can have
phytotoxic effects on plants by decreasing photosynthesis and inducing visible foliar injury.
Since that time, very little new research has been performed on these phytotoxic effects at
concentrations currently observed in the U.S. (see second draft ISA, Appendix section 3.6.2).
Additionally, according to the second draft ISA, although PAN continues to persist as an
important component of photochemical pollutant episodes, there is little evidence in recent years
to suggest that PAN poses a significant risk to vegetation in the U.S. (see second draft ISA,
Appendix section 3.3)
At the time of the 2008 NOx/SOx ISA, it was known that gaseous SO2 caused foliar
injury as well as decreased photosynthesis, growth, and yield in plants. Additionally, it was
known that SO2 caused mortality in lichens. Due to declines in SO2 emissions, few additional
studies evaluating direct gaseous effects of S02have been conducted in the U.S, and these
studies generally focused on recovery (see second draft ISA, Appendix section 3.2).
Additionally, per the second draft ISA, there is no clear evidence of acute foliar injury below the
level of the current SO2 standard (see second draft ISA, Appendix section 3.6.1).
The 2008 NOx/SOx ISA reported that experimental exposure of HNO3 resulted in
damage to the leaf cuticle of pine and oak seedlings, which could predispose those plants to other
stressors such as drought, pathogens, and other air pollutants. Since the 2008 ISA, one study
(Padgett et al., 2009) investigated dry deposition of HNO3 on the foliage, with findings that
supported the earlier research. The 2008 ISA also reported several lines of evidence in lichen
studies, including transplant and controlled exposure studies, indicating that HNO3
concentrations contributed to the decline in lichen species in the Los Angeles basin. Since that
time, several new studies have been published that continue to support this evidence (see second
draft ISA, Appendix section 3.4).
Limited new evidence regarding direct effects of NO, NO2, PAN, and SO2 has become
available since the 2008 NOx/SOx ISA. Additionally, there are no data and/or tools available to
support a quantitative assessment of direct gaseous effects. Hence, we do not intend to conduct a
quantitative assessment for these effects in the REA.
27 The causality statements in the second draft ISA relate to gas phase SO2, NO, NO2, PAN, and HNO3 and injury to
vegetation.
3-10
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
3.4 NON-NITROGEN AND NON-SULFUR EFFECTS OF PM
Since publication of the 2009 PM ISA, there has been some new evidence to support the
understanding of ecological effects associated with PM components, particularly for metals and
organics. The causality determination from the last review remains unchanged, finding that there
is a likely causal relationship between PM deposition and a variety of effects on individual
organisms and ecosystems. While the new studies do provide some additional evidence for
community-level responses to PM deposition, the second draft ISA notes that uncertainties
remain due to the difficulty in quantifying relationships between ambient concentrations of PM
and ecosystem response (see second draft ISA, Chapter 1, section 1.10). Additionally, there are
currently no data and/or tools available for evaluating these effects. Hence, we do not intend to
conduct a quantitative assessment for these effects in the REA.
3.5 FRESHWATER ACIDIFICATION AND NITROGEN ENRICHMENT
Since the 2008 NOx/SOx ISA, new studies have been published on the effects of N and S
deposition on freshwater ecosystems. This evidence includes new CLs for freshwater ecosystems
and has led to the development of expanded causality determinations in the second draft ISA
from the last review. These CLs provide national coverage and could significantly expand the
scope for analyses since the last review, and for N enrichment, to fill a major gap from the last
review. The sections that follow provide more details about the new scientific and technical
information pertaining to freshwater acidification and N enrichment that are available for
consideration in the REA.
3.5.1 Freshwater Acidification
While the causality determination for biogeochemistry changes remains largely
unchanged,28 the causality determination for changes in biota have been expanded in the second
draft ISA, having found that the body of evidence is sufficient to infer a causal relationship
between acidifying atmospheric deposition and changes in biota, including physiological
impairment and alteration of species richness, community composition, and biodiversity in
freshwater ecosystems (rivers, lakes, and streams).
At the time of the 2008 NOx/SOx ISA, it was known that atmospheric N deposition can
alter the pools and fluxes of the C and N cycles, causing nitrification and denitrification and
NO3" leaching to surface waters, increasing acidity. Atmospheric deposition of S directly adds
SO42" to soil leachate and surface waters, increasing acidity. The processes of acidification and
28 See Table 1-1 of the second draft ISA for a side-by-side comparison of causality statements from the last review
and this review.
3-11
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
the chemical indicators that describe it were well characterized at the time of the 2008 NOx/SOx
ISA, and newer studies have only further described and quantified some of these relationships
(second draft ISA, Appendix section 7.1.7). Effects of N and/or S deposition on changes in biota
are linked to chemical indicators in surface water, which include water pH, ANC, and the
concentrations of SO42", dissolved inorganic aluminum (Al), and base cations. The strongest
evidence for a causal relationship between acidifying deposition and aquatic biogeochemistry
comes from studies of changes in surface water chemistry (second draft ISA, Chapter 1, section
1.6.1). Consistent evidence from multiple studies spanning several decades shows that in
acidified waters, changes in surface water chemistry can cause the loss of acid-sensitive
biological species. Biological effects are primarily attributable to low pH and elevated
concentrations of inorganic Al (see second draft ISA, Appendix section 8.1). ANC integrates
chemical components of acidification, and surface water acidification models calculate ANC as a
proxy that relates to pH and inorganic Al concentrations. However, ANC does not directly alter
the health of biota (see second draft ISA, Chapter 1, section 1.6.3).
Since the last review, there is new evidence for chemical recovery from acidification in
freshwater ecosystems (see second draft ISA, Appendix section 7.1.5.1), but evidence for
biological recovery has been more limited. Biological recovery in a freshwater ecosystem can
occur only if chemical recovery is sufficient to allow reproduction, growth, and survival of acid-
sensitive plants and animals to occur (see second draft ISA, Appendix 8, Section 8.4). Also,
chemical recovery of ANC or pH may not necessarily lead to recovery of the ecosystem to its
previous condition before acidification, due to changes in relationships among ANC, pH, DOC,
and Al; depletion of soil base cation pools; and/or partially reversible (or irreversible) S
adsorption on soils. In the 2008 NOx/SOx ISA, studies of biological recovery generally indicated
that the time required for full biological recovery is uncertain and that responses in biota lag
behind chemical recovery and may take decades from the onset of chemical recovery. New
studies documented in the second draft ISA of multiple trophic levels continue to support these
findings.
As mentioned earlier in Chapter 2, the 2009 NOx/SOx REA related ANC (as an indicator
that relates well to pH) to fish species richness and species-level responses to acidification in
lakes and streams in two case study areas, and calculated CLs for those areas. Since that time,
new CLs using ANC as a chemical indicator for freshwater acidification have become available
in the peer reviewed literature. Some of these new data are based on studies that used SMB
models to calculate CLs (see second draft ISA, Appendix section 8.5.4). McDonnell et al.
(2014), for example, developed a regional dataset that provides CLs for continuous coverage of
streams for the southern Appalachian Mountains region from Georgia to southern Pennsylvania,
and from eastern Kentucky and Tennessee to central Virginia and western North Carolina. In
3-12
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
addition, Shaw et al. (2014) developed steady-state CLs for lakes in the Sierra Nevada Mountain,
where N deposition is the stronger driver of acidification. New CLs have also been determined
through the use of biogeochemical watershed models such as MAGIC and the photosynthesis
evapotranspiration biogeochemical (Pnet-BGC) model (e.g., second draft ISA, Appendix section
8.5.4; Lawrence et al., 2015; Fakhraei et al., 2014; Fakhraei et al., 2016) for the Adirondacks
Mountains, Appalachian Mountains, and for the Great Smoky Mountains National Park. When
these CLs, as well others in the National Critical Loads Database of Sulfur and Nitrogen (NCLD
v3.0)29, are considered together, they approximate broad, national coverage since sites are
represented across the contiguous United States (CONUS).
In addition to new CLs, new information has become available through the National
Aquatic Resource Surveys (NARS) program.30 Through the National Rivers and Streams
Assessment (NRSA), 1.2 million miles of rivers and streams were assessed for their ecological
condition and through the National Lakes Assessment (NLA), more than 1,000 lakes, ponds, and
reservoirs (representing nearly 50,000 water bodies) were sampled for their water quality and
biological and habitat conditions using comparable field and laboratory protocols. Although the
surveys were not specifically focused on acid-base chemistry, ANC and pH were among the
chosen chemical indicators used to assess biological integrity. However, the NLA and NRSA
were designed to be representative of all lakes and streams, so many of the lakes and streams that
were sampled are not sensitive to acidification.
The NLA and NRSA datasets are robust and comprehensive. There is potential to use
these datasets to evaluate risk of acidification to lakes and streams from exposure to deposition
from oxides of nitrogen, oxides of sulfur and particulate matter. However, air quality metrics
were not collected as part of the NARS program which limits their use in NAAQS
applications. Data from the 2007 and 2012 NLA have been incorporated into a database of lake
nutrient chemistry data for the western United States (Williams and Labou, 2017). These data
29 Since the last review, the National Atmospheric Deposition Program (NADP) Executive Committee formed the
CLs of Atmospheric Deposition Science Committee (CLAD) in April 2010. The CLAD has developed the
NCLD, which is a compilation of CLs data and supporting information from many regional- and national-scale
projects within the US. The focus of this database is on CL of N and S deposition and the effects on terrestrial and
aquatic environments. The database is updated through periodic "calls for data" and corrections. The most recent
version of the database, version 3.0, was released in October 2017. We note that studies in the NCLDv.3 are
included in the second draft ISA.
30 This program is designed to assess the status of and changes in quality of the nation's coastal waters, lakes and
reservoirs, rivers and streams, and wetlands. These are statistical surveys first conducted in 2007 and the data
were not available at the time of the last review. The NARS are made up of four individual surveys that are
implemented on a rotating basis; National Lake Assessment (NLA), National River and Stream Assessment
(NRSA), National Coastal Condition Assessment (NCCA), and National Wetland Condition Assessment
(NWCA). The NLA and NRSA are discussed in this section while the NWCA and NCCA are discussed in
Section 3.7.
3-13
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
were used by Williams et al. (2017) along with data from the 1985 EPA Western Lake Survey, a
precursor to the NLA used by Nanus et al. (2017) to develop the CLs discussed below. However,
no studies are currently available linking the NLA and NRSA (nor the NWCA and NCCA) to
atmospheric deposition to develop CLs or exposure-response functions. Further analysis of the
NARS data for purposes of this NAAQS review would require regression-based approaches,
linking ambient air quality data to the biological or chemical metrics in the data to develop new
CLs or exposure-response functions. Such new, deposition-based analyses are an area of interest
for future research, but fall outside the scope of this REA.
Given the newly available CLs related to changes in pH and ANC, as well as the
important changes in N and S deposition since the last review, we judge that an updated analysis
of freshwater acidification would inform our understanding of how the associated risks have
potentially changed since the last review. In addition, the availability of new datasets and
improvements in assessment tools will likely reduce some of the uncertainties and limitations
identified in our analysis of freshwater acidification in the last review.
3.5.2 Freshwater Nitrogen Enrichment
Nitrogen enrichment from atmospheric deposition of N to freshwater ecosystems leads to
increased productivity of algae and aquatic plants, altered nutrient ratios, and sometimes
decreased oxygen levels (see second draft ISA, Chapter 1, section 1.6). The previous review
found the relationship to be causal for changes in biogeochemistry and it remains largely
unchanged for this review.31 However, based on new scientific evidence since the last review,
the causality determination for changes in biota, which included species richness, community
composition, and biodiversity, has been expanded to include altered growth as an endpoint.
As discussed in Chapter 2, the 2009 NOx/SOx REA did not include a quantitative
assessment of freshwater N enrichment effects due to lack of data. For this review, several of the
new studies available include CLs related to biogeochemical changes as well as altered growth,
consistent with the expansion of the causality determination in the second draft ISA. Baron et al.
(2011) assessed lakes in the Sierra Nevada Mountains, Rocky Mountains, and the northeastern
U.S. CL values were estimated by comparing total N deposition (estimated from a combination
of the Parameter-elevation Regression on Independent Slopes Model (PRISM) and NADP data
for wet deposition and CMAQ modelled dry deposition) to N03" concentrations in the 3 regions.
The CL was estimated as the point within the data where N03" concentrations begin to increase,
indicating that deposition is exceeding the N uptake of the watershed and leading to N
enrichment in the lake. In this study, N03" concentrations were not related to a specific biological
31 See Table 1-1 of the second draft ISA for a side-by-side comparison of causality statements from the last review
and this review.
3-14
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
endpoint, but N enrichment is connected to a wide range of ecological effects (see second draft
ISA, Chapter 1, section 1.6). This analysis resulted in non-lake specific CL estimates of 2, 3 and
6 kg N/ha/yr for the Sierra Nevada's, Rocky Mountains and northeast, respectively.
Two additional studies used regression models to link N deposition with changes in water
chemistry that can then provide some insight into potential ecological effects. Williams et al.
(2017) assessed lakes in the Georeferenced Lake Nutrient Chemistry (GLNC) database
(Williams and Labou, 2017). The Lakes included in the study were located at elevations above
1200 meters and on federally protected lands (National Park Service (NPS) and U.S. Forest
Service (USFS). The CL was estimated to be a point at which there was a fundamental change in
the lake biogeochemical cycle and used a logistic regression approach to relate total N deposition
to the N03" concentration above which there is > 50% chance of the phytoplankton growth in the
lake having shifted from N to P nutrient limitation. This resulted in a CL value of 4.1 kg N/ha/yr
(with a range of 2.8 - 5.2 kg N/ha/yr), and a lower estimate of 2 kg N/ha/yr to protect against
false positive (lakes where deposition does not exceed the CL, but lake chemistry data indicate
there has been a shift to P limitation).
Nanus et al. (2017) determined CLs for lakes in the greater Yellowstone area (including 2
National Parks - Yellowstone and Grand Teton) which refined CLs estimates from earlier work
by Nanus et al. (2012) for the broader Rocky Mountains area. CLs were estimated using a multi-
linear regression approach to predict surface water N03- concentrations, using basin
characteristics and TN deposition. CLs ranged from <1.5 + 1.0 kg N/ha/yr to >4 + 1.0 kg
N/ha/yr. These values were based on a N03- threshold of 1.0 (amol/L based on growth
characteristics of indicator diatom species, which are indicative of shifts from oligotrophic to
mesotrophic or eutrophic conditions (Saros et al., 2005). CLs for N03- thresholds of 0.5 |imol/L,
1.6 |imol/L and 2.0 |imol/L were also reported.
The availability of new CLs in this review for freshwater N enrichment potentially fills a
data gap from the last review and informs an assessment of how changes in deposition could
impact changes in water chemistry. We also note that the availability of the new data through the
NARS program could possibly be used in this assessment to identify, and prioritize for further
assessment, impacted lakes and streams. At this time, however, we are unsure whether a new
quantitative assessment of freshwater N enrichment will be included in the REA since there is
uncertainty as to whether it would provide additional insight into potential ecological effects, be
broadly applicable in their application, and/or improve our understanding of pollutant-
attributable risks since the last review, and as such, substantively inform our ability to assess
national standards for freshwater N enrichment.
3-15
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
3.6 TERRESTRIAL ACIDIFICATION AND NITROGEN ENRICHMENT
In terrestrial ecosystems, the acidifying effects from N and S, and the enrichment effects
from N can occur simultaneously. However, the strength of one process over the other (i.e.,
acidification versus N enrichment) can vary depending upon several factors, including soil pH
and buffering capacity, the presence and abundance of sensitive species, and the degree of
nitrogen limitation on primary production.
Since the 2008 NOx/SOx ISA, more recent research has confirmed and strengthened our
understanding of terrestrial acidification (see second draft ISA, Appendix section 5.7). While the
causal determinations for changes in biogeochemistry have generally remained unchanged for
the second draft ISA, the causal determinations for changes in biota has been expanded for this
review to include alteration of species richness, community composition, and biodiversity as
endpoints. Additionally, the second draft ISA includes a new causality determination, finding
that there is a causal relationship between atmospheric N and S deposition and the alteration of
the physiology and growth of terrestrial organisms, as well as the productivity of terrestrial
ecosystems.
Similar to acidification, the causality determination for biogeochemical changes from N
enrichment has remained largely unchanged since the last review.32 However, the largest
increase in scientific evidence, over any other effect category, is for terrestrial N enrichment
ecological effects (see second draft ISA, Chapter 1, section 1.2.3). This new evidence has
confirmed and strengthened our understanding of the mechanistic links that inform causal
determinations between atmospheric N deposition, biogeochemistry, and biota in terrestrial
ecosystems. The second draft ISA includes a new causal determination for N enrichment, having
determined that there is a causal relationship between atmospheric N deposition and the
alteration of the physiology and growth of terrestrial organisms and the productivity of
terrestrial ecosystems. The new data also improves our ability to quantify dose response
relationships between N deposition and ecological response. New studies include investigations
of plant and microbial physiology, long-term ecosystem-scale N addition experiments, regional
and continental-scale monitoring studies, and syntheses. Additionally, since the last review, new
research techniques have been developed to understand community composition, a larger number
of communities have been surveyed, and new regional and continental-scale studies been
conducted (see second draft ISA, Chapter 1, section 1.2.3).
The sections that follow provide more details about the new scientific and technical
information available for consideration in the REA.
32 See Table 1-1 of the second draft ISA for a side-by-side comparison of causality statements from the last review
and this review.
3-16
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
3.6.1 Forests
New studies since the 2008 NOx/SOx ISA indicate that terrestrial acidification is an
ongoing and widespread issue in areas of elevated acidic deposition. For example, new studies
have observed the ongoing depletion of exchangeable base cations in forest soils of the
northeastern U.S. (see second draft ISA, Appendix section 4.3.4). There is considerable evidence
that N and/or S-driven acidification in forests likely contributes to the mobilization of toxic
forms of A1 and decreases the availability of nutrients to plants due to leaching of base cations
from soil and interference with uptake (see second draft ISA, section 5.2.1). These soil changes
can affect tree regeneration (e.g., Sullivan et al., 2013 sugar maple study; Lieb et al., 2011).
Since the 2008 NOx/SOx ISA, there is new evidence of species-specific effects of
atmospheric N deposition on tree growth and mortality in the U.S. (e.g., Thomas et al., 2010;
Dietze and Moorcroft et al., 2011); but it is uncertain whether changes in growth or mortality are
driven by N-enrichment or terrestrial acidification effects. Additionally, Simkin et. al (2016),
suggests that species richness of forest understory plant communities generally has unimodal
relationships with atmospheric N deposition. This unimodal relationship was found to be pH-
dependent, with Simkin et al. (2016) reporting that plant species richness was more likely to
decline with increasing atmospheric N deposition if the herb community occurred on more acidic
soils.
The 2009 NOx/SOx REA calculated CLs for soil acidification in two forest ecosystems
using the SMB model and estimates of base cation weathering from the clay substrate method.
Since that document was completed, new CLs have become available from studies that used the
SMB model with the clay substrate method for estimating base cation weathering (BCw) rates,
and relied on a Bc/Al ratio of either 1 or 10, based on the precedent for use of these numbers by
the scientific community.33 (For a summary of these studies, see second draft ISA, Appendix
section 5.5.3).
Most new studies explicitly examining acidification from N and S since the last review
have built upon earlier approaches that modeled soil acidification CLs using one of several
different models (i.e., SMB, STA, MAGIC, and ForSAFE-VEG)34,which are summarized in the
second draft of the ISA (see second draft ISA, Appendix sections 4.5 and 4.6). In addition, there
are several studies (whose CLs are included in the NCLD v3.0) that have evaluated the use of
other models besides the clay substrate method for purposes of estimating BCw rates. For
33 Bc/Al ratios of 1 and 10 have been used as average numbers to represent different forest types in different parts of
the U.S. (e.g., Eastern versus Western, or coniferous versus deciduous forests).
34 For a summary of these models and related considerations, see second draft ISA, Appendix section 4.5.
3-17
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
example, because of the known limitations of the clay substrate method,35 Phelan et al. (2014)
evaluated the use of the dynamic model, PROFILE, for estimating BCw. Phelan et al. (2014)
paired PROFILE with national datasets36 as a method to estimate BCw rates for forests in the
U.S., applying it to 51 forested sites across Pennsylvania.
In the 2008 NOx/SOx ISA, there was consistent evidence thatN additions stimulated
forest productivity. However, responses included both neutral and negative effects of N additions
on tree growth (second draft ISA, Appendix section 6.3.2.1). Since the 2008 NOx/SOx ISA,
there is considerable new evidence from deposition gradient studies, forest modeling, and long-
term N addition experiments that atmospheric N deposition broadly stimulates tree growth and
the productivity of forested ecosystems (see second draft ISA, Appendix section 6.3.2.1);
however, the effects can vary by species. For example, studies have shown that conifer species,
particularly at high elevations, are more likely to exhibit negative growth responses or mortality
in response to added N (second draft ISA, Appendix section 6.2.3.1). One particular study,
Thomas et al (2010) used exposure-response functions to relate N deposition to growth and
mortality impacts on 23 individual tree species in the Northeast and mid-Atlantic.37 Leveraging
forest inventory data from the early 1980s through the mid-1990s, Thomas et al. (2010) reported
declines in growth for three conifer species, consistent with other observational studies (see
second draft ISA, section 6.2.3.1). The study also reported higher mortality rates for 8 tree
species with increasing atmospheric N deposition while only three species showed a positive
relationship. Overall, this analysis was consistent with other multi-variate analyses that explored
the key drivers of tree mortality in the region (e.g., Dietze and Moorcroft, 2011).
3.6.2 Lichens and Mycorrhizal Fungi
Since the 2008 NOx/SOx ISA, there is new evidence of decreases in lichen species
richness as the result of atmospheric N deposition in the U.S., and there are now ambient
observations that atmospheric N deposition in the U.S. is changing mycorrhizal community
composition, (see second draft ISA, Appendix section 6.6.2).
As mentioned in Chapter 2 of this document, the 2009 NOx/SOx REA included a
qualitative assessment of N enrichment effects through a case study for coastal sage scrub and
mixed conifer forests. No quantitative assessment was conducted due to lack of data. Since the
35 See section 2.5 for a summary of these limitations.
36 See Table 2 in Phelan et. al., 2014, for the list of national datasets.
37 We anticipate that this study will be expanded to include 94 species nationally, relating growth and mortality
impacts to N and S deposition (Horn et al., 2018 in review).
3-18
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
last review, an extensive amount of new data with national coverage is available for terrestrial N
enrichment CLs for lichens and mycorrhizal fungi.38
New lichen CLs and response curves are available through two studies conducted in the
Pacific Northwest, for the portions of the North American West Coast Marine Ecoregion in
Oregon and Washington (Geiser et al., 2010) and for the portions of the Northwestern Forested
Mountains Ecoregion in Northern California, Oregon, Washington, and parts of northern Idaho
and Montana (Root et al., 2015). For the lichens dataset, researchers collected information on
two responses - lichen community composition and lichen N concentrations - and related those
responses to atmospheric concentrations of N (Root et al., 2015) and/or measures of atmospheric
deposition (Geiser et al., 2010). The resulting equations constitute exposure-response curves and
were used to derive CLs for lichen community composition and lichen N concentrations. These
data have been extrapolated for application to the CONUS for the NCLDv3.0.
The mycorrhizal fungi dataset is a collection of empirical CLs for total atmospheric N
deposition from various studies included in Pardo et al. (2011). This dataset includes minimum
and maximum CLs for community composition.
3.6.3 Herbs and Shrubs
A national CLs dataset is available for the species richness of herbs and shrubs based on
the Simkin et al. (2016) study. As mentioned earlier, the Simkin et al. (2016) study analyzed
relationships between plant species richness and atmospheric N deposition involving interactions
with soil pH (along with precipitation and temperature). Therefore, the dataset relates to both N
enrichment and acidification effects. The dataset covers nearly 4,000 herb/shrub species over
more than 15,000 plots. This dataset includes a collection of total atmospheric N deposition CLs
from Simkin et al. (2016), with CLs calculated separately for "open canopy" (grasslands, shrub
lands, and woodlands) and "closed canopy" (forested understory plants) ecosystems.
The CLs in Simkin et al. (2016) were derived using regression analyses relating
atmospheric N deposition to species richness. These underlying regression models are equations
that constitute an exposure-response relationship (i.e., for both open and closed canopy systems,
species richness is related to atmospheric N deposition, soil pH, temperature, and precipitation;
see Table 1 in Simkin et al. (2016)). These equations can be used to evaluate potential changes in
effects based on varying levels of atmospheric deposition.
38 For descriptions of these data, see NADP CLAD report "2017 Summary of Critical Load Maps,:
http://nadp.slh.wisc.edu/committees/clad/db/NCLDMapSummarv 2016.pdf
3-19
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
3.6.4 Summary
For this review, new or refined national CLs as well as new national exposure response
curves are available for various terrestrial ecosystems and receptors, which will expand and
enhance our ability to evaluate sensitivity and risk for acidification and N enrichment nationally
for multiple species. This new information also represents many of the new endpoints that have
contributed to the stronger weight of evidence and expanded causality determinations in the
second draft ISA, enabling an evaluation of growth and mortality effects as well as species or
community composition changes that was not available at the time of the last review.
Additionally, several studies have produced refined CLs and/or refined methods for deriving
CLs, which may reduce identified uncertainties from the last review. Given the extent to which
new scientific and technical information are expected to fill the data gaps/limitations from the
last review and add to our understanding of pollutant-driven risk and exposure, we judge that
quantitative assessments evaluating terrestrial acidification and N enrichment effects will provide
insight into potential ecological effects and improve our understanding of pollutant-attributable
risks since the last review.
3.7 OTHER NITROGEN AND SULFUR EFFECTS
3.7.1 Estuarine Nitrogen Enrichment
The causality determination for atmospheric N deposition and biogeochemical changes in
estuarine and near coastal marine ecosystems remains unchanged, with new evidence continuing
to support the findings of a causal relationship in the last review. However, since the last review,
new paleontological studies, observational studies, and experiments have further characterized
the effects of N on phytoplankton growth and community dynamics, macroinvertebrate response,
and other indices of biodiversity. Hence, the causality determination pertaining to atmospheric N
deposition and alteration of species richness, species composition and biodiversity has been
expanded to include altered growth, total primary production, and total algal community
biomass as endpoints.
The 2009 NOx/SOx REA included a case study for the Potomac River and Potomac
Estuary in the Chesapeake Bay and the Neuse River and Neuse River Estuary in Pamlico Sound,
using ASSETS EI and the SPARROW model. The 2009 NOx/SOx REA noted uncertainties in
the inputs and outputs of the SPARROW model, and noted the importance in choosing case
study areas where atmospheric deposition plays a large role in N loading to an estuary. Since the
last review, several modeling studies have estimated the amount and proportion of current and
future N loading expected to result from atmospheric deposition (see second draft ISA, Table 7-
9).
3-20
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
At the time of the last review, CLs were not available. Several new CLs pertaining to salt
marsh community structure, microbial activity, biogeochemistry, and loss of eel grass are now
available from Greaver et al. (2011) based on a review of three studies (Caffrey et al., 2007;
Wigand et al., 2003; and Latimer and Rego, 2010). There are three new studies with threshold
values for eel grasses in estuaries now available (see second draft ISA, Appendix section 10.2.5,
Table 10.4). These new studies are based on total N loading to the estuary. To assess
exceedances of CLs (or threshold levels) related to atmospheric N deposition, watershed-level
estimates of total N loading would need to be calculated from modelling or other watershed-level
estimates.
While new studies are now available to determine which estuaries are predominantly
affected by atmospheric deposition, we note that any modelling to evaluate impacts of N
deposition on ecological effects in an estuary would need to be conducted on an estuary-by-
estuary basis. Additionally, the CLs are limited in geographic scope, with studies conducted
primarily in New England, so would be of limited use in assessing national standards.
New data are also available through the NCCA which provide information on overall
estuarine conditions, but as described in section 3.5.1, these conditions are not specifically
related to atmospheric deposition, but rather to total nutrient loading. This, and the lack of
representativeness of any individual estuarine modelling approach, limits the usefulness of
conducting a quantitative assessment. Given these limitations, such information would be of
limited usefulness in informing decisions on national-scale standards, therefore we do not intend
to conduct quantitative assessments for this effect category.
3.7.2 Wetlands Nitrogen Enrichment
The causality determination for atmospheric N deposition and biogeochemical changes in
wetland ecosystems remains unchanged, with new evidence continuing to support the findings in
the last review. New research on wetland biogeochemistry since 2008, includes a synthesis of
wetland improvements to water quality through denitrification and biological uptake, a meta-
analysis of N addition effects on methane (CH4) and nitrous oxide fluxes, and multiple
observations of changes in belowground C cycling in response to added N. These
biogeochemical shifts may diminish the wetland ecosystem services of long-term carbon storage
and flood protection, as well as reduce the stability and persistence of wetlands on the landscape.
Nitrogen loading effects upon productivity are uneven across species, which may affect wetland
biodiversity and the wetland ecosystem service of provisioning.
New evidence published from observational studies, experimental N addition studies in
the field and in mesocosms, and re-analysis of large data sets supports and extends the
conclusions of the 2008 NOx/SOx ISA. Additionally, since the last review, there is newly
3-21
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
published evidence of atmospheric N deposition effects on endpoints not covered in the 2008
ISA, including alterations to plant physiology and plant architecture (see second draft ISA,
Appendix section 11.5). Given this newly available information, the causality determination for
atmospheric N deposition and alteration of species richness, community composition, and
biodiversity has been expanded to include alteration of growth and productivity as well as
species physiology as endpoints.
The 2009 NOx/SOx REA did not include a quantitative assessment for wetlands due to
lack of available data. Since the last review, new CLs have become available for ombrotrophic
bogs and the purple pitcher plant (Sarraceniapurpurea L.) (Greaver et al., 2011). The CLs in
Greaver et al. (2011) were based on 6 studies (two studies of purple pitcher plant, and four
studies related to net primary production and peat accumulation in ombrotrophic bogs). One
additional study, Crumley et al. (2016), found threshold levels of deposition for purple pitcher
plant consistent with the CLs estimated by Greaver et al. (2011).
To determine the utility of these CLs for purposes of the REA, we considered the
geographic coverage, representativeness and uncertainty tied to these data. We have determined
that application of these CLs has the highest level of certainty at the individual sites included in
the 7 studies. However, only four of the studies (Aldous et al., 2002; Gotelli and Ellison, 2002,
2006; Crumley et al., 2016) included sites within the United States (the remaining studies
included only sites in Canada). This limits the geographic scope of potential analyses to 15 study
sites located in the northeastern U.S. The geographic scope of the CLs estimates could be
expanded by aggregating to Ecoregions, similar to other CLs estimated in Pardo et al. (2011),
which would include some of the study sites located in Canada. However, this would only
expand the coverage to include Ecoregions in the northeast and north central U.S. while
increasing the uncertainty associated with the CLs as values are extrapolated from individual
points to larger areas.
Additionally, data is available through the NWCA. The overall scope of the NARS
datasets is described in section 3.5. Similar to the other NARS datasets, the NWCA did not
include measurements of ambient air quality or atmospheric deposition. While new CLs are
available, given the level of uncertainty and limited geographic representation of these studies
and the reasons discussed in section 3.5.1 regarding the NARS data, we do not intend to conduct
a quantitative assessment for this effect category.
3.7.3 Coastal Acidification
The 2008 NOx/SOx ISA did not address nutrient enhanced coastal acidification. Since
the 2008 ISA, N enrichment has been recognized as a possible contributing factor to increasing
acidification of marine environments. Specifically, N has been recognized as a possible
3-22
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
contributing factor to coastal acidification because the CO2 produced by organic matter
decomposition in eutrophic waters can contribute CChto the water column along with the
dissolution of atmospheric anthropogenic CO2, decreasing the pH (see second draft ISA,
Appendix section 10.5). Given the new scientific information available supporting this effect,
the second draft ISA found that the relationship between atmospheric N deposition and
increased nutrient-enhanced coastal acidification is likely causal. Additionally, the second
draft ISA found that the evidence is suggestive of, but not sufficient to infer, a causal relationship
between atmospheric N deposition and changes in biota including altered physiology, species
richness, community composition, and biodiversity due to nutrient enhanced coastal
acidification. Despite newly available scientific evidence for this effect category, there are
currently no tools or data available to quantitatively assess the risks due to coastal acidification
that are associated with atmospheric N and/or S deposition. Because of this, no quantitative
analyses are supported for this RE A.
3.7.4 S Enrichment in Freshwater and Wetland Ecosystems
3.7.4.1 Mercury Methylation
Recent research since the last review continues to strengthen and inform our scientific
understanding of the relationship between atmospheric S deposition and freshwater MeHg
production. This understanding has expanded since the last review to include the identification
of: 1) additional types of organisms that play a role in the methylation process, 2) additional
macro- and micro-environments in which methylating organisms are found, and 3) additional
areas within the U.S. containing habitats with conditions suitable for methylation. Building on
the body of available science, including that available in 2008, the second draft ISA finds a
causal relationship between S deposition and the alteration of mercury (Hg) methylation in
surface waters, sediment, and soils in wetland and freshwater ecosystems (see second draft ISA,
Appendix section 12.1).
Specifically, this recent research has demonstrated that in addition to sulfur-reducing
bacteria (SRB), certain strains of archaea found in wetland sediments are also active in S
reduction. The current review therefore uses the broader term of sulfur-reducing prokaryotes
(SRP), when appropriate, to reflect the joint role of both groups of organisms in certain wetland
environments (see second draft ISA, Appendix sections 12.1 and 12.3). Further, additional
organisms that possess the ability to methylate mercury have also been identified, due in part to
the discovery of the genes associated with this ability (see second draft ISA, Appendix section
12.3). New evidence has broadened our understanding of where methylation occurs, both in
terms of types of wetland and freshwater ecosystems, as well as the specific areas within these
ecosystems where methylation is most likely to occur. For example, it is now known that
3-23
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
methylation occurs in agricultural wetlands, rivers and streams and that within aquatic
ecosystems, periphyton can play an important role as a host for SRPs. Furthermore, the microbial
communities embedded within the periphyton can be quite complex and diverse, are more
efficient in the methylation of Hg than are SRPs in sediments, and boost Hg methylation rates
within the oxygenated water column of freshwater and wetland ecosystems (see second draft
ISA, Appendix section 12.3).
These scientific advances confirm and highlight the case, as described in the second draft
ISA (second draft ISA, Appendix section 12.3.3) and mentioned in Chapter 2 of this document,
that the relationship between atmospheric S deposition and measured increases in MeHg in
associated wetland and freshwater aquatic systems and biota is complex, in part because it is
mediated by the activity of SRP which are influenced by multiple interacting physical, chemical
and biological variables such as oxygen content, temperature, pH, and labile carbon supply.
There are several controlling factors that can influence the rates of mercury methylation (see
second draft ISA, Appendix 12, Figure 12-9). Given the complexity, temporal and seasonal
variability, and multiple drivers of the MeHg process across the national landscape, no dose-
response functions have been established and there is currently a lack of CLs and assessment
tools available by which to assess the risks of MeHg enhancement from atmospheric deposition
of S in North American ecosystems (see second draft ISA, Appendix 12 sections 12.6 and
12.3.3).
Based on the information presented in the second draft ISA and summarized above, we
note that there are remaining uncertainties associated with the linkages connecting atmospheric S
deposition to aquatic SO42" concentrations to MeHg production in wetland and freshwater aquatic
ecosystems. Additionally, there are no known tools or data available for quantitative
assessments. Therefore, we do not intend to conduct a quantitative assessment for this effect.
3.7.4.2 Sulfide Phytotoxicity
The 2008 NOx/SOx ISA only included information regarding sulfide phytotoxicity in
European systems, and in mesocosm studies that showed sulfide toxicity reduced biomass of
wetland plants and aquatic macrophytes under exposure levels higher than those that occur in
U.S. regions with high atmospheric S deposition. Newer research, however, shows sulfide
toxicity occurring in the field within multiple wetland ecosystems under ambient exposure
conditions in North America (see second draft ISA, Appendix sections 12.2.3 and 12.7.3). This
new information is sufficient to support a new causal determination in the second draft ISA,
finding that the body of evidence is sufficient to infer a causal relationship between S deposition
and changes in biota due to sulfide phytotoxicity, including alteration of growth and
productivity, species physiology, species richness, community composition, and biodiversity in
3-24
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
wetland and freshwater ecosystems. Sulfide interferes with nutrient uptake in plant roots and
studies show elevated sulfide can result in decreased seed mass, seed viability, seedling
emergence rates, decreased seedling height, decreased seedling survival rates, and reductions in
total plant cover, all which can lead to shifts in plant community composition (see second draft
ISA, Appendix section 12.2.3).
The relationship between atmospheric S deposition and sulfide toxicity is complex and
influenced by multiple factors. Since sulfide is the product of microbial SO42" reduction, its
concentration in water (including sediment pore water) is heavily dependent upon environmental
factors that influence microbial activity, particularly the availability of dissolved organic carbon
(DOC). In addition, the degree to which aquatic vegetation is exposed to phytotoxic
concentrations of sulfide depends on its residence time as a free ion in water. When sulfide
binds with iron it precipitates out of solution and is no longer available to plants. Thus, the
phytotoxicity of sulfide is regulated in part by the availability of iron in the wetland or freshwater
system.
The largest uncertainty that remains regarding sulfide effects, as with methymercury
discussed in 3.7.4.1. above, is that associated with the linkages connecting atmospheric S
deposition to aquatic sulfate concentrations in wetland and freshwater aquatic ecosystems. We
are currently unable to characterize how deposition amounts or changes in deposition affect
aquatic concentrations of SO42". Because of this fundamental uncertainty, we do not intend to
conduct a quantitative assessment for this effect.
3.8 ECOSYSTEM SERVICES
There are several ways in which ecological effects can be related to public welfare. The
most comprehensive model is that of ecosystem services, which can provide information on the
linkages between changes in ecological effects and known or anticipated effects to public
welfare. Ecosystem services can be generally defined as the benefits that individuals and
organizations obtain from ecosystems. The EPA has defined ecological goods and services as the
outputs of ecological functions or processes that directly or indirectly contribute to social welfare
or have the potential to do so in the future. Conceptually, changes in ecosystem services may be
used to aid in characterizing a known or anticipated adverse effect on public welfare. Ecosystem
services, those related to Class I areas and endangered species, and analyses of both non-use and
use values across larger areas and in the human economy, including commercial uses, can be
used separately or together to help inform public welfare decisions.
Appendix 14 of the second draft ISA includes summaries of several new papers
published since the 2008 NOx/SOx ISA that connect the effects of N and S deposition to
ecosystem services. Some of these studies present valuation of costs and benefits of nitrogen
3-25
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
loading across the nation for several ecosystem services such as coastal fish harvests,
recreational uses of waterbodies, and lakefront hedonic values. The second draft ISA also
includes descriptions of five papers that resulted from a joint workshop between the NPS, EPA,
the USFS, and members of academia. The papers present a concise conceptual model linking
CLs to final ecosystem services and a framework for assessing the scientific strength of those
linkages (Bell et al., 2017; Clark et al., 2017; Irvine et al.; 2017, O'Dea et al., 2017; Rhodes et
al., 2017). The second draft ISA also presents an evaluation of studies conducted in Europe and a
global-scale analysis of N cycling and impacts on ecosystem services. Additionally, the second
draft ISA includes profiles of several threatened or endangered species and their related
ecosystem services.
Since the prior review, the NPS has published a list of threatened or endangered species
in each of the National Parks.39 A national study has also been conducted, which identifies 78
threatened and endangered species for which N deposition is a contributing stressor (Hernandez
et al., 2016). Many of these species are located in National Parks. The second draft ISA includes
case studies for several National Parks (including Rocky Mountain, Acadia, Great Smoky
Mountains, Sequoia and Kings Canyon National Parks), where new scientific evidence for
deposition-related effects is highlighted.
A separate literature search, from 2009 to 2017, and review for ecosystem services was
conducted for this REA plan to identify papers related to ecosystem services analysis that were
not included in the second draft ISA because they do not always directly relate nitrogen and or
sulfur deposition to changes in ecosystem services. The papers can be loosely binned into
categories including: freshwater acidification, freshwater nitrogen enrichment, coastal/estuary,
coral reefs, wetlands, forests and wilderness, grasslands and deserts, fires, lichens, and
biodiversity. A list of these papers is included in the Appendix. These papers provide linkages
from deposition to ecological effects and finally to ecosystem services, and provide methods that
either describe the potential qualitative risk to the services covered or possible avenues for
quantitative evaluation of the changes in services related to N and/or S deposition. Since the last
review, information in the recent literature identified by the search increases the number of
ecosystem services and ecosystems potentially available for analysis and/or expands and
improves previously available methods.
In addition to the results of the literature search new information is available from
updated databases such as the Recreational Values Database,40 government reports such as the
39 See https://irma.nps.gov/NPSpecies/
40 See http://recvaluation.forestrv.oregonstate.edu/
3-26
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Internet Research Information Series 41(IRIS) reports from the USFS, and new analyses done by
EPA's National Center for Environmental Economics for the Chesapeake Bay Program.42 These
databases and reports expand and improve upon analyses included in the previous review which
could be updated for the current review.
There are a limited number of papers and methods available that provide national-scale
estimates of ecosystem services affected by atmospheric deposition. The majority of relevant
work has been place-based studies that are useful for case studies although there are some few
that could be amenable to a benefit transfer approach to applying the study results to larger
geographic areas. Given the limited geographical scope of the available information, as well as
the inherent uncertainties associated with quantifying specific ecosystem service effects using
the available methods, we do not intend to conduct any specific quantitative analyses of
ecosystem services in the REA. Instead, we plan to use the information in the available literature
and databases to describe how the public values specific ecosystem services and to link that
information qualitatively to the policy relevance of the assessed changes in ecological risks.
Availability of this type of information can also help prioritize the selection of case study areas
for quantitative assessment (see section 4.4 for more information).
3.9 CONCLUSIONS
The discussion above reflects the EPA staff assessment of the degree to which currently
available information, including newly available information since the last review (e.g., as
summarized in the second draft ISA), might be expected to appreciably change our
understanding of risk and exposures beyond the insights gained from the assessments from the
last review. A critical consideration is the extent to which use of newly available information or
approaches in new or updated quantitative assessments would provide risk and exposure
estimates that are appreciably different or have the potential to reduce uncertainty or limitations
in the previous review, and that indicates that a new assessment of risk and exposure is
warranted to provide an adequate characterization of ecological risk and exposure, particularly
with regard to the current standards.43
Based on these considerations, we have concluded that an REA is warranted to
quantitatively evaluate acidification and nitrogen enrichment effects within freshwater and
41 See https://www.srs.fs.usda.gov/trends/iris/recstats.html
42 See https://www.epa.gov/environmental-economics/research-environmental-economics-ncee-working-paper-
series
43 In considering this point, the EPA staff additionally recognize that such a characterization of risk and exposure, in
addition to the currently available evidence, will be considered in the PA in terms of both evidence-based
considerations and risk/exposure-based considerations.
3-27
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
terrestrial ecosystems. Given the lack of information and/or concerns about limitations or
uncertainties as described in the sections above, we do not intend to quantitatively evaluate the
other effect categories discussed in this chapter. More specifically, we conclude that it is
appropriate to focus the REA on the following analyses, for which important new information is
available in this review:
• Air Quality Information. Since the previous review, advances in scientific methods and
the changing levels and spatial distribution of N and S deposition have appreciably
changed our understanding of the linkages between ambient air concentrations,
atmospheric deposition, and ecosystem exposures. New techniques are now available for
combining these measurements and modeling outputs to estimate total deposition with
lower uncertainty. The spatial variability and distribution of deposition has changed in
recent years, as reduced forms of nitrogen deposition are now the largest source of N
deposition in many places. Finally, because concentrations and deposition of oxidized N
and sulfur have declined in recent years, there are new opportunities to better assess the
linkage between a change in concentration and a change in deposition, while also better
quantifying the uncertainties.
• Freshwater Acidification and N Enrichment. Since the 2008 NOx/SOx ISA, new
studies have been published on the effects of N and S deposition on freshwater
ecosystems. This evidence includes new CLs for freshwater ecosystems and has created a
greater weight of evidence and led to the development of expanded causality
determinations in the second draft ISA from the previous review. These CLs provide
national coverage and are expected to significantly expand the scope for analyses since
the last review, and for N enrichment, to fill a major gap from the last review. Therefore,
we intend to conduct quantitative assessments evaluating freshwater acidification and N
enrichment effects using CLs.
• Terrestrial Acidification and Nitrogen Enrichment. A substantial body of new
information is available for terrestrial acidification and N enrichment for this review.
New information pertaining to acidification and N enrichment of terrestrial ecosystems
includes studies that evaluate endpoints (herb/shrub and mycorrhizal community
composition, lichen species richness, and tree growth and mortality) through CLs and/or
response curves that were not available at the time of the last review. Many of these CLs
and response curves are national in scope and therefore expected to significantly expand
and enhance our ability to evaluate sensitivity and risk for acidification and N
enrichment. The availability of this information fills a gap in the last review.
Additionally, several new studies are now available that evaluate potential methods for
deriving CLs and in particular, for estimating BCw rates in soils. These studies are
expected to inform updated CLs assessment in the REA, potentially reducing
uncertainties from the prior review. Therefore, we intend to conduct quantitative
assessments evaluating terrestrial acidification and N enrichment effects in this review.
In summary, a new REA that utilizes the new information and approaches summarized
above to provide a more precise characterization of exposure and risks, and one with reduced
uncertainty, would inform the current review. Therefore, we conclude that it is appropriate to
3-28
-------
1 develop a risk and exposure assessment, based on the newly available air quality and ecological
2 information, to inform the current review. More information regarding our plans for analyses is
3 included in Chapter 4.
3-29
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
4 PLAN FOR QUANTITATIVE ASSESSMENT
This Chapter describes the proposed quantitative analysis and approach for the REA to
characterize ecological risk and exposure associated with NOx, SOx, and PM for current air
quality conditions, as well as for when air quality is just meeting the current standards, and if
appropriate, when meeting potential alternative air quality standards. The REA will focus
particularly on the contribution of these pollutants to deposition-driven ecological effects on
freshwater and terrestrial ecosystems based on the conclusions in Chapter 3. These ecosystems
warrant further assessment to evaluate information newly available since the last review that
might be expected to appreciably change our understanding of risk and exposures or have the
potential to reduce uncertainty or limitations in the previous review. Given the lack of
information and/or concerns about limitations or uncertainties, we do not intend to quantitatively
evaluate the other effect categories described in Chapter 3.
4.1 ANALYTICAL FRAMEWORK
The proposed analysis approach for the REA is shown in Figure 4-1 and discussed in
more detail below. In general, we intend to use air quality data from monitors and chemical
transport models to estimate atmospheric deposition of N and S associated with ambient
concentrations of NOx, SOx, and PM. This information will also be used to estimate deposition
response factors that predict the relationship between ambient concentrations and atmospheric
deposition in certain areas of the country and can be used to adjust deposition to represent just
meeting the current standards, as well as potential alternative standards, as appropriate. This air
quality information can then be used to assess ecological effects at a national scale for current
conditions, as well as within selected case study areas for air quality conditions that just meet
current and potential alternative standards. Assessment of these effects for freshwater and
terrestrial ecosystems can generally be grouped into two main categories: (1) assessment of
exceedances using CLs; and (2) assessment of changes in biological and chemical responses
using exposure-response curves. In these analyses, CLs and exposure-response curves would be
used to relate N and S deposition to changes in biogeochemistry and changes in species-level or
community-level ecological or biological responses. The resulting information would then be
used in the consideration of the overall ecological risk for the review. Included below is a brief
discussion of some of the main components of the analytical framework, including air quality
and exposure, CLs, and exposure-response curves, as well as the general application of these
components in the REA. More specific information about the assessments for the REA is
included in the sections that follow.
4-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
Overview of REA Planned for this Review
Critical Loads
(sections 4.1.2, 4.1.3)
Exposure-Response
Functions
(sections 4.2.4, 4.2.5)
Ecological and
Chemical Survey
Data
Deposition Response Factors
(section 4.2.1)
Exposure and Risk-Related Considerations in Review of Standard
Output: CLs exceedances for
different levels of deposition
Critical Load
Exceedances
Ambient Air Quality
(Monitors + CMAQ)
(sections 4.1.1. 4.2.2)
Output: Estimated biological and
chemical changes at different levels
of deposition
Changes in Ecological
Responses
recent conditions + adjusted AQ
scenarios
(section 4.2.1)
National atmospheric
deposition for recent conditions
(2014-2016) (TDEP)
(section 4.1.1)
Figure 4-1. Analytical Framework for the REA
4.1.1 Ambient Air Quality and Atmospheric Deposition
The objective for the REA is to characterize risk and exposure associated with just
meeting the current standards, as well as any potential alternati ve standards under consideration.
For this review, we plan to consider both the secondary and primary standards as part of the suite
of current standards. In the REA, we plan to include a national-scale assessment of ecological
risk that will be based on estimates of current air quality concentrations and deposition in the
U.S. Applying the same methodology as used to create the TDEP datasets (Schwede and Lear,
2014), recent air quality information would be used to create a national-scale gridded surface of
N and S deposition, with separate estimates of the reduced and oxidized nitrogen contributions to
4-2
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
total N deposition.44 This surface will represent average deposition over three years to
incorporate emissions and meteorological variations across recent years, with the individual
years also available for use. Because much of the U.S. is already meeting the secondary and
primary standards for NO2, SO2 and PM, this surface can be used to assess ecological risk in
most areas for air quality conditions that are at or below the current standards. To further assess
ecological effects for air quality conditions under different air quality scenarios, we propose to
adjust air quality and create new gridded surfaces of N and S deposition for each of the study
areas that reflect deposition levels when just meeting the current standards, and any potential
alternative standards, as appropriate. The focus on air quality adjustments in smaller areas of the
country will help focus the analyses on specific air quality conditions and reduce uncertainty in
the adjustment methodology being applied.
4.1.2 Critical Loads and Exceedances
The second draft ISA uses the CL concept as an organizing principle to relate
atmospheric deposition to ecological endpoints that indicate impairment (see second draft ISA,
Section 1.2.2.3). "A critical load is formally defined as a quantitative estimate of exposure to one
or more pollutants below which significant harmful effects on specified sensitive elements of the
environment do not occur according to present knowledge" (Nilsson and Grennfelt, 1988,
UNECE, 2004). CL estimates reflect the current state of knowledge and the selected indicators
and responses (see second draft ISA, section 1.2.2.3). It is important to recognize that there is no
single "definitive" CL for an ecological effect, nor is there a single definition of "harm" across
CLs datasets. Given the heterogeneity of ecosystems affected by N and S deposition, published
CL values for locations in the U.S. vary depending on both biological and physical factors. In
fact, it is not uncommon for there to be multiple CLs available for a given pollutant at a single
location due to the nested sequence of disturbances, receptors, and biological indicators
considered for a given pollutant, (see second draft ISA, Chapter 1, section 1.2.2.3).
CLs are point estimates of when harmful effects begin to occur. Given the breadth of CL
data available through new studies, we intend to use the CLs to provide a national-scale picture
of sensitivity45 and risk46 (when used in conjunction with atmospheric deposition estimates to
calculate exceedances). The CLs can also be useful in identifying risks of concern under
different levels of atmospheric deposition and at different spatial scales. Therefore, we intend to
use CLs and exceedances as tools for identifying potential case study areas for quantitative
44 This information allows for better understanding of the portion of deposition that is controllable under the CAA.
44 Sensitivity is as the degree to which an ecosystem is affected by atmospheric N and/or S deposition.
45 Risk is defined as the potential that adverse ecological effects may occur, or are occurring, as a result of exposure
to one or more stressors
4-3
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
assessments, including designated areas that are considered important to the public (e.g., Class I
areas).
To calculate exceedances of CLs in the REA, we plan to compare CLs to the estimated
levels of deposition in the national-scale gridded surface for current air quality. Put simply, this
means: Dep - CL = X, if X >0, then it is exceeding. This concept will be applied to the national-
scale surface of gridded deposition estimates. While this is a straight-forward comparison,
further work is required to understand the potential effect of an estimated exceedance. Additional
considerations may also include estimates of the contribution of N and S in the system from air
deposition, estimates of the time scale of the effect and potential changes in the effect, associated
uncertainties in both the computation of the CL and the analytical application of the CL in the
REA, and judgements of the potential adversity of the impact on public welfare.
4.1.3 Changes in Ecological Responses
Exposure-response curves will be a key component of the assessments in the REA. These
curves can show predicted ecological responses to specific levels of N or S deposition, as well as
demonstrate the responsiveness of individual species and/or communities to changes in
atmospheric deposition. These curves are primarily generated by observations of ecological
response to experimental additions (e.g., N and S addition studies) or by observation of
ecological response along a deposition gradient. We intend to use exposure-response curves to
understand impacts under current conditions at a national-scale. In addition, we also plan to use
the exposure-response curves to relate different air quality conditions, and associated deposition
levels, in case study locations to potential changes in ecological effects.
4.2 AIR QUALITY ANALYSES
The sections that follow include information regarding the air quality and deposition
datasets and analytical approaches that we intend to use to evaluate ecological effects in this
REA.
4.2.1 Consideration of Current Conditions for National-scale Air Quality Concentrations
and Deposition
Since the previous review, N and S deposition has changed and it is important to develop
the most up-to-date datasets for the assessment of atmospheric deposition to capture these
changes. We propose to rely on measurements of atmospheric concentration and deposition
where available, and chemical transport model simulations to provide data for chemical species
and locations where measurements are not available.
Accordingly, this review proposes to calculate air concentrations and deposition using the
most recent CMAQ chemical transport model (version 5.2.1), with the most up to date
4-4
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
meteorological and emission datasets, which will likely be for the calendar years 2014-2016.
While there are other photochemical models available, CMAQ has been shown to have relatively
low bias in estimating annual average wet deposition, with normalized mean bias reported as
7.9% for SO42", -12.8% for NH4+, and -15% for NO3" when compared with NADP measurements
(Appel et al., 2011). In addition, while historically the CMAQ chemical transport model has
reported dry deposition of NH3 and particulate NH4+ separately, as these happen via different
physical processes, the model has been recently augmented to be able to track the relative
contribution of particulate NH4+ and gas-phase NH3 to wet deposition of reduced nitrogen. This
recent augmentation can provide data that will be useful in informing the contribution of
particulate NH4+ to deposition for this review. For further analyses, we propose to evaluate the
CMAQ simulations using the NTN observations of wet deposition as well as the CASTNet,
CSN, IMPROVE, and AMoN observations of oxidized nitrogen, particulate chemical
composition, and NH3, respectively. The comparison of CMAQ results with these measurements
can then be used to inform the uncertainty analysis.
The best available assessment of atmospheric deposition generally requires combining
data from ambient measurements and computer model outputs. To do this, we propose to use the
process described in Schwede and Lear (2014) to develop TDEP datasets for 2014, 2015 and
2016 that cover the continental U.S. at 4-km horizontal resolution. These gridded surfaces will
also be combined to provide an estimate of average deposition for 2014-2016. The TDEP
method estimates wet deposition by spatially interpolating NTN wet deposition measurements
and PRISM precipitation observations. These estimates would then be combined with the dry
deposition estimates from the 2014-2016 CMAQ simulations, projected from the 12-km CMAQ
model resolution onto the 4-km TDEP grid. For further analyses of the potential error and
uncertainty in the wet deposition dataset, we intend to use cross validation methods, such as
reserving some of the wet deposition observations for evaluation.
The result of this analysis will be a spatially complete data set of 3-yr average deposition
across the continental U.S. for 2014-2016, including nitrogen and sulfur deposition, wet and dry
deposition, as well as the relative contribution from gas-phase NH3 and particulate-phase NH4+.
These data will be used to assess ecosystem effects at a national-scale and under current
conditions, where much of the U.S. is meeting the current NO2, SO2 and PM standards.
4.2.2 Consideration of Air Quality Scenarios
The goal of the REA is to consider the level of risk under air quality conditions that just
meet the current standards, as well as any potential alternative standards under consideration. To
do so, we plan to conduct quantitative analyses in study area locations with air quality adjusted
to reflect just meeting the current NAAQS, and just meeting other air quality scenarios, as
4-5
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
needed. Accordingly, the sections below describe the proposed methodology for adjusting
ambient concentrations to reflect specific air quality scenarios and statistically relating those
levels of ambient concentrations to deposition levels of N and S. Table 4-1 lists the current
NAAQS for PM, N02, and S02.
Table 4-1. Current National Ambient Air Quality Standards for PM, NO2, and SO2
Pollutant
Primary/
secondary
Averaging
time
Level
Form
N02
Primary
1 hour
100 ppb
98th percentile of 1-hour daily maximum
concentration, averaged over 3 years
Primary &
secondary
1 year
53 ppb
Annual mean
PM2.5
Primary
1 year
12 |jg/m3
Annual mean, averaged over 3 years
Secondary
1 year
15 |jg/m3
Annual mean, averaged over 3 years
Primary &
secondary
24 hours
35 |jg/m3
98th percentile, averaged over 3 years
PM10
Primary &
secondary
24 hours
150 |jg/m3
Not to be exceeded more than once per year
on average over 3 years
S02
Primary
1 hour
75 ppb
98th percentile of 1-hour daily maximum
concentration, averaged over 3 years
Secondary
3 hours
0.5 ppm
Not to be exceeded more than once per year
4.2.2.1 Estimating Ambient Concentrations
The goal of this analysis is to create spatially consistent datasets that represent scenarios
where, in well-defined case study areas, concentrations are adjusted to just meet the current
standards. A similar methodology would likely be followed if adjustments are needed to reflect
just meeting potential alternative standards. These air quality scenarios will be used to estimate
N and S deposition in the study area and to quantify ecological effects associated with the
atmospheric deposition.
There are a number of unique technical challenges in adjusting air quality to reflect just
meeting multiple standards, particularly given that the indicator pollutants for those standards are
4-6
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
related differently through emissions, chemistry and transport. The first challenge is that, SO2,
NO2, and PM, all have different levels of spatial variability. SO2 and NOx are directly emitted
from sources. As SO2 and NOx are transported from the emission sources, they are diluted by
atmospheric mixing, are chemically oxidized to other forms, including particles, and removed
from the atmosphere by deposition. Locations further from sources have low SO2 and NOx
concentrations and higher concentrations of their oxidation products, like particulate sulfate and
particulate nitrate. Because the sulfate and nitrate components of PM2.5 are not directly emitted
but are instead formed in the atmosphere, these components of PM2.5 are often more spatially
widespread. A scenario that adjusts SO2 or NOx must consider downwind impacts to sulfate and
nitrate PM2.5, and a scenario that adjusts PM2.5 must consider if such changes would require SO2
or NOx concentrations that are greater than the standards in upwind areas.
Second, many sensitive ecological areas are located far from large emission sources and
are affected by emissions in a large area. A recent study by Lee et al. (2016) examined eight
Class I areas in the U.S., and found that 50% of the nitrogen deposition could be attributed to
emission sources within 500 km and 90% could be attributed to emission sources as far away as
1500 km. The scenarios should not only consider concentration changes in an isolated emission
sources area, but also in the upwind areas whose emissions contribute to deposition.
Finally, a related challenge is that there are additional components of PM that are not
contributors to sulfate and nitrate deposition. For example, concentrations in the South Coast Air
Basin are greater than the primary PM2.5 NAAQS, but not the primary SO2 NAAQS. Only a
small portion of PM2.5 in Los Angeles is comprised of particulate sulfate. If SO2 concentrations
were increased, sulfate would also increase, pushing PM2.5 concentrations further beyond "just
meeting" the standard. Because of the mix of emission sources and contributing components to
PM2.5, at many locations, there is not a physically realistic way to adjust SO2, NO2, and PM2.5
concentrations where concentrations are just meeting all the relevant standards at the same time.
Instead, the REA will need to identify the controlling standard, and associated pollutant, for the
study area. In this analysis, we define the controlling standard as one where, when it is met, any
increases in the concentrations of other N and S pollutants would cause the controlling standard
to not be met.
Noting these challenges, we propose to first identify the area of influence for a particular
study area location based on available information. The study area location is likely to be around
100 km in diameter. As a default, we may assume an area of influence with a radius of 1500 km,
based on the findings of Lee et al. (2016). Within this area of influence, we will identify the
controlling standard and associated pollutant (i.e. SO2, NOx, or PM) by evaluating information
from ambient measurements and air quality modeling and considering current levels, as well as
historical relationships between emissions and ambient concentrations. Then, using this
4-7
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
information, we intend to develop an approach to adjust air quality in the study area location to
reflect just meeting the identified controlling standard under consideration. Based on the
analytical evidence in the recent SO2, NO2 and PM NAAQS reviews (U.S. EPA, 2018; U.S.
EPA, 2017c; U.S. EPA, 2010b), we note the successful application of various statistical
adjustment approaches in those reviews to estimate realistic changes in air quality concentration
to reflect changes in emissions of SO2, NO2 and PM. Similar approaches will also be considered
for this review, though in some areas we may also need additional information from air quality
modeling to better quantify the contribution from an important emissions source. Using the
approach developed, we intend to adjust air quality concentrations in the study area to develop an
air quality scenario that just meets the current standards at the locations most relevant for the
case study area. A similar approach would also be used to adjust air quality in the study area
location to reflect just meeting any potential alternative standards under consideration. Using the
approaches described in the next section, the changes in air concentration would be related to a
change in deposition.
4.2.2.2 Relating Changing Levels of Atmospheric Concentration to Deposition
For our proposed analyses, relating air concentrations to welfare effects requires that a
change in ambient concentration be related to a change in atmospheric deposition for input into
an equation or model that then relates a deposition of N and/or S to an ecosystem effect. This
section discusses the approach under consideration to estimate "deposition response factors," or
factors that can be used to relate changes in ambient concentrations of measured N and S species
to changes in atmospheric deposition of N and S in areas across the U.S.
The previous review introduced the transference ratio, TR, where: Deposition = TR x
concentration. In that review, the TR was estimated by dividing the annual average deposition by
concentration in a single CMAQ simulation and averaging over an Ecoregion. There were also
noted several uncertainties as described in Chapters 2 and 3, including:
• averaging over an ecoregion introduces uncertainty since the transference ratio is lower
near emissions sources and larger far from emission sources, especially as NOx is
oxidized to form compounds that deposit more quickly;
• different chemical transport models report similar estimates of wet deposition but report
very different transference ratios; and
• no quantitative assessment of uncertainty was conducted.
We propose to use an updated approach in this review to estimate deposition response
factors for each form of N and S deposition, including both dry and wet deposition. For this
approach, we note that in most areas of the U.S., concentrations and deposition of both oxidized
nitrogen and sulfur have declined since the last review and that measurements at dozens of co-
4-8
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
located CASTNet and NTN sites over the last 25 years are available to provide an observational
record of how a change in ambient concentration is related to a change in atmospheric wet
deposition. In addition, a 20-year CMAQ chemical transport model simulation (Zhang et al.,
2018) provides a modeled estimate of the relationship between ambient concentrations and both
wet and dry deposition. Taking advantage of these available modeled and measured
concentration and deposition data, we propose using statistical models to estimate deposition
response factors that relate a change in air concentration to a change in deposition and
quantitatively assess the uncertainty in various study locations that will be selected for this
analysis. These deposition response factors would then be used in conjunction with the adjusted
air quality scenarios (section 4.2.2.1) to estimate the N and S deposition associated with just
meeting the current standards, as well as any potential alternative standards under consideration.
In doing this, we propose to identify several pollutants or combinations of pollutants that
can be measured and used to relate ambient concentrations to atmospheric deposition. We will
refer to these as "deposition response factors." Criteria for a robust deposition predictor are (z) it
can be measured in ambient air with known accuracy and (z'z) it can be used to predict
atmospheric deposition with low error. We note that these deposition response factors may be
different than the indicators for the current standards (i.e., NO2, SO2, PM2.5 and PM10) and when
this is true, we plan to also assess the relationship between the deposition predictor and the
indicator for the current standards (including consideration of form, averaging time and level of
the standards).
We have listed our potential deposition response factors in Table 4-1. For some
deposition response factors, co-located measurements of ambient concentration and wet
deposition from 1990 to the present day exist and can provide a dataset to estimate the change in
wet deposition related to a change in air quality concentration. For dry deposition measurements
that are not routinely available, the CMAQ modeling information can be used to estimate the
relationship between dry deposition and air quality concentrations. Using these datasets, the
annual average concentrations and deposition calculated by a 20-year CMAQ chemical transport
model simulation will be used to fit a statistical model for each deposition predictor at the grid-
level scale (36 km grid cells), as well as at larger spatial scales. Measurements of ambient
concentration and wet deposition and their 20-year trends will be used to evaluate the chemical
transport model and statistical model results. Each of the steps in this analysis is listed below.
(1) Using a CMAQ simulation for 1990-2010 (Zhang et al., 2018), calculate the annual
average deposition (wet, dry, and total) for oxidized N, reduced N, and sulfur at each grid
cell location. Also calculate the annual average air concentration of NO2, NOy, HNO3,
particulate NO3", NH3, particulate NH4 , SO2, and particulate SO42" at each grid cell
location.
4-9
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
(2) For a specific grid cell, use the 20 years of concentration and deposition data to estimate a
statistical model that uses annual average atmospheric concentration — one of the
proposed deposition response factors — to predict the annual average dry deposition. For
example, the sum of particulate SO42" and SO2 is a proposed deposition predictor for dry
deposition of sulfur.
(3) Record the residual - the difference between the deposition predicted by the statistical
model and the deposition calculated by CMAQ - for each of the predicted annual
deposition. Calculate the average residual. This is used as an assessment of the magnitude
of error when using this deposition predictor.
(4) Repeat (3) for each grid cell and create a map of the average residual and best fit
coefficient. Examine how these vary.
(5) Repeat (3-4) for each deposition predictor listed in Table 4-1. For example, for oxidized
N, this could include HNO3, NOy, or total nitrate (HNO3 + particulate nitrate). Compare
the residual for each different deposition predictor. Compare both the national residual
average as well as use the map to make sure there are no areas where the error is
exceptionally large.
(6) Repeat (3-5), but instead fit a statistical model to predict wet deposition and total (sum of
wet and dry) deposition. Where possible, compare with historical measurements of wet
deposition and atmospheric concentration at co-located NADP and CASTNet sites.
Examine if the residual lower when estimating wet and dry deposition individually or
when added together as total deposition.
(7) Repeat (3-6), but rather than fit a different model for each grid cell, use all the grid cells,
excluding those over oceans, to fit a statistical model. Create a map of the average
residual for each grid cell. Explore different averaging areas for use in case studies to find
a balance that is representative and does not obscure variability in the transference ratio.
(8) Compare the residuals and select a deposition predictor that has low average residual and
is nationally relevant. Use the distribution of residuals as an estimate of uncertainty.
4-10
-------
1 Table 4-2. Initial Assessment of Available Deposition Response Factors
Deposition
predictor
Description
Available measurements
Reliably related to deposition?
Total nitrate:
HN03(g) +
PM2.5 NO3"
The sum of gas-
phase nitric acid
and particulate
nitrate
Measured at CASTNet at
many locations since 1990
Co-located NTN and CASTNet monitors
show oxidized N wet deposition can be
reliably predicted with total NO3- air
concentration measurements (Sickles
and Shadwick, 2013)
Total sulfate:
SC>2(g) + PM2.5
S042"
The sum of gas-
phase SO2 and
particulate SO42"
Measured at CASTNet at
many locations since 1990.
Co-located NTN and CASTNet monitors
show S wet deposition can be reliably
predicted with total sulfate air
concentration measurements
NOy
Sum of oxidized N
in the gas phase
and particulate
nitrate
Measured at NCore sites,
although most are in urban
areas.
Most NOy measurements are not co-
located with deposition measurements,
which makes it difficult to evaluate.
Sickles and Shadwick (2013)
demonstrated that this sometimes has
(and sometimes does not have) low
error for predicting deposition.
PM2.5 nh4+
Particulate
ammonium
Measured at CASTNet and
CSN but measurements are
biased and known to be
lower than atmospheric
concentrations.
Co-located NTN and CASTNet monitors
show reduced N deposition cannot be
reliably predicted using PM2.5 NH4+.
This could be explained partially by bias
in CASTNet NH4+measurements.
Another consideration is most of the US
is thought to have excess NH3 relative to
sulfate and nitrate, which suggests that
NH4+concentrations are more controlled
by sulfate and nitrate levels, rather than
NH3 levels.
NHX
The sum of gas-
phase NH3 and
particulate NH4+
Co-located CASTNet, NTN,
and AMoN sites measure
gas phase NH3 and
particulate NH4+. New
approaches to measure NHX
more robustly are under
evaluation.
Co-located CASTNet, NTN, and AMoN
sites suggest that NH4+ wet deposition
can be reliably predicted using NHX
measurements. However, NH3 is not
part of this review and therefore the
scope of this review would need to be
expanded before this deposition
predictor could be considered.
N02, S02,
PM2.5, PM10
Existing primary
NAAQS are based
on these
compounds
Nation-wide measurement
networks, but mostly located
near urban areas or large
emission sources. Few are
co-located with deposition
measurements.
The analysis described above will
examine how reliably these compounds
can be used to predict deposition.
4-11
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Based on preliminary analyses, total nitrate (defined as the sum of nitric acid and
particulate nitrate) could be the most promising deposition predictor for oxidized nitrogen
deposition. Analysis of co-located CASTNet and NTN measurements have found total nitrate to
be well correlated with oxidized nitrogen wet deposition (Figure 4-2) in the Eastern U.S. where
year-to-year variability in precipitation is relatively low. Analyses also suggest that the sum of
SO2 and particulate sulfur is well correlated with sulfur wet deposition. However, preliminary
results suggest that in many locations, particulate NH4+ is not a good predictor of reduced
nitrogen deposition, as recent measured trends show declining particulate NH4+ concentrations
with increasing or stable reduced-nitrogen deposition.
This approach addresses several of the uncertainties inherent in the approach used to
relate concentrations and deposition in the previous review. By using a statistical approach and
the 20-year time series, we can quantify and compare the level of uncertainty for different
deposition response factors. Rather than average over ecoregions, we will assess the spatial
variability in the transference ratio and select an averaging area best suited to the ecological
effect of interest. As described in Section 4.5, we will conduct comparisons with two chemical-
transport models. More analysis (discussed in Section 4.5) is planned to identify sources of
variability and quantitatively assess the uncertainty in using each of these deposition response
factors.
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
Ammonium correlation 0.08
Nitrate correlation 0.76
1999 2001 2003 2005 2007 2009 20112013 2015
NH4depasN NH4conc
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
3.0
2.5
2.0
1.5
1.0
0.5
0.0
19992001 20032005 2007 2009 201120132015
N03depasN TN03conc
0.4
0.4
0.3
0.3
0.2
0.2
0.1
0.1
0.0
Figure 4-2. NADP National Trends Network wet deposition (blue, left axis, kg/ha N) and
CASTNet air concentration measurements (orange, right axis, jig/m3 N) for Acadia
National Park from 1999-2016 annual average. CASTNet measurement of particulate
NH4+ is not well correlated with reduced N wet deposition (r = 0.08, left figure), while
total nitrate (gas-phase nitric acid plus particulate nitrate) air concentration is well
correlated (r = 0.76) with oxidized N wet deposition (right figure).
4-12
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
4.3 ECOLOGICAL EFFECTS ASSESSMENT
4.3.1 Aquatic Acidification and Nitrogen Enrichment
4.3.1.1 Freshwater Acidification
The connection between SO2 and NOx emissions, atmospheric deposition of N and/or S,
and the acidification of surface waters is well documented with several decades of evidence for
the eastern U.S. (see second draft ISA, Appendix section 7.1.1.2 and Driscoll et al., 2016). As
mentioned in Chapter 2, the 2009 NOx/SOx REA evaluated freshwater acidification in the
Adirondacks and Shenandoah, using ANC as the chemical indicator to derive CLs. In the REA
for this review, we intend to use a similar approach to evaluate freshwater acidification on a
national scale, and to better understand the linkages between chemical changes (changes in
ANC) in freshwater ecosystems attributable to acidifying deposition of N and/or S and changes
in biological effects on receptors, such as fish. To accomplish this, we intend to use a
combination of water quality measurements, geology, and surface water steady-state CLs data to
define the spatial distribution of acid-sensitive ecosystems across the U.S.
Biogeochemical dynamic models have also been used to assess impacts on water quality
(e.g., pH and ANC) compared with pre-acidification (i.e., pre-industrial) water quality conditions
(see second draft ISA, Appendix section 7.1.5.1, and Sullivan et al., 2011). Atmospheric N and
S deposition in recent decades have shown marked decreases due to significant power sector,
industrial, and mobile emissions reductions, allowing for the chemical and biological recovery of
water bodies (see Section 1.11.1 of the second draft ISA for more information on recovery).
Therefore, the evaluation of current ecological conditions will need to be done in the context that
aquatic ecosystems are on a recovery trajectory in response to decreases in atmospheric N and/or
S deposition. We intend to evaluate current conditions of aquatic ecosystems and their response
to atmospheric N and/or S deposition using steady-state and target CLs and model output from
biogeochemical models. We intend to use CLs for aquatic acidification from the NCLD v3.0 as
well as any relevant new peer-reviewed publications (e.g., Blett et al., 2014; Sullivan et al., 2015;
Fakhraei et al., 2014; Fakhraei et al., 2016; Shaw et al., 2014; Lawrence et al., 2016), where the
data are not yet included in the NCLD v3.0. In developing the REA, consideration will also be
given to any new CLs in scientific literature, but a comparison of all CLs will be completed to
more fully understand the differences.
ANC is an important chemical indicator, defined as the total amount of strong base ions
minus the total amount of strong acid anions:
ANC = (Ca2+ + Mg2+ + K+ + Na+ + NH4+) - (S042" + NO3" + CI")
4-13
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
While ANC is not directly linked to the physiological impacts of aquatic biota, it is
commonly used because it integrates overall acid status. Unlike pH, it is not affected by
dissolved carbon dioxide (CO2), and water quality models are generally better at modeling ANC
than other indicators (see second draft ISA, Appendix, section 8.1). In addition, ANC has been
extensively related to the health of biota and other surface water constituents like pH and A1 (see
second draft ISA, Appendix, sections 8.3.6.3). In the case where DOC plays an important role in
acid-base balance of the surface water, the measure of base cation surplus (BCS) that adjusts for
the organic acid status of surface waters (Lawrence, 2007) will be considered. BCS is a metric
like ANC, but accounts for the charge balance natural organic acidity contributes to the ionic
concentration of the surface water (see second draft ISA, Appendix section 8.1).
Given the connection between atmospheric N and S deposition to biogeochemical
changes that may induce harmful biological effects, we intend to use ecological literature
included in the second draft ISA, Appendix section 8.3 and survey data from various sources to
explore revising the aquatic status ANC categories of potential effects that were used in the 2009
NOx/SOx REA (e.g., Table 4.2-1 from the 2009 NOx/SOx REA). We will also work to refine
the relationship between ANC and the biological responses. Compiled information for major
taxonomic groups and other biogeochemical and ecological information (e.g. pH, habitat
condition) will be used to inform the aquatic status ANC categories of potential effects described
in the second draft ISA (see second draft ISA, Appendix section 8.3). Where data are available,
we intend to develop exposure-response relationships (total species-richness vs. pH or ANC,
biomass vs. pH or ANC, etc.) by building on the research of Layer et al. (2013), Baldigo et al.
(2018), and other studies listed in the second draft ISA, Appendix section 8.3. Additionally, we
intend to evaluate the CLs with respect to range of ANC thresholds of 20 to 100, which represent
the range of ecosystem protection for biota as described by the aquatic status ANC categories.
This falls within the range that is most commonly used in the NCLD v3.0 and among CL
assessments in the U.S. (see second draft ISA, Appendix Table 8-8).
For the case study areas, we intend to use the MAGIC model to evaluate potential
ecological and biological changes due to changes in atmospheric deposition. Knowing the pre-
acidification water quality conditions provides a benchmark to measure how near chemical
recovery is to the pre-acidification level. The time it takes to reach chemical recovery will also
be determined using the MAGIC model. A host of environmental factors and the severity of
atmospheric deposition-driven acidification influence how fast aquatic ecosystems respond to
reduction in atmospheric N and/or S deposition. In sensitive ecosystems that are strongly
impacted chemical recovery varies significantly, but can take decades. The selection of case
study areas will aim to include a subset of aquatic ecosystems from various sensitive regions
4-14
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
with long-term water quality records with supporting environmental data to be modeled using the
biogeochemical models.
4.3.1.2 Freshwater Nitrogen Enrichment
We intend to further consider the extent of the analyses for the REA that use new CLs
data from Baron et al. (2011), Williams et al. (2017), and Nanus et al. (2017), as described in
Chapter 3, to evaluate aquatic N enrichment effects in freshwater lakes and streams in the REA.
These CLs use N03 concentrations as an indicator of N enrichment, associated with varying
biological and chemical criteria as discussed in section 3.5.2 and the second draft ISA Appendix
8, section 8.5.4.147. The potential analyses are described below but, as we develop the REA, we
plan to further consider which, if any, of these analyses will be included in the REA with the
judgement being based on whether the results can substantively inform our ability to assess
national standards for freshwater N enrichment.
The Baron et al. (2011) and Williams et al. (2017) studies provide CLs with broad
geographic scope, which apply to all lakes within a given study region. We could potentially
apply these CLs in the specific regions covered under each of the studies as follows: apply the
CLs from Baron et al. (2011) to lakes included in the study in the Sierra Nevada Mountains,
Rocky Mountains, and the northeastern U.S. to assess risk of elevated NO3 in surface waters48;
apply the CLs from the Williams et al. (2017) study to western lakes that fall within the sampled
criteria (elevation > 1200m), with a higher level of confidence attributed to lakes that were
included in GLNC to assess risk of shifts from N to P-limitation of phytoplankton growth. The
CLs from the Nanus et al. (2017) study provide CLs that are specific to each lake within the
greater Yellowstone area. We could apply these CLs to the greater Yellowstone area as defined
in the study to potentially assess the risk of changes to diatom growth and community
composition. We could also explore the application of these CLs to the larger Rocky Mountain
area assessed by Nanus et al. (2012) and potentially to other areas of the western U.S. In
addition, the regression parameters in Nanus et al. (2017) constitute exposure-response functions
that potentially relate the NO3 levels in individual lakes to atmospheric N deposition as well as
other watershed level factors, including slope and amount of barren ground. These functions
could be used to calculate expected NO3 levels based on different atmospheric deposition
patterns. The resulting NO3 levels could be compared the categories of potential effects for NO3
described above. The combination of these response functions and categories of potential effects
47 We note that elevated NO3 concentrations in lakes and streams are also indicative of possible N saturation in
adjacent terrestrial ecosystems leading to leaching of NO3 into the waterbodies (see second draft ISA Appendix 4,
section 4.3.2 and Appendix 7, section 7.1.2).
48 Lake data in Baron et al. (2011) were taken from Linthurst et al. (1986) for eastern lakes and Eilers et al. (1987)
for western lakes
4-15
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
could potentially be used to assess the risk posed by different levels of atmospheric N deposition
to freshwater lakes. Depending on the study being applied, risk may be assessed at individual
lakes and/or aggregated at larger areas.
4.3.2 Terrestrial Effects
The connections between emissions, atmospheric deposition of N and/or S, and the
alteration of plant community composition and species richness in terrestrial ecosystems is well
documented (see second draft ISA, Appendix section 5). For terrestrial ecosystems, the main
findings of the current ISA are that N and S deposition cause alteration of soil biogeochemistry
and effects on multiple levels of biological organization ranging from physiological processes to
shifts in biodiversity and ecological function (ISA ES, p. lxxii).
Thus, this section describes potential quantitative analyses for soil biogeochemistry and
groups of organisms. Some data are available at the level of biogeochemical effects while other
datasets reflect growth and mortality effects on individual species of plants or effects to
community composition and biodiversity. The sections below provide details on the studies that
we intend to use, as well as any specific modifications and decisions that will be made or
considered, when conducting the quantitative assessments.
It is important to note that in terrestrial ecosystems, the acidifying effects from N and S,
and the enrichment effects from N can occur simultaneously within the geographic boundaries of
an ecosystem. However, the dominance of one process over the other (i.e., acidification versus N
enrichment) can vary depending upon several factors, including soil pH and buffering capacity,
the presence and abundance of sensitive species, and the influence of nitrogen limitation on
primary production. Because acidification and N enrichment effects are intertwined, it is often
difficult in large scale gradient studies to isolate mechanisms (i.e., acidification versus N
enrichment), as both are co-occurring across the landscape. When possible, the mechanism of
effect will be noted in the description of the datasets.
4.3.2.1 Biogeochemical Effects in Soils
As mentioned in Chapter 2, the 2009 NOx/SOx REA used a case study approach for
sugar maple and red spruce, deriving CLs using Bc/Al as the chemical indicator for acidification
of forested soil that related to tree health. In the REA for this review, we intend to use a similar
approach, but now will expand the CLs nationally. We intend to use the Bc/Al ratio as a soil
chemical indicator, given that it is commonly used, relates well to the Ca/Al ratio, and serves as
an important input into the SMB model for calculating soil CLs (see second draft ISA, Appendix
section 5.2.1).
There is precedent for using Bc/Al levels between 1 and 10, because these values are
documented to cause deleterious effects to trees, as the chemical limit in CLs calculated using
4-16
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
the SMB model. For example, McNulty et al. (2007) chose Bc/Al levels of 1 (for coniferous
forests) and 10 (for deciduous forests). Additionally, Duarte et. al (2013) and Phelan et. al (2014)
conducted studies in the Northeastern U.S., each using a Bc/Al level of 10. For the REA, we
intend to calculate national CLs using the SMB model and Bc/Al levels of 1 and 10 to represent
different forest types in different parts of the U.S.. We will also explore in the REA whether
there are levels between 1 and 10 that should be applied, considering any new scientific evidence
for relating Bc/Al to categories of ecological effects. This approach is similar to the approach
proposed for the use of ANC as a chemical indicator for freshwater acidification and NO3 as a
chemical indicator of freshwater N enrichment.
BCw is one of the most influential yet difficult to estimate parameters for deriving the
forest soil acidification CLs (see second draft ISA, Appendix section 4.5.1). There are two
primary methods available for estimating BCw: the clay-substrate method and the PROFILE
model. The clay-substrate method has been applied nationally, but as noted in the 2009
NOx/SOx REA, has known issues especially when estimating BCw in unglaciated soils. Given
the results of Phelan et al. (2014) in the Pennsylvania case study, we intend to explore the use of
PROFILE modelling to obtain more accurate national estimates of BCw rates for the REA,
considering uncertainties identified in Whitfield et al. (2018). Additionally, researchers have
recommended applying PROFILE in different areas of the U.S. to improve our understanding of
how the model performs and to compare the BCw rates estimated by PROFILE to those
estimated using the clay correlation-substrate model, (see second draft ISA, Appendix section
5.5). We will use the results from Phelan et al. (2014) to explicitly quantify the relationship
between PROFILE estimated BCw and clay-substrate-method quantified BCw.
4.3.2.2 Species-Level Effects
4.3.2.2.1 Trees
We intend to conduct quantitative analyses to understand growth and mortality effects on
individual tree species across the contiguous U.S under current air quality conditions using the
exposure-response curves described in Horn et al. (in review).49 These curves are national in
scope and represent statistical equations that relate growth and mortality responses to N
deposition, S deposition, and other factors including temperature, precipitation, tree size, stand
basal area.
There are 156 species assessed in Horn et al. (in review). Additionally, the dataset
includes benchmarks for the points on the curve where a species experiences various levels of
reductions in growth and survival - 1%, 2%, 5%, 10%. However, the equations themselves can
49 In developing the REA, if Horn et al. (in review) is not published, we plan to revisit the proposed plan for the
REA and consider analyses that focus on other published literature (e.g., Thomas et al., 2010).
4-17
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
be used to calculate any percentage desired. Trees were sampled at United States Forest Service
(USFS) FIA plots, with at least 2,000 trees sampled for each species constituting a dataset of
over 50,000 plots and more than 1.2 million trees. There were 94 species that met the N>2,000
sample size cutoff for both growth and survival (highlighted in the publication) and an additional
14 species that met the cutoff for survival only50. TDEP data years used for the deposition
estimates overlap with the period of re-measurement for each plot. Samples at each plot were
collected every 5 years on average. Data years for the dataset span from 2000-2016, with each
plot linked with the average deposition experienced between the first and last measurement
period.
There are two methods under consideration for spatially associating the response curves
from Horn et al. (in review) to the specific locations of the tree species being evaluated. One
approach considers analytics based on the USFS FIA plot data, and the other based on the USFS
continuous surface maps (Wilson et al., 2013). Both methods have their advantages and
disadvantages, which are summarized more below.
• USFS FIA plot data. Our confidence is high in this data because it is based on empirical
field measurements. However, although the density of plots is high when viewed at the
continental scale, at smaller scales the density may be low enough that inferences across
areas are limited. Thus, because there is one sample taken every 6,000 acres across the
landscape according to the FIA plot layout, there are several areas that are important to
public welfare, such as Class I Wilderness Areas and National Parks, that are composed
of only a few FIA plots.
• Wilson et al. (2013) USFS maps: These maps are based on the FIA data that provides
continuous surface maps of predicted abundances for 322 species in 250 m grid cells
across the contiguous U.S. (including all 94 tree species included in Horn et al. (in
review)). These abundances are based on statistical relationships with various factors
(i.e., phenology, temperature, precipitation, topography, and ecoregion). Abundance in
Wilson et al. (2013) is based on basal area, which is a common metric in forestry that
represents the area of tree trunks within an area of land (ft2 ha"1). The accuracy of these
predictions varies by species and the scale of inference, with higher accuracy for more
abundant species and larger areas. Wilson et al. (2013) and Riemann et al. (2010) provide
several metrics to assess the accuracy of projections for all 322 tree species. All tree
species in Horn et al. (in review) are included in Wilson et al. (2013).
For the analyses, we will determine the most appropriate methodology for applying
response curves across the landscape in the REA by considering the benefits and limitations of
both datasets. These data will be compared to the N and sulfur deposition values from the
national-scale TDEP surface for current air quality conditions. This information will help to
50 Note the study focused on the 94 species that met the criteria for both, though all results are provided in the
Appendix of the paper.
4-18
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
further identify sensitive areas of the U.S. that are impacted by recent deposition levels and may
be good candidates for the case study approach.
4.3.2.3 Community-Level Effects
4.3.2.3.1 Mycorrhizal Fungi
The mycorrhizal fungi dataset is a collection of empirical CLs for total atmospheric N
deposition from various studies included in Pardo et al. (2011) as well as newer studies reviewed
by the ISA (e.g. Allen et al., 2016). This dataset includes minimum and maximum CLs for
changes in community composition. CLs for mycorrhizal fungi will be used to assess the
potential for current levels of deposition to affect mycorrhizal community composition in forests.
4.3.2.3.2 Lichen Community
The underlying analyses in Geiser et al. (2010) and Root et al. (2015) include response-
functions relating N deposition to lichen community composition. The functions described are
currently based on data from the Pacific Northwest, so extrapolation to other areas of the country
would have higher levels of uncertainty. While the NCLD v.3.0 includes other lichens CLs from
Pardo et al. (2010) based on individual site-specific studies, we intend to use only the CLs from
Root et al. (2015) and Geiser et al. (2010) for assessing risk because they use more consistent
methodologies. Furthermore, we note that new work is underway to extend the work of Root et
al. (2015) and Geiser et al. (2010) using similar methods on a national scale (personal
communications with Linda Geiser, 2018). If this work becomes available, it may be used to
broaden the geographic scope of the lichen data considered in the REA.
Because of the spatially restricted nature of the lichen study from Geiser et al. (2010) and
Root et al. (2015), we will likely restrict the application of these exposure-response curves to the
Western U.S. (specifically, to the two Ecoregions in the west where data were collected: North
American West Coast Marine Ecoregion in Oregon and Washington and the Northwestern
Forested Mountains Ecoregion in Northern California, Oregon, Washington, and parts of
northern Idaho and Montana). We will consider extrapolation to other portions of these
Ecoregions, but not nationally. Using the relationships in Geiser et al. (2010) and Root et al.
(2015), we will calculate sensitivity and risk to lichen communities using the CLs and the
exposure-response curve.
4.3.2.3.3 Herbs/Shrubs
Given their national coverage, we intend to use the CLs from Simkin et al. (2016) to
evaluate changes in species richness for herbs/shrubs communities. We note that while the
NCLD v.3.0 also includes a collection of herbs/shrubs empirical CLs from Pardo et al. (2011),
we have chosen to use those derived from Simkin et al. (2016) because the study is national in
4-19
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
coverage, has explicitly quantified uncertainty, and used a common methodology in contrast with
those from Pardo et al. (2011). The statistical equations in Simkin et al. (2016) constitute
exposure-response curves that relate the species richness of herbs/shrubs at a site to N deposition
as well as other factors, including soil pH, temperature, and precipitation. We will use these
response curves to assess the risk posed by N deposition to herb/shrub community species
richness. Risk will be assessed at individual sites or aggregated at larger areas. We will also
explore the use of these curves to examine loadings that would minimize reductions in species
richness at individual points/sites and across the country. However, avoiding reductions in total
numbers of species does not avoid changes in composition and losses of individual species (e.g.,
a gain for 2 species and loss for 2 with no change in total species richness). Thus, if additional
information becomes available at the species level, we will incorporate it into the analysis in the
REA.
4.3.2.3.4 Forests and Grasslands
While we intend to conduct quantitative analyses to understand growth and mortality
effects on individual tree species across the contiguous U.S under current air quality conditions
using the exposure-response curves described in Horn et al. (in review), we also plan to apply the
exposure-response curves to consider changes in the growth and survival of multiple species in
the selected study areas for various air quality scenarios that include just meeting the current
standards, as well as any potential alternative standards, as appropriate. In doing so, we intend to
evaluate potential effects on tree species within a forested or grassland area and, as such,
potential changes in the species composition for that area. We note that in our evaluation that, in
some areas, we may find that some species are especially abundant and responsive to N and S
deposition, and drive much of the changes in forest or grassland growth and survival rates. On
the other hand, another area may be dominated by unresponsive species, and only locally rare
species are estimated to be impacted. Therefore, the analyses will also provide information
regarding the extent to which there are differences in effects for various species within each
study area and between the set of study area locations.
4.4 CASE STUDY AREA SELECTION
A case study approach is being proposed in the REA to provide the ability to include
more refined analyses than can be done at the national scale and that can reflect various
combinations of changing air quality and deposition, as needed and as appropriate. For the case
study component of this REA, we intend to characterize ecological risk and exposure in 5-10
study area locations for when air quality is just meeting the current standards, and if appropriate,
when meeting potential alternative air quality standards. While we do not intend for the study
area analyses to provide a comprehensive national assessment, they are intended to provide
4-20
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
assessment for a small varied set of study areas that will be informative to EPA's consideration
of potential risks and exposures that may be associated with various air quality scenarios.
In selecting the case study locations, we intend to consider specific criteria, with the goal
of including case studies in the REA that are particularly informative to the review. Based on
currently available data, tools and information and considering the proposed approaches to the
analyses, as discussed above, we propose the following set of criteria for consideration in the
selection of the study area locations:
• Influence of current air quality on deposition-related effects. As recognized above,
historic air quality (and associated deposition) has had appreciable impacts on
ecosystems in many areas. In such areas, the continued role of this historic deposition in
the area's deposition-related impacts poses challenges to us in assessing the effects of
current deposition or deposition associated with the current standards. Therefore, it is
particularly important for study area selection to understand, as well as possible, the
influence of current air quality (compared to historic air quality) on deposition-related
effects. In areas with relatively greater influence of current air quality, we would expect
the deposition-related risk and exposures to be responsive to changes in deposition. The
impacts for various air quality scenarios in such areas would be expected to be
particularly informative to consideration of the adequacy of the current standards. Thus,
case study locations in which the deposition-related effects are not dominated by the
influence of historic air quality would be good case study candidates. Accordingly,
features of an area to be considered will include geology, historical deposition levels,
recovery potential, and has the area experienced historically high deposition levels such
that, for example, a terrestrial system is N saturated.
• Sensitivity. Related to the prior criterion is one that considers the extent of ecosystem
sensitivity. To insure consideration of the depositional impacts on at-risk ecosystems, it
will be important to consider including areas for which ecosystems or receptors are
sensitive to N and/or S deposition, at risk due to such deposition, and specifically
sensitive to changes in deposition.
• Importance to public welfare. To assess impacts on public welfare (in order to inform
consideration of the adequacy of public welfare protection afforded by the current
standards), good candidate areas will be areas for which there is a connection of potential
impacts to ecosystems and effects on public welfare. For example, pertinent
considerations may include: is the area in a National Park or Class I area where there is a
clear connection to public welfare? Does the area include endangered or threatened
species? Are there culturally or economically significant species present?
• Diversity and occurrence. A good set of candidate areas would encompass a set of
diverse conditions occurring across the country. Given this, we propose to consider
selecting a set of areas comprising a broad array of NOx/SOx/PM-related chemical
environments, while also ensuring inclusion of areas that reflect current atmospheric
deposition patterns (e.g. areas dominated by oxidized as opposed to reduced forms of N).
Similarly, we propose to consider areas where the resident ecological receptors also occur
more broadly (e.g. tree species in the area are widespread and abundant in locations
outside of the study area) as well as areas that encompass several different ecosystems.
4-21
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
• Data availability and adequacy. It is important for the case study locations to have
adequate data available, thus reducing uncertainty in the risk and exposure analyses.
- Air quality data: Given that our objective relates to the relationship between air
quality concentrations and deposition, a good candidate study area would be one
in which: air concentration and wet deposition measurements for multiple N and S
species are available for recent years and the monitors are either co-located or
located close to each other; CMAQ model performance in predicting
concentrations and deposition of N and S species is reasonable; and the statistical
model fit to predict wet deposition and total deposition has reasonably low
uncertainty. An additional related consideration is the magnitude of air quality
adjustment that would be required for the area's air quality to just meet the current
standards, with the understanding that smaller air quality adjustments will likely
be less uncertain.
- Freshwater and terrestrial ecosystem data: Given the reliance on CLs and
exposure-response function data in our proposed analyses, a good candidate area
would have: CLs and exposure-response function data available; CLs data
available for multiple endpoints or effects (sections 4.1.2 and 4.1.3); and
exposure-response function data available for multiple receptors (sections 4.2.4
and 4.2.5) and for multiple tree species (section 4.2.5.2). Additionally, in a good
candidate area, CLs data would have sufficient individual site measurements and
tree species with exposure-response functions to allow robust quantitative
analyses.
• Availability of additional information. An additional criterion for consideration is the
availability of other pertinent data. For example, the second draft ISA included six case
studies, chosen because they were areas for which there was substantial amount of
published work on ecological response to N and/or S deposition available (second draft
ISA, Appendix 16) 51. A consideration may be whether the qualitative scientific and air
quality information in one or more of the case studies in the second draft ISA be used
when considering the criteria above. Other information may also be available that makes
an area a good candidate study location. For example, we may want to consider whether
there is any other qualitative information that is useful (e.g., ecosystem services
information or survey data).
4.5 CONSIDERATION OF VARIABILITY AND UNCERTAINTY
To characterize ecological risks, exposure and risk assessors commonly use an iterative
process of gathering data, developing models, and estimating exposures and risks, given the
goals of the assessment, scale of the assessment performed, and limitations of the input data
available. However, significant uncertainty often remains and emphasis is then placed on
characterizing the nature of that uncertainty and its impact on exposure and risk estimates.
51 The locations included were the northeastern U.S., Adirondack State Park, southeastern Appalachia, Tampa Bay,
Rocky Mountain National Park, and southern California. The case studies identified current acidification and
nutrient status, as well as empirical and modelled CLs.
4-22
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Below we summarize the most important uncertainties potentially affecting the risk and
exposure estimates and how we can address some of them using newly available information.
Conclusions drawn from these characterizations based on new information and any new
evaluations performed in the current REA will be synthesized following the approach outlined
by the World Health Organization (WHO, 2008). This synthesis will be used to identify,
evaluate, and prioritize the most important uncertainties relevant to the estimated exposure and
risk outcomes. We intend to use both a qualitative approach along with quantitative approaches
where possible in the REA as discussed below.
The approach to be used here varies from that described by WHO (2008) in that a greater
focus will be placed on evaluating the direction and the magnitude52 of the uncertainty; that is,
qualitatively rating how the source of uncertainty, in the presence of alternative information, may
affect the estimated exposures and ecological risk results. Following the identification of key
uncertainties, staff will subjectively scale the overall impact of the uncertainty by considering the
relationship between the source of uncertainty and the exposure concentrations (e.g., low,
moderate, or high potential impact). Also to the extent possible, staff will also include an
assessment of the direction of influence, indicating how the source of uncertainty could affect
estimated exposures or risk estimates (e.g., the uncertainty could lead to over- or
underestimates). Further, and consistent with the WHO (2008) guidance, staff will discuss the
uncertainty in the knowledge base (e.g., the accuracy of the data used, acknowledgement of data
gaps) and decisions made where possible (e.g., selection of model forms). The output of the
uncertainty characterization will be a summary describing, for each identified source of
uncertainty, the magnitude of the impact and the direction of influence the uncertainty may have
on the exposure and risk characterization results. The discussion below illustrates some of the
main considerations for uncertainty and variability in air quality, CLs datasets broadly, and
specific uncertainty considerations for specific effect categories.
4.5.1 Air Quality
The review intends to examine uncertainty and variability in (z) our approach to assess
current conditions and (z'z) our approach for estimating the deposition response factors as the
change in deposition related to a change in air concentration. To assess uncertainty in our
estimate of current conditions, specifically, the merged model and measurement datasets of air
concentrations and deposition from 2014 to 2016, we intend to examine the difference between
modeled values and measurements. The spatially-interpolated wet deposition dataset will be
assessed using cross validation, such as "leave-p-out" methods where a subset of the
52 This is synonymous with the "level of uncertainty" discussed in WHO (2008), section 5.1.2.2.
4-23
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
observations is reserved to assess the uncertainty in the interpolation. The CMAQ-calculated dry
deposition could also be evaluated with measurements where possible, but given the scarcity of
dry deposition measurements, we propose examining three sensitivity studies. First, we could
conduct Comprehensive Air Quality Model with extensions (CAMx) simulations with similar
model configurations and the same emissions and meteorological inputs as the 2016 CMAQ
simulations. Earlier work and CASAC committee have noted differences between CMAQ and
CAMx deposition estimates from previous model versions, and a large fraction of the air quality
modeling community (including Federal Land Managers) use CAMx in their work. Second, we
could conduct simulations using CMAQ augmented to include the Zhang dry deposition scheme
used in CAMx. This simulation is to understand uncertainty arising specifically from alternative
model configurations for dry deposition. Third, we could conduct simulations using
CMAQv5.2.1 with mosaic land cover dry deposition to quantify the variability in deposition for
different land cover types. Forests, grasslands, and lakes each support ecosystems with differing
sensitivity to nitrogen and sulfur. Each of the land cover types are also physically different
surfaces, which influences the micrometeorological conditions that govern dry deposition. For
example, the dry deposition velocity of nitric acid over rougher forest areas can be a factor of 4
larger than over water areas. The purpose of this simulation would be to examine the effect of
considering deposition specific to each land cover type.
We propose to also pursue multiple approaches to assess uncertainty in our estimate of the
deposition response factors - the change in deposition related to a change in air concentration.
First, because we are estimating the deposition response factors by fitting a statistical model, we
can assess the range of errors expected by analyzing the residuals, or the difference between the
statistical model predictions and the data used to fit the statistical model. Second, since the
statistical model is fit using CMAQ model results, we can also fit a statistical model to co-
located measurements of air concentration and wet deposition. We have greater certainty when
our estimate of the deposition response factors is similar when using either CMAQ model data or
co-located measurements to fit the statistical model. Next, we can examine different approaches
to spatial averaging of the air concentration and deposition data before fitting the statistical
model, in order to develop an approach with lower uncertainty that can still capture the spatial
variability needed to assess ecosystem impacts. Finally, we can examine our findings collectively
from these uncertainty assessments and develop a framework to integrate them into a
quantitative assessment of the total uncertainty in our estimate of the deposition response factors.
4.5.2 Critical Loads Data
Potential uncertainty/limitations for the CLs datasets depend upon several factors: 1)
whether the dataset includes CLs from one study (i.e., that used a consistent methodology)
4-24
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
versus multiple studies (i.e., where different methods may have been applied); 2) whether CLs
were extrapolated from study areas to broader geographical areas; 3) whether there is statistical
uncertainty in inputs or calculations within modeled CLs.
CLs are also built using different air quality surfaces and projections and this inherently
introduces a degree of uncertainty in the CL estimates. This uncertainty manifests in several
ways. If the air quality estimates have quantified uncertainties, but do not have a directional bias,
then the CL should be considered an average estimate, with some range of uncertainty. However,
if uncertainty in the air quality estimates is known to be biased, then the CL may be considered
either a high or low estimate based on the directionality of the bias. For example, a CL based on
air quality estimates that are known to be low should be considered the low end of the CL
estimates, with a range of uncertainty that is greater than the estimate. An important component
of uncertainty related to air quality with respect to CLs will be in the comparison of deposition
surfaces developed in the REA to CLs developed with different air quality and how the
uncertainties inherent in both the CL and the deposition surface relate to each other.
The following table (Table 4-2) summarizes the major sources of uncertainty for the CLs
included in this plan based on the considerations outlined above.
Table 4-3. Major Sources of Uncertainty for the CLs Considered for the REA.
Critical Load
Consistent
Methodology?
Extrapolation?
Quantified Statistical
Uncertainty
Air Quality
Mycorrhizal
fungi
No (multiple studies)
Yes
No
Multiple AQ inputs
Lichens
Yes within Root et al.
(2015) and Geiser et
al. (2010), some
differences between.
Yes
No (in the NCLD) or
Yes (source
publications)
IMPROVE + CMAQ +
NADP (Geiser et al.
2010), IMPROVE (Rootet
al. 2015)
Herbaceous
species
richness
Yes
No (point based)
or Yes
(ecoregion-
based).
Yes
CMAQ + NADP
Forest soil
acidification
Yes
No
No (in the NCLD) or
Yes (source
publications)
Not an input
Freshwater
acidification
Yes
No
No (in the NCLD)
Not an Input
Freshwater N
enrichment
Yes within Baron et al.
(2011), Williams et al.
(2017) and Nanus et
al. (2017), different
between.
Yes within Baron
et al. (2011) and
Williams et al.
(2017). No within
Nanus et al.
(2017)
Yes within Williams et
al. (2017) and Nanus et
al. (2017). No within
Baron et al. (2011)
PRISM+NADP+CMAQ
(Baron et al. 2011),
CMAQ (Williams et al.
2017),
NADP+RMS+PRISM
(wet) +TDEP (dry)
(Nanus et al. 2017)
4-25
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Given these considerations, we expect uncertainty for the mycorrhizal fungi dataset to be
high as compared to the other datasets given that the data was built by combining the results
from multiple studies and approaches, it includes extrapolation, and there is no explicit
quantification of uncertainty included in the NCLD. Additionally, we expect uncertainty for the
lichens dataset to be moderate given that data from two studies in Washington and Oregon were
extrapolated nationally, and at least in the NCLD v.3 there is no inclusion of uncertainty. Within
the areas/ecoregions of study, the uncertainty in the lichen CLs will be lower. For herbs/shrubs,
we intend to use CLs from the Simkin et al (2016) study, which included an explicit
quantification of uncertainty that was found to be low. In addition, researchers reported that the
CLs from Simkin et al. (2016) were consistent with CLs derived from other studies. Therefore,
there is high confidence in the data.
The CLs for freshwater N enrichment have differing uncertainties. We intend to apply the
CLS for Baron et al. (2011) and Williams et al. (2017) across the study areas, which includes
higher uncertainty due to extrapolation. In both datasets, we intend to also apply the CLs only to
lakes included in the calculation of the CLs, which will reduce the uncertainty. Additionally, the
Williams et al. (2017) study included quantification of uncertainty. We intend to apply the CLs
from Nanus et al. (2017) across the study area. Similar to the Simkin et al. (2016) study, Nanus
et al. (2017) included quantification of uncertainty as well as being consistent with CLs derived
from other studies.
The CLs for freshwater and terrestrial acidification were modelled and empirical values
which did not include air quality as a model input. Both datasets have lower levels of uncertainty
based on the criteria discussed above, however, other uncertainties apply to these CLs which are
discussed below.
An important component of uncertainty analysis is sensitivity analysis. Sensitivity
analysis has been used extensively to look at critical load model parameters uncertainty ranges
(Skeffington, 2006; Heywood et al., 2007; Fakhraei et al., 2016; Fakhraei et al., 2017). For
example, Skeffington (2006) and Heywood et al. (2007) used a Monte Carlo probabilistic
method to assess the degree of confidence in the CL exceedance and the coefficient of variation
of the CL. They concluded that CL uncertainties are typically surprisingly small because of a
"compensation of errors," despite the input parameters often being poorly known. While natural
variation and uncertainty in model parameters are important considerations, existing Monte
Carlo sensitivity methods/tools are widely available, with many case study examples that can to
be used in the REA to provide CL confidence, uncertainty, and CL exceedance probability to
assess the risk to ecosystems from atmospheric N and/or S deposition. We propose using a
probabilistic Monte Carlo approach to describe uncertainty and to describe the likelihood of the
CL to be exceeded.
4-26
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
4.5.2.1 Freshwater and Terrestrial Acidification CLs
4.5.2.1.1 Freshwater
Freshwater ecosystems are complex and dynamic environmental systems and the impacts
of ambient air pollution and the resulting deposition is influenced by a host of environmental
processes that interact with the spatial variability of soils, geology, and biota. Given this
complexity, CL models and even more complex biogeochemical models are an inevitable
simplification of the natural system, which leads to inherent uncertainty in these assessments
(Morgan and Henri on, 1990). The most noted simplification is the steady-state assumption for
such dynamic ecosystems. Steady-state CLs are often derived from mathematical mass-balance
models under assumed or modeled equilibrium conditions based on present water quality
measurements. While steady-state CL models vary in complexity about the processes
represented, a fundamental aspect of this approach is the steady-state assumption which rarely
occurs in natural ecosystems since the ambient environment and atmosphere are continuously
changing. Future weather modification may compound this limitation given that recovery of
aquatic ecosystems from acidification will be affected by physical, chemical, and biological
modification as annual mean temperature and precipitation levels lead to changes in runoff and
bedrock and soil weathering (Greaver, 2016). Nevertheless, the steady-state assumption still
provides an approximation that allows a quantitative risk assessment of atmospheric N and/S
deposition impacts on aquatic ecosystems.
The models and their parameters used to determine aquatic CLs have been extensively
examined over the past 30 years (e.g., Skeffington, 2006; Rapp and Bishop, 2009). In all cases,
the strength of the aquatic CL estimate and the exceedance calculation relies on the ability of the
method to estimate the watershed-average base cation supply (i.e., input of base cations from
weathering of bedrock, soil and deposition), runoff volume, base cation and acidifying
deposition, and NO3" leaching. Runoff volume and base cation deposition make smaller
contributions to uncertainty as compared to base cation supply. It is the base cation supply that
buffers the inputs of the acid anions from atmospheric deposition and is the factor that has the
most influence on the CL calculation and the largest uncertainty (Skeffington, 2006; Li and
McNulty, 2007; Fakrhaei et al., 2017).
Compounding the difficulty in estimating the base cation supply is the lack of methods to
directly measure bedrock and soil weathering rates. In most cases, estimates of watershed-
average base cation supply is determined indirectly by simple mass-balance differencing using
present watershed water chemistry measurements or by more complex biogeochemical
watershed models that still use water chemistry measurements as the bases for the determination.
The CL is based on the present-day water chemistry and a model that drives the pre-industrial
condition.
4-27
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
A host of different methods have been used to determine base cation supply, from more
complex biogeochemical models (e.g., MAGIC, PnET-BGC, PROFILE) to simple models, such
as the F-factor as part of the SSWC model (Hernriksen et al., 1990). While the importance of
base cation supply to the estimate of CLs is well recognized, limited comparisons across models
have been completed with the focus being on comparing dynamic biogeochemical models
(Tominage et al., 2010). However, the use of dynamic biogeochemical models is more difficult
because they require substantial input data, whereas simple models like the F-factor along with
the SSWC model are often more widely used given the fewer inputs needed. Although the
SSWC/F-factor critical load model has been widely published around the world (e.g., Henriksen
and Posch et al., 2001; Dupont et al., 2005; Shaw et al., 2014), many model limitations have
been noted. Rapp and Bishop (2009) extensively examined the F-factor approach and found key
limitations related to the use of present-day water chemistry data to parameterize the F-factor.
They found that water chemistry data from the period where the watershed is recovering from
acidification can lead to uncertain estimates of the base cation supply. In addition, the F-factor
was sensitive to short-term variation in the water chemistry data that reduced the confidence in
the CL. However, the use of long-term averaged water chemistry data help to minimize this
impact. Another limitation is the influence of increasing DOC concentrations in lakes and
streams observed in recent water quality data (Monteith et al., 2007) and its impact on the
accuracy of the F-factor value on the CL estimate, because DOC as a strong influence on the
acid/base status of the water. The models used to derive the CLs in the NCLD v.3.0 vary from
simple (e.g., SSWC or FAB) to more complex models (e.g., biogeochemical models); hence, we
intend to complete a comprehensive comparison of CL models to account for their differences
and biases.
4.5.2.1.2 Terrestrial
Given that we intend to update the soils acidification dataset and explore the use of
PROFILE for purposes of estimating BCw rates, the level of uncertainty for this dataset it not yet
known. However, we provide a summary below of the uncertainties that will be considered
further in the REA. There are two main sources of uncertainty in the forest soil acidification
CLs: the Bc/Al ratio as a chemical indicator and the BCw rates as an input into the SMB model.
As mentioned earlier, the Bc/Al ratio is commonly used as an indicator because it relates well to
the Ca/Al ratio. It should be noted, however, that while soil Bc/Al and Ca/Al ratios are
recognized as good indicators of soil acidity risk in most U.S. forest ecosystems, these metrics
are not static and the responses of sensitive receptors to such changes are variable depending on
many outside factors. For example, the ratios themselves change depending on spatial scale and
seasonal fluctuations in soil conditions. In addition, the methods for both sampling and
4-28
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
laboratory analyses of base cations in soils can be inconsistent between studies as reported in
Appendix 5 of the second draft ISA. For example, Bc/Al or Ca/Al ratios reported in studies are
often based on seedling studies in controlled environments, and these responses are often less
consistent for seedlings or trees growing outside such controlled environments.
As mentioned in Chapter 2, the 2009 NOx/SOx REA included the use of the clay
substrate method for calculating BCw rates. The estimated BCw rates were identified in that
review as a major source of uncertainty. Given the results of Phelan et al. (2014) in the
Pennsylvania case study, we intend to explore the use of PROFILE modelling to obtain more
accurate national estimates of BCw rates for the REA, considering uncertainties identified in
Whitfield et al. (2018).
4.5.3 Exposure-Response Functions
The studies that have developed exposure-response functions (Horn et al. in review,
Nanus et al. 2017, Geiser et al. 2010, Root et al. 2015, Simkin et al. 2016) all include some
analysis of uncertainty and variability. For example, the functions were developed using air
quality estimates averaged across several years. This potentially misses inter-annual variability in
atmospheric deposition and introduces some uncertainty into the exposure -response function.
Also, the functions in Horn et al (in review) were developed across the entire range of a given
tree species. This potentially misses within species variability, which we intend to assess for a
few important species. Finally, like CLs, an important component of uncertainty related to air
quality with respect to exposure-response functions will be in the comparison of deposition
surfaces developed in the REA to exposure-response functions developed with different air
quality and how the uncertainties inherent in both the exposure-response function and the
deposition surface relate to each other. Given these uncertainties and limitations for the
exposure-response datasets, we will further consider their potential impact on the risk and
exposure results in the REA.
4-29
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
5 REFERENCES
Abbott, J., Hayman, G., Vincent, K., Metcalfe, S., Dore, T., Skeffington R. (2003). Uncertainty
in acid deposition modelling and critical load assessments, R&D Technical Report P4-
083(5)/l, AEA Technology pic, Culham Science Centre, Abingdon, Oxfordshire, OX13
3ED.
Aldous, AR. (2002). Nitrogen translocation in Sphagnum moss: effects of atmospheric nitrogen
deposition. New Phytologist 156: 241-253.
Appel, KW; Foley, KM; Bash, JO; Pinder, RW; Dennis, RL; Allen, DJ; Pickering, K. (2011). A
multi-resolution assessment of the Community Multiscale Air Quality (CMAQ) model
v4.7 wet deposition estimates for 2002 — 2006. Geoscientific Model Development 4: 357-
371.
Baldigo, B.P., Kulp, M.A., Schwartz, J.S. (2018). Relationships between indicators of acid-base
chemistry and fish assemblages in streams of the Great Smoky Mountains National Park.
Ecol Indicators 88: 465-484.
Banzhaf, HS, Burtraw, D., Evans, D., Krupnick, AJ. (2006). Valuation of natural resource
improvements in the Adirondacks. Land Econ 82: 445-464.
Baron, J.S., Driscoll, C.T., Stoddard, J.L., Richer, E.E. (2011). Empirical Critical Loads of
Atmospheric Nitrogen Deposition for Nutrient Enrichment and Acidification of Sensitive
US Lakes. Bioscience, 61(8): 602-613.
Bell, M. D., Phelan, J., Blett, T. F., Landers, D., Nahlik, A. M., Van Houtven, G., Davis, C.,
Clark, C. M, Hewitt, J. (2017). A framework to quantify the strength of ecological links
between an environmental stressor and final ecosystem services. Ecosphere 8(5).
Blett, T.F., Lynch, J.A., Pardo, L.H., Huber, C., Haeuber, R., Pouyat, R. (2014). FOCUS: A pilot
study for national-scale critical loads development in the United States. Environ Sci Pol
38: 225-236.
Bricker, S., Longstaff, B., Dennison, W., Jones, A., Boicourt, K., Wicks, C., Woerner, J. (2007).
Effects of nutrient enrichment in the nation's estuaries: A decade of change: National
estuarine eutrophication update. Silver Spring, MD: National Centers for Coastal Ocean
Science.
Caffrey, J.M., Murrell, M.C., Wigand, C., McKinney, R. (2007). Effect of nutrient loading on
biogeochemical and microbial processes in a New England salt marsh. Biogeochemistry
82: 251-264.
Clark, C.M., Cleland, E.E., Collins, S.L., Fargione, J.E., Gough, L., Gross, K.L., Pennings, S.C.,
Suding, K.N., Grace, J.B. (2007). Environmental and plant community determinants of
species loss following nitrogen enrichment. Ecol Lett 10: 596-607.
5-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Clark, C. M., Bell, M. D., Boyd, J. W., Compton, J. E., Davidson, E. A., Davis, C., Fenn, M. E.,
Geiser, L., Jones, L., Blett, T. F. (2017). Nitrogen-induced terrestrial eutrophication:
cascading effects and impacts on ecosystem services. Ecosphere 8(7).
Conley, D.J., Paerl, H.W., Howarth, R.W., Boesch, D.F., Seitzinger, S.P., Havens, K.E., Lancelot,
C., Likens, G.E. (2009). Controlling eutrophication: Nitrogen and phosphorus. Science
323:1014-1015.
Cosby, B.J., Wright R.F., Hornberger G.M., and Galloway J.N. (1985a). Modelling the effects of
acid deposition: Assessment of a lumped parameter model of soil water and streamwater
chemistry. Water Resources Research 21:51-63.
Cosby, B.J., Wright R.F., G.M. Hornberger, J.N. Galloway. (1985b). Modelling the effects of acid
deposition: Estimation of long-term water quality responses in a small forested catchment.
Water Resources Research 21:1591-1601.
Cosby, B.J., Hornberger, G.M., Galloway, J.N., Wright, R.F. (1985c). Time scales of catchment
acidification: a quantitative model for estimating freshwater acidification. Environ Sci
Technol, 19, 1144-1149.
Cosby, B. J., Ferrier, R. C., Jenkins, A., Wright, R. F. (2001) Modelling the effects of acid
deposition: refinements, adjustments and inclusion of nitrogen dynamics in the MAGIC
model. Hydrology and Earth System Sciences Discussions, Copernicus Publications, 5 (3),
pp.499-518.
Cronan, C.S., Grigal, D.F. (1995). Use of calcium/aluminum ratios as indicators of stress in forest
ecosystems. J Environ Qual 24: 209-226.
Crumley, K.M., Teece, M.A., Crandall, J.B., Sauer, A.K., Driscoll, C.T. (2016) Effects of
nitrogen deposition on nitrogen acquisition by Sarraceniapurpurea in the Adirondack
Mountains, New York, USA. Jr Torrey Botanical Soc 143: 8-20.
Dietze, M.C., Moorcroft, P.R. (2011). Tree mortality in the eastern and central United States:
Patterns and drivers. Global Change Biol 17: 3312-3326.
Driscoll, C.T., Lawrence, G.B., Bulger, A.J., Butler, T.J., Cronan, C.S., Eagar, C., Lambert, K.F.,
Likens, G.E., Stoddard, J.L., Weathers, K.C. (2001). Acidic deposition in the
northeastern United States: Sources and inputs, ecosystem effects, and management
strategies. Bioscience 51: 180-198.
Driscoll, C.T., Driscoll, K.M., Fakhraei, H., Civerolo, K. (2016). Log-term temporal trends and
spatial patterns in the acid-base chemistry of lakes in the Adirondack region of New York
in response to decreases in acidic deposition. Atmos Environ 146:5-14.
Duarte, N., Pardo, L.H., Robin-Abbott, M.J. (2013). Susceptibility of forests in the northeastern
USA to nitrogen and sulfur deposition: Critical load exceedance and forest health. Water
Air Soil Pollut 224: 1355.
5-2
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Dupont, J., Clair, T.A., Gagnon, C., Jeffries, D.S., Kahl, J.S., Nelson, S.J., Peckenham, J.M.
(2005). Estimation of critical loads of acidity for lakes in northeastern United States and
eastern Canada. Environ Monit Assess 109:275-291.
Eilers, J.M., Kanciruk, P., McCord, R.A., Overton, W.S., Hook, L., Blick, D.J., Brakker, D.F.,
Kellar, P.E., DeHaan, M.S., Silverstein, M.E., Landers, D.H. (1987). Characteristics of
Lakes in the Western United States. Volume II: Data Compendium of Selected Physical
and Chemical Variables. US Environmental Protection Agency. Report no. EPA-600/3-
86/054b.
Elser, J.J., Andersen, T., Baron, J.S., Bergstrom, A.K., Jansson, M., Kyle, M., Nydick, K.R.,
Steger, L., Hessen, D.O. (2009a). Shifts in lake N:P stoichiometry and nutrient limitation
driven by atmospheric nitrogen deposition. Science, 326(5954): 835-837.
Elser, J.J., Kyle, M., Steger, L., Nydick, K.R., Baron, J.S. (2009b). Nutrient availability and
phytoplankton nutrient limitation across a gradient of atmospheric nitrogen deposition.
Ecology 90: 3062-3073.
Eshleman, K.N., Sabo, R.D., Kline, K.M. (2013). Surface water quality is improving due to
declining atmospheric N deposition. Environ Sci Technol 47: 12193-12200.
Fakhraei, H., Driscoll, C.T., Selvendiran, P., DePinto, J.V., Bloomfield, J., Quinnc, S., Rowell, C.
(2014). Development of a total maximum daily load (TMDL) for acid-impaired lakes in
the Adirondack region of New York. Atmospheric Environment, 95: 277-287.
Fakhraei, H., Driscoll, C.T., Kulp, M.A., Renfro, J.R., Blett, T.F., Brewer, P.F., Schwartz J.S.
(2016). Critical loads and exceedances for nitrogen and sulfur atmospheric deposition in
Great Smoky Mountains National Park, United States. Ecosphere 7(10): 1-28.
Fakhraei, H., Driscoll, C.T., Kulp, M.A., Renfro, J.R., Blett, T,F„ Brewer, P.F., Schwartz J.S.
(2017). Sensitivity and uncertainty analysis of PnET-BGC to inform the development of
Total Maximum Daily Loads (TMDLs) of acidity in the Great Smoky Mountains
National Park. Environmental Modelling & Software 95: 156-167.
Fenn, M.E., Jovan, S., Yuan, F., Geiser, L., Meixner, T., Gimeno, B.S. (2008). Empirical and
simulated critical loads for nitrogen deposition in California mixed conifer forests.
Environ Pollut 155: 492-511.
Finlay, J.C., Small, G.E., Sterner, R.W. (2013). Human influences on nitrogen removal in lakes.
Science 342: 247-250.
Forsius, M., Posch, M., Aherne, J., Reinds, G.J., Christensen, J., Hole, L. (2010). Assessing the
impacts of long-range sulfur and nitrogen deposition on arctic and sub-arctic ecosystems.
Ambio 39: 136-147.
Geiser, L.H., Jovan, S.E., Glavich, D.A., Porter, M.K. (2010). Lichen-based critical loads for
atmospheric nitrogen deposition in Western Oregon and Washington Forests, USA.
Environ Pollut 158: 2412-2421.
5-3
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Gotelli, N.J, Ellison, A.M. (2002). Nitrogen deposition and extinction risk in the northern pitcher
plant, Sarraceniapurpurea. Ecology 83: 2758-2765.
Gotelli, N.J, Ellison, A.M. (2006). Forecasting extinction risk with nonstationary matrix models.
Cont Stat and Ecol 16: 51-61.
Grantz, E.M., Haggard, B.E., Scott, J.T. (2014). Stoichiometric imbalance in rates of nitrogen and
phosphorus retention, storage, and recycling can perpetuate nitrogen deficiency in highly-
productive reservoirs. Limnol Oceanogr 59: 2203-2216.
Greaver, T.L., Liu, L., Bobbink, R. (2011). Wetlands. In Assessment of nitrogen deposition
effects and empirical critical loads of nitrogen for ecoregions of the United States (General
Technical Report NRS-80) (pp. 193-208). Newtown Square, PA: U.S. Department of
Agriculture, Forest Service, Northern Research Station.
Greaver, T.L., Sullivan, T.J., Herrick, J.D., Barber, M.C., Baron, J.S., Cosby, B.J., Deerhake,
M.E., Dennis, R.L., Dubois, J.B., Goodale, C.L., Herlihy, A.T., Lawrence, G.B., Liu, L.,
Lynch, J. A., Novak, K.J. (2012). Ecological effects of nitrogen and sulfur air pollution in
the US: What do we know? Front Ecol Environ 10: 365-372.
Hautier, Y., Niklaus, P. A., Hector, A. (2009). Competition for light causes plant biodiversity loss
after eutrophication. Science 324: 636-638.
Henriksen, A. and Posch, M. (2001). Steady-state models for calculating critical loads of acidity
for surface waters. Water, Air Soil Pollut: Focus, 1, 375-398.
Hernandez, D.L., Vallano, D.M., Zavaleta, E.S., Tzankova, Z., Pasari, J.R., Weiss, S., Selmants,
P.C., Morozumi, C. (2016). Nitrogen pollution is linked to US listed species declines.
Bioscience 66: 213-222.
Heywood, L., Skeffington, R., Whitehead, P., Reynolds, B. (2007). Water Comparison of
Critical Load Exceedance and Its Uncertainty Based on National and Site-specific Data
Air Soil Pollut: Focus 7:391-397.
Homyak, P.M., Sickman, J.O., Miller, A.E., Melack, J.M., Meixner, T., Schimel, J.P. (2014).
Assessing nitrogen-saturation in a seasonally dry Chaparral watershed: Limitations of
traditional indicators of N-saturation. Ecosystems 17: 1286-1305.
Horn, K.J., Thomas, R.Q., Clark, C.M., Pardo, L.H., Fenn, M.E., Lawrence, G.B., Perakis, S.S.,
Smithwick, E.A.H., Baldwin, D., Braun, S., Nordin, A., Perry, C.H., Phelan, J.N.,
Schaberg, P.G., St. Clair, S.B., Warby, R., Watmough, S. Growth and survival
relationships of 94 tree species to nitrogen and sulfur deposition across the conterminous
U.S. Submitted to PLoS One, July 2018.
Howarth, R.W., Boyer, E.W., Pabich, W.J., Galloway, J.N. (2002). Nitrogen use in the United
States from 1961-2000 and potential future trends. Ambio 31: 88-96.
5-4
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Hundey, E.J., Moser, K.A., Longstaffe, F.J., Michelutti, N, Hladyniuk, R. (2014). Recent changes
in production in oligotrophic Uinta Mountain lakes, Utah, identified using
paleolimnology. Limnol Oceanogr 59: 1987-2001.
Irvine, I. C., Greaver, T., Phelan, J., Sabo, R. D., Van Houtven G. (2017). Terrestrial acidification
and ecosystem services: effects of acid rain on bunnies, baseball, and Christmas trees.
Ecosphere 8(6).
Koo, B., Piyachaturawat, P., Morris, R., Knipping, E. (2012) Evaluation of the Variability in
Chemical Transport Model Performance for Deposition and Ambient Concentrations of
Nitrogen and Sulfur Compounds. Atmosphere, 3(3), 400-418.
Koo. B., Kumar, N., Knipping, E., Nopmongcola, U., Sakulanontvittvaa, T., Odmanc, M.T.,
Russell, A., Yarwooda, G. (2015) Chemical transport model consistency in simulating
regulatory outcomes and the relationship to model performance. Atmospheric
Environment, 116, pp 159-171.
Landers, D.H., Nahlik, A.M. (2013). Final ecosystem goods and services classification system
(FEGS-CS) [EPA Report], (EPA/600/R-13/ORD-004914). Washington, DC: U.S.
Environmental Protection Agency.
Latimer, J.S., Charpentier, M.A. (2010). Nitrogen inputs to seventy-four southern New England
estuaries: Application of a watershed nitrogen loading model. Estuar Coast Shelf Sci 89:
125-136.
Latimer, J.S. Rego, S.A. (2010). Empirical relationship between eelgrass extent and predicted
watershed-derived nitrogen loading for shallow New England estuaries. Estuar Coast
Shelf Sci 90: 231-240.
Lawrence, G.B., Sullivan, T.J., Burns, D.A., Bailey, S.A., Cosby, B.J., Dovciak, M., Ewing,
H.A., McDonnell, T.C., Minocha, R., Quant, J., Rice, K.C., Siemion, J., Weathers, K.C.
(2015). Acidic deposition along the Appalachian Trail corridor and its effects on acid-
sensitive terrestrial and aquatic resources: Results of the Appalachian Trail atmospheric
deposition effects MEGA-transect study, Natural Resource Report
NPS/NRSS/ARD/NRR—2015/996. National Park Service, Fort Collins, Colorado, 241 p.
Layer, K., Hildrew, A.G., Woodward, G. (2013). Grazing and detritivory in 20 stream food webs
across abroad pHgradient. Oecologia 171: 459-471.
Lee, H.-M., Paulot, F., Henze, D.K., Travis, K., Jacob, D.J., Pardo, L.H., Schicthel, B.A. (2016)
Sources of nitrogen deposition in Federal Class I areas in the U.S. Atmos. Chem. Phys.,
16,525-540
Lewis, W.M., Wurtsbaugh, W.A., Paerl, H.W. (2011). Rationale for control of anthropogenic
nitrogen and phosphorus to reduce eutrophication of inland waters. Environ Sci Technol
45: 10300-10305.
5-5
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Li, H., McNulty, S.G. (2007). Uncertainty analysis on simple mass balance model to calculate
critical loads for soil acidity. Environ Pollut 149: 315-326.
Lieb, A.M., Darrouzet-Nardi, A., Bowman, W.D. (2011). Nitrogen deposition decreases acid
buffering capacity of alpine soils in the southern Rocky Mountains. Geoderma 164: 220-
224.
Linthurst, R.A., D.H. Landers, J.M. Eilers, D.F. Brakke, W.S. Overton, E.P. Meier and R.E.
Crowe. (1986). Characteristics of Lakes in the Eastern United States. Volume 1.
Population Description and Physico-Chemical Relationships. US Environmental
Protection Agency. Report no. EPA/600/4-86/007a.
Lopez, C.B., Jewett, E.B., Dortch, Q., Walton, B.T., Hudnell, H.K. (2008). Scientific assessment
of freshwater harmful algal blooms. Washington, DC: Interagency Working Group on
Harmful Algal Blooms, Hypoxia, and Human Health of the Joint Subcommittee on Ocean
Science and Technology.
McCrackin, M.L., Harrison, J.A., Compton, J.E. (2013). A comparison ofNEWS and SPARROW
models to understand sources of nitrogen delivered to US coastal areas. Biogeochemistry
114: 281-297.
McCune, B., Geiser, L. (1997). Macrolichens of the Pacific Northwest. In B McCune; L Geiser
(Eds.). Corvallis, OR: Oregon State University Press, USDA Forest Service.
McDonnell, T.D., Sullivan, T.J., Hessburg, P.F., Reynolds, K.M., Povak, N.A., Cosby, B.J.,
Jackson, W., Salter, R.B. (2014). Steady-state sulfur critical loads and exceedances for
protection of aquatic ecosystems in the U.S. southern Appalachian Mountains. Journal of
Environmental Management 146 (2014): 407-419.
McNulty, S.G., Boggs, J.L. (2010). A conceptual framework: Redefining forest soil's critical
acid loads under a changing climate. Environ Pollut 158: 2053-2058.
McNulty, S.G., Cohen, E.C., Myers, J.A.M., Sullivan, T.J., Li, H. (2007). Estimates of critical
acid loads and exceedances for forest soils across the conterminous United States. Environ
Pollut 149: 281-292.
McNulty, S.G., Boggs, J., Aber, J.D., Rustad, L., Magill, A. (2005). Red spruce ecosystem level
changes following 14 years of chronic N fertilization. For Ecol Manage 219: 279-291.
MEA (Millennium Ecosystem Assessment). (2005). Ecosystem and Human Well-Being Volume
1: Current State and Trends: Findings of the Condition and Trends Working Group. In
Hassan R., Scholes R.J., Ash N., (Eds), Millennium Ecosystem Assessment. Island Press,
Washington DC, US. 917 pp.
Monteith, D.T., Stoddard, J.L., Evans, C.D., de Wit, H.A., Forsius, M., H0gasen, T., Wilander,
A., Skjelkvale, B.L., Jeffries, D.S., Vuorenmaa, J., Keller, B., Kopacek, J., Vesely, J.
(2007). Dissolved organic carbon trends resulting from changes in atmospheric
deposition chemistry. Nature 450: 537-540.
5-6
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Morgan, M. and Henri on, M. (1990). Uncertainty: A guide dealing with uncertainty in
quantitative risk and policy analysis. Cambridge University Press, Cambridge, United
Kingdom. 332pp.
Nanus, L., Clow, D.W., Saros, J.E., Stephens, V.C., Campbell, D.H. (2012). Mapping critical
loads of nitrogen deposition for aquatic ecosystems in the Rocky Mountains, USA.
Environ Pollution 166: 125-135.
Nanus, L., McMurray, J.A., Clow, D.W., Saros, J.E., Blett, T., Gurdak, J.J. (2017). Spatial
variation of atmospheric nitrogen deposition and critical loads for aquatic ecosystems in
the Greater Yellowstone Area. Environmental Pollution 223: 644-656.
Nasr, M., Castonguay, M., Ogilvie, J., Raymond, B.A., Arp, P.A. (2010). Modelling and
mapping critical loads and exceedances for the Georgia Basin, British Columbia, using a
zero base-cation depletion criterion. Journal of Limnology 69: 181-192.
Nilsson, J., Grennfelt, P. (1988). Critical loads for sulphur and nitrogen: Report from a workshop
held at Skokloster, Sweden, 19-24 March 1988. Skokloster, Sweden: Nordic Council of
Ministers: Copenhagen.
Newingham, B.A., Vanier, C.H., Charlet, T.N., Smith, S.D. (2012). Effects of enhanced summer
monsoons, nitrogen deposition and soil disturbance on Larrea tridentata productivity and
subsequent herbivory in the Mojave Desert. J Arid Environ 87: 19-28.
Ochoa-Hueso, R., Manrique, E. (2010). Nitrogen fertilization and water supply affect
germination and plant establishment of the soil seed bank present in a semi-arid
Mediterranean scrubland. Plant Ecol 210: 263-273.
O'Dea, C. B., Anderson, S., Sullivan, T., Landers, D., Casey, C. F. (2017). Impacts to ecosystem
services from aquatic acidification: using FEGS-CS to understand the impacts of air
pollution. Ecosphere 8(5):e01807. 10.1002/ecs2.1807.
Omernik, J.M. (1987). Ecoregions of the conterminous United States. Map (scale 1:7,500,000).
Annals of the Association of American Geographers 77(1): 118-125.
Padgett, P.E., Parry, S.D., Bytnerowicz, A., Heath, R.L. (2009). Image analysis of epicuticular
damage to foliage caused by dry deposition of the air pollutant nitric acid. J Environ
Monit 11: 63-74.
Paerl, H.W., Hall, N.S., Peierls, B.L., Rossignol, K.L. (2014). Evolving paradigms and challenges
in estuarine and coastal eutrophication dynamics in a culturally and climatically stressed
world. Estuaries Coasts 37: 243-258.
Pardo, L.H., Geiser, L.H., Fenn, M.E., Driscoll, C.T., Goodale, C.L., Allen, E.B., Baron, J.S.,
Bobbink, R., Bowman, W.D., Clark, C.M., Emmett, B., Gilliam, F.S., Greaver, T.L.,
Hall, S.J., Lilleskov, E.A., Liu, L., Lynch, J.A., Nadelhoffer, K., Perakis, S.S., Robin-
Abbott, M.J., Stoddard, J.L., Weathers, K.C. (2011). Synthesis. In Assessment of
nitrogen deposition effects and empirical critical loads of nitrogen for ecoregions of the
5-7
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
United States (General Technical Report NRS-80) (pp. 229-284). (NRS-80). Newtown
Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station.
Phelan, J., Belyazid, S., Kurz, D., Guthrie, S., Cajka, J., Sverdrup, H., Waite, R. (2014).
Estimation of soil base cation weathering rates with the PROFILE model to determine
critical loads of acidity for forested ecosystems in Pennsylvania, USA: Pilot application
of a potential national methodology. Water Air Soil Pollut 225: 2109-2128.
Rapp, L. and Bishop, K. (2009). Surface water acidification and critical loads: exploring the F-
factor, Hydrol. Earth Syst. Sci., 13: 2191-2201.
Rao, L.E., Allen, E.B. (2010). Combined effects of precipitation and nitrogen deposition on
native and invasive winter annual production in California deserts. Oecologia 162: 1035-
1046.
Rao, L.E., Matchett, J.R., Brooks, M.L., Johnson, R.F., Minnich, R.A., Allen, E.B. (2015).
Relationships between annual plant productivity, nitrogen deposition and fire size in low-
elevation California desert scrub. International Journal of Wildland Fire 24: 48-58.
Rhodes, C., Bingham, A., Heard, A. M., Hewitt, J., Lynch, J., Waite, R., Bell M. D. (2017).
Diatoms to human uses: linking nitrogen deposition, aquatic eutrophication, and
ecosystem services. Ecosphere 8(7):e01858. 10.1002/ ecs2.1858.
Riddell, J; Nash, TH, III; Padgett, P. (2008). The effect of HN03 gas on the lichen Ramalina
menziesii. Flora 203: 47-54.
Riemann, Rachel, Barry Tyler Wilson, Andrew Lister, and Sarah Parks. (2010). An effective
assessment protocol for continuous geospatial datasets of forest characteristics using
USFS Forest Inventory and Analysis (FIA) data. Remote Sensing of Environment 114,
no. 10: 2337-2352.
Root, H.T., Geiser, L.H., Jovan, S., Neitlich, P. (2015). Epiphytic macrolichen indication of air
quality and climate in interior forested mountains of the Pacific Northwest, USA. Ecol
Indicat 53: 95-105.
Rueth, H, Baron, J.S. (2002). Differences in Englemann spruce forest biogeochemistry east and
west of the continental divide in Colorado, USA. Ecosystems 5: 45-57.
Saros, J.E., Rose, K.C., Clow, D.W., Stephens, V.C., Nurse, A.B., Arnett, H.A., Stone, J.R.,
Williamson, C.E., Wolfe, A.P. (2010). Melting Alpine glaciers enrich high-elevation
lakes with reactive nitrogen. Environ Sci Technol 44: 4891-4896.
Schwede, D.B., Lear, G.G. (2014). A novel hybrid approach for estimating total deposition in the
United States. Atmos Environ 92: 207-220.
Seinfeld, J.H., Pandis, S.N. (1998). Atmospheric chemistry and physics: From air pollution to
climate change. In Atmospheric Chemistry and Physics: From Air Pollution to Climate
Change. New York: John Wiley & Sons.
5-8
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Shaw, G.D., Cisneros, R., Schweizer, D., Sickman, J.O., Fenn, M.E._(2014). Critical loads of acid
deposition for wilderness lakes in the Sierra Nevada (California) estimated by the Steady-
State Water Chemistry model. Water Air Soil Pollut 225: 1804.
Sheibley, R.W., Enache, M., Swarzenski, P.W., Moran, P.W., Foreman, J.R. (2014). Nitrogen
deposition effects on diatom communities in lakes from three National Parks in
Washington state. Water Air Soil Pollut 225: 1857.
Sickles, J.E, II, Shadwick, D.S., Kilaru, J.V., Appel, K. (2013). "Transference ratios" to predict
total oxidized sulfur and nitrogen deposition - Part II, modeling results. Atmos Environ
77: 1070-1082.
Simkin, S.M., Allen, E.B., Bowman, W.D., Clark, C.M., Belnap, J., Brooks, M.L., Cade, B.S.,
Collins, S.L., Geiser, L.H., Gilliam, F.S., Jovan, S.E., Pardo, L.H., Schulz, B.K., Stevens,
C.J., Suding, K.N., Throop, H.L., Waller, D.M. (2016). Conditional vulnerability of plant
diversity to atmospheric nitrogen deposition across the United States. Proc Natl Acad Sci
USA 113: 4086-4091.
Skeffington, R., Hall, J., Heywood, L., Wadsworth, R., Whitehead, P., Reynolds, B. (2006).
Estimating uncertainty in critical load assessment models. [Acidification and annual
audits], Final report to the Environment Agency, R&D Project P4-120/4, CEH Monks
Wood.
Skiba, U., Drewer, J., Tang, Y.S., van Dijk, N., Helfter, C., et al. (2009). Biosphere-atmosphere
exchange of reactive nitrogen and greenhouse gases at the NitroEurope core flux
measurement sites: Measurement strategy and first data sets. Agriculture, Ecosystems &
Environment 133: 139-149.
Slemmons, K.E.H., Saros, J.E., Simon, K. (2013). The influence of glacial meltwater on alpine
aquatic ecosystems: A review. 15: 1794-1806.
Slemmons, K.E.H., Saros, J.E., Stone, J.R., McGowan, S., Hess, C.T., Cahl, D. (2015). Effects
of glacier meltwater on the algal sedimentary record of an alpine lake in the central US
Rocky Mountains throughout the late Holocene. J Paleolimnol 53: 385-399.
Sterner, R.W. 2008. On the Phosphorus Limitation Paradigm for Lakes. International Review of
Hydrobiology, 93(4-5): 433-445.
Suding, K.N., Collins, S.L., Gough, L., Clark, C., Cleland, E.E., Gross, K.L., Milchunas, D.G.,
Pennings, S. (2005). Functional- and abundance-based mechanisms explain diversity loss
due to N fertilization. Proc Natl Acad Sci USA 102: 4387-4392.
Sullivan, T.J., Cosby, B.J., Driscoll, C.T., McDonnell, T.C., Herlihy, A.T. (2011). Target loads of
atmospheric sulfur deposition protect terrestrial resources in the Adirondack Mountains,
New York against biological impacts caused by soil acidification. Journal of
Environmental Studies and Sciences 1: 301-314.
5-9
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Sullivan, T.J., Lawrence, G.B., Bailey, S.W., McDonnell, T.C., Beier, C.M., Weathers, K.C.,
McPherson, G.T., Bishop, D.A. (2013). Effects of acidic deposition and soil acidification
on sugar maple trees in the Adirondack Mountains, New York. Environ Sci Technol 47:
12687-12694.
Sullivan, T.J. (2015). Air pollutant deposition and effects on natural resources in New York
State. In Air Pollutant Deposition and Effects on Natural Resources in New York State.
Ithaca, NY: Comstock Publishing Associates.
Sverdrup, H., Warfvinge, P. (1993). The effect of soil acidification on the growth of trees, grass
and herbs as expressed by the (Ca+ Mg+ K)/A1 ratio (2nd ed.). Lund, Sweden: KF-
Sigma.
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. Nat Geosci 3: 13-17.
UNECE (United Nations Economic Commission for Europe). (2004). Manual on Methodologies
and Criteria for Modeling and Mapping Critical Loads and Levels and Air Pollution
Effects, Risks, and Trends. Convention on Long-Range Transboundary Air Pollution,
Geneva, Switzerland.
U.S. EIA (Energy Information Administration). (2018). Annual Energy Outlook 2018.
https://www.eia.gov/outlooks/aeo/pdf/AEO2018.pdf. Washington, DC; U.S. Energy
Information Administration; Office of Energy Analysis; U.S. Department of Energy.
U.S. EPA (Environmental Protection Agency). (2006). 2002 National Emissions Inventory: Data
and Documentation for the final 2002 point source national emissions inventory.
http://www.epa.gov/ttn/chief/net/2002inventory.html. Research Triangle Park, NC;
Emission Inventory and Analysis Group; Air Quality and Analysis Division; U.S.
Environmental Protection Agency.
U.S. EPA (U.S. Environmental Protection Agency). (2008). Integrated science assessment for
oxides of nitrogen and sulfur: Ecological criteria [EPA Report], (EPA/600/R-08/082F).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Research
and Development, National Center for Environmental Assessment- RTP Division.
U.S. EPA. (U.S. Environmental Protection Agency). (2009) Integrated Science Assessment (ISA)
for Particulate Matter. U.S. Environmental Protection Agency, Washington, DC,
EPA/600/R-08/139F.
U.S. EPA. (U.S. Environmental Protection Agency) (2009). Risk and exposure assessment for
review of the secondary National Ambient Air Quality Standards for oxides of nitrogen
and oxides of sulfur [EPA Report], (EPA/452/R-09/008A). Research Triangle Park, NC:
U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards,
Health and Environmental Impacts Division.
U.S. EPA. (U.S. Environmental Protection Agency). (2010a). Particulate Matter Urban-Focused
Visibility Assessment - Final Report. (EPA-452/R-10-004). Research Triangle Park, NC:
5-10
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards,
Health and Environmental Impacts Division.
U.S. EPA. (U.S. Environmental Protection Agency). (2010b). Quantitative Health Risk
Assessment for Particulate Matter. (EPA-452/R-10-005). Research Triangle Park, NC:
U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards,
Health and Environmental Impacts Division.
U.S. EPA. (U.S. Environmental Protection Agency). (2011a). Policy assessment for the review of
the secondary national ambient air quality standards for oxides of nitrogen and oxides of
sulfur [EPA Report], (EPA-452/R-1 l-004a). Washington, DC.
U.S. EPA. (U.S. Environmental Protection Agency) (201 lb). Policy Assessment for the Review
of the Particulate Matter National Ambient Air Quality Standards. (EPA-452/R-11-003).
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards, Health and Environmental Impacts Division.
U.S. EPA (U.S. Environmental Protection Agency (2016). Integrated Review Plan for the
National Ambient Air Quality Standards for Particulate Matter [EPA Report] (EPA-
452/R-16-005), Research Triangle Park, NC: U.S. Environmental Protection Agency,
Office of Air Quality Planning and Standards, Health and Environmental Impacts
Division.
U.S. EPA. (U.S. Environmental Protection Agency). (2017a). Integrated Science Assessment for
Oxides of Nitrogen, Oxides of Sulfur, and Particulate Matter— Ecological Criteria (First
External Review Draft) [EPA Report], (EPA/600/R-16/372). Research Triangle Park,
NC: U.S. Environmental Protection Agency, Office of Research and Development.
U.S. EPA. (U.S. Environmental Protection Agency). (2017b). Integrated Review Plan for the
Secondary National Ambient Air Quality Standards for Ecological Effects of Oxides of
Nitrogen, Oxides of Sulfur and Particulate Matter, (final draft) [EPA Report], (EPA-
452/R-17-002). Research Triangle Park, NC: U.S. Environmental Protection Agency,
Office of Research and Development and Office of Air Quality Planning and Standards.
U.S. EPA. (U.S. Environmental Protection Agency) (2017c). Policy Assessment for the Review
of the Primary National Ambient Air Quality Standards for Oxides of Nitrogen [EPA
Report], (EPA-452/R-17-003). Research Triangle Park, NC: U.S. Environmental
Protection Agency, Office of Air Quality Planning and Standards, Health and
Environmental Impacts Division.
U.S. EPA. (U.S. Environmental Protection Agency) (2018). Risk and exposure assessment for
review of the National Ambient Air Quality Standards for Sulfur Oxides [EPA Report],
(EPA-452/R-18-003). Research Triangle Park, NC: U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards, Health and Environmental
Impacts Division.
5-11
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Vourlitis, G.L., Zorba, G., Pasquini, S.C., Mustard, R. (2007). Carbon and nitrogen storage in
soil and litter of southern Californian semi-arid shrublands. Journal of Arid Environments
70:164-173.
Whitfield, C.J., Phelan, J.N., Buckley, J., Clark, C.M., Guthrie, S., Lynch, J.A. (2018).
Estimating Base Cation Weathering Rates in the USA: Challenges of Uncertain Soil
Mineralogy and Specific Surface Area with Applications of the PROFILE Model. Water,
Air, & Soil Pollution 229, no. 3: 61.
Wigand, C., McKinney, R.A., Charpentier, M.A., Chintala, M., Thursby ,G.B. (2003).
Relationships of nitrogen loadings, residential evelopment, and physical characteristics
with plant structure in New England salt marshes. Estuaries 26: 1494-1504.
Williams, J.J., Lynch, J. A., Saros, J.E., Labou, S.G. (2017). Critical loads of atmospheric N
deposition for phytoplankton nutrient limitation shifts in western U.S. mountain lakes.
Ecosphere 8(10).
Williams, J., Labou, S.G. (2017). A database of georeferenced nutrient chemistry data for
mountain lakes of the Western United States. Scientific Data 4:170069.
Wilson, B.T., Lister, A.J., Riemann, R.I., Griffith, D.M. (2013). Live tree species basal area of
the contiguous United States (2000-2009). Newtown Square, PA: USDA Forest Service,
Rocky Mountain Research Station.
World Health Organization. (2008). (2008). Uncertainty and data quality in exposure
assessment. International Programme on Chemical Safety. Geneva: World Health
Organization, http://vvvvvv.vvho.int/iris/handle/10665/44017
Xing, J., Pleim, J., Mathur, R., Pouliot, G., Hogrefe, C., Gan, C.-M., and Wei, C. (2013).
Historical gaseous and primary aerosol emissions in the United States from 1990 to 2010,
Atmos. Chem. Phys. 13: 7531-7549
Yu, X, -Y., Lee, T., Ayers, B., Kreidenweis, S.M., Malm, W., Collett, J.L., Jr. (2006). Loss of
fine particle ammonium from denuded nylon filters. Atmos Environ 40: 4797-4807.
Zhang, Y., Mathur, R., Bash, J.O., Hogrefe, C., Xing, J., Roselle, S.J. (2018). Long-term trends
in total inorganic nitrogen and sulfur deposition in the U.S. from 1990 to 2010. Atmos.
Chem. Phys. 18: 9091-9106.
5-12
-------
6 APPENDIX
Table 6-1. Results of Literature Search for Ecosystem Services Articles 2009 - 2016
Author
Short Title
Location Notes
Amiraslany et al. (2012)
Economic value of fire conditions and the effects of wildfire on
hiking
NM
Beier et al (2017)
Loss of ecosystem services due to chronic pollution
Adirondacks
Caputo et al. (2017)
Impacts of acidification and potential recovery recreational
fisheries
Adirondacks
Caputo et al. (2016)
Effects of soil acidification on long-term ecological and economic
outcomes
Adirondacks
Christman etal. (2014)
Willingness to Pay to Reduce Wild Fire Risk in Wildland-Urban
Interface
NV
Englin et al (2008)
Wildfire and the Economic Value of Wilderness Recreation.
CA
Holmes et al. (2013)
The effects of personal experience on preferences for wildfire
protection
FL
Huang et al. (2013)
Toward full economic valuation of forest fuels-reduction
treatments.
AZ
Kaval et al. (2008)
Using GIS to test homeowner willingness to pay to reduce wildfire
CO
Loomis et al. (2008)
Contingent valuation of fuel hazard reduction treatments.
CA, FL
Loomis et al. (2009)
Willingness to pay function for fuel treatments to reduce wildfire
acreage
CA, FL
Loomis et al. (2010)
Forest Service Use of Nonmarket Valuation in Fire Economics
US Wide
Loomis et al. (2015)
Are WTP Estimates for Wildfire Risk Reductions Transferrable
CA, FL
Mueller et al. (2009)
Hedonic Analysis of the Short and Long-Term Effects of Repeated
Wildfires
Southern CA
Sanchez et al. (2016)
Valuing hypothetical wildfire impacts on recreation demand.
Southern CA
Stetleret al. (2010)
The effects of wildfire and environmental amenities on property
values
MT
Azevedo et al. (2015)
Willingness to pay for Clear Lake cleanup.
Clear Lake, IA
Banzhaf et al. (2016)
Policy Analysis: Valuation of Ecosystem Services in the S.
Appalachian Mnts
Appalachians
Bin et al. (2013)
Impact of Measures of Water Quality on Coastal Waterfront
Property Values
South FL
Carteret al. (2012)
The economic value of catching and keeping or releasing
saltwater sport fish
Southeast US
Cho et al. (2011)
Negative externalities on property values resulting from water
impairment
NC
Egan et al. (2009)
Valuing Water Quality as a Function of Water Quality Measures
IA
Foster (2008)
Valuing preferences for water quality improvement
FL
Freeman et al (2008)
Statistical analysis of drinking water treatment plant costs
Worldwide
Guignetet al. (2014)
The Implicit Price of Aquatic Grasses
Chesapeake Bay
Hindlesy et al. (2011)
Addressing onsite sampling in recreation site choice models
Southeastern US
Huang et al. (2010)
Quantifying the economic effects of hypoxia on a fishery for brown
shrimp
NC
6-1
-------
Author
Short Title
Location Notes
Huang et al. (2011)
Management of an annual fishery in the presence of ecological
stress
Southeast
Jackson et al. (2012)
Economic value of stream degradation across the central
Appalachians
MD, PA,VA,WV
Jacobsen et al. (2009)
Income Effects on Global Willingness to Pay for Biodiversity
Conservation
Worldwide
Ji et al. (2016)
Water-based Recreation/Water Quality Indices Revealed
Preference Approach
IA
Keeleret al. (2013)
Advances in measuring the value of water quality to people
MN, IA
Keeleret al. (2015)
Recreational demand for clean water
MN, IA
Liao et al. (2016)
The Effects of Ambient Water Quality and watermilfoil on Property
Values
ID
Liu et al. (2014)
Estimating the impact of water quality on surrounding property
values
OH
Martin-Lopez (2008)
Economic Valuation of Biodiversity Conservation: the Meaning of
Numbers
Worldwide
Melstrom etal. (2013)
Valuing recreational fishing in the Great Lakes
Great Lakes
Meyer (2012)
Intertemporal valuation of river restoration
MN
Moore et al. (2015)
A stated preference study of the Chesapeake Bay and watershed
lakes
Chesapeake Bay
Morgan et al. (2010)
Water Quality and Residential Property Values: A Natural
Experiment
VA
Nelson et al. (2015)
Estimate the total economic value of improving water quality
UT
Netusil et al. (2014)
Valuing water quality in urban watersheds
OR and WA
Ramachandran et al.
(2015)
Validating Spatial Hedonic Modeling with a Behavioral Approach
Cape Cod, MA
Richardson et al. (2009)
The total economic value of threatened, endangered and rare
species
Worldwide
Roberts et al. (2008)
Preferences for environmental quality under uncertainty
Tulsa, OK
Tuttleetal. (2011)
Hedonic Studies: Valuing Environmental Amenities in Northern
New York.
Northern NY
Tuttleetal. (2015)
A hedonic analysis of lake water quality in the Adirondacks
Adirondacks
Van Houtven et al.
(2014)
Expert elicitation and stated preference methods to value
ecosystem service
VA
Viscusi et al. (2008)
The economic value of water quality
US Wide
Walsh et al. (2010)
Spatial Extent of Water Quality Benefits in Urban Housing Markets
FL
Walsh et al. (2011)
The spatial extent of water quality benefits in urban housing
markets
FL
Walsh et al. (2015b)
Modeling the Property Price Impact of Water Quality
Chesapeake Bay
Welle etal. (2008)
Property owners' willingness to pay for restoring impaired waters
MN
Welle etal. (2011)
Property owners' willingness to pay for water quality
improvements
MN
Phaneufetal. (2013)
Measuring nutrient reduction benefits for policy analysis
Southeastern US
6-2
-------
FULL CITATIONS FROM Table 6-1
Amiraslany, A., et al. (2012). "Comparing the economic value of fire conditions and the effects
of wildfire on hiking in New-Mexico recreation sites using contingent the valuation
method and travel cost method." American-Eurasian Journal of Agricultural &
Environmental Sciences 12(9): 1196-1204.
Azevedo, C.D., Herriges Sr, J. A. and Kling, C.L., 2015. Willingness to pay for Clear Lake
cleanup. Iowa Ag Review, 7(3), p.2.
Banzhaf, H.S., Burtraw, D., Criscimagna, S.C., Cosby, B.J., Evans, D.A., Krupnick, A.J. and
Siikamaki, J.V., 2016. Policy Analysis: Valuation of Ecosystem Services in the Southern
Appalachian Mountains. Environmental science & technology, 50(6), pp.2830-2836.
Beier, C.M., Caputo, J., Lawrence, G.B. and Sullivan, T.J., 2017. Loss of ecosystem services due
to chronic pollution of forests and surface waters in the Adirondack region (USA).
Journal of Environmental Management, 191, pp. 19-27.
Bin, O., and Czajkowski, J.. 2013. The Impact of Technical and Non-technical Measures of
Water Quality on Coastal Waterfront Property Values in South Florida. Marine Resource
Economics, 28(1): 43-63.
Caputo, J., Beier, C.M., Fakhraei, H. and Driscoll, C.T., 2016. Impacts of acidification and
potential recovery on the expected value of recreational fisheries in Adirondack lakes
(USA). Environmental Science & Technology.
Caputo, J., Beier, C.M., Sullivan, T.J. and Lawrence, G.B., 2016. Modeled effects of soil
acidification on long-term ecological and economic outcomes for managed forests in the
Adirondack region (USA). Science of the Total Environment, 565, pp.401-411.
Carter, D.W. and Liese, C., 2012. The economic value of catching and keeping or releasing
saltwater sport fish in the Southeast USA. North American Journal of Fisheries
Management, 32(4), pp.613-625.
Cho, S.-H., Roberts, R.K., Kim S. G. 2011. Negative externalities on property values resulting
from water impairment: The case of the Pigeon River Watershed, Ecological Economics,
Volume 70, Issue 12, 15 October 2011, Pages 2390-2399, ISSN 0921-8009.
Christman, L., et al. (2014). "Willingness to Pay to Reduce Wild Fire Risk in Wild land-Urban
Interface: A Comparative Analysis of Public Programs and Private Actions." Agricultural
and Applied Economics Association>2014 Annual Meeting, July 27-29, 2014,
Minneapolis, Minnesota http://purl.umn.edu/170703: 23p.
Egan, K. J., Herriges, J. A., Kling, C. L. and Downing, .n Journal of Agricultural Economics,
Vol. 91, No. l,pp. 106-123.
Englin, J., et al. (2008). Wildfire and the Economic Value of Wilderness Recreation. The
Economics of Forest Disturbances: Wildfires, Storms, and Invasive Species. T. P.
6-3
-------
Holmes, J. P. Prestemon and K. L. Abt, Forestry Sciences series, vol. 79. New York:
Springer: 191-208.
Foster C. (2008). "Valuing preferences for water quality improvement in the Ichetucknee
Springs system: a case study from Columbia county, FL." Dissertation
Freeman et al. (2008). "Statistical analysis of drinking water treatment plant costs, source water
quality and land cover characteristics." United States Trust for Public Land
Guignet, D., Griffiths, C., Klemick, H. and Walsh, P., 2014. The Implicit Price of Aquatic
Grasses. US EPA National Center for Environmental Economics Working paper
Series, 6.
Hindsley, P., Landry, C.E. and Gentner, B., 2011. Addressing onsite sampling in recreation site
choice models. Journal of Environmental Economics and Management, 62(1), pp. 95-110.
Holmes, T. P., et al. (2013). "The effects of personal experience on choice-based preferences for
wildfire protection programs." International Journal of Wildland Fire 22(2): 234-245.
Huang L., Smith M.D. (2011) Management of an annual fishery in the presence of ecological
stress: The case of shrimp and hypoxia. Ecol Econ 70: 688-697.
Huang L., Smith M.D., Craig J.K. (2010) Quantifying the economic effects of hypoxia on a
fishery for brown shrimp Farfantepenaeus aztecus. Marine and Coastal Fisheries:
Dynamics, Management, and Ecosystem Science 2: 232-248.
Huang, C. H., et al. (2013). "Toward full economic valuation of forest fuels-reduction
treatments." Journal of Environmental Management 130: 221-231.
Jackson, L. E., Rashleigh, B., and McDonald, M. E. (2012). "Economic value of stream
degradation across the central Appalachians." The Journal of Regional Analysis &
Policy 42(3): 188-197.
Jacobsen, J. B., Hanley, N. (2009). "Are There Income Effects on Global Willingness to Pay for
Biodiversity Conservation? ." Environmental and resource economics 43(2): 137-160.
Ji, Y., Keiser, D. Water-based Recreation and Water Quality Indices: A Revealed Preference
Approach. No. 235886. Agricultural and Applied Economics Association, 2016.
Kaval, P., Loomis, J. (2008). Using GIS to test the relationship between homeowner willingness
to pay to reduce wildfire and landscape characteristics. Proceedings of the second
international symposium on fire economics, planning, and policy: a global view. Gen.
Tech. Rep. PSW-208. Albany, CA: US Department of Agriculture, Forest Service,
Pacific Southwest Research Station
Keeler, B.L., Wood, S.A., Polasky, S., Kling, C., Filstrup, C.T., Downing, J.A., 2015.
Recreational demand for clean water: evidence from geotagged photographs by visitors
to lakes. Frontiers in Ecology and the Environment, 13(2), pp.76-81.
6-4
-------
Keeler, B. L. Water and well-being: Advances in measuring the value of water quality to people.
Diss. University of Minnesota, 2013.
Liao, F.H., Wilhelm, F.M., Solomon, M., 2016. The Effects of Ambient Water Quality and
Eurasian Watermilfoil on Lakefront Property Values in the Coeur d'Alene Area of
Northern Idaho, USA. Sustainability, 8(1), p.44.
Liu, H., Gopalakrishnan, S., Browning, D., Herak, P., Sivandran, G. 2014. Estimating the impact
of water quality on surrounding property values in Upper Big Walnut Creek Watershed in
Ohio for dynamic optimal control. Agricultural and Applied Economics Association,
2014 Annual Meeting, July 27-29, 2014, Minneapolis, Minnesota.
Loomis, J., Gonzalez-Caban A. (2010). "Forest Service Use of Nonmarket Valuation in Fire
Economics: Past, Present, and Future." Journal of Forestry 108(8): 389-396.
Loomis, J., Gonzalez-Caban A (2008). Contingent valuation of fuel hazard reduction treatments.
The economics of forest disturbances: wildfires, storms and invasive species / edited by
Thomas P. Holmes, Jeffrey P. Prestemon, Karen L. Abt, Dordrecht; London : Springer,
2008: 229-243.
Loomis, J. B., et al. (2009). "Willingness to pay function for two fuel treatments to reduce
wildfire acreage burned: A scope test and comparison of White and Hispanic
households." Forest Policy and Economics 11(3): 155-160.
Loomis, J., et al. (2015). "Are WTP Estimates for Wildfire Risk Reductions Transferable from
Coast to Coast? Results of a Choice Experiment in California and Florida." Agricultural
and Applied Economics Association 2015 AAEA & WAEA Joint Annual Meeting, July
26-28, San Francisco, California http://purl.umn.edu/202661: 25p.
Martin-Lopez, B., et al. (2008). "Economic Valuation of Biodiversity Conservation: the Meaning
of Numbers " Conservation biology the journal of the Society for Conservation Biology
22(3): 624-635.
Melstrom, Richard T., Lupi, F. (2013) "Valuing recreational fishing in the Great Lakes." North
American Journal of Fisheries Management 33.6: 1184-1193.
Meyer, A. (2012). "Intertemporal valuation of river restoration." Environmental Resource
Economics. DOI 10.1007/sl0640-012-9580-4
Moore et al. (2015). "A stated preference study of the Chesapeake Bay and watershed lakes."
National Center for Environmental Economics Working Paper Series 15-06.
Morgan, O. A., Hamilton, S. E., Chung,V.. 2010. Water Quality and Residential Property
Values: A Natural Experiment Approach. Appalachian State University, Department of
Economics Working Paper.
Mueller, J., et al. (2009). "Do Repeated Wildfires Change Homebuyers' Demand for Homes in
High-Risk Areas? A Hedonic Analysis of the Short and Long-Term Effects of Repeated
6-5
-------
Wildfires on House Prices in Southern California." Journal of Real Estate Finance and
Economics 38(2): 155-172.
Nelson, N.M., Loomis, J.B., Jakus, P.M., Kealy, M.J., von Stackelburg, N., Ostermiller, J., 2015.
Linking ecological data and economics to estimate the total economic value of improving
water quality by reducing nutrients. Ecological Economics, 118, pp. 1-9.
Netusil, N. R., Kincaid, M, Chang, H. (2014). "Valuing water quality in urban watersheds: A
comparative analysis of Johnson Creek, Oregon, and Burnt Bridge Creek, Washington."
Water Resources Research 50(5): 4254-4268.
Phaneuf, D. J., von Haefen, R. H., Mansfield, C. A., Van Houtven, G. L. (2013). Measuring
nutrient reduction benefits for policy analysis using linked non-market valuation and
environmental assessment models: Final report on stated preference surveys. Prepared for
U.S. Environmental Protection Agency, Office of Water.
Ramachandran, M. (2015). Validating Spatial Hedonic Modeling with a Behavioral Approach:
Measuring the Impact of Water Quality Degradation on Coastal Housing Markets.
Richardson, L. A., et al. (2012). "The hidden cost of wildfires: Economic valuation of health
effects of wildfire smoke exposure in Southern California." Journal of Forest Economics
18(1): 14-35.
Richardson, L., Loomis, J. (2009). The total economic value of threatened, endangered and rare
species: an updated meta-analysis. Ecological Economics, 68(5), pp.1535-1548.
Richardson, L., et al. (2013). "Valuing Morbidity from Wildfire Smoke Exposure: A Comparison
of Revealed and Stated Preference Techniques." Land Economics 89(1): 76-100.
Roberts, D.C., Boyer, T.A., Lusk, J.L. (20080. Preferences for environmental quality under
uncertainty. Ecological Economics, 66(4), pp.584-593.
Sanchez, J. J., et al. (2016). "Valuing hypothetical wildfire impacts with a Kuhn-Tucker model
of recreation demand." Forest Policy and Economics 71: 63-70.
Stetler, K. M., et al. (2010). "The effects of wildfire and environmental amenities on property
values in northwest Montana, USA." Ecological Economics 69(11): 2233-2243.
Tuttle, C.M. and Heintzelman, M.D. (2015). A loon on every lake: A hedonic analysis of lake
water quality in the Adirondacks. Resource and Energy Economics, 39, pp. 1-15.
Tuttle, C.M., 2011. Hedonic Studies: Valuing Environmental Amenities in Northern New York.
Van Houtven, G., Mansfield, C., Phaneuf, D. J., von Haefen, R., Milstead, B., Kenney, M. A.,
Reckhow, K. H. (2014). Combining expert elicitation and stated preference methods to
value ecosystem services from improved lake water quality. Ecological Economics, 99,
40-52.
6-6
-------
Viscusi W.K., Huber J., Bell J. (2008) The economic value of water quality. Environ Resour
Econ 41: 169-187.
Walsh P. J., Milon J.W., Scrogin D.O. (2011) The spatial extent of water quality benefits in urban
housing markets. Land Econ 87: 628-644.
Walsh, P, et al. (2010). The Spatial Extent of Water Quality Benefits in Urban Housing Markets.
Working Paper Series. U.S. Environmental Protection Agency: National Center for
Environmental Economics, Washington, DC: 35.
Walsh, P., Griffiths, C., Guignet, D., Klemick, H. (2015). Modeling the Property Price Impact of
Water Quality in 14 Chesapeake Bay Counties Patrick Walsh, Charles Griffiths.
Welle, P. G., Hodgson, J. (2008). "Property owners' willingness to pay for water quality
improvements: contingent valuation estimates in two central Minnesota watersheds."
Journal of Applied Business and Economics 12(1): 81-94.
6-7
-------
United States Office of Air Quality Planning and Publication No. EPA-452/D-
Environmental Protection Standards 18-001
Agency Health and Environmental Impact Division August 2018
Research Triangle Park, NC
6-8
------- |