1 2 &EPA United! S*K Enviraimerilfll PiutmBmi Agamy 7 Policy Assessment for the Review of the s Secondary National Ambient Air Quality 9 Standards for NOX and SOX 10 11 12 First External Review Draft 13 ------- 1 EPA-452/P-10-006 2 March 2010 O 4 5 6 7 8 Policy Assessment for the Review of the 9 Secondary National Ambient Air Quality Standards 10 for NOX and SOX: 11 12 13 First External Review Draft 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 3 8 Office of Air and Radiation 39 Office of Air Quality Planning and Standards 40 Health and Environmental Impacts Division 41 Research Triangle Park, North Carolina 27711 42 43 44 ------- 1 DISCLAIMER 2 O 4 This document has been reviewed by the Office of Air Quality Planning and Standards, 5 U.S. Environmental Protection Agency (EPA), and approved for publication. This draft 6 document has been prepared by staff from the Office of Air Quality Planning and Standards, 7 U.S. Environmental Protection Agency. Any opinions, findings, conclusions, or 8 recommendations are those of the authors and do not necessarily reflect the views of the EPA. 9 Mention of trade names or commercial products is not intended to constitute endorsement or 10 recommendation for use. This document is being provided to the Clean Air Scientific Advisory 11 Committee for their review, and made available to the public for comment. Any questions or 12 comments concerning this document should be addressed to Dr. Bryan Hubbell, U.S. 13 Environmental Protection Agency, Office of Air Quality Planning and Standards, C504-02, 14 Research Triangle Park, North Carolina 27711 (email: hubbell.bryan@epa.gov ). 15 ------- 1 TABLE OF CONTENTS 2 List of Figures iv 3 List of Tables vii 4 List of Acronyms and Abbreviations ix 5 List of Key Terms xii 6 1. Introduction 1 7 1.1 Definitions of NOX and SOX for this Assessment 3 8 1.2 Policy Objectives 4 9 1.3 Critical Policy Elements 6 10 1.4 Historical Context 8 11 1.4.1 History of NOX and SOXNAAQS Review 8 12 1.4.2 History of Related Assessments and Agency Actions 10 13 1.5 Proposed Conceptual Framework for Combined NOX SOX Standards 13 14 1.6 Policy Relevant Questions 16 15 2. Known or anticipated ecological effects 22 16 2.1 Acidification: Evidence of effects on structure and function of terrestrial and 17 freshwater ecosystems 23 18 2.1.1 What is the nature of acidification related ecosystem responses to 19 reactive nitrogen and/ sulfur deposition? 24 20 2.1.2 What types of ecosystems are sensitive to such effects? In which ways 21 are these responses affected by atmospheric, ecological, and landscape 22 factors? 26 23 2.1.3 What is the magnitude of ecosystem responses to acidifying deposition? 26 24 2.1.4 What are the key uncertainties associated with acidification? 35 25 2.2 Nitrogen enrichment: Evidence of effects on structure and function of terrestrial 26 and freshwater ecosystems 37 27 2.2.1 What is the nature of terrestrial and freshwater ecosystem responses to 28 reactive nitrogen and/ sulfur deposition? 37 29 2.2.2 What types of ecosystems are sensitive to such effects? How are these 30 responses affected by atmospheric, ecological, and landscape factors 39 31 2.2.3 What is the magnitude of ecosystem responses to nitrogen deposition? 40 32 2.2.4 What are the key uncertainties associated with nutrient enrichment? 48 33 2.3 What Ecological effects are associated with gas-phase NOX and SOX? 49 34 2.3.1 What is the nature of ecosystem responses to gas-phase nitrogen and 35 sulfur? 50 36 2.3.2 What types of ecosystems are sensitive to such effects? How are these 37 responses affected by atmospheric, ecological, and landscape factors? 50 38 2.3.3 What is the magnitude of ecosystem responses to gas phase effects of 39 NOxandSOx? 51 40 2.4 Summary 51 41 2.5 References 52 March 2010 i Draft - Do Not Quote or Cite ------- 1 3. Considerations of Adversity to Public Welfare 62 2 3.1 How do we characterize adversity to public welfare? What are the relevant 3 factors and how are they addressed in this document? 62 4 3.1.1 What are the benchmarks for adversity from other sources? 62 5 3.1.2 Other EPA Programs and Federal Agencies 65 6 3.2 What are ecosystem services and how does this concept relate to public 7 welfare? 69 8 3.3 What is the role of economics? 75 9 3.4 What is the evidence for effects on ecosystem services? How do we link 10 ecological indicators to services? 78 11 3.5 References 89 12 4 Addressing the Adequacy of the Current Standards 101 13 4.1 Are the structures of the current NOX and SOX secondary standards based on 14 relevant ecological indicators such that they are adequate to determine and 15 protect public welfare against adverse effects on ecosystems? 101 16 4.2 To what extent are the structures of the current NOX and SOX secondary 17 standards meaningfully related to relevant ecological indicators of public 18 welfare effects? 103 19 4.3 To what extent do current monitoring networks provide a sufficient basis for 20 determining the adequacy of current secondary NOX and SOX standards? 106 21 4.3.1 What does the NADP monitoring network provide and what are the 22 major limitations? Ill 23 4.3.2 How do we characterize deposition through Monitoring and Models? 112 24 4.4 What is our best characterization of atmospheric concentrations of NOy and 25 SOX, and deposition of N and S? 114 26 4.4.1 What are the current atmospheric concentrations of reactive nitrogen, 27 NOy, reduced nitrogen, NHX, sulfur dioxide, SO2, and sulfate, SO4? 115 28 4.5 Are adverse effects on the public welfare occurring under current air quality 29 conditions for NO2 and 802 and would they occur if the nation met the current 30 secondary standards? 130 31 4.5.1 To what extent do the current NOX and SOX secondary standards provide 32 protection from adverse effects associated with deposition of 33 atmospheric NOX, and SOX which results in acidification in sensitive 34 aquatic and terrestrial ecosystems? 133 35 4.5.2 To what extent does the current NOX secondary standard provide 36 protection from adverse effects associated with deposition of 37 atmospheric NOX, which results in nutrient enrichment effects in 38 sensitive aquatic and terrestrial ecosystems? 138 39 4.5.3 Aquatic Nutrient Enrichment 139 40 4.5.4 Terrestrial Nutrient Enrichment 141 41 4.6 To what extent do the current NOX and/or SOX secondary standards provide 42 protection from other ecological effects (e.g., mercury methylation) associated 43 with the deposition of atmospheric NOX, and/or SOX? 142 44 4.7 References 143 45 5. Conceptual Design of an Ecologically Relevant Multi-pollutant Standard 145 46 5.1 Components of the design 145 March 2010 ii Draft - Do No Quote or Cite ------- 1 5.1.1 For which effects is there sufficient information to support setting 2 standards? 146 3 5.2 Ecological Components of the Standard: Aquatic Acidification 147 4 5.2.1 Conceptual design considerations from the ISA and REA 149 5 5.2.2 Design options for aquatic acidification 157 6 5.3 Ecological Components of the Standard: Terrestrial Acidification, Terrestrial 7 Nutrient Enrichment and Surface water Nutrient Enrichment 167 8 5.3.1 Terrestrial Acidification 167 9 5.3.2 Terrestrial and surface water nutrient enrichment 168 10 5.3.3 Summary 169 11 5.4 Linking Deposition to Atmospheric Concentration 169 12 5.4.1 Background 169 13 5.4.2 Aggregation Issues 170 14 5.4.3 AirQuality Simulation Models 171 15 5.4.4 Oxidized Sulfur and Nitrogen Pollutant Species 172 16 5.4.5 Example Calculations 173 17 5.5 Example calculation for the conceptual design and derivation of AAPI 177 18 5.5.1 Example calculation for the conceptual design 177 19 5.5.2 Derivation of the Atmospheric Acidification Potential Index (AAPI): 185 20 5.6 References 188 21 6. Options for Elements of the Standard 190 22 6.1 What atmospheric indicators of oxidized nitrogen and sulfur are appropriate for 23 use in a secondary NAAQS that provides protection for public welfare from 24 exposure related to deposition of N and S? What averaging times and statistics 25 for such indicators are appropriate to consider? 191 26 6.2 What is the appropriate averaging time for the air quality indicators NOy and 27 SOX to provide protection of public welfare from adverse effects from 28 acidification? 193 29 6.3 What form(s) of the standard are most appropriate to provide protection of 30 sensitive ecosystems from the effects of acidifying deposition related to ambient 31 NOX and SOX concentrations? 194 32 6.4 What are the appropriate spatial extents of the boundaries for evaluating AAPI? 33 Within those boundaries, what are the appropriate statistics to use in calculating 34 the parameters of the AAPI, e.g. G, VNoy, Vs, and NHX? Within those 35 boundaries, what s the appropriate spatial averaging for the air quality indicators 36 NOy and SOX to provide protection of public welfare from adverse effects from 37 acidification? 203 38 6.5 What are the options for specifying the targets for the ecological indicator for 39 aquatic acidification? 203 40 6.5.1 What levels of impairment are related to alternative levels of ANC? 204 41 6.6 What are the appropriate ambient air monitoring methods to consider in 42 developing the standards? 208 43 6.6.1 What measurements would be used to characterize NOy and SOX 44 ambient air concentrations for the purposes of the AAPI based standard? 208 45 6.6.2 What sampling frequency would be required? 208 March 2010 iii Draft - Do No Quote or Cite ------- 1 6.6.3 What are the spatial scale issues associated with monitoring for 2 compliance, and how should these be addressed? 209 3 6.7 Taking into consideration information about ecosystem services and other 4 factors related to characterizing adversity for the ecological effects being 5 assessed in this review, what is an appropriate range of alternative standards for 6 the Agency to consider? 210 7 7. Co-protection for Other Effects Using Standards to Protect Against Acidification 213 8 7.1 To what extent would a standard specifically defined to protect against aquatic 9 acidification likely provide protection from terrestrial acidification? 213 10 7.2 To what extent would a standard specifically defined to protect against aquatic 11 acidification likely provide protection from terrestrial nutrient enrichment? 214 12 7.3 To what extent would a standard specifically defined to protect against aquatic 13 acidification likely provide protection from aquatic nutrient enrichment? 215 14 8. Consideration of Issues Regarding Reduced and Oxidized Forms of Nitrogen 216 15 9.1 Conclusions 219 16 9.2 Summary of key uncertainties and research recommendations related to setting 17 a secondary standard forNOx and SOX 223 18 9.2.1 Research Needs to Reduce Uncertainty in the Next Review (focused on 19 aquatic acidification) 223 20 9.2.2 Data Needs to Reduce Uncertainty in the Next Review (focused on 21 aquatic acidification) 223 22 23 LIST OF FIGURES 24 Figure 1-1. Framework of an alternative secondary standard 16 25 Figure 2-1. Ecological Effects Associated with Alternative Levels of Acid Neutralizing 26 Capacity (ANC) 28 27 Figure 2-2. Average NOs" concentrations (orange), SO42" concentrations (red), and ANC 28 (blue) across the 44 lakes in the Adirondack Case Study Area modeled 29 using MAGIC for the period 1850 to 2050 29 30 Figure 2-3. ANC concentrations of preacidification (1860) and current (2006) conditions 31 based on hindcasts of 44 lakes in the Adirondack Case Study Area 32 modeled using MAGIC. [Note: in this map, the symbol for red is 33 reversed and should be < 0. The figure will be revised in the next draft.] 30 34 Figure 2-4. Critical loads of acidifying deposition that each surface water location can 35 receive in the Adirondack Case Study Area while maintaining or 36 exceeding an ANC concentration of 50 ueq/L based on 2002 data. 37 Watersheds with critical load values <100 meq/m2/yr (red and orange 38 circles) are most sensitive to surface water acidification, whereas 39 watersheds with values >100 meq/m2/yr (yellow and green circles) are 40 the least sensitive sites 31 41 Figure 2-5. Average NOs" concentrations orange), SO42"concentrations (red), and ANC 42 (blue) levels for the 60 streams in the Shenandoah Case Study Area 43 modeled using MAGIC for the period 1850 to 2050 32 March 2010 iv Draft - Do No Quote or Cite ------- 1 Figure 2-6. ANC levels of 1860 (preacidification) and 2006 (current) conditions based on 2 hindcasts of 60 streams in the Shenandoah Case Study Area modeled 3 using MAGIC 33 4 Figure 2-7. Critical loads of surface water acidity for an ANC of 50 [j,eq/L for 5 Shenandoah Case Study Area streams. Each dot represents an estimated 6 amount of acidifying deposition (i.e., critical load) that each stream's 7 watershed can receive and still maintain a surface water ANC >50 ueq/L. 8 Watersheds with critical load values <100 meq/m2/yr (red and orange 9 circles) are most sensitive to surface water acidification, whereas 10 watersheds with values >100 meq/m2/yr (yellow and green circles) are 11 the least sensitive sites 34 12 Figure 2-8. Benchmarks of atmospheric nitrogen deposition for several ecosystem 13 indicators with the inclusion of the diatom changes in the Rocky 14 Mountain lakes (REA 5.3.1.2) 42 15 Figure 2-9 (from REA figure 5.3-9). Observed effects from ambient and experimental 16 atmospheric nitrogen deposition loads in relation to using CMAQ 2002 17 modeling results and NADP monitoring data. Citations for effect results 18 are from the ISA, Table 4.4 (U.S. EPA, 2008) 43 19 Figure 3-1. Common anthropogenic stressors and the essential ecological attributes they 20 affect. Modified from Young and Sanzone (2002) 64 21 Figure 3-2. Representation of the benefits assessment process indicating where some 22 ecological benefits may remain unrecognized, unquantified, or 23 unmonetized. (Source: EBASP USEPA 2006) 71 24 Figure 3-3. Conceptual model showing the relationships among ambient air quality 25 indicators and exposure pathways and the resulting impacts on 26 ecosystems, ecological responses, effects and benefits to characterize 27 known or anticipated adverse effects to public welfare. [This figure to be 28 revised for Second Draft Policy Assessment Document] 73 29 Figure 3-4. Locations of Eastern U.S. National Parks (Class I areas) relative to deposition 30 of Nitrogen and Sulfur in sensitive aquatic areas 74 31 Figure 3-5. Location of Western U.S. National Parks (Class I areas) relative to deposition 32 of Nitrogen and Sulfur 75 33 Figure 3-6. Conceptual model linking ecological indicator (ANC) to affected ecosystem 34 services 79 35 Figure 4-1. Routinely operating surface monitoring stations measuring forms of 36 atmospheric nitrogen 107 37 Figure 4-2. Routinely operating surface monitoring stations measuring forms of 38 atmospheric sulfur 108 39 Figure 4-3. Anticipated network of surface based NOy stations based on 2009 network 40 design plans. The NCore stations are scheduled to be operating by 41 January, 2011 110 42 Figure 4-4. Location of approximately 250 National Atmospheric Deposition Monitoring 43 (NADP) National Trends Network (NTN) sites illustrating annual 44 ammonium deposition for 2005. Weekly values of precipitation based 45 nitrate, sulfate and ammonium are provided by NADP 112 March 2010 v Draft - Do No Quote or Cite ------- 1 Figure 4-5. 2005 CMAQ modeled annual average NOy (ppb). These maps will be 2 replaced with full CONUS maps in the next draft 117 3 Figure 4-6. 2005 CMAQ modeled annual average total reduced nitrogen (NHX) (as ng/m3 4 nitrogen) 118 5 Figure 4-7. 2005 CMAQ modeled annual average ammonia, NHs, (as ng/m3 N) 119 6 Figure 4-8. 2005 CMAQ modeled annual average ammonia, NH4, (as ng/m3 N) 120 7 Figure 4-9. 2005 CMAQ modeled annual average SOX, (as ng/m3 S from SO2 and SO4) 121 8 Figure 4-10. 2005 CMAQ modeled annual average SO2 (as ng/m3 S) 122 9 Figure 4-11. 2005 CMAQ modeled annual average SO4 (as ng/m3 S) 123 10 Figure 4-12. 2005 annual average sulfur dioxide concentrations based on CASTNET 11 generated by the Visibility Information Exchange Web Sysytem 12 (VIEWS) 124 13 Figure 4-13. 2005 annual average sulfate concentrations based on CASTNET generated 14 by the Visibility Information Exchange Web Sysytem (VIEWS) 124 15 Figure 4-14. Annual average 2005 NOy concentrations from reporting stations in AQS 125 16 Figure 4-15. 2005 CMAQ modeled Oxidized Nitrogen Deposition (kgN/Ha/Yr) 126 17 Figure 4-16. 2005 CMAQ modeled Oxidized Sulfur Deposition (kgS/Ha/Yr) 127 18 Figure 4-17. Three hour average maximum 2005 SO2 concentrations based on the 19 SLAMS reporting to EPA's Air Quality System (AQS) data base. The 20 current SO2 secondary standard based on a the maximum 3 hour average 21 value is 500 ppb, a value not exceeded. While there are obvious spatial 22 gaps, the majority of these stations are located to capture maximum 23 values generally in proximity to major sources and high populations. 24 Lower relative values are expected in more remote acid sensitive areas 128 25 Figure 4-18. Annual average 2005 NO2 concentrations based on the SLAMS reporting to 26 EPA's Air Quality System (AQS) data base. The current NO2 secondary 27 standard is 53 ppb, a value well above those observed. While there are 28 obvious spatial gaps, the stations are located in areas of relatively high 29 concentrations in highly populated areas. Lower relative values are 30 expected in more remote acid sensitive areas 129 31 Figure 4-19. 2005 CMAQ derived annual average ratio of (NOy - NO2)/NOy. The 32 fraction of NO2 contributing to total NOy generally is less than 50% in 33 the Adirondack and Shenandoah case study areas. The ratio reflects the 34 relative air mass aging associated with transformation of oxidized 35 nitrogen beyond NO and NO2 as one moves from urban to rural 36 locations 130 37 Figure 4-20. National map highlighting the 9 case study areas evaluated in the REA 133 38 Fig 5-1. Schematic diagram of the conceptual design of the standard 146 39 Fig 5-2. Schematic diagram of the conceptual design of the standard based on aquatic 40 acidification. From left to right, if a desired level of ANC is known, then 41 the concentration of the atmospheric indicators that will cause that level 42 may be calculated. From right to left, if the if the concentration of the air 43 quality indicators are known than the ANC that will be caused may be 44 calculated 148 March 2010 vi Draft - Do No Quote or Cite ------- 1 Figure 5-3. The depositional load function 158 2 Fig 5-4. A map of acid sensitive areas of the Eastern U.S. developed from a lithology- 3 based five-unit geologic classification system after methods in Sullivan 4 etal. (2007) 163 5 Figure 5-5. VS/N values for each grid cell in the eastern (right) and western (left) U.S. 6 domains. The top maps are for sulfur and the bottom are for nitrogen 174 7 Figure 5-6. Schematic Diagram illustrating the procedure for converting deposition 8 tradeoff curves of sulfur and nitrogen to atmospheric concentrations of 9 SOxandNOx 175 10 Figure 5-7. Inter-annual coefficients of variation (CV) of a) nitrogen and b) sulfur VS/N 11 values, based on a series of 2002-2005 CMAQ v4.7 simulation 176 12 Figure 5-8. Tradeoff curve for S and N deposition to protect from aquatic acidification in 13 the Adirondacks using Neco equation 2 181 14 Figure 5-9. Tradeoff curve for S and N deposition to protect from aquatic acidification in 15 the Adirondacks using Neco equation 3 181 16 Figure 5-10. Tradeoff curve for S and NOy deposition to protect from aquatic 17 acidification in the Adirondacks using Neco equation 2 183 18 Figure 5-11. Tradeoff curve for S and NOy deposition to protect from aquatic 19 acidification in the Adirondacks using Neco equations 183 20 Figure 5-12. Tradeoff curve for atmospheric concentration of SOX and NOy to protect 21 from aquatic acidification in the Adirondacks using Neco equation 2 184 22 Figure 5-13. Tradeoff curve for atmospheric concentration of SOX and NOy to protect 23 from aquatic acidification in the Adirondacks using Neco equation 3 185 24 Figure 6-1. Ecosystems sensitive to acidifying deposition in the Eastern U.S. (Note that 25 Florida represents a special case where high levels of natural 26 acidification exist unrelated to deposition) This map does not include all 27 sensitive areas in the U.S. Certain mountainous areas of the Western 28 U.S. are also sensitive to acidifying deposition 202 29 Figure 6-2. Number offish species per lake or stream versus ANC level and aquatic 30 status category (colored regions) for lakes in the Adirondack Case Study 31 Area (Sullivan et al., 2006) 206 32 33 LIST OF TABLES 34 Table 3-1. Crosswalk between Ecosystem Services and Public Welfare Effects 70 35 Table 5-1. Illustration of how selected models and water chemistry data were used to 36 calculate critical loads in the REA 151 37 Table 5-2. Summary of the ecological components of design option 1 166 38 Table 5-3. Oxidized sulfur and nitrogen species currently available in CMAQ 39 simulations. Note that PNA concentrations are not available in current 40 CMAQ extractions 174 41 Table 5-4. Example Calculations for Determining the Percent of Water Bodies Achieving 42 Target ANC Levels 180 March 2010 vii Draft - Do No Quote or Cite ------- 1 Table 5-5. Values for N and S deposition tradeoff curves for ANC = 50, protecting 32 2 and 50% of the population, in Adirondacks case study area as illustrated 3 on Fig 5-8 and Fig 5-9. Units are in meq/m2/yr unless noted otherwise 180 4 Table 5-6. Values for NOy and S deposition tradeoff curves for ANC = 50, protecting 32 5 and 50% of the population in Adirondacks case study area as illustrated 6 on Fig 5.10 and Fig 5.11. Units are in meq/m2/yr unless noted otherwise 182 7 Table 7-1. Results of comparing aquatic ANC50 critical loads to average terrestrial 8 watershed area Bc:Al ratios. Left numbers in each column are the 9 number of lakes or streams that had a lower critical load than the 10 terrestrial calculated critical load. Right numbers in each column are the 11 number of lakes that had a higher critical load than the watershed 12 calculated terrestrial critical loads 214 13 14 15 March 2010 viii Draft - Do No Quote or Cite ------- LIST OF ACRONYMS AND ABBREVIATIONS 2 O 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 44 45 AAPI ADR A13+ ANC AQCD AQRV ASSETS El Bc/Al C Ca/Al n 2+ Ca CAA CASAC CASTNet CCS Chi a CLE CMAQ CSS CWA DIN DO DOT EMAP EPA FHWAR FIA FWS GIS GPP H+ H2O H2SO4 ha HAB HFC Hg+2 Hg° HNO3 HONO HUC IMPROVE ISA K+ Atmospheric Acidification Potential Index Adirondack Mountains of New York aluminum acid neutralizing capacity Air Quality Criteria Document air quality related values Assessment of Estuarine Trophic Status eutrophi cation index Base cation to aluminum ratio, also Be: Al carbon calcium to aluminum ratio calcium Clean Air Act Clean Air Scientific Advisory Committee Clean Air Status and Trends Network coastal sage scrub chlorophyll a critical load exceedance Community Multiscale Air Quality model coastal sage scrub Clean Water Act dissolved inorganic nitrogen dissolved oxygen U.S. Department of Interior Environmental Monitoring and Assessment Program U.S. Environmental Protection Agency fishing, hunting and wildlife associated recreation survey Forest Inventory and Analysis National Program Fish and Wildlife Service geographic information systems gross primary productivity hydrogen ion water vapor sulfuric acid hectare harmful algal bloom hydrofluorocarbon reactive mercury elemental mercury nitric acid nitrous acid hydrologic unit code Interagency Monitoring of Protected Visual Environments Integrated Science Assessment potassium March 2010 IX Draft - Do No Quote or Cite ------- 1 2 O 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 44 45 46 47 kg/ha/yr km LRMP LTER LTM MAGIC MCF MEA Mg2+ N N2 N2O N203 N204 N205 Na+ NAAQS NADP NAPAP NAWQA NEEA NEP NH3 NH4+ (NH4)2SO4 NHX NO NO2 NO2- MV NOAA NOX NOy NPP NFS NRC NSWS NTN NTR 03 OAQPS OW PAN PFC pH ppb ppm kilograms per hectare per year kilometer Land and Resource Management Plan Long Term Ecological Monitoring and Research Long-Term Monitoring Model of Acidification of Groundwater in Catchments Mixed Conifer Forest Millennium Ecosystem Assessment magnesium nitrogen gaseous nitrogen nitrous oxide nitrogen trioxide nitrogen tetr oxide dinitrogen pentoxide sodium National Ambient Air Quality Standards National Atmospheric Deposition Program National Acid Precipitation Assessment Program National Water Quality Assessment National Estuarine Eutrophi cation Assessment net ecosystem productivity ammonia gas ammonium ion ammonium sulfate category label for NH3 plus NH4+ nitric oxide nitrogen dioxide reduced nitrite reduced nitrate National Oceanic and Atmospheric Administration nitrogen oxides total oxidized nitrogen net primary productivity National Park Service National Research Council National Surface Water Survey National Trends Network organic nitrate ozone Office of Air Quality Planning and Standards Office of Water peroxyacyl nitrates perfluorocarbons relative acidity parts per billion parts per million March 2010 Draft - Do No Quote or Cite ------- 1 2 O 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 ppt PSD REA REMAP S S203 S207 SAV SF6 SMP SO S02 SO3 so32- SO4 SO42" SOM sox SPARROW SRB STORE! TIME TMDL TP USFS USGS ueq/L Hg/m3 parts per trillion prevention of significant deterioration Risk and Exposure Assessment Regional Environmental Monitoring and Assessment Program sulfur thiosulfate heptoxide submerged aquatic vegetation sulfur hexafluoride Simple Mass Balance sulfur monoxide sulfur dioxide sulfur trioxide sulfite wet sulfate sulfate ion soil organic matter sulfur oxides SPAtially Referenced Regressions on Watershed Attributes sulfate-reducing bacteria STORage and RETrieval Temporally Integrated Monitoring of Ecosystems total maximum daily load total phosphorus U.S. Forest Service U.S. Geological Survey microequivalents per liter micrograms per cubic meter 30 March 2010 XI Draft - Do No Quote or Cite ------- 1 LIST OF KEY TERMS 2 Acidification: The process of increasing the acidity of a system (e.g., lake, stream, forest soil). 3 Atmospheric deposition of acidic or acidifying compounds can acidify lakes, streams, 4 and forest soils. 5 Air Quality Indicator: The substance or set of substances (e.g., PM2.5, NC>2, 862) occurring in 6 the ambient air for which the National Ambient Air Quality Standards set a standard level 7 and monitoring occurs. 8 Alpine: The biogeographic zone made up of slopes above the tree line, characterized by the 9 presence of rosette-forming herbaceous plants and low, shrubby, slow-growing woody 10 plants. 11 Acid Neutralizing Capacity: A key indicator of the ability of water to neutralize the acid or 12 acidifying inputs it receives. This ability depends largely on associated biogeophysical 13 characteristics, such as underlying geology, base cation concentrations, and weathering 14 rates. 15 Arid Region: A land region of low rainfall, where "low" is widely accepted to be less than 250 16 mm precipitation per year. 17 Base Cation Saturation: The degree to which soil cation exchange sites are occupied with base 18 cations (e.g., Ca2+, Mg2+, K+) as opposed to A13+ and H+. Base cation saturation is a 19 measure of soil acidification, with lower values being more acidic. There is a threshold 20 whereby soils with base saturations less than 20% (especially between 10%-20%) are 21 extremely sensitive to change. 22 Ecologically Relevant Indicator: A physical, chemical, or biological entity/feature that 23 demonstrates a consistent degree of response to a given level of stressor exposure and 24 that is easily measured/quantified to make it a useful predictor of ecological risk. 25 Critical Load: A quantitative estimate of an exposure to one or more pollutants, below which 26 significant (as defined by the analyst or decision maker) harmful effects on specified 27 sensitive elements of the environment do not occur, according to present knowledge. 28 Denitrification: The anaerobic reduction of oxidized nitrogen (e.g., nitrate or nitrite) to gaseous 29 nitrogen (e.g., N2O or N2) by denitrifying bacteria. 30 Dry Deposition: The removal of gases and particles from the atmosphere to surfaces in the 31 absence of precipitation (e.g., rain, snow) or occult deposition (e.g., fog). 32 Ecological Risk: The likelihood that adverse ecological effects may occur or are occurring as a 33 result of exposure to one or more stressors (U.S. EPA, 1992). 34 Ecological Risk Assessment: A process that evaluates the likelihood that adverse ecological 35 effects may occur or are occurring as a result of exposure to one or more stressors (U.S. 36 EPA, 1992). 37 Ecosystem: The interactive system formed from all living organisms and their abiotic (i.e., 38 physical and chemical) environment within a given area. Ecosystems cover a hierarchy of 39 spatial scales and can comprise the entire globe, biomes at the continental scale, or small, 40 well-circumscribed systems such as a small pond. 41 Ecosystem Benefit: The value, expressed qualitatively, quantitatively, and/or in economic terms, 42 where possible, associated with changes in ecosystem services that result either directly 43 or indirectly in improved human health and/or welfare. Examples of ecosystem benefits 44 that derive from improved air quality include improvements in habitats for sport fish 45 species, the quality of drinking water and recreational areas, and visibility. March 2010 xii Draft - Do No Quote or Cite ------- 1 Ecosystem Function: The processes and interactions that operate within an ecosystem. 2 Ecosystem Services: The ecological processes or functions having monetary or non-monetary 3 value to individuals or society at large. These are (1) supporting services, such as 4 productivity or biodiversity maintenance; (2) provisioning services, such as food, fiber, or 5 fish; (3) regulating services, such as climate regulation or carbon sequestration; and (4) 6 cultural services, such as tourism or spiritual and aesthetic appreciation. 7 Eutrophication: The process by which nitrogen additions stimulate the growth of autotrophic 8 biota, usually resulting in the depletion of dissolved oxygen. 9 Nitrogen Enrichment: The process by which a terrestrial system becomes enhanced by nutrient 10 additions to a degree that stimulates the growth of plant or other terrestrial biota, usually 11 resulting in an increase in productivity. 12 Nitrogen Saturation: The point at which nitrogen inputs from atmospheric deposition and other 13 sources exceed the biological requirements of the ecosystem; a level beyond nitrogen 14 enrichment. 15 Occult Deposition: The removal of gases and particles from the atmosphere to surfaces by fog 16 or mi st. 17 Semi-arid Regions: Regions of moderately low rainfall, which are not highly productive and are 18 usually classified as rangelands. "Moderately low" is widely accepted as between 100- 19 and 250-mm precipitation per year. 20 Sensitivity: The degree to which a system is affected, either adversely or beneficially, by NOX 21 and/or SOX pollution (e.g., acidification, nutrient enrichment). The effect may be direct 22 (e.g., a change in growth in response to a change in the mean, range, or variability of 23 nitrogen deposition) or indirect (e.g., changes in growth due to the direct effect of 24 nitrogen consequently altering competitive dynamics between species and decreased 25 biodiversity). 26 Total Reactive Nitrogen: This includes all biologically, chemically, and radiatively active 27 nitrogen compounds in the atmosphere and biosphere, such as NFL?, NH4+, NO, NO2, 28 HNOs, N2O, NO3-, and organic compounds (e.g., urea, amines, nucleic acids). 29 Valuation: The economic or non-economic process of determining either the value of 30 maintaining a given ecosystem type, state, or condition, or the value of a change in an 31 ecosystem, its components, or the services it provides. 32 Variable Factors: Influences which by themselves or in combination with other factors may 33 alter the effects on public welfare of an air pollutant (section 108 (a)(2)) 34 (a) Atmospheric Factors: Atmospheric conditions that may influence transformation, 35 conversion, transport, and deposition, and thereby, the effects of an air pollutant on 36 public welfare, such as precipitation, relative humidity, oxidation state, and co-pollutants 37 present in the atmosphere. 38 (b) Ecological Factors: Ecological conditions that may influence the effects of an air 39 pollutant on public welfare once it is introduced into an ecosystem, such as soil base 40 saturation, soil thickness, runoff rate, land use conditions, bedrock geology, and 41 weathering rates. 42 Vulnerability: The degree to which a system is susceptible to, and unable to cope with, the 43 adverse effects of NOX and/or SOX air pollution. 44 Welfare Effects: The effects on soils, water, crops, vegetation, man-made materials, animals, 45 wildlife, weather, visibility, and climate; as well as damage to and deterioration of 46 property, hazards to transportation, and the effects on economic values and on personal March 2010 xiii Draft - Do No Quote or Cite ------- 1 comfort and well-being, whether caused by transformation, conversion, or combination 2 with other air pollutants (Clean Air Act Section 302[h]). 3 Wet Deposition: The removal of gases and particles from the atmosphere to surfaces by rain or 4 other precipitation. 5 6 7 March 2010 xiv Draft - Do No Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 1. INTRODUCTION 2 The U.S. Environmental Protection Agency (EPA) is presently conducting a review of 3 the secondary National Ambient Air Quality Standards (NAAQS) for oxides of nitrogen (NOX) 4 and oxides of sulfur (SOX). The EPA's overall plan and schedule for this review were presented 5 in the Integrated Review Plan for the Secondary National Ambient Air Quality Standards for 6 Nitrogen Dioxide and Sulfur Dioxide (US EPA, 2007). The Integrated Review Plan (IRP) 7 outlined the Clean Air Act (CAA or the Act) requirements related to the establishment and 8 reviews of the NAAQS, the process and schedule for conducting the current review, and the key 9 components in the NAAQS review process: an Integrated Science Assessment (ISA), Risk and 10 Exposure Assessment (REA), and policy assessment/rulemaking. It presented key policy - 11 relevant issues to be addressed in this review as a series of questions that frames our 12 consideration of whether the current secondary (welfare-based) NAAQS for NOX and SOX should 13 be retained or revised. 14 As part of this review, staff in the U.S. Environmental Protection Agency's (EPA) Office 15 of Air Quality Planning and Standards (OAQPS) prepared this first draft Policy Assessment.1 16 The objective of this assessment is to evaluate the policy implications of the key scientific 17 information contained in the document Integrated Science Assessment for Oxides of Nitrogen 18 and Sulfur-Ecological Criteria (USEPA, 2008; henceforth referred to as the ISA), prepared by 19 EPA's National Center for Environmental Assessment (NCEA) and the results from the analyses 20 contained in the Risk and Exposure Assessment for Review of the Secondary National Ambient 21 Air Quality Standards for Oxides of Nitrogen and Oxides of Sulfur (U.S. EPA, 2009; henceforth 22 referred to as the REA). This first draft also presents preliminary staff conclusions on a range of 23 policy options that we believe are appropriate for the Administrator to consider concerning 24 whether, and if so how, to revise the secondary (welfare-based) NOX and SOX NAAQS. 1 Preparation of a PA by OAQPS staff reflects Administrator Jackson's decision to modify the NAAQS review process that was presented in the IRP. See http://www.epa.gov/ttn/naaqs/review.html for more information on the current NAAQS review process. February, 2010 1 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 This policy assessment is intended to help "bridge the gap" between the scientific 2 assessment contained in the ISA and the judgments required of the EPA Administrator in 3 determining whether it is appropriate to retain or revise the secondary NAAQS for NOX and SOX. 4 This policy assessment considers the available scientific evidence and quantitative risk-based 5 analyses, together with related limitations and uncertainties, and focuses on the basic elements of 6 air quality standards: indicators2, averaging times, forms3, and levels. These elements, which 7 serve to define each standard, must be considered collectively in evaluating the welfare 8 protection afforded by the secondary NOX and SOX NAAQS. Our development of this policy 9 assessment is based on the assessment and integrative synthesis of information presented in the 10 ISA and on staff analyses and evaluations presented in this document, and is further informed by 11 comments and advice received from an independent scientific review committee, the Clean Air 12 Scientific Advisory Committee (CASAC), in their review of the previous integrated science and 13 risk and exposure assessments. The Policy Assessment is further informed by comments 14 submitted by the public4. To view related documents developed as part of the planning, science, 15 and risk assessment phases of this review see 16 http://www.epa.gov/ttn/naaqs/standards/no2so2sec/index.html. 17 This document is organized around a conceptual framework for a combined NOX and SOX 18 secondary NAAQS and is focused on answering key policy questions related to the 19 implementation of that conceptual framework. Chapter 2 provides a summary of ecological 20 effects from the deposition of ambient NOX and SOX to sensitive ecosystems, drawing from the 21 ISA and REA. Chapter 3 places those ecological effects within the context of "public welfare" 22 by linking effects to ecosystem services or other benchmarks of public welfare. Chapter 4 23 addresses the adequacy of the current NOX and SOX secondary NAAQS in addressing the impacts 24 on public welfare from ecological effects. Chapter 5 develops the conceptual design for 25 ecologically relevant multi-pollutant standards. Chapter 6 presents options for developing critical 26 elements of a secondary NAAQS necessary to implement the conceptual design. Chapter 7 27 describes how secondary NAAQS designed to protect a specific ecological endpoint may also 28 provide protection for other ecological endpoints. Chapter 8 provides a consideration of issues The "indicator" of a standard defines the chemical species or mixture that is to be measured in determining whether an area attains the standard. 3 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. 4 Summary information on public comments will be provided in a later draft of the policy assessment March 2010 2 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 regarding reduced and oxidized forms of nitrogen. Chapter 9 concludes with preliminary staff 2 conclusions regarding ranges of options for pollutant indicators, averaging times, forms, and 3 levels for the secondary NOX and SOX NAAQS, including a discussion of staff initial conclusions 4 on what levels of the secondary NAAQS might be requisite to protect public welfare. 5 In this document we consider how the available scientific evidence and quantitative risk- 6 based analyses, together with related limitations and uncertainties, inform the review of each 7 element of the NAAQS: indicator, averaging times, forms, and levels. These elements must be 8 considered collectively in evaluating the welfare protection afforded by the secondary NAAQS 9 standards. This draft document does not contain final staff conclusions as to all the necessary 10 components of an alternative secondary standard for NOX and/or SOX but rather describes the 11 current state of thinking with regard to potential policy options and provides an appropriate 12 context of information for the Administrator to consider in making decisions regarding the 13 standards. 14 While this policy assessment should be of use to all parties interested in the secondary 15 NOX and SOX NAAQS review, it is written with an expectation that the reader has some 16 familiarity with the technical discussions contained in the ISA and REA. 17 EPA will be preparing a second draft Policy Assessment subsequent to receiving advice 18 from the CASAC. The second draft will incorporate responses to comments received from 19 CASAC, as well as comments submitted by the public. The second draft will also provide a more 20 complete development of the conceptual model, and will provide a more complete set of staff 21 conclusions on critical elements of the standards. EPA's final Policy Assessment will address 22 additional CASAC comments on the second draft, and will include sufficient information to 23 inform the Administrator on critical elements of the standards, and staff conclusions regarding 24 alternative levels of the standards. 25 1.1 DEFINITIONS OF NOX AND SOX FOR THIS ASSESSMENT 26 As discussed in detail in the REA (REA 1.3.1), in the atmospheric science community 27 NOX is typically referred to as the sum of nitrogen dioxide (NC^), and nitric oxide (NO). From a 28 Clean Air Act perspective, the family of NOX includes any gaseous combination of nitrogen and 29 oxygen (e.g., NC>2, NO, nitrous oxide pSPzO], nitrogen trioxide [N^Os], nitrogen tetroxide PS^O^, 30 and dinitrogen pentoxide pS^Os]). The term used by the scientific community to represent the March 2010 3 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 complete set of oxidized nitrogen compounds, including those listed in CAA Section 108(c), is 2 total oxidized nitrogen (NOy). NOy includes all nitrogen oxides, including e.g. total reactive 3 oxidized atmospheric nitrogen, defined as NOX (NO and NO2) and all oxidized NOX products: 4 NOy = NO2 + NO + HNO3 + PAN +2N2O5 + HONO+ NO3 + organic nitrates + paniculate NO3 5 (Finlayson-Pitts and Pitts, 2000). In this document, unless otherwise indicated, we use the term 6 NOX interchangeably with NOy to refer to the complete set of oxidized nitrogen compounds. 7 For this assessment, SOX is defined to include all oxides of sulfur, including multiple 8 gaseous substances (e.g., SO2, sulfur monoxide [SO], sulfur trioxide [SO3], thiosulfate [S2O3], 9 and heptoxide [S2O?], as well as particulate species, such as ammonium sulfate [(NFL^SOJ). 10 Throughout this text we refer to sulfate as SO4 and nitrate as NO3, recognizing that they have 11 charges of -2 for sulfate and -1 for nitrate. 12 1.2 POLICY OBJECTIVES 13 In conducting this periodic review of the NOX and SOX secondary NAAQS, EPA has 14 decided to jointly assess the scientific information, associated risks, and standards relevant to 15 protecting the public welfare from adverse effects associated with oxides of nitrogen and sulfur. 16 Although EPA has historically adopted separate secondary standards for oxides of nitrogen 17 (NOX) and oxides of sulfur (SOX), EPA is conducting a joint secondary review of these standards 18 because NOX, SOX, and their associated transformation products are linked from an atmospheric 19 chemistry perspective, as well as from an environmental effects perspective. The National 20 Research Council (NRC) has recommended that EPA consider multiple pollutants, as 21 appropriate, in forming the scientific basis for the NAAQS (NRC, 2004). There is a strong basis 22 for considering these pollutants together, building upon EPA's and CAS AC's past recognition of 23 the interactions of these pollutants and on the growing body of scientific information that is now 24 available related to these interactions and associated ecological effects. 25 EPA sets secondary standards for two criteria pollutants related to NOX and SOX: ozone 26 and particulate matter (PM). NOX is a precursor to the formation of ozone in the atmosphere, and 27 under certain conditions, can combine with atmospheric ammonia to form ammonium nitrate, a 28 component of fine PM. SOX is a precursor to the formation of particulate sulfate, which is a 29 significant component of fine PM in many parts of the U.S. While there are a number of welfare 30 effects associated with ozone and fine PM, including ozone damage to vegetation, and visibility March 2010 4 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 degradation related to PM, protection against those effects is provided by the ozone and fine PM 2 standards. This review focuses on evaluation of the protection provided by NOX and SOX 3 secondary standards for effects associated with direct atmospheric concentrations of NOX and 4 SOX, and effects associated with deposition of NOX and SOX to ecosystems, including deposition 5 in the form of particulate nitrate and sulfate in their component forms. 6 The ISA highlights the ecological effects associated with deposition of ambient NOX and 7 SOX to ecosystems other than commercially managed forests and agricultural lands. This 8 assessment evaluates information on gas-phase effects of NOX and SOX via stomatal exposure on 9 vegetation, but primarily focuses on the effects of gas-phase NOX and SOX exposure via 10 deposition on multiple ecological receptors. Highlighted effects include those associated with 11 acidification and nitrogen nutrient enrichment. Based on these highlighted effects, EPA's policy 12 objective is to develop a framework for NOX and SOX standards that incorporate factors that will 13 lead to standards that are ecologically relevant, and that recognizes the interactions between the 14 two pollutants as they deposit to sensitive ecosystems, with an ultimate goal of setting standards 15 that, based on the ecological criteria described in the ISA, and consistent with the requirements 16 of the Clean Air Act, "are requisite to protect the public welfare from any known or anticipated 17 adverse effects associated with the presence of such air pollutant in the ambient air." 18 In presenting policy options for the Administrator's consideration, we note that the final 19 decision on retaining or revising the current secondary standards for NOX and SOX is largely a 20 public welfare policy judgment based on the Administrator's informed assessment of what 21 constitutes requisite protection against adverse effects to public welfare. A final decision should 22 draw upon scientific information and analyses about welfare effects, exposure and risks, as well 23 as judgments about the appropriate response to the range of uncertainties that are inherent in the 24 scientific evidence and analyses. The ultimate determination as to what level of damage to 25 ecosystems and the services provided by those ecosystems is adverse to public welfare is not 26 wholly a scientific question, although it is informed by scientific studies linking ecosystem 27 damage to losses in ecosystem services, and economic information on the value of those losses in 28 ecosystem services. Our approach to informing these judgments, as discussed below, is 29 consistent with the requirements of the NAAQS provisions of the Clean Air Act and with how 30 EPA and the courts have historically interpreted the Act. These provisions require the 31 Administrator to establish secondary NAAQS that, in the Administrator's judgment, are requisite March 2010 5 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 to protect public welfare from any known or anticipated adverse effects associated with the 2 presence of NOX and SOX in the ambient air. In so doing, the Administrator seeks to establish 3 standards that are neither more nor less stringent than necessary for this purpose. 4 For this first draft policy assessment, we have chosen to focus much of our discussion on 5 the effects of ambient NOX and SOX on ecological impacts associated with acidifying deposition 6 of nitrogen and sulfur, which is a transformation product of ambient NOX and SOX. We have the 7 greatest confidence in the causal linkages between NOX and SOX and aquatic acidification effects, 8 and we have the most complete information available with which to develop an ecologically 9 meaningful structure for the standards. In future drafts, we expect to be able to explore whether 10 and how the standards can be expanded to directly address effects of acidification on terrestrial 11 ecosystems, and to address the effects of nutrient enrichment in terrestrial and aquatic 12 ecosystems. 13 1.3 CRITICAL POLICY ELEMENTS 14 Our policy objective is guided by the information in the ISA and REA, framed within the 15 legislative requirements of the CAA. This framing leads us to focus on critical policy elements 16 (CPE) consistent with elements of Clean Air Act language. 17 Sections 108 and 109 of the CAA govern the establishment and periodic review of the 18 NAAQS and of the air quality criteria upon which the standards are based. The NAAQS are 19 established for pollutants that are listed under section 108, based on three criteria, including 20 whether emissions of the air pollutant cause or contribute to air pollution which may reasonably 21 be anticipated to endanger public health or welfare and whose presence in the ambient air results 22 from numerous or diverse mobile or stationary sources. The NAAQS are based on air quality 23 criteria that reflect the latest scientific knowledge, useful in indicating the types and extent of 24 identifiable effects on public health or welfare that may be expected from the presence of the 25 pollutant in ambient air. The criteria refer to criteria issued pursuant to §108 of the Clean Air 26 Act, which include "(A) those variable factors (including atmospheric conditions) which of 27 themselves or in combination with other factors may alter the effects on public health or welfare 28 of such air pollutant; (B) the types of air pollutants which, when present in the atmosphere, may 29 interact with such pollutant to produce an adverse effect on public health of welfare; and (C) any 30 known or anticipated adverse effects on welfare." March 2010 6 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 The following critical policy elements for the design of ecologically relevant secondary 2 standards for NOX and SOX are identified: 3 (CPE 1) An evaluation of the effects of ambient NOX and SOX on ecosystems, and the 4 relationship between those effects and the measure of dose in the ecosystem, 5 indicated by the deposit!onal loadings of N and S. 6 (CPE 1.1) Evaluation of the relationship between response of ecological receptors, e.g. 7 changes in diversity offish species, and the response related to public welfare, 8 e.g. loss in recreational fishing services. 9 (CPE 1.2) Evaluation of the extent to which identified effects are occurring under recent 10 conditions, and the extent to which meeting the current standards would 11 provide protection against these effects. 12 (CPE 2) An assessment of how best to characterize, in defining the standards, the 13 variable ecosystem factors that affect the relationship between ecological 14 effects and deposit! onal loadings of N and S. 15 (CPE 2.1) Specification of potential indicators of ecological effects, e.g. acid 16 neutralizing capacity (ANC) that incorporates variability in ecosystem factors. 17 (CPE 3) Characterization of the complex atmospheric transformations between 18 ambient concentrations of NOX and SOX and deposition of N and S in the 19 specification of a standard. 20 (CPE 4) Specification of those factors, such as precipitation, which interact with 21 ambient NOX and SOX to produce adverse effects on welfare, by affecting 22 deposition of N and S. 23 (CPE 5) Specification of the form for the standard(s), including ambient atmospheric 24 indicators for NOX and SOX, with consideration of averaging times, and 25 options for levels of the standard(s). 26 The development of the conceptual framework for the NOX and SOX standards described 27 in Section 1.4 will be motivated by these critical policy elements. However, in order to provide a 28 historical context for this new framework, the next section provides a brief history of previous 29 reviews of the NOX and SOX secondary NAAQS, as well as other relevant historical reviews of 30 welfare effects associated with these pollutants. March 2010 7 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 1.4 HISTORICAL CONTEXT 2 1.4.1 History of NOX and SOX NAAQS Review 3 1.4.1.1 NOx NAAQS 4 EPA began the most recent previous review of the NOX secondary standards in 1987 and 5 in November 1991, EPA released an updated draft AQCD for CAS AC and public review and 6 comment (56 FR 59285). This draft document provided a comprehensive assessment of the 7 available scientific and technical information on health and welfare effects associated with NO2 8 and other NOX. CAS AC reviewed the draft document at a meeting held on July 1, 1993, and 9 concluded in a closure letter to the Administrator that the document "provides a scientifically 10 balanced and defensible summary of current knowledge of the effects of this pollutant and 11 provides an adequate basis for EPA to make a decision as to the appropriate NAAQS for NO2" 12 (Wolff, 1993). The AQCD Air Quality Criteria for Oxides of Nitrogen was then finalized (U.S. 13 EPA, 1993). EPA also prepared a Staff Paper that summarized and integrated the key studies and 14 scientific evidence contained in the revised NOX AQCD and identified the critical elements to be 15 considered in the review of the NO2 NAAQS. CASAC reviewed two drafts of the Staff Paper and 16 concluded in a closure letter to the Administrator that the document provided a "scientifically 17 adequate basis for regulatory decisions on nitrogen dioxide" (Wolff, 1995). In October 1995, the 18 Administrator announced her proposed decision not to revise either the primary or secondary 19 NAAQS for NO2 (60 FR 52874; October 11, 1995). A year later, the Administrator made a final 20 determination not to revise the NAAQS for NO2 after careful evaluation of the comments 21 received on the proposal (61 FR 52852; October 8, 1996). The level for both the existing primary 22 and secondary NAAQS for NO2 is 0.053 ppm (100 micrograms per cubic meter [jjg/ms] of air), 23 annual arithmetic average, calculated as the arithmetic mean of the 1-hour NO2 concentrations. 24 1.4.1.2 SOX NAAQS 25 Based on the 1970 SOX criteria document (DHEW, 1970), EPA promulgated primary and 26 secondary NAAQS for SO2 on April 30, 1971 (36 FR 8186). The secondary standards included a 27 standard at 0.02 ppm in an annual arithmetic mean and a 3-hour average of 0.5 ppm, not to be 28 exceeded more than once per year. These secondary standards were established solely on the 29 basis of evidence of adverse effects on vegetation. In 1973, revisions made to Chapter 5 March 2010 8 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 ("Effects of Sulfur Oxide in the Atmosphere on Vegetation") of Air Quality Criteria for Sulfur 2 Oxides (U.S. EPA, 1973) indicated that it could not properly be concluded that the vegetation 3 injury reported resulted from the average SO2 exposure over the growing season, rather than 4 from short-term peak concentrations. Therefore, EPA proposed (38 FR 11355) and then finalized 5 (38 FR 25678) a revocation of the annual mean secondary standard. At that time, EPA was aware 6 that SOX have other public welfare effects, including effects on materials, visibility, soils, and 7 water. However, the available data were considered insufficient to establish a quantitative 8 relationship between specific ambient SOX concentrations and effects (38 FR 25679). 9 In 1979, EPA announced that it was revising the Air Quality Criteria Document (AQCD) 10 for sulfur oxides concurrently with that for particulate matter and would produce a combined 11 particulate matter and sulfur oxides criteria document. Following its review of a draft revised 12 criteria document in August 1980, CAS AC concluded that acid deposition was a topic of 13 extreme scientific complexity because of the difficulty in establishing firm quantitative 14 relationships among (1) emissions of relevant pollutants (e.g., SC>2 and oxides of nitrogen), (2) 15 formation of acidic wet and dry deposition products, and (3) effects on terrestrial and aquatic 16 ecosystems. CAS AC also noted that acid deposition involves, at a minimum, several different 17 criteria pollutants: oxides of sulfur, oxides of nitrogen, and the fine particulate fraction of 18 suspended particles. CAS AC felt that any document on this subject should address both wet and 19 dry deposition, since dry deposition was believed to account for at least one half of the total acid 20 deposition problem. 21 For these reasons, CAS AC recommended that a separate, comprehensive document on 22 acid deposition be prepared prior to any consideration of using the NAAQS as a regulatory 23 mechanism for the control of acid deposition. CASAC also suggested that a discussion of acid 24 deposition be included in the AQCDs for nitrogen oxides and PM and SOX. Following CASAC 25 closure on the AQCD for SC>2 in December 1981, EPA's Office of Air Quality Planning and 26 Standards published a Staff Paper in November 1982, but the paper did not directly assess the 27 issue of acid deposition. Instead, EPA subsequently prepared the following documents: The 28 Acidic Deposition Phenomenon and Its Effects: Critical Assessment Review Papers, Volumes I 29 and II (U.S. EPA, 1984a, b), and The Acidic Deposition Phenomenon and Its Effects: Critical 30 Assessment Document (U.S. EPA, 1985) (53 FR 14935 -14936). These documents, though they March 2010 9 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 were not considered criteria documents and did not undergo CASAC review, represented the 2 most comprehensive summary of relevant scientific information completed by EPA at that point. 3 On April 26, 1988 (53 FR 14926), EPA proposed not to revise the existing primary and 4 secondary standards for SC>2. This proposal regarding the secondary SC>2 NAAQS was due to the 5 Administrator's conclusions that (1) based upon the then-current scientific understanding of the 6 acid deposition problem, it would be premature and unwise to prescribe any regulatory control 7 program at that time, and (2) when the fundamental scientific uncertainties had been reduced 8 through ongoing research efforts, EPA would draft and support an appropriate set of control 9 measures. 10 1.4.2 History of Related Assessments and Agency Actions 11 In 1980, the Congress created the National Acid Precipitation Assessment Program 12 (NAPAP) in response to growing concern about acidic deposition. The NAPAP was given a 13 broad 10-year mandate to examine the causes and effects of acidic deposition and to explore 14 alternative control options to alleviate acidic deposition and its effects. During the course of the 15 program, the NAPAP issued a series of publicly available interim reports prior to the completion 16 of a final report in 1990 (NAPAP, 1990). 17 In spite of the complexities and significant remaining uncertainties associated with the 18 acid deposition problem, it soon became clear that a program to address acid deposition was 19 needed. The Clean Air Act Amendments of 1990 included numerous separate provisions related 20 to the acid deposition problem. The primary and most important of the provisions, the 21 amendments to Title IV of the Act, established the Acid Rain Program to reduce emissions of 22 SC>2 by 10 million tons and NOX emissions by 2 million tons from 1980 emission levels in order 23 to achieve reductions over broad geographic regions. In this provision, Congress included a 24 statement of findings that led them to take action, concluding that (1) the presence of acid 25 compounds and their precursors in the atmosphere and in deposition from the atmosphere 26 represents a threat to natural resources, ecosystems, materials, visibility, and public health; (2) 27 the problem of acid deposition is of national and international significance; and (3) current and 28 future generations of Americans will be adversely affected by delaying measures to remedy the 29 problem. March 2010 10 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Second, Congress authorized the continuation of the NAPAP in order to assure that the 2 research and monitoring efforts already undertaken would continue to be coordinated and would 3 provide the basis for an impartial assessment of the effectiveness of the Title IV program. 4 Third, Congress considered that further action might be necessary in the long term to 5 address any problems remaining after implementation of the Title IV program and, reserving 6 judgment on the form that action could take, included Section 404 of the 1990 Amendments 7 (Clean Air Act Amendments of 1990, Pub. L. 101-549, § 404) requiring EPA to conduct a study 8 on the feasibility and effectiveness of an acid deposition standard or standards to protect 9 "sensitive and critically sensitive aquatic and terrestrial resources." At the conclusion of the 10 study, EPA was to submit a report to Congress. Five years later, EPA submitted its report, 11 entitled Acid Deposition Standard Feasibility Study: Report to Congress (U.S. EPA, 1995) in 12 fulfillment of this requirement. The Report concluded that establishing acid deposition standards 13 for sulfur and nitrogen deposition may at some point in the future be technically feasible, 14 although appropriate deposition loads for these acidifying chemicals could not be defined with 15 reasonable certainty at that time. 16 Fourth, the 1990 Amendments also added new language to sections of the CAA 17 pertaining to the scope and application of the secondary NAAQS designed to protect the public 18 welfare. Specifically, the definition of "effects on welfare" in Section 302(h) was expanded to 19 state that the welfare effects include effects ".. .whether caused by transformation, conversion, or 20 combination with other air pollutants." 21 In 1999, seven Northeastern states cited this amended language in Section 302(h) in a 22 petition asking EPA to use its authority under the NAAQS program to promulgate secondary 23 NAAQS for the criteria pollutants associated with the formation of acid rain. The petition stated 24 that this language "clearly references the transformation of pollutants resulting in the inevitable 25 formation of sulfate and nitrate aerosols and/or their ultimate environmental impacts as wet and 26 dry deposition, clearly signaling Congressional intent that the welfare damage occasioned by 27 sulfur and nitrogen oxides be addressed through the secondary standard provisions of Section 28 109 of the Act." The petition further stated that "recent federal studies, including the NAPAP 29 Biennial Report to Congress: An Integrated Assessment, document the continued-and increasing- 30 damage being inflicted by acid deposition to the lakes and forests of New York, New England 31 and other parts of our nation, demonstrating that the Title IV program had proven insufficient." March 2010 11 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 The petition also listed other adverse welfare effects associated with the transformation of these 2 criteria pollutants, including impaired visibility, eutrophication of coastal estuaries, global 3 warming, and tropospheric ozone and stratospheric ozone depletion. 4 In a related matter, the Office of the Secretary of the U.S. Department of Interior 5 requested in 2000 that EPA initiate a rulemaking proceeding to enhance the air quality in 6 national parks and wilderness areas in order to protect resources and values that are being 7 adversely affected by air pollution. Included among the effects of concern identified in the 8 request were the acidification of streams, surface waters, and/or soils; eutrophication of coastal 9 waters; visibility impairment; and foliar injury from ozone. 10 In a Federal Register notice in 2001, EPA announced receipt of these requests and asked 11 for comment on the issues raised in them. EPA stated that it would consider any relevant 12 comments and information submitted, along with the information provided by the petitioners and 13 DOI, before making any decision concerning a response to these requests for rulemaking (65 FR 14 48699). 15 The most recent 2005 NAPAP report states that"... scientific studies indicate that the 16 emission reductions achieved by Title IV are not sufficient to allow recovery of acid-sensitive 17 ecosystems. Estimates from the literature of the scope of additional emission reductions that are 18 necessary in order to protect acid-sensitive ecosystems range from approximately 40-80% 19 beyond full implementation of Title IV.... The results of the modeling presented in this Report to 20 Congress indicate that broader recovery is not predicted without additional emission reductions" 21 (NAPAP, 2005).5 22 Given the state of the science as described in the ISA and in other recent reports, such as 23 the NAPAP's above, EPA believes it is appropriate, in the context of evaluating the adequacy of 24 the current NC>2 and SC>2 secondary standards in this review, to revisit the question of the 25 appropriateness and the feasibility of setting a secondary NAAQS to address remaining known 26 or anticipated adverse public welfare effects resulting from the acidic and nutrient deposition of 27 these criteria pollutants 5 Note that a new NAPAP report is expected to be released later in 2010. The findings of that report will be considered in the final policy assessment. March 2010 12 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 1.5 PROPOSED CONCEPTUAL FRAMEWORK FOR COMBINED NOX 2 SOX STANDARDS 3 There is a strong basis for considering NOX and SOX together at this time, building upon 4 EPA's and CASAC's past recognition of the interactions of these pollutants and on the growing 5 body of scientific information that is now available related to these interactions and associated 6 ecological effects. The REA introduced a conceptual framework for ecologically meaningful 7 secondary standards that recognized the complex processes by which ecosystems are exposed to 8 ambient NOX and SOX. That framework provided a flow from ambient concentrations exposures 9 via deposition to ecological indicators and effects (see Figure ES-2 in the REA Executive 10 Summary). This sequence represents the process by which we can determine the risks associated 11 with ambient concentrations of NOX and SOX. However, for the purposes of discussing a 12 conceptual framework for design of standards to protect against those risks, a modified version 13 of the risk frame work is needed. 14 Figure 1-1 depicts the framework by which we are considering the structure of an 15 ecologically meaningful secondary standard. It is a conceptual diagram that illustrates how a 16 level of protection related to an indicator of ecological effect(s) equates to atmospheric 17 concentrations of NOX and SOX indicators. This conceptual diagram illustrates the linkages 18 between ambient air concentrations and resulting deposition metrics, and between the deposition 19 metric and the ecological indicator of concern. The Atmospheric Deposition Transformation 20 Function translates ambient atmospheric concentrations of NOX and SOX to nitrogen and sulfur 21 deposition metrics, while the Ecological Effect Function transforms the deposition metric into 22 the ecological indicator. 23 Development of a form for the standard that reflects this structure is a critical step in the 24 overall standard setting process. The atmospheric levels of NOX and SOX that satisfy a particular 25 level of ecosystem protection are those levels that result in an amount of deposition that is less 26 than the amount of deposition that a given ecosystem can accept without excessive degradation 27 of the ecological indicator for a targeted effect. 28 The details of this conceptual framework are discussed in Chapter 5, including 29 discussions of modifying factors that alter the relationship between ambient atmospheric 30 concentrations of NOX and SOX and depositional loads of nitrogen and sulfur, and those that 31 modify the relationship between deposition loads and the ecological indicator. March 2010 13 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 In setting NAAQS to protect public health and welfare, EPA has historically established 2 standards which require the comparison of monitored concentrations of an air pollutant against a 3 numerical metric of atmospheric concentration that does not vary geographically. This approach 4 has appropriately protected public health as at-risk populations are widely distributed throughout 5 the nation. As more is learned about the effects of pollutants such as NOX and SOX and the 6 environment, however, such an approach may not be appropriate to provide the requisite level of 7 protection to public welfare from effects on sensitive ecosystems. EPA is considering in this 8 review of the secondary standard for NOX and SOX whether a standard that takes into account 9 variable factors, such as atmospheric variables and ecosystem sensitivities, is the appropriate 10 approach to protect the public welfare from the effects associated with the presence of these 11 pollutants in the ambient air. 12 EPA must undertake a thorough review of the air quality criteria for the pollutant at issue 13 in reviewing a secondary NAAQS, and determine whether a current standard is requisite to 14 protect the public welfare. Under section 108 of the CAA, air quality criteria are to "reflect the 15 latest scientific knowledge useful in indicating the kind and extent of all identifiable effects" 16 associated with the presence of the pollutant in the ambient air. It is clear from the language of 17 the CAA that where the state of the science provides a basis for considering such effects, the 18 review of the air quality criteria should encompass a broad analysis of "any" known or 19 anticipated adverse effects, as well as the ways in which variable conditions such as atmospheric 20 conditions may impact the effect of a pollutant and the ways in which other air pollutants may 21 interact with the criteria pollutant to produce adverse effects. Specifically, section 108(a)(2) of 22 the CAA provides that: 23 Air quality criteria for an air pollutant shall accurately reflect the latest scientific 24 knowledge useful in indicating the kind and extent of all identifiable effects on public health or 25 welfare which may be expected from the presence of such pollutant in the ambient air, in varying 26 quantities. The criteria for an air pollutant to the extent practicable, shall include information on: 27 • (A) those variable factors (including atmospheric conditions) which of themselves or 28 in combination with other factors may alter the effects on public health or welfare of such 29 air pollutants; 30 • (B) the types of air pollutants which, when present in the atmosphere, may interact 31 with such pollutants to produce an adverse effect on public health or welfare; and March 2010 14 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 • (C) any known or anticipated adverse effects on welfare. 2 Based on this extensive review of the air quality criteria for an air pollutant, the 3 Administrator is required to review and to revise, as appropriate, the secondary standard to 4 ensure that the standard "is requisite to protect public welfare from any known or anticipated 5 adverse effects associated with the presence of such air pollutant in the ambient air." CAA § 6 109(b) & (d). "Effects on welfare," in turn, is defined to include a broad array of effects, 7 including effects on soil, water, crops, vegetation, and manmade materials, "whether caused by 8 transformation, conversion, or combination with other air pollutants." CAA § 302(h). Thus, as 9 with the sections of the CAA describing the issuance of air quality criteria, the CAA uses 10 expansive language in describing the scope of EPA's responsibility and the range of effects that 11 EPA should take into account in setting a standard that is requisite to protect public welfare. The 12 term "requisite," however, indicates that section 109 is not open-ended. In considering the 13 meaning of the term "requisite" in the context of the primary standards, the Supreme Court has 14 agreed with EPA that such a standard is one that is "sufficient, but not more than necessary" to 15 protect public health. Whitman v. American Trucking, 531 U.S. 457, 473 (2001). 16 While EPA has most often considered the results of direct exposure to an air pollutant in 17 the ambient air in assessing effects on public health and welfare, such as the health effects on 18 humans when breathing in an air pollutant or the effects on vegetation through the uptake of air 19 pollutants from the ambient air through leaves, EPA has also considered, where appropriate, the 20 effects of exposure to air pollutants through more indirect mechanisms. For example, both in 21 1978 and in 2008, EPA established a NAAQS for lead that addressed the health effects of 22 ambient lead whether the lead particles were inhaled or were ingested after deposition on the 23 ground or other surfaces. 73 FR 66964 (November 12, 2008), Lead Industries v. EPA, 647 F.2d 24 1130 (DC Cir. 1980) (1978 NAAQS). The deposition of ambient NOX and SOX to terrestrial and 25 aquatic environments can impact ecosystems through both direct and indirect mechanisms, as 26 discussed in the REA and this document. Given Congress' instruction to set a standard that "is 27 requisite to protect the public welfare from "any known or anticipated adverse effects associated 28 with the presence of such air pollutant in the ambient air," 42 U.S.C. § 109 (b)(2) (emphasis 29 added), this review appropriately attempts to take into consideration widely acknowledged 30 effects, such as acidification and nutrient enrichment, which are associated with the presence of 31 ambient SOX and NOX. March 2010 15 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 In this review, EPA is also attempting to develop a standard that takes into account the variability in effects from ambient levels of SOX and NOX. The CAA requires EPA to establish "national" standards, based on the air quality criteria, that provide the requisite degree of protection, but does not clearly address how to do so under the circumstances present here. One approach is to develop a secondary standard such as the one discussed in this Policy Assessment Document. Such a standard is designed to provide a generally uniform degree of protection throughout the country by allowing for varying concentrations of allowable ambient NOX and SOX, depending on atmospheric conditions and other variabilities, to achieve that degree of protection. Such a standard protects sensitive ecosystems wherever such ecosystems are found. This approach recognizes that setting a standard that is sufficient to protect the public welfare but not more than is necessary calls for consideration of a standard such as the one discussed in this document. Structure of an Ecologically-based Standard Variable/Fixed Factors: Atmospheric Landscape Atmospheric Deposition Transformation Function Form of the Standard Level of the Standard Figure 1-1. Framework of an alternative secondary standard. 1.6 POLICY RELEVANT QUESTIONS In this policy assessment, a series of general questions frames our approach to identifying a range of policy options for consideration by the Administrator regarding secondary NAAQS for NOX and SOX. These questions are drawn from our Integrated Review Plan with March 2010 16 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 modifications based on further consideration by staff and comments from CAS AC and the 2 public. Our policy assessment begins by characterizing "known or anticipated adverse effects" 3 on public welfare within our conceptual model (CPE 1). As noted earlier, this review is focusing 4 on effects in unmanaged ecosystems (not commercial forests or agricultural lands6) resulting 5 from ambient concentrations of NOX and SOX through deposition of N and S. In Chapter 2, we 6 draw from the information and conclusions presented in the ISA and REA to address the 7 following questions: 8 1. What are the nature and magnitude of ecosystem responses to reactive nitrogen and 9 sulfur deposition? 10 a. How are these responses affected by landscape factors? 11 b. What types of ecosystems are sensitive to such responses? 12 2. To what extent can ecosystem responses to nitrogen deposition be separated into 13 responses related to oxidized and reduced forms of reactive nitrogen compounds? 14 In Chapter 3, we address the following questions related to linking effects to measures of 15 adversity (CPE 1.1): 16 1. How do we characterize adversity to public welfare? What are the sources of 17 potentially relevant characterization for this policy assessment? 18 2. What is the evidence of effects on ecosystem services, and how can those ecosystem 19 services be linked to ecological indicators? 20 3. To what extent are identified ecosystem effects important from a public welfare 21 perspective, and what are the important uncertainties associated with estimating such 22 effects? 23 Once we have described ecological effects, we then provide an assessment of the 24 adequacy of the existing NOX and SOX standards (CPE 1.2). We begin this assessment by 25 drawing from the information and conclusions presented in the ISA and REA to address in 26 Chapter 4 the following questions, which allow us to identify whether the structure of the current 27 standards is appropriate relative to the key ecological effects assessed in the ISA and REA, 6 The decision to focus on unmanaged ecosystems is based on the weight of evidence of effects in those ecosystems. The majority of the scientific evidence regarding acidification and nutrient enrichment is based on studies in unmanaged ecosystems. Non-managed terrestrial ecosystems tend to have a higher fraction of N deposition resulting from atmospheric N (ISA 3.3.2.5). In addition, the ISA notes that agricultural and commercial forest lands are routinely fertilized with amounts of N (100 to 300 kg N/ha) that exceed air pollutant inputs even in the most polluted areas (ISA 3.3.9) March 2010 17 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 including acidification and excess nutrient enrichment and whether there is adequate information 2 and analyses available at this time to assess the extent to which potentially adverse effects on 3 aquatic and terrestrial ecosystems can be associated with current levels of atmospheric reactive 4 nitrogen, accounting for the contributions of oxidized and reduced forms, and SOX and with 5 levels that are at or below the current secondary standards: 6 1. To what extent are effects that could reasonably be judged to be adverse to public 7 welfare occurring under current conditions and would such effects occur if the nation 8 met the current standards? To what extent do the current NOX and SOX secondary 9 standards provide protection from effects associated with deposition of: 10 a. Sulfur and oxidized nitrogen from atmospheric NOX, and SOX which results in 11 acidification in sensitive aquatic and terrestrial ecosystems? 12 b. Oxidized nitrogen from atmospheric NOX, which results in nutrient enrichment 13 effects in sensitive aquatic and terrestrial ecosystems? 14 c. Sulfur and oxidized nitrogen from atmospheric NOX and SOX which results in 15 other ecological effects (e.g. mercury methylation)? 16 2. In what way are the structures of the current NOX and SOX secondary standards 17 inadequate to protect against public welfare effects? 18 In Chapter 5, we follow our adequacy assessment by developing in greater detail the 19 conceptual framework for the design of ecologically relevant multi-pollutant standards 20 introduced in Section 1.4 above. To the extent that the available information calls into question 21 the adequacy of protection afforded by the current standards and/or the appropriateness of the 22 structure of the standards, we explore the extent to which available information supports 23 consideration of alternative standards, in terms of atmospheric and ecological indicators and 24 related averaging times, forms, and levels. This conceptual framework is designed to focus on 25 resolving the following questions: 26 1. (CPE 2.1) Does the available information provide support for the use of ecological 27 indicators to characterize the responses of aquatic and terrestrial ecosystems to 28 oxidized nitrogen and sulfur deposition? 29 2. (CPE 1) Does the available information provide support for the development of 30 appropriate ecological response to deposition relationship(s) that meaningfully relates 31 oxidized nitrogen and sulfur deposition to relevant ecological indicators? Does a March 2010 18 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 quantified relationship exist between the level of a relevant ecological indicator and 2 an amount of nitrogen and sulfur deposition? 3 3. (CPE 2) What are the important variables in the ecological response to deposition 4 relationship(s)? Are these relationships applicable nationally? What are the 5 appropriate temporal scales for these relationships? 6 a. How does ecological response to deposition relationship(s) depend upon spatially 7 heterogeneous geologic factors (e.g. bedrock type, weathering rates) that govern 8 sensitivity? 9 b. How do we consider areas with high natural background acidification or nutrient 10 loadings? 11 4. (CPE 3) Does the available information provide support for the development of 12 appropriate functions that characterize the relationships between atmospheric NOX 13 and SOX and the wet and dry deposition of total reactive nitrogen and sulfur? (CPE 4) 14 How do these relationships depend upon relevant atmospheric factors (e.g., reduced 15 forms of nitrogen, meteorological factors) and landscape factors? 16 a. What deposition function is appropriate to use for the purpose of relating an 17 amount of nitrogen and/or sulfur deposition in sensitive ecosystems to ambient 18 concentrations of atmospheric reactive nitrogen, including oxides and reduced 19 forms, and/or sulfur? What are the important variables in such a function? What 20 are appropriate spatial and temporal scales to use in specifying such variables? 21 Based on the conceptual framework for the structure of the ecologically relevant multi- 22 pollutant standards, we then address in Chapter 6 the elements of the standard needed to develop 23 options for consideration by the Administrator. Development of these options will focus on 24 addressing the following questions: 25 1. (CPE 2.1) What ecological indicators are appropriate to use for the purpose of 26 developing an alternative standard for the various ecological effects assessed in this 27 review? 28 2. (CPE 5) What indicators of oxides of nitrogen and sulfur are appropriate to use for 29 the purpose of determining whether the resultant deposition is within the target values 30 needed to achieve the desired degree of protection? What averaging times and forms 31 are appropriate to consider? March 2010 19 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 3. (CPE 4) What approaches are available to specify non-atmospheric elements of the 2 standard, e.g. weathering rates? Are there approaches that can simplify the structure 3 of the standard by using discrete representations (bins) of continuous variables? 4 4. What are the available approaches for accounting for reduced N in the structure of the 5 standard? 6 5. What is the most appropriate form for the standards to reflect the relationships 7 between ambient NOX and SOX, acidifying deposition, and the ecological indicator for 8 acidification? 9 Several follow-up questions derive from our assessment of options for specifying the 10 elements of a multipollutant standard. In Chapter 7, we address the questions: 11 1. To what extent would a standard specifically defined to protect against one ecological 12 effect (i.e., aquatic acidification) likely provide protection from other relevant 13 ecological effects? 14 2. What are the available approaches for combining multiple indicators into a single 15 standard, e.g. using nitrogen effects to bound the tradeoff curve for NOX/SOX for 16 aquatic acidification effects 17 3. What are the available approaches to integrate potential standards for aquatic and 18 terrestrial acidification and/or aquatic and terrestrial N enrichment? 19 In Chapter 8, we plan to address in the second draft policy assessment issues regarding 20 the adequacy of the current definitions of oxides of nitrogen and sulfur in specifying standards 21 for protection against effects associated with deposition of nitrogen and sulfur. This discussion 22 will be focused on the following questions: 23 1. To what extent are effects associated with atmospheric nitrogen deposition reduced 24 when NOX related deposition is reduced? 25 2. To what extent can appropriate protection from relevant ecological effects be 26 achieved by specifying indicators of atmospheric reactive nitrogen and sulfur 27 compounds in terms of gas- and particle-phase nitrogen oxides and/or sulfur oxides? 28 3. To what extent does the available information on welfare effects provide a basis for 29 considering expanding the list of criteria pollutants to include all reactive nitrogen or 30 gas-phase ammonia? What are the relative merits of listing total reactive nitrogen 31 versus gas phase ammonia for protection of public welfare effects? March 2010 20 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 We conclude with a discussion of a range of options to consider in selecting pollutant 2 indicators, averaging times, forms, and levels for the secondary NOX and SOX standards, 3 including a discussion of staff initial conclusions on what levels of the standard for NOX and SOX 4 would be requisite to protect public welfare against adverse ecological effects. This discussion is 5 informed by a consideration of the role of ecosystem services in helping to characterize what 6 adversity to public welfare, focused on the following questions: 7 1. (CPE 5) What are the risks of ecosystem service impairment under alternative levels 8 of potential standards for NOX and SOX? 9 2. (CPE 5) To what extent can information about ecosystem services be used to help 10 characterize the extent to which differing levels of relevant ecological indicators 11 reflect impacts that can reasonably be judged to be adverse from a public welfare 12 perspective? 13 3. (CPE 5) Are there relevant benchmarks for adversity to public welfare that can be 14 derived from other sources? 15 4. (CPE 5) Taking into consideration information about ecosystem services and other 16 factors related to characterizing adversity to public welfare for the ecological effects 17 being assessed in this review, what is an appropriate range of levels of protection to 18 be achieved by alternative standards for the Agency to consider? 19 March 2010 21 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2. KNOWN OR ANTICIPATED ECOLOGICAL EFFECTS 2 This chapter addresses Critical Policy Element 1, evaluation of the effects of ambient 3 NOX and SOX on ecosystems, and the relationship between those effects and the measure of dose 4 in the ecosystem, indicated by the deposit!onal loadings of N and S. In section 302(h) of the 5 Clean Air Act, welfare effects addressed by a secondary NAAQS include, but are not limited to, 6 "effects on soils, water, crops, vegetation, man-made materials, animals, wildlife, weather, 7 visibility and climate, damage to and deterioration of property, and hazards to transportation, as 8 well as effects on economic values and on personal comfort and well-being". Of these welfare 9 effects categories, the effects of NOX and SOX on aquatic and terrestrial ecosystems, which 10 encompass soils, water, vegetation, wildlife, and contribute to economic value and well-being, 11 are of most concern at concentrations typically occurring in the U.S. Direct effects of NOX and 12 SOX on vegetation are also discussed in this chapter, and have been the focus of previous 13 reviews. However, for this review, the focus of this chapter is on the known and anticipated 14 effects to ecosystems caused by exposure to NOX and SOX through deposition. 15 The information presented here is a concise summary of conclusions from the ISA and 16 the REA. This chapter focuses on effects on specific ecosystems with a brief discussion on 17 critical uncertainties associated with acidification and nutrient enrichment; Chapter 3 evaluates 18 those effects within the context of alternative definitions of, including assessments of potential 19 impacts on ecosystem services. Effects are broadly categorized into acidification and nutrient- 20 enrichment in the proceeding sections. This is background information intended to support new 21 approaches for the design of ecologically relevant secondary NOX and SOX standards which are 22 protective of U.S. ecosystems. More detailed information on the conceptual design and specific 23 options for the proposed standards are presented in Chapters 5 and 6 of this policy assessment 24 document. While we provide a summary of effects for all four of the primary effects categories, 25 we reiterate that the focus of this first draft policy assessment is on effects related to aquatic 26 acidification, without downplaying the potential significance of effects in other categories. March 2010 22 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2.1 ACIDIFICATION: EVIDENCE OF EFFECTS ON STRUCTURE AND 2 FUNCTION OF TERRESTRIAL AND FRESHWATER 3 ECOSYSTEMS 4 Sulfur oxides (SOX) and nitrogen oxides (NOX) compounds in the atmosphere undergo a 5 complex mix of reactions and thermodynamic processes in gaseous, liquid, and solid phases to 6 form various acidic compounds. These acidic compounds are removed from the atmosphere 7 through deposition: either wet (e.g., rain, snow), fog or cloud, or dry (e.g., gases, particles). 8 Deposition of these acidic compounds leads to ecosystem exposure and effects on ecosystem 9 structure and function. Following deposition, these compounds can, in some instances, leach out 10 of the soils in the form of sulfate (SC>42") and nitrate (N(V), leading to the acidification of surface 11 waters. The effects on ecosystems depend on the magnitude of deposition, as well as a host of 12 biogeochemical processes occurring in the soils and waterbodies (REA 2.1). The chemical forms 13 of nitrogen that may contribute to acidifying deposition include both oxidized and reduced 14 species. 15 When sulfur or nitrogen leaches from soils to surface waters in the form of SC>42" or N(V, 16 an equivalent amount of positive cations, or countercharge, is also transported. This maintains 17 electroneutrality. If the countercharge is provided by base cations, such as calcium (Ca2+), 18 magnesium (Mg2+), sodium (Na+), or potassium (K+), rather than hydrogen (H+) and dissolved 19 inorganic aluminum, the acidity of the soil water is neutralized, but the base saturation of the soil 20 is reduced. Continued SC>42 or N(V leaching can deplete the base cation supply of the soil. As 21 the base cations are removed, continued deposition and leaching of SO42" and/or NO3" (with 22 H+and A13+) leads to acidification of soil water, and by connection, surface water. A watershed's 23 ability to neutralize acidic deposition is determined by a host of biogeophysical factors, including 24 base cation concentrations, weathering rates, uptake by vegetation, rate of surface water flow, 25 soil depth, and bedrock. (REA 2.1) Some of these factors such as vegetation and soil depth are 26 highly variable over small spatial scales, but others vary over larger spatial scales like geology. 27 For the purpose of a national secondary standard, the most relevant characteristics are those that 28 are less variable over small scales. 29 Acidifying deposition of NOX and SOX and the chemical and biological responses 30 associated with these inputs vary temporally. Chronic or long-term deposition processes result in March 2010 23 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 increases of N and S and the associated effects of acidifying deposition in the time scale of years 2 to decades. Episodic or short term (i.e., hours or days) deposition refers to events in which the 3 level of the acid neutralizing capacity (ANC) of a lake or stream is temporarily lowered. In 4 aquatic ecosystems, short-term (i.e., hours or days) episodic changes in water chemistry can have 5 significant biological effects. Episodic chemistry refers to conditions during precipitation or 6 snowmelt events when proportionately more drainage water is routed through upper soil horizons 7 that tend to provide less acid neutralizing than was passing through deeper soil horizons (REA 8 4.2). Some streams and lakes may have chronic or base flow chemistry that is suitable for aquatic 9 biota, but may be subject to occasional acidic episodes with lethal consequences. 10 The following summary is a concise overview of the known or anticipated effects caused 11 by acidification to ecosystems within the United States. Acidification affects both terrestrial and 12 freshwater aquatic ecosystems. Terrestrial and aquatic processes are often linked; therefore 13 responses to the following questions address both types of ecosystems unless otherwise noted. 14 2.1.1 What is the nature of acidification related ecosystem responses to reactive 15 nitrogen and/ sulfur deposition? 16 The ISA concluded that deposition of SOX, NOX, and NHX leads to the acidification of 17 ecosystems (EPA 2008). In the process of acidification, geochemical components of terrestrial 18 and freshwater aquatic ecosystems are altered in a way that leads to effects on biological 19 organisms. Deposition to terrestrial ecosystems often moves through the soil and eventually 20 leaches into adjacent water bodies, moreover deposition to the land effects the water as well. 21 The scientific evidence is sufficient to infer a causal relationship between acidifying 22 deposition and effects on biogeochemistry and biota in aquatic ecosystems (ISA 4.2.2). The 23 strongest evidence comes from studies of surface water chemistry in which acidic deposition is 24 observed to alter sulfate and nitrate concentrations in surface waters, sum and surplus of base 25 cations, acid, ANC, inorganic aluminum, calcium, and surface water pH (ISA 3.2.3.2). 26 Consistent and coherent documentation from multiple studies on various species from all major 27 trophic levels of aquatic systems shows that geochemical alteration caused by acidification can 28 result in the loss of acid-sensitive biological species (ISA 3.2.3.3). For example, in the 29 Adirondacks, of the 53 fish species recorded in Adirondack lakes about half (26 species) were 30 absent from lakes with pH below 6.0 (Baker et al., 1990b). Biological effects are linked to March 2010 24 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 changes in water chemistry including ANC, inorganic Al, and pH. Decreases in ANC and pH 2 and increases in inorganic Al concentration contribute to declines in taxonomic richness of 3 zooplankton, macroinvertebrates, and fish, which often are sources of food for birds and other 4 animal species in the ecosystem, as well as serving as a source of food and recreation for 5 humans. Acidification of ecosystems has been shown to disrupt food web dynamics causing 6 alteration to the diet, breeding distribution and reproduction of certain species of birds (ISA 7 4.2.2.2. and Table 3-9). For example, breeding distribution of the common goldeneye 8 (Bucephala clangula) an insectivorous duck, may be affected by changes in acidifying deposition 9 (Longcore and Gill, 1993). Similarly, reduced prey diversity and quantity have been observed to 10 create feeding problems for nesting pairs of loons on low-pH lakes in the Adirondacks (Parker 11 1988). 12 In terrestrial ecosystems, the evidence is sufficient to infer a causal relationship between 13 acidifying deposition and changes in biogeochemistry (ISA 4.2.1.1). The strongest evidence 14 comes from studies of forested ecosystems, with supportive information on other plant 15 communities, including shrubs and lichens (ISA 3.2.2.1.). Three useful indicators of chemical 16 changes and acidification effects on terrestrial ecosystems, showing consistency and coherence 17 among multiple studies: soil base saturation, Al concentrations in soil water and soil C:N ratio 18 (ISA 3.2.2.2). 19 In soils with base saturation less than about 15 to 20% exchange ion chemistry is 20 dominated by Al (Reuss, 1983). Under this condition, responses to inputs of sulfuric acid and 21 nitric acid largely involve the release and mobilization of inorganic Al through cation exchange. 22 The effect can be neutralized by weathering from geologic parent material or base cation 23 exchange. The Ca2+ and Al in soils are strongly influenced by soil acidification and both have 24 been shown to have quantitative links to tree health, including Al interference with Ca2+uptake 25 and Al toxicity to roots (Parker et al., 1989; U.S. EPA, 2009). Effects of nitrification and 26 associated acidification and cation leaching have been consistently shown to occur only in soils 27 with a C:N ratio below about 20 to 25 (Aber et al., 2003; Ross et al., 2004). 28 Acidification has been shown to cause decreased growth and increased susceptibility to 29 disease and injury in sensitive tree species. Red spruce (Picea rubens) dieback or decline has 30 been observed across high elevation areas in the Adirondack, Green and White mountains 31 (DeHayes et al., 1999). The frequency of freezing injury to red spruce needles has increased over March 2010 25 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 the past 40 years, a period that coincided with increased emissions of S and N oxides and 2 increased acidifying deposition (DeHayes et al., 1999). Acidifying deposition may be 3 contributing to episodic dieback in Sugar maple {Acer saccharum) through depletion of nutrient 4 cations from marginal soils (Horsley et al., 2000; Bailey et al., 2004). Grasslands are likely less 5 sensitive to acidification than forests (Blake et al., 1999; Kocky and Wilson 2001). 6 2.1.2 What types of ecosystems are sensitive to such effects? In which ways are 7 these responses affected by atmospheric, ecological, and landscape factors? 8 The intersection between current deposition loading, historic loading, and sensitivity 9 defines the ecological vulnerability to the effects of acidification. Freshwater aquatic and 10 terrestrial ecosystems are the ecosystem types which are most sensitive to acidification. The ISA 11 reports that the principal factor governing the sensitivity of terrestrial and aquatic ecosystems to 12 acidification from sulfur and nitrogen deposition is geology (particularly surficial geology). 13 Geologic formations having low base cation supply generally underlie the watersheds of acid- 14 sensitive lakes and streams. Other factors that contribute to the sensitivity of soils and surface 15 waters to acidifying deposition include topography, soil chemistry, land use, and hydrologic 16 flowpath. Episodic and chronic acidification tends to occur at relatively high elevation in areas 17 that have base-poor bedrock, high relief, and shallow soils (ISA 3.2.4.1). 18 2.1.3 What is the magnitude of ecosystem responses to acidifying deposition? 19 Terrestrial and aquatic ecosystems differ in their response to acidifying deposition. 20 Therefore the magnitude of ecosystem response is described separately for aquatic and terrestrial 21 ecosystems in the following sections. The magnitude of response refers to both the severity of 22 effects and the spatial extent of the U.S. which is affected. 23 2.1.3.1 Aquatic 24 Freshwater ecosystem surveys and monitoring in the eastern United States have been 25 conducted by many programs since the mid-1980s, including EPA's Environmental Monitoring 26 and Assessment Program (EMAP), National Surface Water Survey (NSWS), Temporally 27 Integrated Monitoring of Ecosystems (TIME) (Stoddard, 1990), and Long-term Monitoring 28 (LTM) (Ford et al., 1993; Stoddard et al., 1996) programs. Based on analyses of surface water 29 data from these programs, New England, the Adirondack Mountains, the Appalachian Mountains March 2010 26 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 (northern Appalachian Plateau and Ridge/Blue Ridge region), and the Upper Midwest contain 2 the most sensitive lakes and streams (i.e., ANC less than about 50 ueq/L) since the 1980s. 3 Portions of northern Florida also contain many acidic and low-ANC lakes and streams, although 4 the role of acidifying deposition in these areas is less clear. The western U.S. contains many of 5 the surface waters most sensitive to potential acidification effects, but with the exception of the 6 Los Angeles Basin and surrounding areas, the levels of acidifying deposition are low in most 7 areas. Therefore acidic surface waters are uncommon in the western U.S., and the extent of 8 chronic surface water acidification that has occurred in that region to date has likely been very 9 limited (ISA 3.2.4.2 and REA 4.2.2). 10 There are a number of species including fish, aquatic insects, other invertebrates and 11 algae that are sensitive to acidification and cannot survive, compete, or reproduce in acidic 12 waters (ISA 3.2.3.3). Decreases in ANC and pH have been shown to contribute to declines in 13 species richness and abundance of zooplankton, macroinvertebrates, and fish (Keller and Gunn 14 1995; Schindler et al., 1985). Reduced growth rates have been attributed to acid stress in a 15 number offish species including Atlantic salmon (Salmo salar), Chinook salmon (Oncorhynchus 16 tshawytscha\ lake trout (Salvelinus namaycush\ rainbow trout (Oncorhynchis mykiss), brook 17 trout (Salvelinus Fontinalis\ and brown trout (Salmo trutta) (Baker et al., 1990). In response to 18 small to moderate changes in acidity, acid-sensitive species are often replaced by other more 19 acid-tolerant species, resulting in changes in community composition and richness. The effects of 20 acidification are continuous, with more species being affected at higher degrees of acidification. 21 At a point, typically a pH <4.5 and an ANC <0 ueq/L, complete to near-complete loss of many 22 classes of organisms occur, including fish and aquatic insect populations, whereas others are 23 reduced to only a few acidophilic forms. These changes in species integrity are because energy 24 cost in maintaining physiological homeostasis, growth, and reproduction is high at low ANC 25 levels (Schreck, 1981, 1982; Wedemeger et al., 1990; REA appendix 2.3). Decreases in species 26 richness related to acidification have been observed in the Adirondack Mountains and Catskill 27 Mountains of New York (Baker et al., 1996), New England and Pennsylvania (Haines and Baker, 28 1986), and Virginia (Bulger et al., 2000). 29 From the sensitive areas identified by the ISA, further "case study" analyses on aquatic 30 ecosystems in the Adirondack Mountains and Shenandoah National Park were conducted to 31 better characterize ecological risk associated with acidification (REA Chapter 4). March 2010 27 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 In the literature, ANC is the most widely used indicator of acid sensitivity and has been 2 found in various studies to be the best single indicator of the biological response and health of 3 aquatic communities in acid-sensitive systems (Lien et al., 1992; Sullivan et al., 2006; ISA). In 4 the REA, surface water trends in SC>42" and NCV concentrations and ANC levels were analyzed 5 to affirm the understanding that reductions in deposition could influence the risk of acidification. 6 ANC values were categorized according to their effects on biota, as shown in Figure 2-1. 7 Monitoring data from the EPA-administered TEVIE/LTM and EMAP programs were assessed for 8 the years 1990 to 2006, and past, present, and future water quality levels were estimated by both 9 steady-state and dynamic biogeochemical models. 10 11 12 13 14 15 16 Category Label ANC Levels' Expected Ecological Effects Acute Concern Severe Concern Elevated Concern Moderate Concern Low Concern <0 |.ieq.'L (Acidic) 0-20 20-50 ueq/L 50-100 :-100 ueq/L Near complete loss offish population? is expected. Planktonic communities have extremely low diversity and are dominated by acidophihc forms. The number of individuals in plankton species that are present is greatly reduced. Highly sensitive to episodic acidification. During episodes of high acidifying deposition, brook trout populations may experience lethal effects. Diversity and distribution of zooplankton communities decline sharply. Fish species richness is greatly reduced (i.e., more than half of expected species can be missing). On average, brook trout populations experience sublethal effects, including loss of health, reproduction capacity, and fitness. Diversity and distribution of zooplankton communities decline. Fish species richness begins to decline (i.e., sensitive species are lost from lakes). Brook trout populations are sensitive and variable, with possible sublethal effects. Diversity and distribution of zooplankton communities also begin to decline as species that are sensitive to acidifying deposition are affected. Fish species richness may be unaffected, Reproducing brook trout populations are expected where habitat is suitable. Zooplankton communities are unaffected and exhibit expected diversity and distribution. Figure 2-1. Ecological Effects Associated with Alternative Levels of Acid Neutralizing Capacity (ANC) The analyses of the Adirondack Case Study Area indicated that although wet deposition rates for 862 and NOX have been reduced since the mid-1990s, current concentrations are still well above pre-acidification (1860) conditions. Modeling predicts NCV and SO42" are 17- and 5- March 2010 28 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 4 5 6 7 9 10 11 12 fold higher today, respectively. The estimated average ANC across the 44 lakes in the Adirondack Case Study Area is 62.1 ueq/L (± 15.7 ueq/L); 78 % of all monitored lakes in the Adirondack Case Study Area have a current risk of Elevated, Severe, or Acute. Of the 78%, 31% experience episodic acidification, and 18% are chronically acidic today (REA 4.2.4.2). Based on a deposition scenario that maintains current emission levels to 2020 and 2050, the simulation forecast indicates no improvement in water quality in the Adirondack Case Study Area. The percentage of lakes within the Elevated to Acute Concern classes remains the same in 2020 and 2050. o 140 ^ 120 Id100 80 o 40 20 0 1850 1900 1950 2000 2050 Figure 2-2. Average N(V concentrations (orange), SC>42" concentrations (red), and ANC (blue) across the 44 lakes in the Adirondack Case Study Area modeled using MAGIC for the period 1850 to 2050. March 2010 29 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 4 5 ANC Preacidification (1860) and Current Condition (2006) Preacidification (1860) ANC Source: EPA 2009 >0 0-20 20-50 50-100 >100 Current (2006) Figure 2-3. ANC concentrations of preacidification (1860) and current (2006) conditions based on hindcasts of 44 lakes in the Adirondack Case Study Area modeled using MAGIC. [Note: in this map, the symbol for red is reversed and should be < 0. The figure will be revised in the next draft.] March 2010 30 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX Current Condition of Acidity and Sensitivity Criticial Load meq/m2/yr • Highly Sensitive: < 50 Moderately Sensitive: 51 -100 Low Sensitivity: 101 -200 • Not Sensitive: > 201 | | Adirondack Boundary Source: EPA 2009 1 2 Figure 2-4. Critical loads of acidifying deposition that each surface water location 3 can receive in the Adirondack Case Study Area while maintaining or exceeding 4 an ANC concentration of 50 ueq/L based on 2002 data. Watersheds with critical 5 load values <100 meq/m2/yr (red and orange circles) are most sensitive to surface 6 water acidification, whereas watersheds with values >100 meq/m2/yr (yellow and 7 green circles) are the least sensitive sites. 8 It is important to note that studies on fish species richness in the Adirondacks Case Study 9 Area demonstrated the effect of acidification; of the 53 fish species recorded in Adirondack Case 10 Study Area lakes, only 27 species were found in lakes with a pH <6.0. The 26 species missing 11 from lakes with a pH <6.0 include important recreational species, such as Atlantic salmon, tiger 12 trout (Salmo trutta X Salvelinusfontinalis), redbreast sunfish (Lepomis auritus), bluegill 13 (Lepomis macrochims), tiger musky (Esox masquinongy X Indus), walleye (Sander vitreus), 14 alewife (Alosapseudoharengus), and kokanee (Oncorhynchus nerkd) (Kretser et al., 1989), as 15 well as ecologically important minnows that are commonly eaten by sport fish. A survey of 16 1,469 lakes in the late 1980s found 346 lakes to be devoid offish. Among lakes with fish, there 17 was a relationship between the number offish species and lake pH, ranging from about one 18 species per lake for lakes having a pH <4.5 to about six species per lake for lakes having a pH March 2010 31 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 4 5 6 7 22 23 24 25 >6.5 (Driscoll et al., 2001; Kretser et al., 1989). In the Adirondacks, a positive relationship exists between the pH and ANC in lakes and the number offish species present in those lakes (ISA 3.2.3.4). Since the mid-1990s, streams in the Shenandoah Case Study Area have shown slight declines in N(V and SC>4 2" concentrations in surface waters. Current concentrations are still above pre-acidification (1860) conditions. MAGIC modeling predicts surface water concentrations of NCV and SC>42" arelO- and 32-fold higher today, respectively. The estimated 8 average ANC across 60 streams in the Shenandoah Case Study Area is 57.9 ueq/L (± 4.5 ueq/L). 9 55% of all monitored streams in the Shenandoah Case Study Area have a current risk of 10 Elevated, Severe, or Acute. Of the 55%, 18% experience episodic acidification, and 18% are 11 chronically acidic today (REA 4.2.4.3) 12 Based on a deposition scenario that maintains current emission levels to 2020 and 2050, 13 the simulation forecast indicates that a large number of streams still have Elevated to Acute 14 problems with acidity. In fact, from 2006 to 2050, the percentage of streams with Acute Concern 15 increases by 5%, while the percentage of streams in Moderate Concern decreases by 5%. 16 Biological effects of increased acidification documented in the Shenandoah Case Study 17 Area include a reduction in the condition factor in Blacknose Dace (Dennis and Bulgar 1995, 18 Bulgar et al., 1999) and a decrease in fish biodiversity associated with decreasing stream ANC 19 (Bulger et al., 1995; Dennis and Bulger, 1995; Dennis et al., 1995; MacAvoy and Bulger, 1995, 20 Bulgar et al., 1999). On average, the fish species richness is lower by one fish species for every 21 21 ueq/L decrease in ANC in Shenandoah National Park streams (ISA 3.2.3.4). 120 1850 1900 1950 Years 2000 2050 2- Figure 2-5. Average NOs" concentrations orange), SO4 "concentrations (red), and ANC (blue) levels for the 60 streams in the Shenandoah Case Study Area modeled using MAGIC for the period 1850 to 2050. March 2010 32 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 2 3 4 5 ANC Preacidification (1860) and Current Condition (2006) Pre-acidification (1860) Current (2006) Source: EPA 2009 ANC <0 0-20 20-50 50 - 100 >100 Figure 2-6. ANC levels of 1860 (preacidification) and 2006 (current) conditions based on hindcasts of 60 streams in the Shenandoah Case Study Area modeled using MAGIC. March 2010 33 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX Current Condition of Acidity and Sensitivity Criticial Load meq/m2/yr • Highly Sensitive: < 50 Moderately Sensitive: 51-100 Low Sensitivity: 101 -200 • Not Sensitive: > 201 Source: EPA 2009 1 2 Figure 2-7. Critical loads of surface water acidity for an ANC of 50 ueq/L for 3 Shenandoah Case Study Area streams. Each dot represents an estimated amount 4 of acidifying deposition (i.e., critical load) that each stream's watershed can 5 receive and still maintain a surface water ANC >50 ueq/L. Watersheds with 6 critical load values <100 meq/m2/yr (red and orange circles) are most sensitive to 7 surface water acidification, whereas watersheds with values >100 meq/m2/yr 8 (yellow and green circles) are the least sensitive sites. 9 2.1.3.2 Terrestrial Acidification 10 The ISA identified a variety of indicators that can be used to measure the effects of 11 acidification in soils. Tree health has been linked to base cations (Be) in soil (such as Ca2+, Mg2+ 12 and potassium), as well as soil Al content. Tree species show similar sensitivities to Ca/Al and 13 Bc/Al soil solution ratios, therefore these are good chemical indicators because they directly 14 relate to the biological effects. Critical Bc/Al ratios for a large variety of tree species ranged 15 from 0.2 to 0.8 (Sverdrup and Warfvinge, 1993, a meta-data analysis of laboratory and field 16 studies). This range is similar to critical ratios of Ca/Al. Plant toxicity or nutrient antagonism 17 was reported to occur at Ca/Al ratios ranging from 0.2 to 2.5 (Cronan and Grigal, 1995; meta- 18 data assessment) (REA pg 4-54, REA Appendix 5). March 2010 34 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 There has been no systematic national survey of terrestrial ecosystems to determine the 2 extent and distribution of terrestrial ecosystem sensitivity to the effects of acidifying deposition. 3 However, one preliminary national evaluation estimated that -15% of forest ecosystems in the 4 U.S. exceeds the estimated critical load based on soil chemistry for S and N deposition by >250 5 eq ha"1 yr"1 (McNulty et al., 2007). Forests of the Adirondack Mountains of New York, Green 6 Mountains of Vermont, White Mountains of New Hampshire, the Allegheny Plateau of 7 Pennsylvania, and high-elevation forest ecosystems in the southern Appalachians are the regions 8 most sensitive to terrestrial acidification effects from acidifying deposition (ISA 3.2.4.2). While 9 studies show some recovery of surface waters, there are widespread measurements of ongoing 10 depletion of exchangeable base cations in forest soils in the northeastern U.S. despite recent 11 decreases in acidifying deposition, indicating a slow recovery time. 12 In the REA, a critical load analysis was performed for sugar maple and red spruce forests 13 in the eastern United States by using Bc/Al ratio in acidified forest soils as an indicator to assess 14 the impact of nitrogen and sulfur deposition on tree health. These are the two most commonly 15 studied species in North America for effects of acidification. At a Bc/Al ratio of 1.2, red spruce 16 growth can be reduced by 20%. Sugar maple growth can be reduced by 20% at a Bc/Al ratio of 17 0.6. The REA analysis determined the health of at least a portion of the sugar maple and red 18 spruce growing in the United States may have been compromised with acidifying total nitrogen 19 and sulfur deposition in 2002. Specifically, total nitrogen and sulfur deposition levels exceeded 20 three selected critical loads for tree growth in 3% to 75% of all sugar maple plots across 24 21 states. For Red Spruce, total nitrogen and sulfur deposition levels exceeded three selected critical 22 loads in 3% to 36% of all red spruce plots across eight states. 23 2.1.4 What are the key uncertainties associated with acidification? 24 There are different levels of uncertainty associated with relationships between deposition, 25 ecological effects and ecological indicators. In Chapter 7 of the REA, key uncertainties are 26 characterized as follows to evaluate the strength of the scientific basis for setting a national 27 standard to protect against a given effect (REA 7.0): 28 • Data Availability: high, medium or low quality. This criterion is based on the availability 29 and robustness of data sets, monitoring networks, availability of data that allows for 30 extrapolation to larger assessment areas, and input parameters for modeling and March 2010 35 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 developing the ecological effect function. The scientific basis for the ecological indicator 2 selected is also incorporated into this criterion. 3 • Modeling Approach: high, fairly high, intermediate, or low confidence. This value is 4 based on the strengths and limitations of the models used in the analysis and how accepted 5 they are by the scientific community for their application in this analysis. 6 • Ecological Effect Function: high, fairly high, intermediate, or low confidence. This 7 ranking is based on how well the ecological effect function describes the relationship 8 between atmospheric deposition and the ecological indicator of an effect. 9 2.1.4.1 Aquatic Acidification 10 The REA concludes that the available data are robust and considered high quality. There 11 is high confidence about the use of these data and their value for extrapolating to a larger 12 regional population of lakes. The EPA TIME/LTM network represents a source of long-term, 13 representative sampling. Data on sulfate concentrations, nitrate concentrations and ANC from 14 1990 to 2006 used for this analysis as well as EPA EMAP and REMAP surveys, provide 15 considerable data on surface water trends. 16 There is fairly high confidence associated with modeling and input parameters. 17 Uncertainties in water quality estimates (.i.e. ANC) from MAGIC was derived from multiple site 18 calibrations. The 95% confidence interval for pre-acidification of lakes was an average of 15 19 |j,eq/L difference in ANC concentrations or 10% and 8 |j,eq/L or 5% for streams (REA 7.1.2) The 20 use of the critical load model used to estimate aquatic critical loads is limited by the uncertainties 21 associated with runoff and surface water measurements and in estimating the catchment supply 22 of base cations from the weathering of bedrock and soils (McNulty et al., 2007). To propagate 23 uncertainty in the model parameters, Monte Carlo methods were employed to develop an inverse 24 function of exceedences. There is high confidence associated with the ecological effect function 25 developed for aquatic acidification. In calculating the ANC function, the depositional load for N 26 or S is fixed by the deposition of the other, so deposition for either will never be zero (Figure 27 7.1-6 REA). 28 Terrestrial Acidification 29 The available data used to quantify the targeted effect of terrestrial acidification are 30 robust and considered high quality. The USFS-Kane experimental forest and significant amounts March 2010 36 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 of research work in the Allegheny Plateau have produced extensive, peer-reviewed datasets. A 2 meta-analysis of laboratory studies showed that tree growth was reduced by 20% relative to 3 controls for BC/A1 ratios (ISA 7.2.1 and Figure 7.2-1). Sugar maple and red spruce were the 4 focus of the REA since they are demonstrated to be negatively affected by Ca2+ depletion and 5 high concentrations of available Al, and occur in areas that receive high acidifying deposition, 6 There is high confidence about the use of the REA terrestrial acidification data and their value 7 for extrapolating to a larger regional population of forests. 8 There is high confidence associated with the models, input parameters, and assessment of 9 uncertainty used in the case study for terrestrial acidification. The Simple Mass Balance (8MB) 10 model, a commonly used and widely applied approach for estimating critical loads, was used in 11 the REA analysis (ISA 7.2.2). There is fairly high confidence associated with the ecological 12 effect function developed for terrestrial acidification (REA 7.2.3). 13 2.2 NITROGEN ENRICHMENT: EVIDENCE OF EFFECTS ON 14 STRUCTURE AND FUNCTION OF TERRESTRIAL AND 15 FRESHWATER ECOSYSTEMS 16 The following summary is a concise overview of the known or anticipated effects caused 17 by nitrogen nutrient enrichment to ecosystems within the United States. Nutrient-enrichment 18 affects terrestrial, freshwater and estuarine ecosystems. Nitrogen deposition is often the main 19 source of anthropogenic nitrogen in terrestrial and freshwater ecosystems. In contrast, nitrogen 20 deposition often contributes to nitrogen-enrichment effects in estuaries, but does not drive the 21 effects. Both oxides of nitrogen and reduced forms of nitrogen, e.g. NHX, contribute to nitrogen 22 deposition. For the most part, nitrogen effects on ecosystems do not depend on whether the 23 nitrogen is in oxidized or reduced form. Thus, this summary focuses on the effects of nitrogen 24 deposition in total. We address the issue of incorporating the relative contributions of oxidized 25 and reduced nitrogen into the standards in Chapters 5, 6, and 8. 26 2.2.1 What is the nature of terrestrial and freshwater ecosystem responses to 27 reactive nitrogen and/ sulfur deposition? 28 The ISA found that deposition of nitrogen, including NOX and NHX leads to the nitrogen 29 enrichment of ecosystems (EPA 2008). In the process of nitrogen enrichment, geochemical March 2010 37 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 components of terrestrial and freshwater aquatic ecosystems are altered in a way that leads to 2 effects on biological organisms. 3 The evidence is sufficient to infer a causal relationship between N deposition and the 4 alteration of biogeochemical cycling in terrestrial ecosystems (ISA 4.3.1.1 and 3.3.2.1). This is 5 supported by numerous observational, deposition gradient and field addition experiments. 6 Stoddard (1994) identified the leaching of N(V in soil drainage waters and the export of N(V in 7 steam water as two of the primary indictors of N enrichment. Several N-addition studies indicate 8 that MV leaching is induced by chronic additional of N (Edwards et al., 2002b; Kahl et al., 9 1999; Peterjohn et al., 1996; Norton et al., 1999). Aber et al. (2003) found that surface water 10 MV concentrations exceeded 1 |j,eq/L in watersheds receiving about 9 to 13 kg N/ha/yr of 11 atmospheric N deposition. N deposition disrupts the nutrient balance of ecosystems with 12 numerous biogeochemical effects. The chemical indicators that are typically measured include 13 NO3- leaching, C:N ratio, N mineralization, nitrification, denitrification, foliar N concentration, 14 and soil water NOs - and NH4+ concentrations. Note that N saturation (N leaching from 15 ecosystems) does not need to occur to cause effects. Substantial leaching of NOs- from forest 16 soils to stream water can acidify downstream waters, leading to effects described in the previous 17 section on aquatic acidification. Due to the complexity of interactions between the N and C 18 cycling, the effects of N on C budgets (quantified input and output of C to the ecosystem) are 19 variable. Regional trends in net ecosystem productivity (NEP) of forests (not managed for 20 silviculture) have been estimated through models based on gradient studies and meta-analysis. 21 Atmospheric N deposition has been shown to cause increased litter accumulation and carbon 22 storage in above-ground woody biomass. In the West, this has lead to increased susceptibility to 23 more severe fires. Less is known regarding the effects of N deposition on C budgets of non- 24 forest ecosystems. 25 The evidence is sufficient to infer a causal relationship between N deposition on the 26 alteration of species richness, species composition and biodiversity in terrestrial ecosystems (ISA 27 4.3.1.2). The most sensitive terrestrial taxa are lichens. Empirical evidence indicates that lichens 28 in the U.S. are affected by deposition levels as low as 3 kg N/ha/yr. Alpine ecosystems are also 29 sensitive to N deposition, changes in an individual species (Carex rupestris) were estimated to 30 occur at deposition levels near 4 kg /ha/yr and modeling indicates that deposition levels near 10 31 kg N/ha/yr alter plant community assemblages. In several grassland ecosystems, reduced species March 2010 38 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 diversity and an increase in non-native, invasive species are associated with N deposition (Clark 2 and Tillman, 2008; Schwinning et al., 2005). 3 In freshwater ecosystems, the evidence is sufficient to infer a causal relationship between 4 N deposition and the alteration of biogeochemical cycling in freshwater aquatic ecosystems (ISA 5 3.3.2.3). N deposition is the main source of N enrichment to headwater streams, lower order 6 streams and high elevation lakes. The most common chemical indicators that were studied 7 included NOs- and dissolved inorganic nitrogen (DIN) concentration in surface waters as well as 8 Chi a:total P ratio. Elevated surface water NOs- concentrations occur in both the eastern and 9 western U.S. Bergstrom and Jansson (2006) report a significant correlation between N deposition 10 and lake biogeochemistry by identifying a correlation between wet deposition and [DIN] and Chi 11 a: Total P. Recent evidence provides examples of lakes and streams that are limited by N and 12 show signs of eutrophication in response to N addition. 13 The evidence is sufficient to infer a causal relationship between N deposition and the 14 alteration of species richness, species composition and biodiversity in freshwater aquatic 15 ecosystems (ISA 3.3.5.3). Increased N deposition can cause a shift in community composition 16 and reduce algal biodiversity, especially in sensitive oligotrophic lakes. 17 2.2.2 What types of ecosystems are sensitive to such effects? How are these 18 responses affected by atmospheric, ecological, and landscape factors 19 The numerous ecosystem types that occur across the U.S. have a broad range of 20 sensitivity to N deposition. Organisms in their natural environment are commonly adapted to a 21 specific regime of nutrient availability. Change in the availability of one important nutrient, such 22 as N, may result in imbalance in ecological stoichiometry, with effects on ecosystem processes, 23 structure and function (Sterner and Elser, 2002). In general, N deposition to terrestrial 24 ecosystems causes accelerated growth rates in some species, which may lead to altered 25 competitive interactions among species and nutrient imbalances, ultimately affecting 26 biodiversity. The onset of these effects occurs with N deposition levels as low as 3 kg N/ha/yr in 27 sensitive terrestrial ecosystems. In aquatic ecosystems, N that is both leached from the soil and 28 directly deposited can pollute surface water. This causes alteration of the diatom community at 29 levels as low as 1.5 kg N/ha/yr in sensitive freshwater ecosystems. March 2010 39 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 The degree of ecosystem effects lies at the intersection of N loading and N-sensitivity. N- 2 sensitivity is predominately driven by the degree to which growth is limited by nitrogen 3 availability. Grasslands in the western United States are typically N-limited ecosystems 4 dominated by a diverse mix of perennial forbs and grass species (Clark and Tilman, 2008; 5 Suding et al., 2005). A meta-analysis by Lebauer and Treseder (2008) indicated that N 6 fertilization increased aboveground growth in all non-forest ecosystems except for deserts. In 7 other words, almost all terrestrial ecosystems are N-limited and will be altered by the addition of 8 anthropogenic nitrogen. Likewise, a freshwater lake or stream must be N-limited to be sensitive 9 to N-mediated eutrophication. There are many examples of fresh waters that are N-limited or N 10 and P co-limited (ISA 3.3.3.2). In a meta-analysis that included 653 datasets, Elser et al. (2007) 11 found that N-limitation occurred as frequently as P-limitation in freshwater ecosystems. 12 Additional factors that govern the sensitivity of ecosystems to nutrient enrichment from N 13 deposition include rates and form of N deposition, elevation, climate, species composition, 14 length of growing season, and soil N retention capacity. (ISA 4.3). Less is known about the 15 extent and distribution of the terrestrial ecosystems in the U.S. that are most sensitive to the 16 effects of nutrient enrichment from atmospheric N deposition compared to acidification. 17 2.2.3 What is the magnitude of ecosystem responses to nitrogen deposition? 18 2.2.3.1 Terrestrial 19 Little is known about the full extent and distribution of the terrestrial ecosystems in the 20 U.S. that are most sensitive to impacts caused by nutrient enrichment from atmospheric N 21 deposition. As previously stated, most terrestrial ecosystems are N-limited, therefore they are 22 sensitive to perturbation caused by N additions (LeBauer and Treseder, 2008). Effects are most 23 likely to occur where areas of relatively high atmospheric N deposition intersect with N-limited 24 plant communities. The alpine ecosystems of the Colorado Front Range, chaparral watersheds of 25 the Sierra Nevada, lichen and vascular plant communities in the San Bernardino Mountains and 26 the Pacific Northwest, and the southern California coastal sage scrub (CSS) community are 27 among the most sensitive terrestrial ecosystems. There is growing evidence that existing 28 grassland ecosystems in the western United States are being altered by elevated levels of N 29 inputs, including inputs from atmospheric deposition (Clark and Tilman, 2008; Suding et al., 30 2005). March 2010 40 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 In the eastern U.S., the degree of N saturation of the terrestrial ecosystem is often 2 assessed in terms of the degree of NOs- leaching from watershed soils into ground water or 3 surface water. Stoddard (1994) estimated the number of surface waters at different stages of 4 saturation across several regions in the eastern U.S. Of the 85 northeastern watersheds examined 5 60% were in Stage 1 or Stage 2 of N saturation on a scale of 0 (background or pretreatment) to 3 6 (visible decline). Of the northeastern sites for which adequate data were available for assessment, 7 those in Stage 1 or 2 were most prevalent in the Adirondack and Catskill Mountains. Effects on 8 individual plant species have not been well studied in the U.S. More is known about the 9 sensitivity of particular plant communities. Based largely on results obtained in more extensive 10 studies conducted in Europe, it is expected that the more sensitive terrestrial ecosystems include 11 hardwood forests, alpine meadows, arid and semi-arid lands, and grassland ecosystems (ISA 12 3.8.2). 13 The REA used published research results (REA 5.3.1 and ISA Table 4.4) to identify 14 meaningful ecological benchmarks associated with different levels of atmospheric nitrogen 15 deposition. These are given by figure 2-8. The sensitive areas and ecological indicators identified 16 by the ISA were analyzed further in the REA to create a national map that illustrates effects 17 observed from ambient and experimental atmospheric nitrogen deposition loads in relation to 18 CMAQ 2002 modeling results and NADP monitoring data. This map, reproduced in Figure 2-9, 19 depicts the sites where empirical effects of terrestrial nutrient enrichment have been observed 20 and site proximity to elevated atmospheric N deposition. 21 March 2010 41 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX Rocky Mountain alpine lakes: shift in diatom community dominance (Baron, 2006) • Southern California: CSS loss (Wood et al., 2006) • San Bernardino Mountains and Sierra Nevada Mountains: acidophytic lichen decline in MCF (Fenn et al., 2008) • Eastern Rocky Mountain Slope: low carbon:nitrogen; low lignin:nitrogen (Baron et al.,2000) • Eastern Rocky Mountain Slope: increased foliar nitrogen; increased mineralization (Baron et al., 2000) • San Bernardino Mountains and Sierra Nevada Mountains: shift from acidophytic to neutral or nitrogen-tolerant lichen in MCF (Fenn et al., 2008) • Minnesota grasslands: decreased plant species (Clark and Tilman, 2008) • Northeast U.S.: NO3 leaching (Aber et al., 2003) Bay Area, CA: Increased cover of nonnative grasses; decreased native grasses (Weiss, 1999) San Bernardino Mountains and Sierra Nevada Mountains: loss of acidophytic lichen in MCF (Fenn et al., 2008) Southern California: shift in mycorrhizal species in CSS (Egerton-Warburton and Allen, 2000) Southern California: shift from native species to invasive grasses in CSS (Allen, 2008) • San Bernardino Mountains: high dissolved organic nitrogen (Meixner and Fenn, 2004) • San Bernardino Mountains: nitrogen saturation (Fenn et al., 2000) • Increased nitrogen in lichen (Fenn et al., 2007) MCF: N03 leaching (Fenn et al., 2008) MCF: 25% decrease in fine-root biomass (Fenn et al., 2008) • Southern California: NO3" leaching (Fenn et al., 2003) • Southern California: high foliar nitrogen (Bytnerowicz and Fenn, 1996) • Los Angeles Basin, California: High NO emissions (Bytnerowicz and Fenn, 1996) Fraser Experimental Forest, CO: increased foliar nitrogen; increased mineralization (Rueth et al., 2003) 1 2 3 0246 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 Nitrogen Deposition, kg/ha/yr Figure 2-8. Benchmarks of atmospheric nitrogen deposition for several ecosystem indicators with the inclusion of the diatom changes in the Rocky Mountain lakes (REA 5.3.1.2) March 2010 42 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 2. 3. 4. 5. 6. 7. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Legend Total N Deposition High S6M7 Low 0.761 ) Loutont I Mitonil P»*i Nitrogen enrichment or eutrophication of lakes (Loch Vale, CO: 0.5 to1.5 kg/ha/yr; Niwot Ridge, CO: 4.71 kg/ha/yr) Alpine lakes increase shift in diatom species (Rocky Mountains, CO: 2 kg/ha/yr) Alpine meadows' elevated NOr levels in runoff (Colorado Front Range: 20, 40, 60 kg/ha/yr) Alpine meadows' shift toward hairgrass (Niwot Ridge, CO: 25 kg/ha/yr) Nitrogen enrichment or nitrogen saturation (e.g., soil and foliar nitrogen concentration) (eastern slope of Rocky Mountains: 1.2, 3.6 kg/ha/yr; Fraser Forest, CO: 3.2 to 5.5 kg/ha/yr) Increased nitrogen mineralization rates and nitrification (Loch Vale, CO (spruce): 1.7 kg/ha/yr) Alpine tundra with increased plant foliage and decreased species richness (Niwot Ridge, CO: 50 kg/ha/yr) Nitrogen saturation, high N0s~ in streamwater, soil, leaves; high nitric oxide (NO) emissions (Los Angeles, CA, air basin: saturation at 24 to 25 kg/ha/yr (dry) and at 0.8 to 45 kg/ha/yr (wet); northeastern U.S.: 3.3 to 12.7 kg/ha/yr) Nitrogen saturation, high NOs- in streamwater (San Bernardino Mountains, CA (coniferous): 2.9 and 18.8 kg/ha/yr) NOs- leaching (New England; Adirondack lakes: 8 to10 kg/ha/yr) Nitrogen saturation, high dissolved inorganic nitrogen (San Bernardino Mountains, San Gabriel Mountains, CA, chaparral, hardwood, coniferous): 11 to 40 kg/ha/yr) Increased tree mortality and beetle activity (San Bernardino Mountains, CA (Ponderosa): 8 and 82 kg/ha/yr) Enhanced growth of black cherry and yellow poplar; possible decline in red maple vigor; increased foliar nitrogen (Fernow Forest, WV: 35.5 kg/ha/yr) Impacts on lichen communities (California MCF: 3.1 kg/ha/yr; Columbia R. Gorge, OR/WA: 11/5 to 25.4) Evidence that threatened and endangered species impacted San Francisco Bay, CA (checkerspot butterfly and serpentinitic grass invasion: 10 to15 kg/ha/yr; Jasper Ridge, CA: 70 kg/ha/yr) Decreased diversity of mycorrhizal communities (Southern California: -10 kg/ha/yr) Decreased abundance of CSS (Southern California: 3.3 kg/ha/yr) Loss of grasslands (Cedar Creek, MN: 5.3 [1.3 to 9.8] kg/ha/yr) Decrease in abundance of desert creosote bush, increase in nonnative grasses (Mojave Desert and Chihuahuan Desert, CA: 1.7 kg/ha/yr and up) Decrease in pitcher plant population growth rate (Hawley Bog, MA and Molly Bog, VA: 10 to14 kg/ha/yr) 1 2 3 4 Figure 2-9 (from REA figure 5.3-9). Observed effects from ambient and experimental atmospheric nitrogen deposition loads in relation to using CMAQ 2002 modeling results and NADP monitoring data. Citations for effect results are from the ISA, Table 4.4 (U.S. EPA, 2008). March 2010 43 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Based on information in the ISA and initial analysis in the REA, further case study 2 analyses on terrestrial nutrient enrichment of ecosystems were developed for the CCS 3 community and Mixed Conifer Forest (MCF) (EPA 2009). Geographic information systems 4 (GIS) analysis supported a qualitative review of past field research to identify ecological 5 benchmarks associated with CSS and mycorrhizal communities, as well as MCF's nutrient- 6 sensitive acidophyte lichen communities, fine-root biomass in Ponderosa pine, and leached 7 nitrate in receiving waters. 8 The ecological benchmarks that were identified for the CSS and the MCF are included in 9 the suite of benchmarks identified in the ISA (ISA 3.3). There are sufficient data to confidently 10 relate the ecological effect to a loading of atmospheric nitrogen. For the CSS community, the 11 following ecological benchmarks were identified: 12 • 3.3 kg N/ha/yr - the amount of nitrogen uptake by a vigorous stand of CSS; above this 13 level, nitrogen may no longer be limiting 14 • 10 kg N/ha/yr - mycorrhizal community changes 15 For the MCF community, the following ecological benchmarks were identified: 16 • 3.1 kg N/ha/yr - shift from sensitive to tolerant lichen species 17 • 5.2 kg N/ha/yr-dominance of the tolerant lichen species 18 • 10.2 kg N/ha/yr-loss of sensitive lichen species 19 • 17 kg N/ha/yr - leaching of nitrate into streams. 20 These benchmarks, ranging from 3.1 to 17 kg N/ha/yr, were compared to 2002 21 CMAQ/NADP data to discern any associations between atmospheric deposition and changing 22 communities. Evidence supports the finding that nitrogen alters CSS and MCF. Key findings 23 include the following: 2002 CMAQ/NADP nitrogen deposition data show that the 3.3 kg N/ha/yr 24 benchmark has been exceeded in more than 93% of CSS areas (654,048 ha). These deposition 25 levels are a driving force in the degradation of CSS communities. Although CSS decline has 26 been observed in the absence of fire, the contributions of deposition and fire to the CSS decline 27 require further research. CSS is fragmented into many small parcels, and the 2002 28 CMAQ/NADP 12-km grid data are not fine enough to fully validate the relationship between 29 CSS distribution, nitrogen deposition, and fire. 2002 CMAQ/NADP nitrogen deposition data 30 exceeds the 3.1 kg N/ha/yr benchmark in more than 38% (1,099,133 ha) of MCF areas, and March 2010 44 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 nitrate leaching has been observed in surface waters. Ozone effects confound nitrogen effects on 2 MCF acidophyte lichen, and the interrelationship between fire and nitrogen cycling requires 3 additional research. 4 2.2.3.2 Freshwater 5 The magnitude of ecosystem response may be thought of on two time scales, current 6 conditions and how ecosystems have been altered since the onset of anthropogenic N deposition. 7 As noted previously, Elser et al. (2008) found that N-limitation occurs as frequently as P- 8 limitation in freshwater ecosystems (ISA 3.3.3.2). Recently, a comprehensive study of available 9 data from the northern hemisphere surveys of lakes along gradients of N deposition show 10 increased inorganic N concentration and productivity to be correlated with atmospheric N 11 deposition (Bergstrom and Jansson 2006). The results are unequivocal evidence of N limitation 12 in lakes with low ambient inputs of N, and increased N concentrations in lakes receiving N 13 solely from atmospheric N deposition (Bergstrom and Jansson, 2006). These authors suggested 14 that most lakes in the northern hemisphere may have originally been N-limited, and that 15 atmospheric N deposition has changed the balance of N and P in lakes. 16 Available data suggest that the increases in total N deposition do not have to be large to 17 elicit an ecological effect. For example, a hindcasting exercise determined that the change in 18 Rocky Mountain National Park lake algae that occurred between 1850 and 1964 was associated 19 with an increase in wet N deposition that was only about 1.5 kg N/ha (Baron, 2006). Similar 20 changes inferred from lake sediment cores of the Beartooth Mountains of Wyoming also 21 occurred at about 1.5 kg N/ha deposition (Saros et al., 2003). Pre-industrial inorganic N 22 deposition is estimated to have been only 0.1 to 0.7 kg N/ha based on measurements from remote 23 parts of the world (Galloway et al., 1995; Holland et al., 1999). In the western U.S., pre- 24 industrial, or background, inorganic N deposition was estimated by (Holland et al., 1999) to 25 range from 0.4 to 0.7 kg/ha/yr. 26 Eutrophication effects from N deposition are most likely to be manifested in undisturbed, 27 low nutrient surface waters such as those found in the higher elevation areas of the western U.S. 28 The most severe eutrophication from N deposition effects is expected downwind of major urban 29 and agricultural centers. High concentrations of lake or streamwaterNO3-, indicative of 30 ecosystem saturation, have been found at a variety of locations throughout the U.S., including the March 2010 45 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 San Bernardino and San Gabriel Mountains within the Los Angeles Air Basin (Fenn et al., 1996), 2 the Front Range of Colorado (Baron et al., 1994; Williams et al., 1996), the Allegheny mountains 3 of West Virginia (Gilliam et al., 1996), the Catskill Mountains of New York (Murdoch and 4 Stoddard, 1992; Stoddard, 1994), the Adirondack Mountains of New York (Wigington et al., 5 1996), and the Great Smoky Mountains in Tennessee (Cook et al., 1994) (ISA 3.3.8). 6 2.2.3.3 Nitrogen Enrichment: Evidence of Effects on Estuaries 1 In contrast to terrestrial and freshwater systems, atmospheric N load to estuaries 8 contributes to the total load but does not necessarily drive the effects. In estuaries, N-loading 9 from multiple anthropogenic and non-anthropogenic pathways leads to water quality 10 deterioration, resulting in numerous effects including hypoxic zones, species mortality, changes 11 in community composition and harmful algal blooms that are indicative of eutrophication. The 12 following summary is a concise overview of the known or anticipated effects of nitrogen 13 enrichment on estuaries within the United States. 14 2.2.3.3.1 What is the nature of estuary responses to reactive nitrogen andsulfur 15 deposition? 16 In the ISA, the evidence is sufficient to infer a causal relationship between Nr deposition 17 and the biogeochemical cycling of N and C in estuaries (ISA 4.3.4.1 and 3.3.2.3). In general, 18 estuaries tend to be nitrogen-limited, and many currently receive high levels of nitrogen input 19 from human activities (REA 5.1.1). It is unknown if atmospheric deposition alone is sufficient to 20 cause eutrophication, however, the contribution of atmospheric nitrogen deposition to total 21 nitrogen load is calculated for some estuaries and can be >40% (REA 5.1.1). 22 The evidence is sufficient to infer a causal relationship between N deposition and the 23 alteration of species richness, species composition and biodiversity in estuarine ecosystems (ISA 24 4.3.4.2 and 3.3.5.4). Atmospheric and non-atmospheric sources of N contribute to increased 25 phytoplankton and algal productivity, leading to eutrophication. Shifts in community 26 composition, reduced hypolimnetic DO, reduced biodiversity, and mortality of submerged 27 aquatic vegetation are associated with increased N deposition in estuarine systems. March 2010 46 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2.2.3.3.2 What types of ecosystems are sensitive to such effects? How are these 2 responses affected by atmospheric, ecological, and landscape factors? 3 Because the productivity of estuarine and near shore marine ecosystems is generally 4 limited by the availability of N, they are susceptible to the eutrophication effect of N deposition 5 (ISA 4.3.4.1). A recent national assessment of eutrophic conditions in estuaries found the most 6 eutrophic estuaries were generally those that had large watershed-to-estuarine surface area, high 7 human population density, high rainfall and runoff, low dilution, and low flushing rates (Bricker 8 et al., 2007). In the REA, the National Oceanic and Atmospheric Administration's (NOAA) 9 National Estuarine Eutrophi cation Assessment (NEEA) assessment tool, Assessment of 10 Estuarine Tropic Status (ASSETS) categorical Eutrophication Index (El) (Bricker et al., 2007) 11 was used to evaluate eutrophi cation due to atmospheric loading of nitrogen. ASSETS El is an 12 estimation of the likelihood that an estuary is experiencing eutrophi cation or will experience 13 eutrophi cation based on five ecological indicators: chlorophyll a, macroalgae, dissolved oxygen, 14 nuisance/toxic algal blooms and submerged aquatic vegetation (SAV) (Bricker et al., 2007). 15 In the REA, two regions were selected for case study analysis using ASSETS El, the 16 Chesapeake Bay and Pamlico Sound. Both regions received an ASSETS El rating of Bad 17 indicating that the estuary had moderate to high pressure due to overall human influence and a 18 moderate high to high eutrophic condition (REA 5.2.4.1 and 5.2.4.2). These results were then 19 considered with SPAtially Referenced Regression (SPARROW) modeling to develop a response 20 curve to examine the role of atmospheric nitrogen deposition in achieving desired reduction load. 21 To change the Neuse River Estuary' s El score from Bad to Poor not only must 100% of the total 22 atmospheric nitrogen deposition be eliminated, but considerably more nitrogen from other 23 sources as well must be reduced (REA section 5.2.7.2). In the Potomac River estuary, a 78% 24 reduction of total nitrogen could move the El score from Bad to Poor (REA 5.2.7.1). The results 25 of this analysis indicated reductions in atmospheric deposition alone could not solve coastal 26 eutrophi cation problems due to multiple non-atmospheric nitrogen inputs (REA 7.3.3). However, 27 by reducing atmospheric contributions, it may help avoid the need for more costly controls on 28 nitrogen from other sources. 29 In general, estuaries tend to be N-limited (Elser et al., 2008), and many currently receive 30 high levels of N input from human activities to cause eutrophi cation (Howarth et al., 1996; 31 Vitousek and Howarth, 1991). Atmospheric N loads to estuaries in the U.S. are estimated to March 2010 47 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 range from 2-8% for Guadalupe Bay, TX on the lowest end to as high as 72% for St Catherines- 2 Sapelo estuary, GA (Castro et al., 2003). The Chesapeake Bay is an example of a large, well- 3 studied and severely eutrophic estuary that is calculated to receive as much as 30% of its total N 4 load from the atmosphere. 5 2.2.3.3.3 What is the magnitude of ecosystem responses to eutrophication? 6 There is a scientific consensus that nitrogen-driven eutrophication in shallow estuaries 7 has increased over the past several decades and that the environmental degradation of coastal 8 ecosystems due to nitrogen, phosphorus, and other inputs is now a widespread occurrence (Paerl 9 et al., 2001). For example, the frequency of phytoplankton blooms and the extent and severity of 10 hypoxia have increased in the Chesapeake Bay (Officer et al., 1984) and Pamlico estuaries in 11 North Carolina (Paerl et al., 1998) and along the continental shelf adjacent to the Mississippi and 12 Atchafalaya rivers' discharges to the Gulf of Mexico (Eadie et al., 1994). 13 A recent national assessment of eutrophic conditions in estuaries found that 65% of the 14 assessed systems had moderate to high overall eutrophic conditions and generally received the 15 greatest N loads from all sources, including atmospheric and land-based sources (Bricker et al., 16 2007). Most eutrophic estuaries occurred in the mid-Atlantic region and the estuaries with the 17 lowest degree of eutrophication were in the North Atlantic (Bricker et al., 2007). Other regions 18 had mixtures of low, moderate, and high degree of eutrophication (ISA 4.3.4.3). 19 The mid-Atlantic region is the most heavily impacted area in terms of moderate or high 20 loss of submerged aquatic vegetation due to eutrophication (ISA 4.3.4.2). Submerged aquatic 21 vegetation is important to the quality of estuarine ecosystem habitats because it provides habitat 22 for a variety of aquatic organisms, absorbs excess nutrients, and traps sediments (ISA 4.3.4.2). It 23 is partly because many estuaries and near-coastal marine waters are degraded by nutrient 24 enrichment that they are highly sensitive to potential negative impacts from nitrogen addition 25 from atmospheric deposition. 26 2.2.4 What are the key uncertainties associated with nutrient enrichment? 27 There are different levels of uncertainty associated with relationships between deposition, 28 ecological effects and ecological indicators. The criteria used in the REA to evaluate the degree 29 of confidence in the data, modeling and ecological effect function are detailed in Chapter 7 of the 30 REA and summarized in section 2.1.4 of this chapter (REA 7.0). March 2010 48 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Aquatic 2 The approach for assessing atmospheric contributions to total nitrogen loading in the 3 REA, was to consider the main-stem river to an estuary (including the estuary) rather than an 4 entire estuary system or bay. The biological indicators used in the NOAA ASSETS El required 5 the evaluation of many national databases including the USGS NAWQA files, EPA's STORET 6 database, NOAA's Estuarine Drainage Areas data, and EPA's water quality standards nutrient 7 criteria for rivers and lakes (REA Appendix 6, Table 1.2.-1). Both the SPARROW modeling for 8 nitrogen loads and assessment of estuary conditions under NOAA ASSETS El, have been 9 applied on a national scale. The REA concludes that the available data are medium quality with 10 intermediate confidence about the use of these data and their values for extrapolating to a larger 11 regional area (REA 7.3.1). Intermediate confidence is associated with the modeling approach 12 using ASSETS El and SPARROW. The REA states there is low confidence with the ecological 13 effect function due to the results of the analysis which indicated that reductions in atmospheric 14 deposition alone could not solve coastal eutrophication problems due to multiple non- 15 atmospheric nitrogen inputs (REA 7.3.3). 16 Terrestrial 17 Ecological thresholds are identified for CSS and MCF and these data are considered to be 18 of high quality, however, the ability to extrapolate these data to larger regional areas is limited 19 (REA 7.4.1). No quantitative modeling was conducted or ecological effect function developed 20 for terrestrial nutrient enrichment reflecting the uncertainties associated with these depositional 21 effects. 22 2.3 WHAT ECOLOGICAL EFFECTS ARE ASSOCIATED WITH GAS- 23 PHASE NOX AND SOX? 24 Acidifying deposition and nitrogen enrichment are the main focus of this policy 25 assessment; however, there are other known ecological effects are attributed to gas-phase NOX 26 and SOX. Acute and chronic exposures to gaseous pollutants such as sulfur dioxide (802), 27 nitrogen dioxide (NO2), nitric oxide (NO), nitric acid (HNO3) and peroxyacetyl nitrite (PAN) are 28 associated with negative impacts to vegetation. The current secondary NAAQS were set to 29 protect against direct damage to vegetation by exposure to gas-phase NOX or SOX, such as foliar March 2010 49 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 injury, decreased photosynthesis, and decreased growth. The following summary is a concise 2 overview of the known or anticipated effects to vegetation caused by gas phase N and S. 3 2.3.1 What is the nature of ecosystem responses to gas-phase nitrogen and sulfur? 4 The 2008 ISA found that gas phase N and S are associated with direct phytotoxic effects 5 (ISA 4.4). The evidence is sufficient to infer a causal relationship between exposure to SC>2 and 6 injury to vegetation (ISA 4.4.1 and 3.4.2.1). Acute foliar injury to vegetation from SC>2 may 7 occur at levels above the current secondary standard (3-h average of 0.50 ppm). Effects on 8 growth, reduced photosynthesis and decreased yield of vegetation are also associated with 9 increased SO2 exposure concentration and time of exposure. 10 The evidence is sufficient to infer a causal relationship between exposure to NO, NO2 11 and PAN and injury to vegetation (ISA 4.4.2 and 3.4.2.2). In sufficient concentrations, NO, NO2 12 and PAN can decrease photosynthesis and induce visible foliar injury to plants. Evidence is also 13 sufficient to infer a causal relationship between exposure to HNOs and changes to vegetation 14 (ISA 4.4.3 and 3.4.2.3). Phytotoxic effects of this pollutant include damage to the leaf cuticle in 15 vascular plants and disappearance of some sensitive lichen species. 16 2.3.2 What types of ecosystems are sensitive to such effects? How are these 17 responses affected by atmospheric, ecological, and landscape factors? 18 Vegetation in ecosystems near sources of gaseous NOX and SOX or where ambient 19 concentrations of SC>2, NO, NO2, PAN and HNOs are higher are more likely to be impacted by 20 these pollutants. Uptake of these pollutants in a plant canopy is a complex process involving 21 adsorption to surfaces (leaves, stems and soil) and absorption into leaves (ISA 3.4.2). The 22 functional relationship between ambient concentrations of gas phase NOX and SOX and specific 23 plant response are impacted by internal factors such as rate of stomatal conductance and plant 24 detoxtification mechanisms, and external factors including plant water status, light, temperature, 25 humidity, and pollutant exposure regime (ISA 3.4.2). 26 Entry of gases into a leaf is dependent upon physical and chemical processes of gas phase 27 as well as to stomatal aperature. The aperature of the stomata is controlled largely by the 28 prevailing environmental conditions, such as humidity, temperature, and light intensity. When 29 the stomata are closed, resistance to gas uptake is high and the plant has a very low degree of 30 susceptibility to injury. Mosses and lichens do not have a protective cuticle barrier to gaseous March 2010 50 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 pollutants or stomata and are generally more sensitive to gaseous sulfur and nitrogen than 2 vascular plants (ISA 3.4.2). 3 The appearance of foliar injury can vary significantly across species and growth 4 conditions affecting stomatal conductance in vascular plants (REA 6.4.1). For example, damage 5 to lichens from SC>2 exposure includes reduced photosynthesis and respiration, damage to the 6 algal component of the lichen, leakage of electrolytes, inhibition of nitrogen fixation, reduced K+ 7 absorption, and structural changes (Belnap et al., 1993; Farmer et al., 1992, Hutchinson et al., 8 1996). 9 2.3.3 What is the magnitude of ecosystem responses to gas phase effects of NOX 10 and SOX? 11 The phytotoxic effects of gas phase NOX and SOX are dependent on the exposure 12 concentration and duration and species sensitivity to these pollutants. Effects to vegetation 13 associated with NOX and SOX, are therefore, variable across the U.S. and tend to be higher near 14 sources of photochemical smog. For example, 862 is considered to be the primary factor 15 contributing to the death of lichens in many urban and industrial areas, with fruticose lichens 16 being more susceptible to 862 than many foliose and crustose species (Hutchinson et al., 1996). 17 The ISA states there is very limited new research on phytotoxic effects of NO, NC>2, PAN 18 and FINOs at concentrations currently observed in the United States with the exception of some 19 lichen species (ISA 4.4). Past and current HNOs concentrations may be contributing to the 20 decline in lichen species in the Los Angeles basin (Boonpragob and Nash 1991; Nash and Sigal, 21 1999; Riddell et al., 2008). PAN is a very small component of nitrogen deposition in most areas 22 of the United States (REA 6.4.2). Current deposition of FINOs is contributing to N saturation of 23 some ecosystems close to sources of photochemical smog (Fenn et al., 1998) such as the MCF's 24 of the Los Angeles basin mountain (Bytnerowicz et al., 1999). 25 2.4 SUMMARY 26 In summary, NOX and SOX in the atmosphere contribute to effects on individual species 27 and ecosystems through direct contact with vegetation, and more significantly through deposition 28 to sensitive ecosystems. The ISA concludes that the evidence is sufficient to conclude causal 29 relationships between acidifying deposition of N and S and effects on freshwater aquatic 30 ecosystems and terrestrial ecosystems, and between nitrogen nutrient enrichment and effects on March 2010 51 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 sensitive terrestrial and freshwater aquatic ecosystems. The ISA also concludes that a causal 2 relationship is supported between nitrogen nutrient enrichment and effects on estuarine 3 ecosystems; however, the contribution of atmospheric oxidized nitrogen relative to reduced 4 nitrogen and non-atmospheric nitrogen is more difficult to determine. 5 The REA provides additional support that under recent conditions, deposition levels have 6 exceeded benchmarks for ecological indicators of acidification and nutrient enrichment that 7 indicate that effects are likely to be occurring in significant numbers of lakes and streams within 8 sensitive ecosystems. 9 2.5 REFERENCES 10 Aber JD; Goodale CL; Ollinger SV; Smith ML; Magill AH; Martin ME; Hallett RA; Stoddard 11 JL. (2003). Is nitrogen deposition altering the nitrogen status of northeastern forests? 12 Bioscience, 53, 375-389. 13 Bailey SW; Horsley SB; Long RP; Hallett RA. (2004). Influence of edaphic factors on sugar 14 maple nutrition and health on the Allegheny Plateau. Soil Sci Soc Am J, 68, 243-252. 15 Baker JP; Bernard DP; Christensen SW; Sale MJ. (1990). Biological effects of changes in 16 surface water acid-base chemistry. (State of science / technology report #13).Washington 17 DC: National Acid Precipitation Assessment Program (NAPAP). 18 Baker JP; Van Sickle J; Gagen CJ; DeWalle DR; Sharpe WE; Carline RF; Baldigo BP; Murdoch 19 PS; Bath DW; Kretser WA; Simonin HA; Wigington PJ Jr. (1996). Episodic acidification 20 of small streams in the northeastern United States: Effects on fish populations. Ecol Appl, 21 6,423-437. 22 Baron JS; Ojima DS; Holland EA; Parton WJ. (1994). Analysis of nitrogen saturation potential 23 in Rocky Mountain tundra and forest: Implications for aquatic systems. Biogeochemistry, 24 27,61-82. 25 Baron JS. (2006). Hindcasting nitrogen deposition to determine ecological critical load. Ecol 26 Appl, 16, 433-439. March 2010 52 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Belnap J; Sigal L; Moir W; Eversman S. (1993). Identification of sensitive species, in lichens as 2 bioindicators of air quality. In: Huckaby LS (Ed.), Lichens as bioindicators of air quality 3 (pp. 67-88). Fort Collins, Colorado: Rocky Mountain Forest and Range 4 Experimental Station, U.S. Forest Service, U. S. Department of Agriculture. 5 Bergstrom A; Jansson M. (2006). Atmospheric nitrogen deposition has caused nitrogen 6 enrichment and eutrophication of lakes in the northern hemisphere. Glob Chang Biol, 12, 7 635-643. 8 Blake L; Goulding KWT; Mott CJB; Johnston AE. (1999). Changes in soil chemistry 9 accompanying acidification over more than 100 years under woodland and grass at 10 Rothamsted Experimental Station, UK. Eur J Soil Sci, 50, 401-412. 11 Boonpragob K; Nash THI. (1991). Physiological responses of the lichen Ramalina menziesii 12 Tayl. to the Los Angeles urban environment. Environ Exp Bot, 31, 229-238. 13 Bricker S; Longstaff B; Dennison W; Jones A; Boicourt K; Wicks C; Woerner J. (2007). Effects 14 of nutrient enrichment in the nation's estuaries: A decade of change. 15 http://ccmaserver.nos.noaa.gov/publications/eutroupdate/. (NOAA Coastal Ocean 16 Program Decision Analysis Series No. 26). Silver Spring, MD: National Centers for 17 Coastal Ocean Science, National Oceanic and Atmospheric Administration(NOAA). 18 Bulger AJ; Dolloff CA; Cosby BJ; Eshleman KN; Webb JR; Galloway JN. (1995). The 19 "Shenandoah National Park: fish in sensitive habitats" (SNP: FISH) Project. An 20 integrated assessment offish community responses to stream acidification. Water Air 21 SoilPollut, 85, 309-314. 22 Bulger AJ; Cosby, BJ; Dolloff, CA; Eshleman, KN; Webb, JR; Galloway, JN. (1999). 23 SNP:FISH, Shenandoah National Park: Fish in sensitive habitats. Project final report - 24 Volume I: Project description and summary of results; Volume II: Stream water 25 chemistry and discharge, and synoptic water quality surveys. Volume III: Basin-wide 26 habitat and population inventories, and behavioral responses to acid in a laboratory 27 stream. Volume IV: Stream bioassays, aluminum toxicity, species richness and stream 28 chemistry, and models of susceptibility to acidification. (Project Completion Report to March 2010 53 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 the National Park Service). Charlottesville, VA, Department of Environmental Sciences; 2 University of Virginia. 3 Bulger AJ; Cosby BJ; Webb JR. (2000). Current, reconstructed past, and projected future status 4 of brook trout (Salvelinus fontinalis) streams in Virginia. Can J Fish Aquat Sci, 57, 1515- 5 1523. 6 Bytnerowicz A; Padgett P; Percy K; Krywult M; Riechers G; Horn J. (1999). Direct effects of 7 nitric acid on forest vegetation. In: Miller PR, McBride J (Eds.), Oxidant air pollution 8 impacts in the Montane forests of Southern California: The San Bernardino Mountains 9 case study, Ecological Series 134. (pp. 270-287). New York: Springer. 10 Castro MS; Driscoll CT; Jordan TE; Reay WG; Boynton WR. (2003). Sources of nitrogen to 11 estuaries in the United States. Estuaries, 26, 803-814. 12 Clark CM; Tilman D. (2008). Loss of plant species after chronic low-level nitrogen deposition to 13 prairie grasslands. Nature, 451, 712-715. 14 Cook RB; Elwood JW; Turner RR; Bogle MA; Mulholland PJ; Palumbo AV. (1994). Acid-base 15 chemistry of high-elevation streams in the Great Smoky Mountains. Water Air Soil 16 Pollut, 72, 331-356. 17 Cronan CS; Grigal DF. (1995). Use of calcium/aluminum ratios as indicators of stress in forest 18 ecosystems. J Environ Qual, 24, 209-226. 19 DeHayes DH; Schaberg PG; Hawley GJ; Strimbeck GR. (1999). Acid rain impacts on calcium 20 nutrition and forest health. Bioscience, 49, 789-800. 21 Dennis TD; MacAvoy SE; Steg MB; Bulger AJ. (1995). The association of water chemistry 22 variables and fish condition in streams of Shenandoah National Park (USA). Water Air 23 Soil Pollut, 85, 365-370. 24 Dennis TE; Bulger AJ. (1999). The susceptibility of blacknose dace (Rhinichthys atratulus).to 25 acidification in Shenandoah National Park. In: Bulger AJ; Cosby BJ; Dolloff CA; 26 Eshleman KN; Galloway JN; Webb. JR (Eds.), Shenandoah National Park: Fish in March 2010 54 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 sensitive habitats. Project final report, Volume IV: Stream bioassays, aluminum toxicity, 2 species richness and stream chemistry, and models of susceptibility to acidification. 3 Chapter 6B. Project completion report to the National Park Service. Charlottesville, VA; 4 Department of Environmental Sciences, University of Virginia. 5 Driscoll CT; Lawrence GB; Bulger AJ; Butler TJ; Cronan CS; Eagar C; Lambert KF; Likens 6 GE; Stoddard JL; Weather KC. (200Ib). Acidic deposition in the northeastern United 7 States: Sources and inputs, ecosystem effects, and management strategies. Bioscience, 8 51, 180-198. 9 Eadie BJ; McKee BA; Lansing MB; Robbins JA; Metz S; Trefry JH. (1994). Records of 10 nutrient-enhanced coastal productivity in sediments from the Louisiana continental shelf. 11 Estuaries, 17, 754-765. 12 Edwards PJ; Kochenderfer JN; Coble DW; Adams MB. (2002). Soil leachate responses during 13 10 years of induced whole-watershed acidification. Water Air Soil Pollut, 140, 99-118. 14 Elser JJ; Bracken MES; Cleland EE; Gruner DS; Harpole WS; Hillebrand IIH; Ngai JT; 15 Seabloom EW; Shurin JB; Smith JE. (2007). Global analysis of nitrogen and phosphorus 16 limitation of primary producers in freshwater, marine, and terrestrial ecosystems. Ecol 17 Lett, 10, 1135-1142 18 Farmer AM; Bates JW; Bell JNB. (1992). Ecophysiological effects of acid rain on bryophytes 19 and lichens. In: Bates JW; Farmer AM (Eds.), Bryophytes and lichens in a changing 20 environment. Oxford, UK: Claredon Press. 21 Fenn ME; Poth MA; Johnson DW. (1996). Evidence for nitrogen saturation in the San 22 Bernardino Mountains in southern California. For Ecol Manage, 82, 211-230. 23 Ford J; Stoddard JL; Powers CF. (1993). Perspectives in environmental monitoring: an 24 introduction to the U.S. EPA long-term monitonring (LTM) project. Water Air Soil 25 Pollut, 67, 247-255. March 2010 55 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Galloway JN; Schlesinger WH; Levy H; Michaels AF; Schnoor JL. (1995). Nitrogen-fixation: 2 anthropogenic enhancement, environmental response. Glob Biogeochem Cycles, 9, 235- 3 252. 4 Gilliam FS; Adams MB; Yurish BM. (1996). Ecosystem nutrient responses to chronic nutrient 5 inputs at Fernow Experimental Forest, West Virginia. Can J For Res, 26, 196-205. 6 Haines TA; Baker JP. (1986). Evidence offish population responses to acidification in the 7 eastern United States. Water Air Soil Pollut, 31, 605-629. 8 Holland EA; Dentener FJ; Braswell BH; Sulzman JM. (1999). Contemporary and pre-industrial 9 global reactive nitrogen budgets. Biogeochemistry, 46, 7-43. 10 Horsley SB; Long RP; Bailey SW; Hallett RA; Hall TJ. (2000). Factors associated with the 11 decline disease of sugar maple on the Allegheny Plateau. Can J For Res, 30, 1365-1378. 12 Howarth RW; Billen G; Swaney D; Townsend A; Jaworski N; Lajtha K; Downing JA; Elmgren 13 R; Caraco N; Jordan T; Berendse F; Freney J; Kudeyarov V; Murdoch PS; Zhao-Liang Z. 14 (1996). Regional nitrogen budgets and riverine N & P fluxes for the drainages to the 15 North Atlantic Ocean: natural and human influences. Biogeochemistry, 35, 75-139. 16 Hutchinson J; Maynard D; Geiser L. (1996). Air quality and lichens - a literature review 17 emphasizing the Pacific Northwest, USA. Pacific Northwest Region Air Resource 18 Management Program; U.S. Forest Service; U.S. Department of Agriculture (USDA). 19 Kahl J; Norton S; Fernandez I; Rustad L; Handley M. (1999). Nitrogen and sulfur input-output 20 budgets in the experimental and reference watersheds, Bear Brook Watershed in Maine 21 (BBWM). Environ Monit Assess, 55, 113-131 22 Keller W; Gunn JM. (1995). Lake water quality improvements and recovering aquatic 23 communities. In: Gunn JM (Ed.), Restoration and recovery of an industrial region: 24 progress in restoring the smelter-damaged landscape near Sudbury, Canada (pp. 67-80). 25 New York, NY: Springer-Verlag. March 2010 56 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Kochy M; Wilson SD. (2001). Nitrogen deposition and forest expansion in the northern Great 2 Plains. JEcol, 89, 807-817. 3 Kretser W; Gallagher J; Nicolette J. (1989). Adirondack Lakes Study, 1984-1987: An Evaluation 4 of Fish Communities and Water Chemistry: Ray Brook, NY; prepared for: Adirondack 5 Lakes Survey (ALS) Corporation. 6 LeBauer DS; Treseder KK. (2008). Nitrogen limitation of net primary productivity in terrestrial 7 ecosystems is globally distributed. Ecology, 89,371-379. 8 Lien L; Raddum GG; Fjellheim A. (1992). Critical loads for surface waters: invertebrates and 9 fish. (Acid rain research report no 21). Oslo, Norway: Norwegian Institute for Water 10 Research. 11 Longcore JR; Gill JD (Eds.), (1993). Acidic depositions: effects on wildlife and habitats. 12 (Wildlife Society technical review no 93-1). Bethesda, MD: The Wildlife Society. 13 MacAvoy SW; Bulger AJ. (1995). Survival of brook trout (Salvelinus fontinalis) embryos and 14 fry in streams of different acid sensitivity in Shenandoah National Park, USA. Water Air 15 Soil Pollut, 85, 445-450. 16 McNulty SG; Cohen EC; Myers JAM; Sullivan TJ; Li H. (2007). Estimates of critical acid loads 17 and exceedances for forest soils across the conterminous United States. Environ Pollut, 18 149,281-292. 19 Murdoch PS; Stoddard JL. (1992). The role of nitrate in the acidification of streams in the 20 Catskill Mountains of New York. Water Resour Res, 28, 2707-2720 21 Nash TH; Sigal LL. (1999). Epiphytic lichens in the San Bernardino mountains in relation to 22 oxidant gradients. In: Miller PR, McBride JR (Eds.), Oxidant air pollution impacts on the 23 montane forests of southern California: A case study of the San Bernardino mountains. 24 Ecological Studies, 134, (pp. 223-234). New York, NY: Springer-Verlag. March 2010 57 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Norton S; Kahl J; Fernandez I. (1999). Altered soil-soil water interactions inferred from stream 2 water chemistry at an artificially acidified watershed at Bear Brook Watershed, Maine 3 USA. Environ Monit Assess, 55, 97-111. 4 Officer CB; Biggs RB; Taft JL; Cronin LE; Tyler MA; Boynton WR. (1984). Chesapeake Bay 5 anoxa. Origin, development and significance. Science, 223, 22-27'. 6 Paerl HW; Boynton WR; Dennis RL; Driscoll CT; Greening HS; Kremer JN; Rabalais NN; 7 Seitzinger SP. (2001). Atmospheric deposition of nitrogen in coastal waters: 8 Biogeochemical and ecological implications. In: Valigura RW; Alexander RB; Castro 9 MS; Meyers TP; Paerl HW; Stacey PE; Turner RE (Eds.), Nitrogen loading in coastal 10 water bodies: an atmospheric perspective. (Coastal and estuarine series; Volume 57, pp 11 11-52). Washington, DC: American Geophysical Union. 12 Paerl H; Pinckney J; Fear J; Peierls B. (1998). Ecosystem responses to internal and watershed 13 organic matter loading: Consequences for hypoxia in the eutrophying Neusse River 14 Estuary, NC, USA. Mar Ecol Prog Ser, 166, 17-25. 15 Parker KE. (1988). Common loon reproduction and chick feeding on acidified lakes in the 16 Adirondack Park, New York. Canadian Journal of Zoology, 66, 804-810. Parker DR; Zelazny LW; Kinraide TB. (1989). Chemical speciation and plant toxicity of aqueous aluminum. In: Lewis TE (Ed.), Environmental chemistry and toxicology of 19 aluminum (pp. 117-145). American Chemical Society. 17 18 20 Peterjohn WT; Adams MB; Gilliam FS. (1996). Symptoms of nitrogen saturation in two central 21 Appalachian hardwood forest ecosystems.Biogeochemistry, 35, 507-522. 22 Reuss JO. (1983). Implications of the calcium-aluminum exchange system for the effect of acid 23 precipitation on soils. J Environ Qual, 12, 591-595. 24 Riddell J; Nash lii TH; Padgett P. (2008). The effect of HNO3 gas on the lichen Ramalina 25 menziesii. Flora - Morphology, Distribution, Functional Ecology of Plants, 203, 47-54. March 2010 58 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Ross DS; Lawrence GB; Fredriksen G. (2004). Mineralization and nitrification patterns at eight 2 northeastern USA forested research sites. For Ecol Manage, 188, 317-335. 3 Saros JE; Interlandi SJ; Wolfe AP; Engstrom DR. (2003). Recent changes in the diatom 4 community structure of lakes in the Beartooth Mountain Range, USA. Arct Antarct Alp 5 Res, 35, 18-23. 6 Schindler DW; Mills KH; Malley DF; Findlay MS; Schearer JA; Davies U; Turner MA; Lindsey 7 GA; Cruikshank DR. (1985). Long-term ecosystem stress: Effects of years of 8 experimental acidification. Science, 228, 1395-1401. 9 Schreck CB. (1981). Stress and rearing of salmonids. Aquaculture, 28, 241-249. 10 Schreck CB. (1982). Stress and compensation in teleostean fishes: response to social and 11 physical factors. In: Pickering AD (Ed.), Stress and fish (pp. 295-321). London: 12 Academic Press. 13 Schwinning S; Starr BI; Wojcik NJ; Miller ME; Ehleringer JE; Sanford RL Jr. (2005). Effects of 14 nitrogen deposition on an arid grassland in theColorado Plateau cold desert. Journal of 15 Rangeland Ecology and Management, 58, 565-574. 16 Sterner RW; Elser JJ. (2002). Ecological stoichiometry: the biology of elements from molecules 17 to the biosphere. Princeton, NJ: Princeton University Press. 18 Stoddard JL. (1990). Plan for converting the NAPAP aquatic effects long-term monitoring 19 (LTM) project to the temporally integrated monitoring of ecosystems (TIME) project. 20 (International Report). Corvallis, OR; U.S. Environmental Protection Agency. 21 Stoddard JL. (1994). Long-term changes in watershed retention of nitrogen: its causes and 22 aquatic consequences. In Baker LA (Ed.), Environmental chemistry of lakes and 23 reservoirs, (pp. 223-284). Washington, D.C.: American Chemical Society. 24 Stoddard JL; Urquhart NS; Newell AD; Kugler D. (1996). The Temporally Interated Monitoring 25 of Ecosystems (TIME) project design 2. Detection of regional acidification trends. Water 26 Resour Res, 32, 2529-2538. March 2010 59 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Suding KN; Collins SL; Gough L; Clark C; Cleland EE; Gross KL; Milchunas DG; Pennings S. 2 (2005). Functional- and abundance-based mechanisms explain diversity loss due to N 3 fertlization. Proc Natl Acad Sci USA, 102, 4387-4392. 4 Sullivan TJ; Driscoll CT; Cosby BJ; Fernandez IJ; Herlihy AT; Zhai J; Stemberger R; Snyder 5 KU; Sutherland JW; Nierzwicki-Bauer SA; Boylen CW; McDonnell TC; Nowicki NA. 6 (2006). Assessment of the extent to which intensively studied lakes are representative of 7 the Adirondack Mountain region. (Final Report no 06-17).Corvallis, OR; prepared by 8 Environmental Chemistry, Inc. for: Albany, NY; Environmental Monitoring Evaluation 9 and Protection Program of the New York State Energy Research and Development 10 Authority (NYSERDA). 11 Sverdrup H; Warfvinge P. (1993). The effect of soil acidification on the growth of trees, grass 12 and herbs as expressed by the (Ca+ Mg+ K)/A1 ratio. Rep in Ecol & Eng, 2, 1993. 13 US EPA (2008) U.S. EPA. Integrated Science Assessment (ISA) for Oxides of Nitrogen and 14 Sulfur Ecological Criteria (Final Report). U.S. Environmental Protection Agency, 15 Washington, D.C., EPA/600/R-08/082F, 2008. 16 US EPA (2009) Risk and Exposure Assessment for Review of the Secondary National Ambient 17 Air Quality Standards for Oxides of Nitrogen and Oxides of Sulfur-Main Content - Final 18 Report. U.S. Environmental Protection Agency, Washington, D.C., EPA-452/R-09-008a 19 Vitousek PM; Howarth RW. (1991). Nitrogen limitation on land and in the sea: how can it 20 occur? Biogeochemistry, 13, 87-115. 21 Williams MW; Baron JS; Caine N; Sommerfeld R; Sanford JR. (1996). Nitrogen saturation in 22 the Rocky Mountains. Environ Sci Technol, 30, 640-646. 23 Wigington PJ Jr; DeWalle DR; Murdoch PS; Kretser WA; Simonin HA; Van Sickle J; Baker JP. 24 (1996b). Episodic acidification of small streams in the northeastern United States: Ionic 25 controls of episodes. Ecol Appl, 6, 389-407. March 2010 60 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Wedemeyer GA; Barton BA; MeLeay DJ. (1990). Stress and acclimation. In: Schreck CB, 2 Moyle PB (Eds.), Methods for fish biology (pp. 178-196). Bethesda, MD: American 3 Fisheries Society. March 2010 61 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX i 3. CONSIDERATIONS OF ADVERSITY TO PUBLIC 2 WELFARE 3 3.1 HOW DO WE CHARACTERIZE ADVERSITY TO PUBLIC 4 WELFARE? WHAT ARE THE RELEVANT FACTORS AND HOW 5 ARE THEY ADDRESSED IN THIS DOCUMENT? 6 The paradigm of looking at adversity to public welfare as deriving from disruptions in 7 ecosystem structure and function has been used broadly by EPA to categorize effects from the 8 cellular to the ecosystem level. An evaluation of adversity to public welfare might consider the 9 type, intensity, and scale of the effect as well as the potential for recovery. 10 Similar concepts were used in past reviews of secondary NAAQS for ozone, PM relating 11 to visibility as well as initial reviews of effects from lead deposition. Because NOX and SOX are 12 deposited from ambient sources into ecosystems where they affect changes to organisms, 13 populations and ecosystems, the concept of adversity to public welfare as related to impacts on 14 the public from alterations in structure and function of ecosystems is appropriate for this review. 15 Other information that may be helpful to consider includes the role of critical loads and 16 ecosystem service impacts as benchmarks or measures of impacts on ecosystems that may affect 17 public welfare. Ecosystem services can be related directly to concepts of public welfare to 18 inform discussions of societal adverse impacts. Subsequent sections will discuss each of these 19 concepts as they relate to adversity. 20 3.1.1 What are the benchmarks for adversity from other sources? 21 3.1.1.1 Ozone and PM NAAQS Reviews 22 The evaluation of adversity from a public welfare perspective in the context of ozone and 23 particulate matter (PM) are relevant to this current review. Both ozone and PM have documented 24 effects on ecological receptors. These criteria pollutants are being reviewed on a schedule as part 25 of the NAAQS process. The ozone secondary standard is currently under reconsideration from 26 the 2008 ruling with a proposal due on January 6, 2010. A draft Policy Assessment for PM is 27 being developed for CASAC and public consultation. March 2010 62 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 3.1.1.1.1 Ozone 2 Welfare effects of ozone are primarily limited to vegetation. These effects begin at the 3 level of the individual cell and accumulate up to the level of whole leaves and plants. If effects 4 occur on enough individual plants within the population, communities and ecosystems may be 5 impacted. Prior to the 2008 ozone review, Ozone vegetation effects were classified as either 6 "injury" or "damage" (FR 72 37889). "Injury" was defined as; encompassing all plant reactions, 7 including reversible changes or changes in plant metabolism, quality or reduced growth that does 8 not impair the intended use of the plant while "damage" includes those injury effects that reach 9 sufficient magnitude as to reduce or impair the intended use of the plant (FR 72 37890). The 10 "intended use" of the plant was imbedded with the concept of adversity to public welfare. 11 Ozone-associated "damage" was considered adverse if the intended use of the plant was 12 compromised (i.e. crops, ornamentals, plants located in Class I areas). Effects of ozone on single 13 plants or species grown in monocultures such as agricultural crops and managed forests were 14 evaluated without consideration of potential effects on natural forests or entire ecosystems. 15 In the 2008 rulemaking, EPA expanded the characterization of adversity to go beyond the 16 individual plant level and this language is continued in the 2010 ozone reconsideration. The 2008 17 final rule and 2010 proposal conclude that a determination of what constitutes an "adverse" 18 welfare effect in the context of secondary NAAQS review can appropriately occur by 19 considering effects at higher ecological levels (populations, communities, ecosystems) as 20 supported by recent literature. The ozone review uses the example of the construct presented in 21 Hogsett et al. (1997) as a model for assessing risks to forests. This study suggests that adverse 22 effects could be classified into one or more of the following categories: (1) economic production, 23 (2) ecological structure, (3) genetic resources, and (4) cultural values". Another recent 24 publication, "A Framework for Assessing and Reporting on Ecological Condition: an SAB 25 report" (Young and Sanzone, 2002) provides additional support for expanding the consideration 26 of adversity beyond the species level and at higher levels by making explicit the linkages 27 between stress-related effects at the species level and at higher levels within an ecosystem 28 hierarchy (See Figure 3.1.1). March 2010 63 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX • Hydrologic alteration Habitat conversion Habitat fragmentation t/K Climate change Ij Invasive non-native species |§ Turbidity/sedimentation uj Pesticides 5 Disease/pest otabHktitt Nutrient pukes Metals Dissolved oxygen depletion Ozone ftropospberic) 1 Hydroiogic aiteratit Habita f conversion Habit* t fragment^ tion Climate chanty Over-harvestmg of vegetation Large-scale invasive species introductions Large-scale disease/pes t outbreaks L Landscape Condition Biotic Condition Hydrologic alteration Habitat conversion Climate change Turbidity/seaimentation Pesticides Nutrient pukes Metals Dissolved oxygen depletion Ozone (tropospherif) Nitroget) oxides Natural Disturbance Hydrologic alteration Habitat conversion (3, Climate change Q Over-harvesting of vegetation tu Disease/pest outbreaks g Altered fire regime <-n Altered flood regime - , 1 Hydrology/ Geomorphology. Ecological Processes Hydrol&gic alteration Habitat conversion ^ Habitat fragmentation P CMmate change § Turbidity/sedimentation I Hydroiogic alteration ff. Habitat conversion g Climate chanty Q Pesticides Disease/pest outbreaks 5 Nutrient pulses Dissohea oxygen depletion -. | Nitrogen oxides 2 Figure 3-1. Common anthropogenic stressors and the essential ecological 3 attributes they affect. Modified from Young and Sanzone (2002) 4 In the 2008 ozone NAAQS review and current ozone NAAQS proposal, the 5 interpretation of what constitutes an adverse effect on public welfare can vary depending on the 6 location and intended use of the plant. The degree to which Os-related effects are considered 7 adverse to public welfare depends on the intended use of the vegetation and its significance to 8 public welfare (73 FR 16496). Therefore, effects on vegetation (e.g., biomass loss, foliar injury, 9 impairment of intended use) may be judged to have a different degree of impact on public 10 welfare depending, for example, on whether that effect occurs in a Class I area, a city park, 11 commercial cropland or private land. 12 In the proposed ozone reconsideration in 2010 the Administrator has found that the types 13 of information most useful in informing the selection of an appropriate range of protective levels March 2010 64 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 is appropriately focused on information regarding exposures and responses of sensitive trees and 2 other native species known or anticipated to occur in protected areas such as Class I areas or on 3 lands set aside by States, Tribes and public interest groups to provide similar benefits to the 4 public welfare, for residents on those lands, as well as visitors to those areas. She further notes 5 that while direct links between Os induced visible foliar injury symptoms and other adverse 6 effects (e.g., biomass loss) are not always found, visible foliar injury in itself is considered by the 7 National Park Service (NFS) to affect adversely air quality related values (AQRV) in Class I 8 areas, while the Administrator recognizes that uncertainty remains as to what level of annual tree 9 seedling biomass loss when compounded over multiple years should be judged adverse to the 10 public welfare, she believes that the potential for such anticipated effects should be considered in 11 judging to what degree a standard should be precautionary (73 FR 16496). The range of 12 proposed levels from 7-15 ppb includes at the maximum level of 15 ppb protection of 13 approximately 75% of seedlings from more than 10% biomass loss. 14 3.1.1.1.2 PM 15 [To be added in the second draft policy assessment based on the draft PM policy 16 assessment] 17 3.1.2 Other EPA Programs and Federal Agencies 18 Various federal laws and policies exist to protect ecosystem health. How other federal 19 agencies and EPA offices consider ecosystem effects in carrying out their programs can help 20 inform the Administrator when she evaluates the adversity of ecosystem impacts on public 21 welfare. For example, an effect may be considered adverse to public welfare if it contributes to 22 the inability of areas to meet water quality objectives as defined by the Clean Water Act. The 23 following federal statutes and policies may prove helpful to consider. 24 EPA Office of Water 25 Section 101 of the Clean Water Act (CWA) (Declaration of Goals and Policy) states that 26 the objective of the CWA is to restore and maintain the chemical, physical, and biological 27 integrity of the Nation's waters and to attain, where possible, water quality that protects fish, 28 shellfish, wildlife and provides for water-based recreation. March 2010 65 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 The CWA also authorizes EPA to develop water quality criteria as a guide for the states 2 to set water quality standards to protect aquatic life. In consideration of acidification effects, 3 EPA's Redbook, Quality Criteria for Water, published originally in 1976, recommends that 4 alkalinity be 20 mg/1 or more as CaCO3 for freshwater aquatic life except where natural 5 concentrations are less. Alkalinity is the sum total of components in the water that tend to elevate 6 the pH of the water above a value of about 4.5. 7 As mentioned in the Redbook, alkalinity is expressed as CaCO3 in mg/1. Alkalinity 8 differs slightly from ANC in that ANC includes other buffering compounds (Na, Mg, and K) as 9 well and includes buffering capacity of particulates in the water sample. Since alkalinity is 10 expressed as mg/1 and ANC is expressed as ueq/1, alkalinity must be multiplied by 20 to be 11 converted to ueq/1. Thus a recommended criterion of 20 mg/1 alkalinity is roughly equivalent to 12 an ANC of 400 ueq/1. 13 The Clean Air Act's Prevention of Significant Deterioration (PSD) program (42 14 U.S.C. 7470) purposes include to "preserve, protect and enhance the air quality in national parks, 15 wilderness areas and other areas of natural, recreational, scenic or historic value . . . ." Also, the 16 PSD program charges the Federal Land Managers, including the NFS, with ". . . an affirmative 17 responsibility to protect the air quality related values . . . "within federal Class I lands. (42 U.S.C. 18 7475(d)(2)(B)). 19 National Park Service 20 The National Park Service (NFS) is responsible for the protection of all resources within 21 the national park system. These resources include those that are related to and/or dependent upon 22 good air quality, such as whole ecosystems and ecosystem components. The NFS, in its Organic 23 Act (16 U.S.C. 1), is directed to conserve the scenery, natural and historic objects and wildlife 24 and to provide for the enjoyment of these resources unimpaired for current and future 25 generations. 26 The Wilderness Act of 1964 asserts wilderness areas will be administered in such a 27 manner as to leave them unimpaired and preserve them for the enjoyment of future generations. 28 NFS Management Policies (2006) guide all NFS actions including natural resources 29 management. In general, the NFS Management Policies reiterate the NFS Organic Act's mandate 30 to manage the resources "unimpaired." March 2010 66 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 U.S. Fish and Wildlife Service 2 On endangered species, Title 16 USC Chapter 35 Section 1531 states "The Congress 3 finds and declares that— these species offish, wildlife, and plants are of esthetic, ecological, 4 educational, historical, recreational, and scientific value to the Nation and its people and that all 5 Federal departments and agencies will use their authorities to conserve threatened and 6 endangered species. 7 The United States Fish and Wildlife Service (FWS) manages the National Wildlife 8 Refuge System lands to "...ensure that the biological integrity, diversity, and environmental 9 health of the Systems are maintained for the benefit of present and future generations of 10 Americans." 16 U.S.C. Section 668dd(a)(4)(B)(1997). 11 U.S. Forest Service 12 The National Forest units are managed consistent with Land and Resource Management 13 Plans (LRMPs) under the provisions of the National Forest Management Act (NFMA). 16 14 §U.S.C. 1604 (1997). LRMPs are, in part, specifically based on recognition that the National 15 Forests are ecosystems and their management for goods and services requires an awareness and 16 consideration of the interrelationships among plants, animals, soil, water, air, and other 17 environmental factors within such ecosystems. 36 C.F.R. §219.1(b)(3) 18 Any measures addressing Air Quality Related Values (AQRV) on National Forest 19 System lands will be implemented through, and be consistent with, the provisions of an 20 applicable LRMP or its revision (16 U.S.C. §1604(i)). Additionally, the Secretary of Agriculture 21 must prepare a Renewable Resource Program that recognizes the need to protect and, if 22 necessary, improve the quality of air resources. 16 U.S.C. §1602(5)(C). 23 AQRVs in Wilderness areas may receive further protection by the previously mentioned 24 1964 Wilderness Act. For Wilderness Areas in the National Forest System, the Act's 25 implementing regulations are found at 36 C.F.R. §293 requiring these Wilderness Areas be 26 administered to preserve and protect [their] wilderness character. 27 Chesapeake Bay Total Maximum Daily Loads 28 Under section 303(d) of the Clean Water Act, states, territories, and authorized tribes are 29 required to develop lists of impaired waters. These are waters that are too polluted or otherwise 30 degraded to meet the water quality standards set by states, territories, or authorized tribes. The March 2010 67 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 law requires that these jurisdictions establish priority rankings for waters on the lists and develop 2 TMDLs for these waters. A Total Maximum Daily Load, or TMDL, is a calculation of the 3 maximum amount of a pollutant that a waterbody can receive and still safely meet water quality 4 standards. EPA is developing a TMDL for the Chesapeake Bay and its tributaries. The 5 Chesapeake Bay Program has modeled the level of nitrogen that can reach the Bay and still meet 6 the Bay's water quality standards. The TMDL, with full public participation, will set waste load 7 allocations for point source discharges and load allocations for nonpoint sources of nitrogen. Air 8 deposition to the Bay and its watershed, as a source category, will have a specific allocation. The 9 allocation can be used to calculate the level of ambient air concentrations of reactive nitrogen 10 that are likely to meet the deposition allocation. To find the NOX portion of the allocation one 11 would subtract the reduced forms from the total allocation. If the total load to the Bay of nitrogen 12 from all the allocated source categories remains below the allocations, then the Bay is expected 13 to meet the water quality standards, which are set to protect the designated uses of the Bay. Since 14 the designated uses are set by the states with public input, not meeting the designated uses can be 15 seen as having an adverse effect. 16 United Nations Economic Commission for Europe (UNECE) 17 [This information will be included in the second draft.] 18 Critical Loads 19 The term critical load is used to describe the threshold of air pollution deposition that 20 causes a specified level of harm to sensitive resources in an ecosystem. A critical load is 21 technically defined as "the quantitative estimate of an exposure to one or more pollutants below 22 which significant harmful effects on specified sensitive elements of the environment are not 23 expected to occur according to present knowledge" (Nilsson and Grennfelt, 1988). The 24 determination of when a harmful effect becomes "significant" may be in the view of a researcher 25 or through a policy development process. Researchers often use the term "critical loads" to 26 describe when particular detrimental effects are realized, as is the case in Figure 2-1. In many 27 European countries a critical loads framework is used to determine a level of damages to 28 ecosystem services from pollution that are legally allowed. These critical loads are determined 29 through a policy process. March 2010 68 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX Harmful effects due to acidification have been defined here as those that occur below a given ANC for aquatic systems and below a given Be: Al ratio for terrestrial systems. However, 3 the level at which an effect becomes harmful in that it causes adverse effects on public welfare is H(^t(^rmm(^H \\\j tVi£> A Hmim ctrcitr\r 1 2 given 4 determined by the Administrator. 5 3.2 WHAT ARE ECOSYSTEM SERVICES AND HOW DOES THIS 6 CONCEPT RELATE TO PUBLIC WELFARE? 7 An additional concept that may be useful in considering the issue of adversity to public 8 welfare is ecosystem services. In the next section the concept of ecosystem services, its 9 relationship to adversity and public welfare within the context of this review are explained. 10 Characterizing a known or anticipated adverse effect to public welfare is an important 11 component of developing any secondary NAAQS. According to the Clean Air Act, welfare 12 effects include: 13 effects on soils, water, crops, vegetation, manmade materials, animals, wildlife, 14 weather, visibility, and climate, damage to and deterioration of property, and 15 hazards to transportation, as well as effect on economic values and on personal 16 comfort and well-being, whether caused by transformation, conversion, or 17 combination with other air pollutants (CAA, Section 302(h)). 18 While the text above lists a number of welfare effects, these effects are not an effect on 19 public welfare in and of themselves. 20 Ecosystem services can be generally defined as the benefits individuals and organizations 21 obtain from ecosystems. Ecosystem services can be classified as provisioning (food and water), 22 regulating (control of climate and disease), cultural (recreational), and supporting (nutrient 23 cycling) (MEA 2005). Conceptually, changes in ecosystem services may be used to aid in 24 characterizing a known or anticipated adverse effect to public welfare. In the context of this 25 review, ecosystem services may also aid in assessing the magnitude and significance to the 26 public of a resource and in assessing how NOX and SOX concentrations and deposition may 27 impact that resource. The relationship between ecosystem services and public welfare effects is 28 illustrated in Table 3.2.1. March 2010 69 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX Table 3-1. Crosswalk between Ecosystem Services and Public Welfare Effects Public Welfare Effect Soils Water Crops Vegetation Wildlife Climate * Personal Comfort and Wellbeing Ecosystem Service Nutrient Cycling Drinking water, Recreation, Aesthetic Food, Fuel Production Food, Recreation, Aesthetic, Nonuse Recreation, Food, Nonuse Climate Control Service Category Supporting Provisioning, Cultural Provisioning Provisioning, Cultural Cultural, Provisioning Regulating 1 *A11 ecosystem services contribute to personal comfort and wellbeing. 2 EPA has defined ecological goods and services for the purposes of a Regulatory Impact 3 Analysis as the "outputs of ecological functions or processes that directly or indirectly contribute 4 to social welfare or have the potential to do so in the future. Some outputs may be bought and 5 sold, but most are not marketed" (US EPA 2006). Though this is not a definition specifically for 6 use in the NAAQS process it may be a useful one in considering the scope of ecosystem services 7 and the effects of air pollutants upon those services. Especially important is the 8 acknowledgement that most of the goods and services supplied by ecosystems cannot be fully 9 measured or monetized. Valuing ecological benefits, or the contributions to social welfare 10 derived from ecosystems, can be challenging as noted in EPA's Ecological Benefits Assessment 11 Strategic Plan (US EPA 2006) and the Science Advisory Board report "Valuing the Protection of 12 Ecological Systems and Services" (US EPA, 2009). It can be informative in characterizing 13 adversity to public welfare to attempt to place an economic valuation on the set of goods and 14 services that have been identified with respect to a change in policy however it must be noted 15 that this valuation will be incomplete and illustrative only. The stepwise concept leading to the 16 valuation of ecosystem services is graphically depicted in Figure 3-2. March 2010 70 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX EPA action ^^•^H Ecosyste Ecological goods and services affected by the policy Planning and problem formulation ' Goods and services identified Ecological analysis Goods and services quantified Economic analysis Goods and services monetized Goods and services not identified Identified goods and services not quantified Quantified goods and services not monetized 2 Figure 3-2. RepresemauoTI 01 me oenenis assessment process indicating where 3 some ecological benefits may remain unrecognized, unquantified, or 4 unmonetized. (Source: EBASP USEPA 2006). 5 A conceptual model integrating the role of ecosystem services in characterizing known or 6 anticipated adverse effects to public welfare is shown in Figure 3-3. Under Section 109 of the 7 CAA, the secondary standard is to specify a level of air quality that is requisite to protect public 8 welfare. For this review, the relevant air quality indicator is interpreted as ambient NOX and SOX 9 concentrations that can be linked to levels of deposition for which there are ecological effects 10 that are adverse to public welfare. The case study analyses (described in Chapters 4 and 5 of the 11 REA and summarized in Chapter 2 of this document) link deposition in sensitive ecosystems 12 (e.g., the exposure pathway) to changes in a given ecological indicator (e.g., for aquatic 13 acidification, changes in acid neutralizing capacity [ANC]) and then to changes in ecosystems 14 and the services they provide (e.g., fish species richness and its influence on recreational 15 fishing). To the extent possible for each targeted effect area, ambient concentrations of nitrogen 16 and sulfur (i.e., ambient air quality indicators) were linked to deposition in sensitive ecosystems 17 (i.e., exposure pathways), and then deposition was linked to system response as measured by a 18 given ecological indicator (e.g., lake and stream acidification as measured by ANC). The March 2010 71 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 ecological effect (e.g., changes in fish species richness, etc.) was then, where possible, associated 2 with changes in ecosystem services and their public welfare effects (e.g., recreational fishing). 3 Knowledge about the relationships linking ambient concentrations and ecosystem 4 services can be used to inform a policy judgment on a known or anticipated adverse public 5 welfare effect. The conceptual model outlined for aquatic acidification in Figure 3-3 can be 6 modified for any targeted effect area where sufficient data and models are available. For 7 example, a change in an ecosystem structure and process, such as foliar injury, would be 8 classified as an ecological effect, with the associated changes in ecosystem services, such as 9 primary productivity, food availability, and aesthetics (e.g., scenic viewing), classified as public 10 welfare effects. Additionally, changes in biodiversity would be classified as an ecological effect, 11 and the associated changes in ecosystem services—productivity, recreational viewing and 12 aesthetics—would be classified as public welfare effects. This information can then be used by 13 the Administrator to determine whether or not the changes described are adverse to public 14 welfare. In subsequent sections these concepts are applied to characterize the ecosystem services 15 potentially affected by nitrogen and/or sulfur for each of the effect areas assessed in the REA. March 2010 72 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX Ambient Air Quality Indicator Exposure Pathway Affected Ecosystem Ecological Response (ecological indicator ) NOX/SOX Concentrations Atmospheric N & S Deposition Aquatic ~ Acidification (lake/stream ANC ) Ecological Effect Change in Ecosystem Structure & Processes (fish species richness ) 1 2 3 4 5 6 9 10 11 12 Ecological Benefit Welfare Effect Change in Ecosystem Services (recreational fishing ) Figure 3-3. Conceptual model showing the relationships among ambient air quality indicators and exposure pathways and the resulting impacts on ecosystems, ecological responses, effects and benefits to characterize known or anticipated adverse effects to public welfare. [This figure to be revised for Second Draft Policy Assessment Document] These concepts can also be applied to the programs described in section 3.1. National parks represent areas of nationally recognized ecological and public welfare significance, which are afforded a higher level of protection. Therefore, staff has also focused on air quality and deposition in the subset of national park sites and important natural areas. Figures 3-4 and 3-5 illustrate the spatial relationships between sensitive regions, Class 1 areas and nitrogen deposition levels. March 2010 73 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 * *;••,••' '•^-.j^ | | Class 1 Areas J Sensitive Aquatic Areas Combined N and S (wet and dry) Value •I High : 1 50003e»008 • Low: 1 01593e+006 •• ~\ ^ Figure 3-4. Locations of Eastern U.S. National Parks (Class I areas) relative to deposition of Nitrogen and Sulfur in sensitive aquatic areas March 2010 74 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX | Class 1 Areas Combined N and S (wet and dry) Value • High : 150.003 kg/ha/yr | Low : 1.016 kg/ha/yr 1 2 Figure 3-5. Location of Western U.S. National Parks (Class I areas) relative to 3 deposition of Nitrogen and Sulfur 4 [Figures 3-4 and 3-5 will be revised for Second Draft policy Assessment Document] 5 3.3 WHAT IS THE ROLE OF ECONOMICS? 6 As discussed earlier in this document, a secondary NAAQS is required to be set at the 7 "level(s) of air quality necessary to protect the public welfare from any known or anticipated 8 adverse effects". As part of the effort to determine the standard, EPA linked the changes in the 9 ambient air concentrations of NOX and SOX to the changes in ecosystem services and ultimately 10 to changes in public welfare (U.S. EPA, 2009). As previously mentioned most ecosystem March 2010 75 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 services are not amenable to monetization a small subset of changes in services can be described 2 by economic valuation methods. And although economics on its own cannot determine which 3 impact on public welfare is "adverse", economics could be helpful in the context of a secondary 4 NAAQS for determining the degree to which improvements are beneficial to public welfare and 5 illustrating and aggregating those impacts.7 6 The Role of Economics in Defining "Adversity" There is neither an economic definition 7 of how much loss in public welfare is adverse nor an economic definition of adversity. While an 8 economist might consider a particular scenario adverse because it might imply some harm or 9 potential for improvement, there is no specific threshold level when a loss in welfare (e.g. loss in 10 dollars) becomes adverse. An individual might be willing to give up some of their resources to 11 avoid a threat or negative outcome (i.e., willing to pay to avoid a particular outcome). According 12 to economic theory, if an individual is willing to give up something to avoid the outcome, then 13 imposing the outcome on the individual must make them worse off, at which point an economist 14 might colloquially describe the outcome as adverse. However, the amount they would have been 15 willing to pay to avoid the outcome might be quite small, and might not rise to a level of harm 16 that the Administrator interprets as "adverse" to public welfare. In summary, economics provides 17 little guidance as to how the Administrator should interpret the word "adverse" in the context of 18 public welfare. 19 Ecosystem Services and Links to Public Welfare An ecosystem service framework 20 provides a structure to measure changes in public welfare from changes in ecosystem functions 21 affected by air pollution. EPA's Risk Assessment for this rulemaking defines ecosystem services 22 as "the ecological processes or functions having monetary or nonmonetary value to individuals 23 or society at large" (EPA 2009.) The discipline of economics provides a useful approach for 24 summarizing how the public values changes in the services provided by the environment. An 25 ecosystem services framework (with or without valuation) can provide measures of changes in 26 public welfare. 7 Section 109 of the Clean Air Act forbids consideration of the compliance costs of reducing pollution when setting a NAAQS. However, there is no prohibition regarding the consideration of the monetized impacts of welfare effects occurring due to levels of pollution above alternative standards in evaluating the adversity of the impacts to public welfare. Ecosystem services can be characterized as a method of monetizing the impacts of the air pollution. Although a separate regulatory document quantifying the costs and benefits of attaining a NAAQS is prepared simultaneously, this document is not considered when selecting a standard. March 2010 76 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Economics as a Framework to Illustrate Changes in Public Welfare Economics can 2 provide a framework to illustrate how public welfare8 changes in response to changes in 3 environmental quality by quantitatively linking changes in ecosystem services to preferences. 4 Economics assumes that the choices that individuals make reflect their preferences over certain 5 outcomes and that, generally speaking, they will make choices that, in expectation, will make 6 them as well off as possible given their resources. In economics revealed and stated preference 7 methods are used to observe the choices individuals make to understand the outcomes 8 individuals prefer. What individuals are willing to give up for an outcome is their willingness-to- 9 pay (WTP) for that outcome. An example of an outcome is an improvement in an ecosystem 10 service. Often, to provide comparability to other goods and services, in economics these 11 tradeoffs are framed relative to dollars for convenience.9 12 Economics could inform the Administrator by valuing and characterizing the changes in 13 public welfare from changes in the quantity and quality of ecosystem services. Overall, this 14 assessment intends to characterize changes in ecosystem services from a scientific perspective 15 using effects on ecosystem structures and functions or ecosystem integrity. Economics then 16 estimates the effect on public welfare of these changes in the quantity and quality of ecosystem 17 services. For example, a decrease in a particular bird species can be characterized by its effect on 18 the ecosystem's structure and function, while from an economic perspective, the effects would 19 be based on the impact on public welfare or the value the public places on that species. A simple 20 example is a comparison between a decrease in a bird species that is relatively unknown 21 compared to a decrease in a very prominent species (e.g. Bald Eagle). The public is likely to 22 have a higher WTP to avoid the latter, and thus the decrease would affect the public welfare 23 more. 24 There are important complications with using preferences to understand the effect of 25 pollution on public welfare. For example, while the field of economics generally assumes that 26 public preferences are the paramount consideration; these preferences may change when the 27 public receives new information. Therefore, if individuals do not understand how pollution will [A discussion of economic interpretation of "Public Welfare" will be included in the second draft] 9 Often groups collectively make choices to engage in activities that improve the collective welfare of the group. For example, a community around an acidified lake might purchase lime and use it to reduce the acidity of the lake. The collective decisions can also be used to understand how people value improvements to ecosystem services. [Additional discussion will be included in the second draft related to collective actions that reveal preferences for improvements in relevant ecosystem services and how these collective actions, and the absence of these actions, can be interpreted.] March 2010 77 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 affect ecosystem services, or even how those ecosystem services affect their quality of life, then 2 they will have a difficult time valuing changes in those services. Similarly, it may be very costly 3 for individuals to learn and understand how changes in particular ecosystem services may affect 4 them, in part because typically there are significant interdependences within an ecosystem. 5 Because of this complexity, individuals may implicitly value a species, or habitat, or ecosystem 6 function because it supports an ecosystem service that they do clearly value. Furthermore, the 7 public also has limited understanding regarding irreversibilities, tipping points, and other more 8 complex aspects of ecosystems, which limits the ability to adequately value these ecosystems.10 9 In addition, where and when a change in an ecosystem takes places is crucial for characterizing 10 the associated change in an ecosystem service, and will also affect the value the public places on 11 that change. 12 3.4 WHAT IS THE EVIDENCE FOR EFFECTS ON ECOSYSTEM 13 SERVICES? HOW DO WE LINK ECOLOGICAL INDICATORS TO 14 SERVICES? 15 The process used to link ecological indicators to ecosystem services is discussed 16 extensively in Appendix 8 of the REA. In brief, for each effect area assessed the ecological 17 indicators were linked to an ecological response that was subsequently linked, to the extent 18 possible, to associated services. For example in the case study for aquatic acidification the 19 chosen ecological indicator is ANC which can be linked to the ecosystem service of recreational 20 fishing as illustrated in the conceptual model shown in Figure 3-6. Although recreational fishing 21 losses are the only service effects that can be quantified or monetized at this time, there are, as 22 can be seen in the Figure, numerous other ecosystem services that may be related to the 23 ecological effects of acidification. 10 While the public may not fully appreciate the interdependencies within ecosystems, they can learn them, but again it may be costly to do so. It is possible for individuals to value outcomes that are irreversible or result in discrete changes (i.e., tipping points) in the quality and quantity of ecosystem services. Avoiding irreversible outcomes should be and are more valued by individuals than outcomes that are not irreversible (Arrow and Fischer, 1974). March 2010 78 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX Acidifying Inputs to Surface Wat&r Impacts on Ecosystem Endpoinrs Affected Ecosystem Services N+S Deposition Surface Water Acidification: Low pH and ANC Declines in Aquatic Biota Declines in •1 Aquatic Biota: I Reduced [ Species J Abundance, Diversity, and Richness • Declines in Terrestnai Nearshore Biota ^^^^^H ^^> * Provisioning Services •production f of commercial and subsistence fishing Cultural Services •recreational fishing •waterfowl hunting •aesthetic enjoyment •no nuse services Regulating services •biological control 2 Figure 3-6. Conceptual model linking ecological indicator (ANC) to affected 3 ecosystem services. 4 The next four sections summarize the current levels of certain ecosystem services for 5 each of the effect areas analyzed in the REA and present results of analyses that have attempted 6 to quantify and monetize the harms to public welfare, as represented by ecosystem services, due 7 to nitrogen and sulfur deposition. 8 Evidence for Adversity Related to Aquatic Acidification 9 Acidification primarily affects the ecosystem services that are derived from the fish and 10 other aquatic life found in these surface waters (REA, Section 5.2.1.3). Food is generally the 11 most important provisioning services provided by inland surface waters (MEA, 2005). In the 12 northeastern United States, the surface waters affected by acidification are not a major source of 13 commercially raised or caught fish; however, they are a source of food for some recreational and 14 subsistence fishers and for other consumers. Although data and models are available for 15 examining the effects on recreational fishing, relatively little data are available for measuring the 16 effects on subsistence and other consumers. For example, although there is evidence that certain 17 population subgroups in the Northeastern United States, such as the Hmong and Chippewa ethnic 18 groups, have particularly high rates of self-caught fish consumption (Hutchison and Kraft, 1994; 19 Peterson et al., 1994), it is not known if and how their consumption patterns are affected by the 20 reductions in available fish populations caused by surface water acidification. 21 Inland surface waters support several cultural services, such as aesthetic and educational 22 services; however, the type of service that is likely to be most widely and significantly affected March 2010 79 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 by aquatic acidification is recreational fishing11. Recreational fishing in lakes and streams is 2 among the most popular outdoor recreational activities in the northeastern United States. Data 3 from the 2006 National Survey of Fishing, Hunting, and Wildlife Associated Recreation 4 (FHWAR) indicate that more than 9% of adults in this part of the country participate annually in 5 freshwater fishing with 140 million freshwater fishing days. Based on studies conducted in the 6 northeastern United States, Kaval and Loomis (2003) estimated average consumer surplus values 7 per day of $35 for recreational fishing (in 2007 dollars). Therefore, the implied total annual value 8 of freshwater fishing in the northeastern United States was $5 billion in 2006. 9 In general, inland surface waters such as lakes, rivers, and streams provide a number of 10 regulating services, such as hydrological regime regulation and climate regulation. There is little 11 evidence that acidification of freshwaters in the northeastern United States has significantly 12 degraded these specific services; however, freshwater ecosystems also provide biological control 13 services by providing environments that sustain delicate aquatic food chains. 14 The toxic effects of acidification on fish and other aquatic life impair these services by 15 disrupting the trophic structure of surface waters (Driscoll et al., 2001). Although it is difficult to 16 quantify these services and how they are affected by acidification, it is worth noting that some of 17 these services may be captured through measures of provisioning and cultural services. For 18 example, these biological control services may serve as "intermediate" inputs that support the 19 production of "final" recreational fishing and other cultural services. 20 What is the value of the impaired recreational fishing services? 21 The previous section describes the ecosystem services that are most likely to be affected 22 by N and S deposition, and it summarizes evidence regarding the current magnitude and values 23 of recreational fishing services; however, it does not measure the degree to which these services 24 are impaired by existing NOX/SOX levels. 25 To address this limitation, the REA (Appendix 8) provides insights into the magnitude of 26 ecosystem service impairments. 27 Specifically, the REA focuses on measuring the benefits of ecosystem service 28 enhancements resulting from the elimination of anthropogenic sources of NOX/SOX. Rather than 29 asking how much public welfare is currently adversely affected relative to a scenario without 11 Banzhaf et al (2006) has shown that non-use services are arguably a more significant source of benefits from reduced acidification than recreational fishing. March 2010 80 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 anthropogenic NOX/SOX, it asks a similar question of how much public welfare would improve if 2 the emissions were eliminated. The REA provides quantitative estimates of selected ecosystem 3 services impairments or enhancements for three main categories of ecosystem effects - aquatic 4 acidification, terrestrial acidification, and aquatic nutrient enrichment12. Within these three 5 categories, the selection of specific ecosystem services for more in-depth analysis depended 6 primarily on the expected magnitude of impairments and on the availability of appropriate data 7 and modeling tools. 8 The analysis of ecosystem service impairments due to aquatic acidification builds on the 9 case study analysis of lakes in the New York Adirondacks. It estimates changes in recreational 10 fishing services, as well as changes more broadly in "cultural" ecosystem services (including 11 recreational, aesthetic, and nonuse services). First, the MAGIC model was applied to 44 lakes to 12 predict what ANC levels would be under both "business as usual" conditions (i.e., allowing for 13 some decline in deposition due to existing regulations) and pre-emission (i.e., background) 14 conditions. When these model runs were initiated staff were interested in a prospective analysis 15 of conditions assuming a 2010 implementation of "zero-out" emissions with a projected lag time 16 to improvement of 10 years thus results were calculated for the year 2020. These predictions 17 were then extrapolated to the full universe of Adirondack lakes. Second, to estimate the 18 recreational fishing impacts of aquatic acidification in these lakes, an existing model of 19 recreational fishing demand and site choice was applied. This model predicts how recreational 20 fishing patterns in the Adirondacks would differ and how much higher the average annual value 21 of recreational fishing services would be for New York residents if lake ANC levels 22 corresponded to background (rather than business as usual) conditions. Aggregating these values 23 across all NY residents implies that acidification of Adirondack lakes due to anthropogenic 24 sources of NOX/SOX would impair annual recreational fishing services of NY residents by $6 25 million to $11 million in 2020. Current annual impairments are most likely of a similar 26 magnitude because, although current NOX/SOX levels are somewhat higher than those expected in 27 2020 (under business as usual - given expected emissions controls associated with Title IV 28 regulations but no additional nitrogen or sulfur controls), the affected NY population is also 29 somewhat smaller (based on U.S. Census Bureau projections). 12 Estimates for terrestrial nutrient enrichments were not generated due to the limited availability of necessary data and models for this effect category. March 2010 81 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Third, to estimate impacts on a broader category of cultural ecosystem services, results 2 from an existing valuation survey of NY residents were adapted and applied to this context. The 3 survey used a contingent valuation approach to estimate the average annual household WTP for 4 future reductions (20% and 45%) in the percent of Adirondack lakes impaired by acidification. 5 These WTP estimates were then (1) rescaled to reflect predicted changes between business-as- 6 usual and background conditions in 2020 (MAGIC lake modeling results indicate that the 7 percentage of impaired lakes would be 22 to 31 points lower under background conditions), and 8 (2) aggregated across NY households. The aggregate annual value to NY residents in 2010 for a 9 reduction in lake acidification to background levels by 2020 was estimated to range $4 million to 10 $300 million in 2007 dollars. For comparison the previous section estimated the value of 11 recreational fishing in the Northeastern states at approximately $5 billion in 2006. These results 12 suggest that the value of avoiding current impairments to ecosystem services from Adirondack 13 lakes are even higher than the estimate, because they occur today rather than in 2020 (i.e., no 14 delayed effect) and because the percent of impaired lakes is slightly higher today than expected 15 in 2020 under business-as-usual. These results imply significant value to the public derived from 16 recreational fishing services. The analysis especially illustrates what may be the scale of all 17 impacts to public welfare when viewed as a subset of all services impacted by acidification. 18 Evidence for Terrestrial Acidification 19 A similar model to Figure 3-6 can be drawn for terrestrial acidification that links Bc:Al 20 ratio to reduced tree growth to decreases in timber harvest although we have less confidence in 21 the significance of this linkage than we do for aquatic acidification. There are numerous services 22 expected to be affected, but the means to adequately describe those losses does not as yet exist. 23 These services include effects to forest health, water quality, and habitat, including decline in 24 habitat for threatened and endangered species, decline in forest aesthetics, decline in forest 25 productivity, increases in forest soil erosion and decreases in water retention (ISA, 2009; REA, 26 2009; Krieger, 2001). 27 Forests in the Northeastern United States provide several important and valuable 28 provisioning services, which are reflected in the production and sales of tree products. 29 Sugar maples are a particularly important commercial hardwood tree species in the 30 United States, producing wood products like timber and maple syrup that provide hundreds of March 2010 82 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 millions of dollars in economic value annually (NASS, 2008). Red spruce is also used in a 2 variety of wood products and provides up to $100 million in economic value annually. 3 Forests in the Northeastern United States are also an important source of cultural 4 ecosystem services, including nonuse (existence value for threatened and endangered species), 5 recreational, and aesthetic services (ISA, 2009; REA, 2009). Red spruce forests are home to two 6 federally listed species. 7 Although we do not have the data to link acidification damages directly to economic 8 values of lost recreational services in forests, these resources are valuable to the public. A recent 9 study suggests that the total annual value of off-road driving recreation was more than $9 billion, 10 total and value of hunting and wildlife viewing was more than $4 billion each in the Northeastern 11 United States in 2006(Kaval and Loomis, 2003). In addition, fall color viewing is a recreational 12 activity that is directly dependent on forest conditions. Sugar maple trees, in particular, are 13 known for their bright colors and are, therefore, an essential aesthetic component of most fall 14 color landscapes. Statistics on fall color viewing are much less available than for the other 15 recreational and tourism activities; however, a few studies have documented the extent and 16 significance of this activity. For example, Spencer and Holecek (2007) found that roughly 30% 17 of residents reported at least one trip in the previous year involving fall color viewing. In a 18 separate study conducted in Vermont, Brown (2002) reported that more than 22% of households 19 visiting Vermont in 2001 made the trip primarily for the purpose of viewing fall colors. 20 Two studies that have estimated values for protecting high-elevation spruce forests in the 21 Southern Appalachians. Kramer et al. (2003) conducted a contingent valuation study estimating 22 households' WTP for programs to protect remaining high-elevation spruce forests from damages 23 associated with air pollution and insect infestation (Haefele et al., 1991; Holmes and Kramer, 24 1995). Median household WTP was estimated to be roughly $29 (in 2007 dollars) for the 25 minimal program and $44 for the more extensive program. Another study by Jenkins, Sullivan, 26 and Amacher (2002) estimated an aggregate annual value of $3.4 billion for avoiding a 27 significant decline in the health of high-elevation spruce forests in the Southern Appalachian 28 region. 29 Forests in the Northeastern United States also support and provide a wide variety of 30 valuable regulating services, including soil stabilization and erosion control, water regulation, 31 and climate regulation (Krieger, 2001). Forest vegetation plays an important role in maintaining March 2010 83 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 soils in order to reduce erosion, runoff, and sedimentation that can adversely impact surface 2 waters. In addition to protecting the quality of water in this way, forests also help store and 3 regulate the quantity and flows of water in watersheds. Finally, forests help regulate climate 4 locally by trapping moisture and globally by sequestering carbon. The total value of these 5 ecosystem services is very difficult to quantify and the magnitude of these impacts is currently 6 very uncertain. 7 What is the value of current ecosystem service impairments? 8 The analysis of ecosystem service impairments associated with terrestrial acidification 9 specifically addresses impacts on the forest product provisioning services from two 10 commercially important tree species - sugar maple and red spruce—that are particularly sensitive 11 to the effects of acidification. Using data from the USFS Forest Inventory and Analysis (FIA) 12 database, an exposure-response relationship was estimated for each species to measure the 13 average negative effect of critical load exceedances (CLEs) of nitrogen and sulfur deposition on 14 annual tree growth. These estimated relationships were then applied to sugar maple and red 15 spruce stocks in the Northeast and North central regions to estimate the average percent increase 16 in annual tree growth that would occur if all CLEs were eliminated. To estimate the aggregate- 17 level forest market impacts of eliminating CLEs starting in the year 2000, the tree-level growth 18 adjustments were applied using the Forest and Agricultural Sector Optimization Model 19 (FASOM), which is a dynamic optimization model of the U.S. forest and agricultural sectors. 20 The public welfare gains linked to these markets from eliminating CLEs was estimated to be 21 $0.69 million per year. These estimates can also be interpreted as the current value of 22 impairments to forest provisioning services due to forest acidification effects from nitrogen and 23 sulfur deposition. 24 Nutrient Enrichment 25 For the purposes of the following sections nutrient enrichment refers only to that due to 26 NOy deposition. Additionally these sections focus on the detrimental effects of that deposition. 27 Staff acknowledges that a certain amount of NOX deposition in managed terrestrial ecosystems 28 may have a beneficial effect. However no attempt has been made to quantify those beneficial 29 effects. March 2010 84 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Evidence for Aquatic Nutrient Enrichment 2 Estuaries in the eastern United States are an important source of food production, in 3 particular fish and shellfish production. The estuaries are capable of supporting large stocks of 4 resident commercial species, and they serve as the breeding grounds and interim habitat for 5 several migratory species (U.S. EPA, 2009). To provide an indication of the magnitude of 6 provisioning services associated with coastal fisheries, from 2005 to 2007, the average value of 7 total catch was $1.5 billion per year in 15 East Coast states. It is not known, however, what 8 percentage of this value is directly attributable to or dependent upon the estuaries in these states. 9 Based on commercial landings in Maryland and Virginia, the values for three key species—blue 10 crab, striped bass, and menhaden- totaled nearly $69 million in 2007 in the Chesapeake Bay 11 alone. 12 Assessing how eutrophication in estuaries affects fishery resources requires bioeconomic 13 models (i.e., models that combine biological models offish population dynamics with economic 14 models describing fish harvesting and consumption decisions), but relatively few exist (Knowler, 15 2002). Kahn and Kemp (1985) estimated that a 50% reduction in SAV from levels would 16 decrease the net social benefits from striped bass by $16 million (in 2007 dollars). In a separate 17 analysis, Anderson (1989) modeled blue crab harvests under baseline conditions and under 18 conditions with "full restoration" of SAV. In equilibrium, the increase in annual producer surplus 19 and consumer surplus with full restoration of SAV was estimated to be $7.9 million (in 2007 20 dollars). Mistiaen, Strand, and Lipton (2003) found that reductions in DO cause a statistically 21 significant reduction in commercial harvest and revenues crab harvests. For the Patuxent River 22 alone, a simulated reduction of DO from 5.6 to 4.0 mg/L was estimated to reduce crab harvests 23 by 49% and reduce total annual earnings in the fishery by $275,000 (in 2007 dollars). 24 In addition, eutrophi cation in estuaries may also affect the demand for seafood. For 25 example, a well-publicized toxic pfiesteria bloom in the Maryland Eastern Shore in 1997 led to 26 an estimated $56 million (in 2007 dollars) in lost seafood sales for 360 seafood firms in 27 Maryland in the months following the outbreak (Lipton, 1999). Surveys by Whitehead, Haab, 28 and Parsons (2003) and Parsons et al. (2006) indicated a reduction in consumer surplus due to 29 eutrophication-related fish kills ranging from $2 to $5 per seafood meal.13 As a result, they 13 Surprisingly, these estimates were not sensitive to whether the fish kill was described as major or minor or to the different types of information included in the survey. March 2010 85 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 estimated aggregate consumer surplus losses of $43 million to $84 million (in 2007 dollars) in 2 the month after a fish kill. 3 As mentioned in the REA (5.2.1.3), estuaries in the eastern United States also provide an 4 important and substantial variety of cultural ecosystem services, including water-based 5 recreational and aesthetic services. For example, FHWAR data indicate that 4.8% of the 6 population in coastal states from North Carolina to Massachusetts participated in saltwater 7 fishing, in 26 million saltwater fishing days in 2006 (U.S. DOT, 2007). Based on estimates in 8 Section 5.2.1.3 of the REA, total recreational consumer surplus value from these saltwater 9 fishing days was approximately $1.3 billion (in 2007 dollars). Recreational participation 10 estimates for several other coastal recreational activities are also available for 1999-2000 from 11 the NSRE. Almost 6 million individuals participated in motorboating in coastal states from North 12 Carolina to Massachusetts. Again, based on analysis in the REA, the aggregate value of these 13 coastal motorboating outings was $2billion per year. Almost 7 million participated in 14 birdwatching, for a total of almost 175 million days per year, and more than 3 million 15 participated in visits to nonbeach coastal waterside areas, for a total of more than 35 million days 16 per year. 17 Estuaries and marshes have the potential to support a wide range of regulating services, 18 including climate, biological, and water regulation; pollution detoxification; erosion prevention; 19 and protection against natural hazards (MEA, 2005c). The relative lack of empirical models and 20 valuation studies imposes obstacles to the estimation of ecosystem services affected by nitrogen 21 deposition. While atmospheric deposition contributes to eutrophication there is uncertainty in 22 separating the effects of atmospheric nitrogen from nitrogen reaching the estuaries from many 23 other sources. 24 What is the value of current ecosystem service impairments? 25 The aquatic nutrient enrichment case study relied on the NOAA Eutrophi cation Index as 26 the indicator, which includes dissolved oxygen, HABs, loss of SAV and loss of water clarity. 27 There are methods available to link some of the components to ecosystem services, most notably 28 loss of SAV and reductions in DO. The REA analysis estimates the change in several ecosystem 29 services including recreational fishing, boating, beach use, aesthetic services and nonuse 30 services. The REA focuses on two major East Coast estuaries - the Chesapeake Bay and the 31 Neuse River. Both estuaries receive between 20%-30% percent of their annual nitrogen loadings March 2010 86 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 through air deposition and both are showing symptoms of eutrophication. The analysis uses and 2 adapts results from several existing studies to approximate effects on several ecosystem services, 3 including commercial fishing, recreation, aesthetic enjoyment, and nonuse values. For example, 4 it is estimated that atmospheric nitrogen reduces the annual benefits of recreational fishing, 5 boating, and beach use in the Chesapeake Bay by $43-$217 million, $3-8 million, and $124 6 million respectively, and reduces annual aesthetic benefits to nearshore residents by $39-102 7 million. In the Neuse River, the value of annual commercial crab fishing services would be 8 between $0.1-1 million higher without the contribution of atmospheric nitrogen, and recreation 9 fishing services in the larger Albermarle Pamlico Sound estuary system (which includes the 10 Neuse) would be $ 1 -8 million greater per year. 11 Evidence for Terrestrial Nutrient Enrichment 12 The ecosystem service impacts of terrestrial nutrient enrichment include primarily 13 cultural and regulating services. In CSS areas, concerns focus on a decline in CSS and an 14 increase in nonnative grasses and other species, impacts on the viability of threatened and 15 endangered species associated with CSS, and an increase in fire frequency. Changes in MCF 16 include changes in habitat suitability and increased tree mortality, increased fire intensity, and a 17 change in the forest's nutrient cycling that may affect surface water quality through nitrate 18 leaching (EPA, 2008). 19 The value that California residents and the U.S. population as a whole place on CSS and 20 MCF habitats is reflected in the various federal, state, and local government measures that have 21 been put in place to protect these habitats. Threatened and endangered species are protected by 22 the Endangered Species Act. The State of California passed the Natural Communities 23 Conservation Planning Program (NCCP) in 1991, and CSS was the first habitat identified for 24 protection under the program (see www.dfg.ca.gov/habcon/nccp). Private organizations such as 25 The Nature Conservancy, the Audubon Society, and local land trusts also protect and restore 26 CSS and MCF habitat. 27 CSS and MCF are found in numerous recreation areas in California. Three national parks 28 and monuments in California contain CSS, including Cabrillo National Monument, Channel 29 Islands National Park, and Santa Monica National Recreation Area. All three parks showcase 30 CSS habitat with educational programs and information provided to visitors, guided hikes, and 31 research projects focused on understanding and preserving CSS. Over a million visitors traveled March 2010 87 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 through these three parks in 2008. MCF is highlighted in Sequoia and Kings Canyon National 2 Park, Yosemite National Park, and Lassen Volcanic National Park, where more than 5 million 3 people visited in 2008. 4 The 2006 FHWAR for California (DOT, 2007) reports on the number of individuals 5 involved in fishing, hunting, and wildlife viewing in California. Millions of people are involved 6 in just these three activities each year. The quality of these trips depends in part on the health of 7 the ecosystems and their ability to support the diversity of plants and animals found in important 8 habitats found in CSS or MCF ecosystems and the parks associated with those ecosystems. 9 Based on analyses in Section 5.3.1.3 of the REA (U.S.EPA, 2009), average values of the total 10 benefits in 2006 from fishing, hunting, and wildlife viewing away from home in California were 11 approximately $947 million, $169 million, and $3.59 billion, respectively. In addition, data from 12 California State Parks (2003) indicate that in 2002, 68.7% of adult residents participated in trail 13 hiking for an average of 24.1 days per year. The analyses in the REA (U.S.EPA, 2009) indicate 14 that the aggregate annual benefit for California residents from trail hiking in 2007 was $11.59 15 billion. 16 CSS and MCF are home to a number of important and rare species and habitat types. CSS 17 displays richness in biodiversity with more than 550 herbaceous annual and perennial species. Of 18 these herbs, nearly half are endangered, sensitive, or of special status (Burger et al., 2003). 19 Additionally, avian, arthropod, herpetofauna, and mammalian species live in CSS habitat or use 20 the habitat for breeding or foraging. Communities of CSS are home to three important federally 21 endangered species. MCF is home to one federally endangered species and a number of state- 22 level sensitive species. The Audubon Society lists 28 important bird areas in CSS habitat and at 23 least 5 in MCF in California (http://ca.audubon.org/iba/index.shtml).14 24 The terrestrial enrichment case study in Section 5.3.1.3 of the REA and Section 3.3.5 of 25 the ISA identified fire regulation as a service that could be affected by nutrient enrichment of the 26 CSS and MCF ecosystems by encouraging growth of more flammable grasses, increasing fuel 27 loads, and altering the fire cycle. Over the 5-year period from 2004 to 2008, Southern California 28 experienced, on average, over 4,000 fires per year burning, on average, over 400,000 acres per 29 year (National Association of State Foresters [NASF], 2009). It is not possible at this time to 30 quantify the contribution of nitrogen depositio, among many other factors, to increased fire risk. 14 Important Bird Areas are sites that provide essential habitat for one or more species of bird. March 2010 88 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 The CSS and MCF were selected as case studies for terrestrial enrichment because of the 2 potential that these areas could be adversely affected by excessive N deposition. To date, the 3 detailed studies needed to identify the magnitude of the adverse impacts due to N deposition 4 have not been completed. Based on available data, this report provides a qualitative discussion of 5 the services offered by CSS and MCF and a sense of the scale of benefits associated with these 6 services. California is famous for its recreational opportunities and beautiful landscapes. CSS 7 and MCF are an integral part of the California landscape, and together the ranges of these 8 habitats include the densely populated and valuable coastline and the mountain areas. Through 9 recreation and scenic value, these habitats affect the lives of millions of California residents and 10 tourists. Numerous threatened and endangered species at both the state and federal levels reside 11 in CSS and MCF. Both habitats may play an important role in wildfire frequency and intensity, 12 an extremely important problem for California. The potentially high value of the ecosystem 13 services provided by CSS and MCF justify careful attention to the long-term viability of these 14 habitats. 15 The terrestrial nutrient enrichment case study relies on benchmark deposition levels for 16 various species and ecosystems as indicators of ecosystem response. While it would be expected 17 that deposition above those levels would have deleterious effects on the provision of ecosystem 18 services in those areas, at this time it is possible only to describe the magnitude of the some of 19 the services currently being provided. Methods are not yet available to allow estimation of 20 changes in services due to nitrogen deposition. 21 3.5 REFERENCES 22 Adams, D., R. Alig, B.A. McCarl, and B.C. Murray. February 2005. FASOMGHG 23 Conceptual Structure, and Specification: Documentation. Available at 24 http://agecon2.tamu.edu/people/faculty/mccarl-bruce/FASOM.html. Accessed on 25 October 22, 2008. 26 Anderson, E. 1989. "Economic Benefits of Habitat Restoration: Seagrass and the Virginia Hard- 27 Shell Blue Crab Fishery." North American Journal of Fisheries Management 9:140-149. 28 Acharya, G., and L.L. Bennett. 2001. "Valuing Open Space and Land-Use Patterns in Urban 29 Watersheds." Journal of Real Estate Finance and Economics 22(2/3):221-237. Arrow K. March 2010 89 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 and A.C. Fisher. 1974. "Environmental Preservation, Uncertainty, and Irreversibility." 2 The Quarterly Journal of Economics., 88(2):312-319. 3 Arrow K. and A.C. Fisher. 1974. "Environmental Preservation, Uncertainty, and Irreversibility." 4 The Quarterly Journal of Economics, 88(2):312-319. 5 Banzhaf, S., D. Burtraw, D. Evans, and A. Krupnick. 2006. "Valuation of Natural Resource 6 Improvements in the Adirondacks." Land Economics 82:445-464. 7 Bartik, TJ. 1988. "Measuring the Benefits of Amenity Improvements in Hedonic Price 8 Models "LandEconomics 64(2): 172-183. 9 Bockstael, N.E., K.E. McConnell, and I.E. Strand. 1988. "Benefits from Improvements in 10 Chesapeake Bay Water Quality, Volume III." Environmental Protection Agency 11 Cooperative Agreement CR-811043-01-0. 12 Bockstael, N.E., K.E. McConnell, and I.E. Strand. 1989. "Measuring the Benefits of 13 Improvements in Water Quality: The Chesapeake Bay." Marine Resource Economics 14 6:1-18. 15 Boyd, J., and S. Banzhaf. 2007. "What are Ecosystem Services? The Need for Standardized 16 Environmental Accounting Units." Ecological Economics 63:616-626. 17 Bricker, S., D. Lipton, A. Mason, M. Dionne, D. Keeley, C. Krahforst, J. Latimer, and 18 J.Pennock. 2006. Improving Methods and Indicators for Evaluating Coastal Water 19 Eutrophication: A Pilot Study in the Gulf of Maine. NOAA Technical Memorandum 20 NOS NCCOS 20. Silver Spring, MD: National Ocean Service, National Centers for 21 Coastal Ocean Science, Center for Coastal Monitoring and Assessment. 22 Bricker, S., Longstaff, B., Dennison, W., Jones, A., Boicourt, K., Wicks, C., and Woerner, J. 23 2007. Effects of Nutrient Enrichment in the Nation's Estuaries: A Decade of Change. 24 NOAA Coastal Ocean Program Decision Analysis Series No. 26. National Centers for 25 Coastal Ocean Science, Silver Spring, MD. March 2010 90 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Bricker, S.B., J.G. Ferreira, and T. Simas. 2003. "An Integrated Methodology for Assessment of 2 Estuarine Trophic Status." Ecological Modelling 169:39-60. 3 Brown, L.H. 2002. Profile of the Annual Fall Foliage Tourist in Vermont: Travel Year 4 2007.Report prepared for the Vermont Department of Tourism and Marketing. 5 Burlington, VT: Vermont Tourism Data Center, University of Vermont. 6 Burger, J.C., R.A. Redak, E.B. Allen, J.T. Rotenberry, and M.F. Allen. 2003. "Restoring 7 Arthropod Communities in Coastal Sage Scrub." Conservation Biology 17(2):460-467. 8 California Department of Fish and Game. n.d. 9 http://www.dfg.ca.gov/habcon/nccp/images/region.gif. 10 Chen, L., and C.T. Driscoll. 2004. "Modeling the Response of Soil and Surface Waters in the 11 Adirondack and Catskill Regions of New York to Changes in Atmospheric Deposition 12 and Historical Land Disturbance." Atmospheric Environment 38:4099-4109. 13 Chesapeake Bay Program, n.d. "Chesapeake Bay Program: A Watershed Partnership." Available 14 at http://www.chesapeakebay.net/. 15 Chesapeake Bay Program. "Bay Grass Abundance (Baywide)." Available at 16 http://www.chesapeakebay.net/ status_baygrasses.aspx?menuitem=19652. Last updated 17 May 21, 2008. 18 Chesapeake Eco-Check. 2007. "Indicator Details: 2007. Water Quality Index." Available at 19 http://www.eco-check.org/reportcard/chesapeake/2007/indicators/water_quality_index/. 20 Environmental Systems Research Institute, Inc. (ESRI). 2002. "Census Block Groups: 21 DC, DE, MD, and VA." [CD-ROM]. ESRI Data & Maps 2002. 22 Cordell, H.K., CJ. Betz, G. Green, and M. Owens. 2005. Off-Highway Vehicle Recreation in the 23 United States, Regions and States: A National Report from the National Survey on 24 Recreation and the Environment (NSRE). Report prepared for the Forest Service's 25 National OHV Policy & Implementation Teams. USDA Forest Service. March 2010 91 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Cordell, K., B. Leeworthy, G.T. Green, C. Betz, and B. Stephens, n.d. The National Survey 2 onRecreation & the Environment. Research Work Unit 4953. Athens, GA: Pioneering 3 Research on Changing Forest Values in the South and Nation USDA Forest Service 4 Southern Research Station. Available at www.srs.fs.fed.us/trends. 5 Daily, G.C., S. Alexander, P.R. Ehrlich, L. Goulder, J. Lubchenco, P.A. Matson, H.A. Mooney, 6 S. Postel, S.H. Schneider, D. Tilman, and G.M. Woodwell. 1997. "Ecosystem 7 Services:Benefits Supplied to Human Societies by Natural Ecosystems." Issues in 8 Ecology 2:1-16 9 DeHayes, D.H., P.G. Schaberg, GJ. Hawley, and G.R. Strimbeck, 1999. "Acidic Rain Impacts 10 on Calcium Nutrition and Forest Health." BioScience 49:789-800. 11 Driscoll, C.T. et al. 2003. "Effects of Acidic Deposition on Forest and Aquatic Ecosystems in 12 New York State." Environmental Pollution 123:327-336. 13 Driscoll, C.T., G.B. Lawrence, AJ. Bulger, TJ. Butler, C.S. Cronan, C. Eagar, K.F. Lamber, 14 G.E. Likens, J.L. Stoddard, and K.C. Weathers. 2001. Acidic Deposition in the 15 Northeastern United States: Sources and Inputs, Ecosystem Effects and Management 16 Strategies. BioScience 51:180-198. 17 Economist.com Dictionary at: 18 http://www.economist.com/research/economics/alphabetic.cfm?letter=E 19 Federal Register (2007) National Ambient Air Quality Standards for Ozone; Proposed Rule. 40 20 CFR 50; Federal Register 72: 37818 21 Federal Register (2008) National Ambient Air Quality Standards for Ozone; Final Rule. 40 CFR 22 parts 50 and 58; Federal Register 73:16436 23 Fenn, M.E., T.G. Huntington, S.B. McLaughlin, C. Eager, A. Gomez, and R.B. Cook. 24 2006."Status of Soil Acidification in North America." Journal of Forest Science 52 25 (special issue):3-13. 26 Greene, W.H. 2003. Econometric Analysis, 5th Ed. New Jersey: Prentice Hall. March 2010 92 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Haefele, M., R.A. Kramer, and T.P. Holmes. 1991. Estimating the Total Value of a Forest 2 Quality in High-Elevation Spruce-Fir Forests. The Economic Value of Wilderness: 3 Proceedings of the Conference. Gen. Tech. Rep. SE-78 (pp. 91-96). Southeastern For 4 Exper. Station. Asheville, NC: USDA Forest Service. 5 Holmes, T.P. 1992. Economic Welfare Impacts of Air Pollution Damage to Forests in the 6 Southern United States. Asheville, NC: U.S. Dept. of Agriculture, Forest Service, 7 Southeastern Forest Experiment Station, pp. 19-26. 8 Holmes, T., and R. Kramer. 1995. "An Independent Sample Test of Yea-Saying and Starting 9 Point Bias in Dichotomous-Choice Contingent Valuation. " Journal of Environmental 10 Economics and Management 28:121-132.Final 11 Hutchison, R., and C.E. Kraft. 1994. "Hmong Fishing Activity and Fish Consumption." Journal 12 of Great Lakes Research 20(2):471-487. 13 Industrial Economics, Inc. June 2008. "The Economic Impact of the Clean Air Interstate Rule on 14 Recreational Fishing in the Adirondack Region of New York State." Prepared for the 15 Clean Air Markets Division, Office of Air and Radiation, U.S. EPA. 16 Irwin, E.G. 2002. "The Effects of Open Space on Residential Property Values." Land Economics 17 78(4):465-480. 18 J. Nilsson and Grennfelt, P. (Eds) (1988) Critical loads for sulphur and nitrogen. UNECE/Nordic 19 Council workshop report, Skokloster, Sweden. March 1988. Nordic Council of Ministers: 20 Copenhagen. 21 Jenkins, D.H., J. Sullivan, and G.S. Amacher. 2002. "Valuing High Altitude Spruce-Fir 22 Forestlmprovements: Importance of Forest Condition and Recreation Activity." Journal 23 of'Forest Economics 8:77'-99. 24 Johnston, R.J., E.Y. Besedin, R. lovanna, C.J. Miller, R.F. Wardwell, and M.H. Ranson. 25 2005."Systematic Variation in Willingness to Pay for Aquatic Resource Improvements 26 and Implications for Benefit Transfer: A Meta-Analysis." Canadian Journal of 27 Agricultural Economics 53:221-248. March 2010 93 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Johnston, R.J., M.H. Ranson, E. Y. Besedin, and B.C. Helm. 2006. "What Determines 2 Willingness to Pay per Fish? A Meta-Analysis of Recreational Fishing Values." Marine 3 Resource Economics 21:1-32. 4 Kahn, J.R., and W.M. Kemp. 1985. "Economic Losses Associated with the Degradation of an 5 Ecosystem: The Case of Submerged Aquatic Vegetation in Chesapeake Bay." Journal of 6 Environmental Economics and Management 12:246-263. 7 Kaoru, Y. 1995. "Measuring Marine Recreation Benefits of Water Quality Improvements by the 8 Nested Random Utility Model." Resource and Energy Economics 17(2):119-136. 9 Kaoru, Y., V. Smith, and J.L. Liu. 1995. "Using Random Utility Models to Estimate the 10 Recreational Value of Estuarine Resources." American Journal of Agricultural 11 Economics 77:141-151. 12 Kaval, P., and J. Loomis. 2003. Updated Outdoor Recreation Use Values With Emphasis On 13 National Park Recreation. Final Report October 2003, under Cooperative Agreement CA 14 1200-99-009, Project number EVIDE-02-0070. 15 Kennedy, P. 2003. A Guide to Econometrics, pp. 312-313. Cambridge, MA: MIT Press. 16 Millennium Ecosystem Assessment (MEA). 2005. Ecosystems and Human Well-being: 17 Wetlands and Water. Synthesis. A Report of the Millennium Ecosystem Assessment. 18 Washington, DC: World Resources Institute. 19 Knowler, D. 2002. "A Review of Selected Bioeconomic Models: With Environmental Influences 20 in Fisheries." Journal of Bioeconomics 4:163-181. 21 Kramer, A., T. Holmes, and M. Haefele. 2003. "Contingent Valuation of Forest Ecosystem 22 Protection." In Forests in a Market Economy., E.O. Sills and K.L. Abt, eds., pp. 303-320. 23 Dordrecht, The Netherlands: Kluwer Academic Publishers. 24 Krieger, DJ. 2001. Economic Value of Forest Ecosystem Services: A Review. Washington, DC: 25 The Wilderness Society. March 2010 94 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Krupnick, A. 1988. "Reducing Bay Nutrients: An Economic Perspective." Mary land Law Review 2 47(2):453-480. 3 Land Trust Alliance. 2006. The 2005 National Land Trust Census Report. Washington, 4 D.C.:Land Trust Alliance, November 30, 2006. 5 Leeworthy, V.R., and P.C. Wiley. 2001. Current Participation Patterns inMarine 6 Recreation.Silver Spring, MD: U.S. Department of Commerce, National Oceanic and 7 Atmospheric Administration, National Ocean Service, Special Projects. 8 Lipton, D. W. 1999. "Pfiesteria's Economic Impact on Seafood Industry Sales and Recreational 9 Fishing." In Proceedings of the Conference, Economics of Policy Options for Nutrient 10 Management andPfiesteria. B. L. Gardner and L. Koch, eds., pp. 35-38. College Park, 11 MD: Center for Agricultural and Natural Resource Policy, University of Maryland. 12 Lipton, D. 2004. "The Value of Improved Water Quality to Chesapeake Bay Boaters." Marine 13 Resource Economics 19:265-270. 14 Lipton, D. 2006. "Economic Impact of Maryland Boating in 2005." University of Maryland Sea 15 Grant Extension Program. Available at 16 ftp://ftp.mdsg.umd.edu/Public/MDSG/rec_boat05.pdf. 17 Lipton, D. 2008. "Economic Impact of Maryland Boating in 2007." University of Maryland Sea 18 Grant Extension Program. Available at 19 ftp://ftp.mdsg.umd.edu/Public/MDSG/rec_boat07.pdf. 20 Lipton, D.W., and R.W. Hicks. 1999. "Linking Water Quality Improvements to Recreational 21 Fishing Values: The Case of Chesapeake Bay Striped Bass." In Evaluating the Benefits of 22 Recreational Fisheries. Fisheries Centre Research Reports 7(2), TJ. Pitcher, ed., pp. 105- 23 110. Vancouver: University of British Columbia. 24 Lipton, D.W., and R.W. Hicks. 2003. "The Cost of Stress: Low Dissolved Oxygen and the 25 Economic Benefits of Recreational Striped Bass Fishing in the Patuxent River." Estuaries 26 26(2A):310-315. March 2010 95 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Loomis, J., D. Griffin, E. Wu, and A. Gonzalez-Caban. 2002. "Estimating the Economic Value 2 of Big Game Habitat Production from Prescribed Fire Using a Time Series Approach. 3 Journal of Forest Economics 2:119-29. 4 Loomis, J.B., and D.S. White. 1996. "Economic Benefits of Rare and Endangered Species: 5 Summary and Meta-Analysis." Ecological Economics 18(3): 197-206. 6 Luzadis, V. A. and E.R. Gossett. 1996. "Sugar Maple." In Forest Trees of the Northeast., 1 J.P.Lassoie, V.A. Luzadis, and D.W. Grover, eds., pp. 157-166. Cooperative Extension 8 Bulletin 235. Cornell Media Services. Available at 9 http://maple.dnr.cornell.edu/pubs/trees.htm. 10 Mansfield, C.A., S.K. Pattanayak, W. McDow, R. MacDonald, and P. Halpin. 2005. "Shades of 11 Green: Measuring the Value of Urban Forests in the Housing Market." Journal of Forest 12 Economics \\(3y.m-\99. 13 Massey, M., S. Newbold, and B. Gentner. 2006. "Valuing Water Quality Changes Using a 14 Bioeconomic Model of a Coastal Recreational Fishery." Journal of 15 EnvironmentalEconomics and Management 52:482-500. 16 Millennium Ecosystem Assessment (MEA). 2005a. Ecosystems and Human Well-being: Current 17 State and Trends, Volume 1. R. Hassan, R. Scholes, and N. Ash, eds. Washington, 18 DCisland Press. Available at 19 http://www.millenniumassessment.org/documents/document.766.aspx.pdf. 20 Millennium Ecosystem Assessment (MEA). 2005b. Ecosystems and Human Well- 21 being:Synthesis. Washington, DC: World Resources Institute. 22 Mistiaen, J.A., I.E. Strand, and D. Lipton. 2003. "Effects of Environmental Stress on Blue Crab 23 (Callinectes sapidus) Harvests in Chesapeake Bay Tributaries." Estuaries 26(2a): 316- 24 322. 25 Montgomery, M. and M. Needelman. 1997. "The Welfare Effects of Toxic Contamination in 26 Freshwater Fish." Land Economics 73(2):211-223. March 2010 96 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Morgan, C., and N. Owens. 2001. "Benefits of Water Quality Policies: The Chesapeake Bay." 2 Ecological Economics 39:271-284. 3 Murray, T., and J. Lucy. 1981. "Recreational Boating in Virginia: A Preliminary Analysis." 4 Special Report in Applied Marine Science and Ocean Engineering No. 251. Available at 5 http://www.vims.edu/GreyLit/VIMS/sramsoe251 .pdf 6 National Agricultural Statistics Service (NASS). June 12, 2008. "Maple Syrup Production Up 30 7 Percent Nationwide." New England Agricultural Statistics, NASS, USDA. Spencer, D.M., 8 and D.F. Holecek. 2007. "Basic Characteristics of the Fall Tourism Market." Joz/ram 9 Management 28:491 -504 10 National Association of State Foresters (NASF). 2009. Quadrennial Fire Review 11 2009.Washington, DC: NASF. Quadrennial Fire and Fuel Review Final Report 2009. 12 National Marine Manufacturers Association (NMMA). 2008. "2007 Recreational Boating 13 Statistical Abstract." Available at 14 http://www.nmma.org/facts/boatingstats/2007/files/Abstract.pdf 15 National Oceanic and Atmospheric Administration (NOAA). (2007, August). "Annual 16 Commercial Landing Statistics." Available at 17 http://www.st.nmfs.noaa.gov/stl/commercial/landings/annual_landings.html. 18 National Oceanic and Atmospheric Administration (NOAA). "Recreational Fisheries." 19 Available at www.st.nmfs.noaa.gov/stl/recreational/overview/overview.html. Last 20 updated February 12, 2009. 21 National Research Council. 2005. Valuing Ecosystem Services: Toward Better Environmental 22 Decision-Making. Washington, DC: National Academies Press. 23 Parsons, G., A.O. Morgan, J.C. Whitehead, and T.C. Haab. 2006. "The Welfare Effects of 24 Pfiesteria-Related Fish Kills: A Contingent Behavior Analysis of Seafood Consumers." 2 5 Agricultural and Resource Economics Review 3 5 (2): 1 -9. March 2010 97 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Peterson, D.E., M.S. Kanarek, M.A. Kuykendall, J.M. Diedrich, H.A. Anderson, P.L. 2 Remington, and T.B. Sheffy. 1994. "Fish Consumption Patterns and Blood Mercury 3 Levels in Wisconsin Chippewa Indians." Archives of Environmental Health 49(1):53 58. 4 Poor, P.I, K.L. Pessagno, and R.W. Paul. 2007. "Exploring the Hedonic Value of Ambient 5 Water Quality: A Local Watershed-Based Study." Ecological Economics 60:797-806. 6 Smith, M.D. 2007. "Generating Value in Habitat-Dependent Fisheries: The Importance of 7 Fishery Management Institutions." Land Economics 83(l):59-73. 8 Sweeney, J. 2007. "Impacts of CAMD 2020 CAIR on Nitrogen Loads to the Chesapeake Bay." 9 University of Maryland, Chesapeake Bay Program Office. 10 U.S. Department of Agriculture, Forest Service. Forest Inventory and Analysis National 11 Program, RPA Assessment Tables. 2002. Available at 12 http://Ncrs2.Fs.Fed.Us/4801/Fiadb/Rpa_Tabler/Draft_RPA_2002_Forest_Resource_Tabl 13 es.pdf. 14 U.S. Census Bureau. 2008a. "Annual Social and Economic (ASEC) Supplement." Available at 15 http://pubdb3.census.gov/macro/032008/hhinc/new02_001.htm. 16 U.S. Census Bureau. 2008b. "Housing Unit Estimates for Counties of MD and VA: April 1/2000 17 to July 1/2007." Available at http://www.census.gov/popest/housing/HU-EST2007- 18 CO.html. 19 U.S. Department of the Interior, Fish and Wildlife Service, and U.S. Department of Commerce, 20 U.S. Census Bureau. 2007. 2006 National Survey of Fishing, Hunting, and Wildlife- 21 Associated Recreation. 22 U.S. EPA. 1998.. EPA/630/R-95/002F. Risk Assessment Forum, Washington, DC. Available 23 from: http://www.epa.gov/ncea/ecorsk.htm. March 2010 98 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 U.S. Environmental Protection Agency (EPA), Office of Air and Radiation. November 1999.The 2 Benefits and Costs of the Clean Air Act 1990 to 2010: EPA Report to Congress.EPA- 3 410-R-99-001. Washington, DC: U.S. Environmental Protection Agency. 4 U.S. Environmental Protection Agency (EPA). 2002. December. Environmental and Economic 5 Benefit Analysis of Final Revisions to the National Pollutant Discharge Elimination 6 System Regulation and the Effluent Guidelines for Concentrated Animal Feeding 7 Operations (EPA-821-R-03-003). Washington, DC: U.S. Environmental Protection 8 Agency, Office of Water, Office of Science and Technology. 9 U. S. EPA (Environmental Protection Agency). 2006. Ecological Benefits Assessment Strategic 10 Plan. EPA-240-R-06-001. Office of the Administrator. Washington, DC. Available at 11 http://www.epa.gov/economics. 12 U.S. Environmental Protection Agency (EPA). 2008. Integrated Science Assessment for Oxides 13 of Nitrogen and Sulfur-Environmental Criteria.EPA/600/R-08/082. U.S. Environmental 14 Protection Agency, Office of Research and Development, National Center for 15 Environmental Assessment - RTF Division, Research Triangle Park, NC 16 US Environmental Protection Agency (EPA). 2009. Risk and Exposure Assessment for the 17 Review of the Secondary national Ambient Air Quality Standards for Oxides of Nitrogen 18 and Oxides of Sulfur. EPA/452/R/09/008a. U.S environmental Protection Agency, Office 19 of Air Quality planning and Standards, Health and Environmental impacts Division, 20 Research Triangle Park, NC. 21 U.S. EPA (Environmental Protection Agency). 2009. Valuing the Protection of Ecological 22 Systems and Services: A report of the EPA Science Advisory Board. EPA-SAB-09-012. 23 Office of the Administrator. Washington, DC. Available at 24 http://yosemite.epa.gov/sab/sabproduct.nsf/WebBOARD/ValProtEcolSys&Serv? 25 Valigura, R.A., R.B. Alexander, M.S. Castro, T.P. Meyers, H.W. Paerl, P.E. Stacy, and 26 R.E.Turner. 2001. Nitrogen Loading in Coastal Water Bodies: An Atmospheric 27 Perspective.Washington, DC: American Geophysical Union. March 2010 99 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Van Houtven, G. and A. Sommer. December 2002. Recreational Fishing Benefits: A Case Study 2 of Reductions in Nutrient Loads to the Albemarle-Pamlico Sounds Estuary. Final Report. 3 Prepared for the U.S. Environmental Protection Agency. Research Triangle Park, 4 NC:RTI International. 5 Van Houtven, G.L., J. Powers, and S.K. Pattanayak. 2007. "Valuing Water Quality 6 Improvements Using Meta-Analysis: Is the Glass Half-Full or Half-Empty for National 7 Policy Analysis?" Resource and Energy Economics 29:206-228. 8 Vaughan, WJ. 1986. "The Water Quality Ladder." Included as Appendix B in R.C. Mitchell, 9 and R.T. Carson, eds. The Use of Contingent Valuation Data for Benefit/Cost Analysis in 10 Water Pollution Control. CR-810224-02. Prepared for the U.S. Environmental Protection 11 Agency, Office of Policy, Planning, and Evaluation. 12 Virginia Department of Conservation and Recreation. 2007. "2006 Virginia Outdoors 13 Survey."Available at 14 http://www.dcr. virginia.gov/recreational_planning/documents/vopsurvey06.pdf 15 Wallace, KJ. 2007. "Classification of Ecosystem Services: Problems and Solutions." Ecological 16 Conservation 139:235-246. 17 Whitehead, J.C., T.C. Haab, and G.R. Parsons. 2003. "Economic Effects of Pfiesteria." Ocean & 18 Coastal Management 46(9-10):845-858. 19 March 2010 100 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 4 ADDRESSING THE ADEQUACY OF THE CURRENT STANDARDS 2 Based on the information in Chapters 2 and 3, we conclude that there is support in the 3 available effects-based evidence for consideration of secondary standards for NOX and SOX that 4 are protective against adverse ecological effects associated with deposition of NOX and SOX to 5 sensitive ecosystems. Having reached this general conclusion, we then to the extent possible 6 evaluate the adequacy of the current NOX and SOX secondary standards by considering to what 7 degree risks to sensitivity ecosystems would be expected to occur in areas that meet the current 8 standards. Staff conclusions regarding the adequacy of the current standards are based on the 9 available ecological effects, exposure and risk-based evidence. In evaluating the strength of this 10 information, staff have taken into account the uncertainties and limitations in the scientific 11 evidence. This chapter addresses key policy relevant questions that inform our determination 12 regarding the adequacy of the structure and levels of the current secondary standards. The 13 chapter begins with a discussion of the structure of the current standards, followed by a 14 presentation of information on recent air quality relative to the existing standards, recent NOX 15 and SOX deposition levels, evaluation of recent deposition levels relative to levels where adverse 16 ecological effects have been observed, and a set of conclusions regarding the adequacy of the 17 current structure and levels of the standards. 18 It is also appropriate in this review to consider whether the current standards are adequate 19 to protect against the direct effects on vegetation resulting from ambient NC>2 and 862 which 20 were the basis for the current secondary standards. We will include a discussion of this issue in 21 the second draft policy assessment. 22 4.1 ARE THE STRUCTURES OF THE CURRENT NOX AND SOX 23 SECONDARY STANDARDS BASED ON RELEVANT 24 ECOLOGICAL INDICATORS SUCH THAT THEY ARE 25 ADEQUATE TO DETERMINE AND PROTECT PUBLIC WELFARE 26 AGAINST ADVERSE EFFECTS ON ECOSYSTEMS? 27 The current secondary NOX and SOX standards are intended to protect against adverse 28 effects to public welfare. For NOX, the current secondary standard was set identical to the March 2010 101 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 primary standard15, e.g. an annual standard set for NC>2 to protect against adverse effects on 2 vegetation from direct exposure to ambient NOX. For SOX, the current secondary standard is a 3- 3 hour standard intended to provide protection for plants from the direct foliar damage associated 4 with atmospheric concentrations of SC>2. 5 The ISA has established that the major effects of concern for this review of the NOX and 6 SOX standards are associated with deposition of N and S associated with atmospheric 7 concentrations of NOX and SOX (see Chapter 2). As such, the current secondary standards do not 8 reflect the conclusions of the ISA in the major areas of indicator, form, or averaging times. By 9 using atmospheric NC>2 and 862, concentrations as indicators the current standards address only 10 a fraction of total atmospheric NOX and SOX, and do not take into account the effects from 11 deposition of total atmospheric NOX and SOX. By addressing short-term concentrations of SO2, 12 the current SC>2 standard, while protective against direct foliar effects from gaseous SOX, does 13 not take into account the findings of effects in the ISA, which notes the relationship between 14 annual deposition of S and acidification effects which are likely to be more severe and 15 widespread than phytotoxic effects under current ambient conditions. Acidification is a process 16 which occurs over time, as the ability of an aquatic system to counteract acidic inputs is reduced 17 as natural buffers are used more rapidly than they can be replaced through geologic weathering. 18 The relevant period of exposure for ecosystems is therefore not the exposures captured in the 19 short averaging time of the current 862 standard. In addition, the ISA has concluded that NOX 20 and SOX and their deposition products jointly affect ecosystems, and as such the current separate 21 structure of the NOX and SOX secondary standards does not take into account the joint ecological 22 effects of the two pollutants. 23 Current standards are specified as allowable single atmospheric concentration levels for 24 NC>2 or SC>2. This type of structure does not take into account variability in the atmospheric and 25 ecological factors that may alter the effects of NOX and SOX on public welfare. Consistent with 26 section 108, the ISA includes in the air quality criteria consideration of how these variable 27 factors impact the effects of ambient NOX and SOX on public welfare. Secondary standards are 28 intended to address a wide variety of effects occurring in different types of environments and 29 ecosystems. Ecosystems are not uniformly distributed either spatially or temporally in their 15 The current primary NO2 standard has recently been changed to the 3 year average of the 98thpercentile of the annual distribution of the 1 hour daily maximum of the concentration of NO2. The current secondary standard remains as it was set in 1971. March 2010 102 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 sensitivity to air pollution. Therefore, failure to account for the major determinants of variability, 2 especially geologic conditions related to sensitivity to acidification and atmospheric conditions 3 which govern rates of deposition, may lead to standards that do not provide appropriate levels of 4 protection across ecosystems. We can state with confidence the current standards were not 5 designed to be protective against those welfare effects tied to deposition of ambient NOX and SOX 6 and thus are not likely to be adequate to protect public welfare against known or anticipated 7 adverse effects from deposition. 8 Because most areas of the U.S. are in attainment with the current NO2 and SOX standards, 9 it is possible to evaluate current conditions, and evaluate the impact on public welfare from the 10 current effects on ecosystems from NOX and SOX deposition in areas that attain the current 11 standards that use NO2 and SO2 as indicators. In addition, this chapter qualitatively addresses the 12 adequacy of the structures of the existing standards relative to ecologically relevant standards for 13 NOX and SOX, and sets up arguments for developing an ecologically relevant structure for the 14 standards as described in Chapter 5. 15 4.2 TO WHAT EXTENT ARE THE STRUCTURES OF THE CURRENT 16 NOX AND SOX SECONDARY STANDARDS MEANINGFULLY 17 RELATED TO RELEVANT ECOLOGICAL INDICATORS OF 18 PUBLIC WELFARE EFFECTS? 19 The current secondary standard for NOX, set in 1971, using NO2 as the atmospheric 20 indicator, is 0.053 parts per million (ppm) (100 micrograms per cubic meter of air [|jg/m3]), 21 annual arithmetic average, calculated as the arithmetic mean of the 1-hour NO2 concentrations. 22 This standard was selected to provide protection to the public welfare against acute injury to 23 vegetation from direct exposure and resulting phytoxicity. During the last review of the NOX 24 standards, impacts associated with chronic acidification and eutrophication from NOX deposition 25 were acknowledged, but the relationships between atmospheric concentrations of NOX and levels 26 of acidification and eutrophication and associated welfare impacts were determined to be too 27 uncertain to be useful as a basis for setting a national secondary standard (USEPA 1995). 28 The current secondary standard for SOX, set in 1971, uses SO2 as the atmospheric 29 indicator, is a 3-hour average of 0.5 ppm, not to be exceeded more than once per year. This 30 standard was selected to provide protection to the public welfare against acute injury to March 2010 103 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 vegetation. In the last review of the SOX secondary standard, impacts associated with chronic 2 acidification were acknowledged, but the relationships between atmospheric concentrations of 3 SOX and levels of acidification, along with the complex interactions between SOX and NOX in 4 acidification processes, were cited as critical uncertainties which made the setting of secondary 5 NAAQS to protect against acidification inappropriate at that time (USEPA 1982). 6 In the previous independent reviews of the NOX and SOX secondary standards, each 7 review acknowledged the additional impacts of NOX and SOX on public welfare through the 8 longer term impact of the pollutants once deposited to ecosystems. However, the previous 9 reviews cited numerous uncertainties as the basis for not addressing those impacts in the setting 10 of the standards. In addition, these previous reviews did not consider the common pathways of 11 impact for the two pollutants acting on the same ecosystem endpoints. 12 Three issues arise that call into question the ecological relevance of the current structure 13 of the secondary standards for NOX and SOX. One issue is the exposure period that is relevant for 14 ecosystem impacts. The majority of deposition related impacts are associated with depositional 15 loads that occur over periods of months to years. This differs significantly from exposures 16 associated with hourly concentrations of NC>2 and 862 as measured by the current standards. 17 Even though the NC>2 standard uses an annual average of NC>2, it is focused on the annual 18 average of 1-hour NC>2 concentrations, rather than a cumulative metric or an averaging metric 19 based on daily or monthly averages. A second issue is the choice of atmospheric indicators. NC>2 20 and SO2 are used as the component of oxides of nitrogen and sulfur that are measured, but they 21 do not provide a complete link to the direct effects on ecosystems from deposition of NOX and 22 SOX as they do not capture all relevant species of oxidized nitrogen and oxidized sulfur that 23 contribute to deposition. The ISA provides evidence that deposition related effects are linked 24 with total nitrogen and total sulfur, and thus all forms of oxidized nitrogen and oxidized sulfur 25 that are deposited will contribute to effects on ecosystems. This suggests that more 26 comprehensive atmospheric indicators should be considered in designing ecologically relevant 27 standards. Further discussions of the need for more ecologically relevant atmospheric indicators 28 as well as the relative contributions to deposition from various species of NOX and SOX can be in 29 found in Chapters 5 and 6. The third issue is that the current standards reflect separate 30 assessments of the two individual pollutants, NC>2 and SC>2, rather than assessing the joint 31 impacts of deposition of NOX and SOX to ecosystems, recognizing the role that each pollutant March 2010 104 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 plays in jointly affecting ecosystem indicators, functions, and services. The clearest example of 2 this interaction is in assessment of the impacts of acidifying deposition on aquatic ecosystems. 3 Acidification in an aquatic ecosystem depends on the total acidifying potential of the 4 deposition of both N and S from both atmospheric deposition of NOX and SOX as well as the 5 inputs from other sources of N and S such as reduced nitrogen and non-atmospheric sources. It is 6 the joint impact of the two pollutants that determines the ultimate effect on organisms within the 7 ecosystem, and critical ecosystem functions such as habitat provision and biodiversity. Standards 8 that are set independently are less able to account for the contribution of the other pollutant. This 9 suggests that interactions between NOX and SOX should be a critical element of the conceptual 10 framework for ecologically relevant standards. There are also important interactions between 11 NOX and SOX and reduced forms of nitrogen, which also contributes to acidification and nutrient 12 enrichment. While the standards do not address reduced forms of nitrogen in the atmosphere, it is 13 important that the structure of the standards address the role of reduced nitrogen in determining 14 the ecological effects resulting from deposition of atmospheric NOX and SOX. Consideration will 15 also have to be given to account for loadings coming from non-atmospheric sources as 16 ecosystems will respond to these sources as well. 17 In addition to the fundamental issues discussed above, the current structures of the 18 standards do not address the complexities in the responses of ecosystems to deposition of NOX 19 and SOX. Ecosystems contain complex grouping of organisms that respond in various ways to the 20 alterations of soil and water that result from deposition of nitrogen and sulfur compounds. 21 Different ecosystems therefore respond in different ways depending on a multitude of factors 22 that control how deposition is integrated into the system. For example, the same levels of 23 deposition falling on limestone dominated soils have a very different effect than those falling on 24 shallow glaciated soils underline with granite. One system may over time display no obvious 25 detriment while the other may experience a catastrophic loss in fish communities. This degree of 26 sensitivity is a function of many atmospheric factors which control rates of deposition as well as 27 ecological factors which control how an ecosystem responds to that deposition. The current 28 standards do not take into account spatial and seasonal variations not only in depositional 29 loadings but also in sensitivity of ecosystems exposed to those loadings. March 2010 105 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 4.3 TO WHAT EXTENT DO CURRENT MONITORING NETWORKS 2 PROVIDE A SUFFICIENT BASIS FOR DETERMINING THE 3 ADEQUACY OF CURRENT SECONDARY NOX AND SOX 4 STANDARDS? 5 There are over 1000 ground level monitoring platforms (Figures 4-1 and 4-2) that provide 6 measurements of some form of atmospheric nitrogen or sulfur. The key pollutants for this 7 assessment are total oxidized nitrogen (NOy), total reduced nitrogen (NHX), and total sulfur (ST). 8 Total reactive oxidized atmospheric nitrogen, NOy, is defined as NOX (NO and NO2) and all 9 oxidized NOX products: NOy = NO2 + NO + HNO3 + PAN +2N2O5 + HONO+ NO3 + organic 10 nitrates + particulate NO3 (Finlayson-Pitts and Pitts, 2000). This definition of NOy reflects the 11 operational principles of standard measurement techniques in which all oxidized nitrogen species 12 are converted to nitrogen oxide (NO) through catalytic reduction and the resulting NO is detected 13 through luminescence. Thus, NOy is truly defined as total oxidized nitrogen as converted to NO. 14 NOy is not a strict representation of the all moles of oxidized nitrogen as the diatomic nitrogen 15 species such as N2Os yield 2 moles of NO. This definition is consistent with the relationship 16 between atmospheric nitrogen and acidification processes as the reported NOy provides a direct 17 estimate of the potential equivalents available for acidification. Total reduced nitrogen (NHX) 18 includes ammonia, NH3, plus ammonium, NH4 (EPA, 2008). Reduced nitrogen plus oxidized 19 nitrogen is referred to as total reactive nitrogen. Total sulfur (ST) includes SO2 gas and 20 particulate sulfate, SO4. Ammonium and sulfate are components of atmospheric particulate 21 matter as well as directly measured and modeled in precipitation as direct deposition 22 components. As discussed in this section, there are only very limited routine measurements of 23 total oxidized and reduced nitrogen. In addition, existing monitoring networks do not provide 24 adequate geographic coverage to fully assess concentrations and deposition of reactive nitrogen 25 and sulfur in and near sensitive ecosystems. 26 March 2010 106 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 MAPI (All N) NCore, NOY(2009), SEARCH, PA MS/SLA MS, CASTNET, IMPROVE •*• NCore Q| Rural NCore » SEARCH Q Rural SEARCH • PAMS_HO- NO2- NQX-NQY_2009 * SLAMS_NO-N02-N OX-HOY » CASTNET-NPS CASTNET-EPA • IMPR QVE_Hrtr3te£_20Q6 Figure 4-1. Routinely operating surface monitoring stations measuring forms of atmospheric nitrogen. March 2010 107 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX MAP 6-1 a (AIIS) NCore, SO2(2008), SEARCH, CASTNET, IMPROVE, and Trends/Supplemental Speciation Sites (2008) - SO2(2008) includes NAMS / SLAMS / PAMS - • SEARCH D Rural SEARCH , CASTNET-NPS CASTNET-EPA A Special! on_Sulfates_2003 • IMPROVE Sulfales 2006 1 2 Figure 4-2. Routinely operating surface monitoring stations measuring forms of 3 atmospheric sulfur. 4 The principal monitoring networks include the regulatory based State and Local Air 5 Monitoring Stations (SLAMS) providing mostly urban-based 862, NO and NOX, the PM2.5 6 chemical speciation networks Interagency Monitoring of Protected visual Environments 7 (IMPROVE) and EPA's Chemical Speciation Network (CSN) providing particle bound sulfate 8 and nitrate, and the Clean Air Status and Trends Network (CASTNET) providing weekly 9 averaged values of SO2, nitric acid, and particle bound sulfate, nitrate and ammonium. The 10 private sector supported SouthEastern Aerosol Research and Characterization (SEARCH) Study 11 network of 4-8 sites in the southeast provides the only routinely operating source of true 12 continuous NO2, ammonia, and nitric acid measurements. SEARCH also provides PM2.5 size 13 fractions of nitrate and sulfate. Collectively, the SLAMS, Photochemical Assessment 14 Measurement Stations (PAMS), SEARCH and NCore networks will provide over 100 sites 15 measuring NOy (Figure 4-3). The NCore network (Scheffe et al., 2009) is a multiple pollutant March 2010 108 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 network with co-located measurements of key trace gases (CO, SO2, 63, NO and NOy), PM2.5 2 and PM(10-2.5) mass and PM2.5 chemical speciation. Additional air pollutants, particularly 3 volatile organic compounds (VOCs), will be measured at those sites that are part of the existing 4 PAMS and National Air Toxics Trends (NATTS) platforms. The NATTS (EPA, 2008) include 5 27 stations across the U.S. that monitor for a variety of hazardous air pollutants and are intended 6 to remain in place to provide a longe term record. Additional measurements of ammonia and 7 possibly true NO2 are under consideration. True NO2 is noted to differentiate from the NO2 8 determined through routine regulatory networks that have known variable positive bias for NO2. 9 The network currently is being deployed and expected to be operational with nearly 75 sites by 10 January, 2011. The sites are intended to serve as central site monitors capturing broadly 11 representative (e.g., not strongly influenced by nearby sources) air quality in a suite of major and 12 mid size cities, and approximately 20 sites are located in rural locations. March 2010 109 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX MAP 4 Currents. Planned Routine NOY Monitoring Sites NCore, NOY(2009), SEARCH 1 2 Figure 4-3. Anticipated network of surface based NOy stations based on 2009 3 network design plans. The NCore stations are scheduled to be operating by 4 January, 2011. 5 There are significant measurement gaps for characterizing NOy, NHX and SC>2 in the 6 nations ambient air observation networks (EPA, 2008) that lead to greater reliance on air quality 7 modeling simulations to describe current conditions. National design of routinely operating 8 ambient air monitoring networks is driven mostly by data uses associated with implementing 9 primary NAAQS, with noted exceptions of the CASTNET and IMPROVE networks In addition 10 to significant spatial gaps in sensitive ecosystem areas that arise from a population oriented 11 network design, the current measurements for primary and secondary nitrogen are markedly 12 different and in some instances of negligible value for secondary NOX and SOX standards. For 13 example, a true NOX (NO plus NO2) measurement typically would capture less than 50% (see 14 discussion below) of the total regional NOy mass in rural locations as the more aged air masses 15 contain significant oxidized nitrogen products in addition to NOX. Note that the NOX monitors March 2010 110 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 used for NAAQS primary compliance purposes do capture varying amounts of transformed 2 nitrogen species; however, the method provides biased low estimates with significant airshed 3 induced variability relative to true NOy. With the exception of the SEARCH network in the 4 southeast, there are virtually no routine networks that measure ammonia, although EPA is 5 considering options for ammonia sampling in CASTNET and NCORE networks. Ammonium is 6 reported in EPA chemical speciation networks, although the values are believed to be biased low 7 due to ammonia volatization. 8 CASTNET provides mostly rural measurements of 862, total nitrate, and ammonium, and 9 affords an existing infrastructure useful for future monitoring in support of a NOX and SOX 10 secondary standard. However, the lack of NOy, SOX and NHX measurements in sensitive 11 ecosystems will require attention in the N/S secondary standard proposal. 12 As a result of the limited monitoring networks for NOy and SOX in sensitive ecosystems, 13 we are unable to use current monitoring data to fully assess whether the current standards have 14 resulted in levels of NOy and SOX in sensitive ecosystems that would result in deposition levels 15 that are or are not causing ecological effects adverse to public welfare. We supplement the 16 available monitoring data with the use of sophisticated atmospheric modeling conducted using 17 EPA'sCMAQmodel. 18 4.3.1 What does the NADP monitoring network provide and what are the major 19 limitations? 20 The National Atmospheric Deposition Program (NADP) includes approximately 250 21 sites (Figure 4-4) across the U.S. providing annual total wet deposition based on weekly 22 averaged measures of wet deposition of nitrate, ammonium and sulfate ions based on the 23 concentrations of these ions in precipitation samples. Meteorological models have difficulty in 24 capturing the correct spatial and temporal features of precipitation events, raising the importance 25 of the NADP as a principal source of precipitation chemistry. The NADP has enabled several 26 organizations to participate in a measurement program with a centralized laboratory affording 27 measurement and analysis protocol consistency nationwide. Virtually every CASTNET site is 28 located at an NADP site and the combined NADP/CASTNET infrastructure is a starting point for 29 discussions addressing future N/S monitoring needs. The Organic bound nitrogen is not analyzed 30 routinely in NADP samples. Consideration might be given to adding NADP sites in locations March 2010 111 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 where ambient air monitoring is conducted to assess compliance with a secondary NOX/SOX 2 standard. 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Ammonium ion wet deposition, 2005 Ammonium as NH,* (kg/ha) Sites not pictured: AK03 0.3 kg/ha VI01 0.4 kgJha National Atmospheric Deposition Program/National Trends Network http://nadp.sws.uiuc.edu 0.5.1.0 1.0-1.5 2.5 - 3.0 3.0 - 3.5 3.5 - 4.0 4.0-4.6 >4,5 Figure 4-4. Location of approximately 250 National Atmospheric Deposition Monitoring (NADP) National Trends Network (NTN) sites illustrating annual ammonium deposition for 2005. Weekly values of precipitation based nitrate, sulfate and ammonium are provided by NADP. 4.3.2 How do we characterize deposition through Monitoring and Models? Routinely available directly measured precipitation to quantify wet deposition of sulfur and nitrogen species are provided through the NADP. Dry deposition is not a directly measured variable in routine monitoring efforts and, for all practical purposes, largely will remain a research endeavor that supports the parameterizations used for estimating dry deposition, as opposed to striving to develop operational methods. Estimates of dry deposition based on observations are provided through the CASTNET program. However, dry deposition is a calculated value represented as the product of ambient concentration (either observed or estimated through air quality modeling) and deposition velocity, Dep®"7 = v®"7 • C^mb March 2010 112 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Deposition velocity is modeled as a mass transfer process through resistance layers 2 associated with the canopy, uptake by vegetation, water and soil which collectively are 3 influenced by micrometeorology, land surface and vegetation types and species specific 4 solubility and reactivity. Dry deposition is calculated through deposition velocity models 5 capturing these features and using species specific ambient air concentrations. This approach 6 conceptually is similar using either observed or modeled air concentrations. Dry deposition 7 estimates from the Community Multi-scale Air Quality (CMAQ) model (EPA, 1999) have been 8 used in this assessment to provide spatially more resolved and extensive estimates of dry 9 deposition for sulfur and all reactive nitrogen (oxidized and reduced) species (CASTNET does 10 not capture important gases such as nitrogen dioxide and peroxyacetyl nitrate). All of the 11 relevant meteorological, land use, vegetation and elevation data required to estimate deposition 12 velocities are generated or accessible in the CMAQ and/or meteorological pre-processors. 13 4.3.2.1 Why are we using CMAQ to model deposition? How are we using it? Why is 14 CMAQ the right model to use? What is the spatial and temporal resolution of 15 CMAQ? What are the model years ? What are the limitations to CMAQ? 16 CMAQ provides a platform that allows for a consistent mass accounting approach across 17 ambient concentrations and dry and wet deposition values. Recognizing the limitations of 18 ambient air networks, CMAQ was used to estimate dry deposition to complement NADP wet 19 deposition for MAGIC modeling and for the FAB critical load modeling. CMAQ promotes 20 analytical consistency and efficiency across analyses of multiple pollutants. EPA's Office of 21 Research and Development continues to enhance the underlying deposition science in CMAQ. 22 For the purposes of this policy assessment, CMAQ provides a consistent platform incorporating 23 the atmospheric and deposition species of interest over the entire United States. The caveats and 24 limitations of the use of model predictions are largely associated with the general reliance on 25 calculated values, rather than measurements. Model evaluation addressing the comparison of 26 predictions with observed values is addressed in the REA. Currently, there are efforts to improve 27 a number of nitrogen related processes in CMAQ, recognizing comparatively less uncertainty 28 with the treatment of sulfur. Active areas of model process improvement are in the treatment of 29 lightning generated NOX and the transference of nitrogen between atmospheric and terrestrial and 30 aquatic media, often referred to as bi-directional flux. Lightning NOX potentially provides a March 2010 113 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 significant contribution to wet deposition as the resulting NOX is rapidly entrained into aqueous 2 cloud processes. Both the thermodynamics of soil processes and mass transfer of nitrogen 3 species across the surface-atmosphere interface is governed by an assortment of temperature, 4 moisture, advection and concentration patterns. These processes and mass transfer relationships 5 are coupled within the emissions, meteorological, and chemical simulation processes and 6 associated surface/vegetation and terrain information incorporated in or accessed by the CMAQ. 7 In addition to research activities to improve the characterization of nitrogen-related processes in 8 CMAQ, efforts are also underway to improve the general characterization of ammonia emissions 9 which remains as an area of large uncertainty due to limited source data and the ubiquitous 10 nature of these emissions. Another challenge for regional/national air quality modeling is 11 properly representing the effects on pollutant concentrations, precipitation and therefore 12 deposition of variable terrain features, particularly steep mountain-valley gradients and the 13 interfaces to wide open basins encountered in the Western United States. 14 The CMAQ was used in this assessment because it is the state of science model for 15 treating simulating sources, formation, and fate of nitrogen and sulfur species. In addition to 16 undergoing periodic independent scientific peer review, CMAQ bridges the scientific and 17 regulatory communities as it is used extensively by EPA for regulatory air quality assessments 18 and rules. CMAQ provides hourly estimates of the important precursor, intermediate and 19 secondarily formed species associated with atmospheric chemistry and deposition processes 20 influencing ozone, particulate matter concentrations and sulfur and nitrogen deposition. 21 Simulations based on horizontal spatial scale resolutions of 12 km and 36 km were used in this 22 PAD for 2002-2005. 23 4.4 WHAT IS OUR BEST CHARACTERIZATION OF ATMOSPHERIC 24 CONCENTRATIONS OF NOY AND SOX, AND DEPOSITION OF N 25 AND S? 26 Air quality models and blending of model results and observations are used to 27 characterize current environmental state conditions due to the relative sparseness of monitoring 28 coverage in sensitive ecosystems as well as gaps in coverage for specific atmospheric species of 29 N and S most relevant to deposition, such as NOy, in available monitoring platforms. March 2010 114 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 4.4.1 What are the current atmospheric concentrations of reactive nitrogen, NOy, 2 reduced nitrogen, NHX, sulfur dioxide, SOi, and sulfate, SC>4? 3 To provide information for use in characterizing the adequacy of the current standards, 4 we assess the best available data for estimating the ambient concentrations of the major sources 5 of atmospheric nitrogen and sulfur across the U.S. Acidification and nutrient enrichment 6 processes are largely dependent on the cycling of total nitrogen and sulfur species. From an 7 atmospheric perspective, it is convenient and consistent with current measurement and modeling 8 frameworks to consider the reduced and oxidized forms of atmospheric nitrogen. Virtually all 9 atmospheric sulfur is considered oxidized sulfur in the forms of particulate bound sulfate and 10 gaseous sulfur dioxide. In order to assess current concentrations of reactive nitrogen and sulfur 11 we evaluated data available from monitoring the existing networks as well as from the CMAQ 12 model. Regarding the monitoring data, there are a number of important issues in understanding 13 the measurements of NOy provided by different monitoring networks. In principle, measured 14 NOy is based on catalytic conversion of all oxidized species to NO followed by 15 chemiluminescence NO detection. We recognize the caveats associated with instrument 16 conversion efficiency and possible inlet losses. The CMAQ treats the dominant NOy species as 17 explicit species while the minor contributing non-PAN organic nitrogen compounds are 18 aggregated. Atmospheric nitrogen and sulfur are largely viewed as regional air quality issues due 19 to the importance of chemical conversion of primary emissions into secondarily formed species; 20 a combination of ubiquitous sources, particularly mobile source emissions of NOX, and elevated 21 emissions of NOX and SO2 that aid pollutant mass dispersal and broader physical transport over 22 large distances. In effect, the regional nature is due to both transport processes as well as the 23 relatively ubiquitous nature of sources combined with chemical processes that tend to form more 24 stable species with extended atmospheric lifetimes. This regionalized effect, particularly 25 throughout the Eastern United States, dominates the overall patterns discussed below of 26 secondarily formed species such as sulfate or NOy, which is an aggregate of species where the 27 more aged air masses consisting largely of chemically processed air is dominated by secondarily 28 formed peroxyacetyl nitrate (PAN), particulate nitrate and nitric acid. 29 Nationwide maps of CMAQ-predicted 2005 annual average NOy, NHX (NH3 and NH4), 30 NH3, NH4, ST, SO/t, and SO2 are provided in figures 4-5 through 4-11 respectively. Given the 31 considerable gaps in air quality observation networks as discussed in the REA and ISA (2008), March 2010 115 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 modeled concentration patterns are used here to illustrate national representations of current air 2 quality conditions for nitrogen and sulfur. The 2005 model year reflects the most recent available 3 simulation for inclusion in this policy assessment. In addition, figures 4-12 and 4-13 provide 4 maps of 2005 annual average SO2 and SO/t, respectively based on CASTNET observations. Site 5 specific annual average 2005 NOy measured concentrations at SLAMS (Figure 4-14) are 6 typically are less than 40 ppb., The spatial patterns for the 2005 modeled and observed NOy and 7 SC>2 concentrations are similar to the 2002 CMAQ-based maps provided in the REA., largely 8 capturing the influence of major source regions throughout the nation. A spreading of the 9 oxidized sulfur fields (Figures 4-5 and 4-6), relative to 862, is consistent with sulfate 10 transformation and associated air mass aging and transport. Ammonia and ammonium 11 concentration patterns (Figure 4-4) are influenced strongly by the ammonia emissions 12 distribution, with marginal spreading associated with the addition of NFLj. The NHX fields are 13 more strongly influenced by source location, relative to sulfur, based on the fast removal of 14 atmospheric ammonia through deposition. Total deposition for nitrogen and sulfur (Figures 4-15 15 and 4-16) basically follow the patterns of ambient air concentrations. 16 Current conditions indicate that the current 862 and NC>2 secondary standards are not 17 exceeded (Figures 4-17 and 4-18) in locations where ecological effects have been observed, and 18 where critical loads of nitrogen and sulfur are exceeded. This is consistent with the fact that NC>2 19 accounts for only a fraction of NOy, and thus reductions in NC>2 emissions would not be expected 20 to fully address concentrations of NOy. The map in Figure 4-19 further illustrates this point by 21 showing that the contribution of NC>2 to NOy is often less than 50% in rural areas. March 2010 116 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 AMAD 2005af CMAQ — NOy (ppb) Figure 4-5. 2005 CMAQ modeled annual average NOy (ppb). These maps will be replaced with full CONUS maps in the next draft. March 2010 117 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX Legend >= 1.0to<3.0 >= 30 to < 5.0 >= 5.0 to < 7.0 =•= 7.0 to < 10.0 >= 10.0 AMAD 2005af CMAQ — NHx (ug/m3) 2 3 4 Figure 4-6. 2005 CMAQ modeled annual average total reduced nitrogen (NHX) (as ng/m3 nitrogen) March 2010 118 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 AMAD 2005af CMAQ — NH3 (ug/m3) Figure 4-7. 2005 CMAQ modeled annual average ammonia, NHs, (as ng/m N) March 2010 119 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 AMAD 2005af CMAQ — NH4 (ug/m3) Figure 4-8. 2005 CMAQ modeled annual average ammonia, NH4, (as ng/m N) March 2010 120 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX AMAD 2005af CMAQ — ST (ug/m3) 2 3 4 5 Figure 4-9. 2005 CMAQ modeled annual average SOX, (as ng/m S from 862 and S04). March 2010 121 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 AMAD 2005af CMAQ — SO2 (ug/m3) Legend I l<05 I | >=0.5lo< 1.0 I |>=1.0tO<3.0 I | >=• 3,0 to < 5.0 I | >= 5,0 to < 7.0 ^H >= 7.0 to * 10.0 ^B ~-'= 10 ° *° *' 20 ° ^H =•= 20.0 Figure 4-10. 2005 CMAQ modeled annual average SO2 (as ng/m S) March 2010 122 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 AMAD 2005af CMAQ — SO4 (ug/m3) Figure 4-11. 2005 CMAQ modeled annual average SO4 (as ng/m S). March 2010 123 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 4 5 6 7 Figure 4-12. 2005 annual average sulfur dioxide concentrations based on CASTNET generated by the Visibility Information Exchange Web Sysytem (VIEWS). Figure 4-13. 2005 annual average sulfate concentrations based on CASTNET generated by the Visibility Information Exchange Web Sysytem (VIEWS). March 2010 124 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 Annual Average NOY Concentrations (2005) Figure 4-14. Annual average 2005 NOy concentrations from reporting stations in AQS. March 2010 125 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 <2,0 >= 2.0 to < 3.0 >=3.0tO<4.0 >- 4.0 to < 50 >= 5.0 to < 7.0 >=7.0to<9.0 >=9.Qto< 14,0 >= 14.0 to < 20.0 >= 20.0 AMAD 200Saf CMAO — Oxidized Nitrogen Deposition ( kgN/Ha/Yr ) Figure 4-15. 2005 CMAQ modeled Oxidized Nitrogen Deposition (kgN/Ha/Yr). March 2010 126 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 AMAD 2005af CMAQ — Oxidized Sulfur Deposition { kgS/Ha/Yr ) „-• Legend »=1.0to<2.0 >= 2.0 lo < 3.0 >=3.0(o<6.0 >=6.0lo< 10.0 >- 10.0 to < 16.0 >= 16.0 to < 24.0 >= 24.0 10 < 30.0 >= 30.0 Figure 4-16. 2005 CMAQ modeled Oxidized Sulfur Deposition (kgS/Ha/Yr). March 2010 127 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 3-hr Max SO2 Concentrations (2005) 1 2 3 4 5 6 7 8 Figure 4-17. Three hour average maximum 2005 SC>2 concentrations based on the SLAMS reporting to EPA's Air Quality System (AQS) data base. The current SC>2 secondary standard based on the maximum 3 hour average value is 500 ppb, a value not exceeded. While there are obvious spatial gaps, the majority of these stations are located to capture maximum values generally in proximity to major sources and high populations. Lower relative values are expected in more remote acid sensitive areas. March 2010 128 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX Annual Average NO2 Concentrations (2005) 1 2 3 4 5 6 7 8 Figure 4-18. Annual average 2005 NC>2 concentrations based on the SLAMS reporting to EPA's Air Quality System (AQS) data base. The current NC>2 secondary standard is 53 ppb, a value well above those observed. While there are obvious spatial gaps, the stations are located in areas of relatively high concentrations in highly populated areas. Lower relative values are expected in more remote acid sensitive areas. March 2010 129 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX Layer 1 (NOY[1]- NO2[2])/ NOY[1] 9 10 11 12 13 14 231 221 211 201 191 181 171 161 151 141 131 - 121 111 101 91 81 71 1 2 3 4 5 6 1 4.5 December 31,0002 00:00:00 UTC Mill (10, 34) = 0.175, Max (5,10) = 0.915 Figure 4-19. 2005 CMAQ derived annual average ratio of (NOy - NO2)/NOy. The fraction of NO2 contributing to total NOy generally is less than 50% in the Adirondack and Shenandoah case study areas. The ratio reflects the relative air mass aging associated with transformation of oxidized nitrogen beyond NO and NO2 as one moves from urban to rural locations. ARE ADVERSE EFFECTS ON THE PUBLIC WELFARE OCCURRING UNDER CURRENT AIR QUALITY CONDITIONS FOR NO2 AND SO2 AND WOULD THEY OCCUR IF THE NATION MET THE CURRENT SECONDARY STANDARDS? The previous sections have established that almost all areas of the U.S. were at concentrations of SO2 and NO2 below the levels of the current standards. In many locations, SO2 and NO2 concentrations are substantially below the levels of the standards. This suggests that levels of deposition and any effects on ecosystems due to deposition of NOX and SOX under March 2010 130 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 recent conditions are occurring even though areas meet or are below current standards. This 2 section focuses on summarizing the evidence of effects occurring at deposition levels consistent 3 with recent conditions. 4 The ISA summarizes the available studies of relative nitrogen contribution and finds that 5 in much of the U.S., NOX contributes from 50 to 75 percent of total atmospheric deposition [ISA 6 Section 2.8.4]. While the proportion of total nitrogen loadings associated with atmospheric 7 deposition of nitrogen varies across locations (N deposition in the Eastern U.S. includes 8 locations with greater than 9 kg N/ha/year, and in the central U.S. high deposition locations with 9 values on the order of 6 to 7 kg N/ha/year), the ISA indicates that atmospheric N deposition is 10 the main source of new anthropogenic N to most headwater streams, high elevation lakes, and 11 low-order streams. Atmospheric N deposition contributes to the total N load in terrestrial, 12 wetland, freshwater, and estuarine ecosystems that receive N through multiple pathways. In 13 several large estuarine systems, including the Chesapeake Bay, atmospheric deposition accounts 14 for between 10 and 40 percent of total nitrogen loadings. 15 Atmospheric concentrations of SOX account for nearly all S deposition in the US. For the 16 period 2004-2006, mean S deposition in the U.S. was greatest east of the Mississippi River with 17 the highest deposition amount, 21.3 kg S/ha/yr, in the Ohio River Valley where most recording 18 stations reported 3 year averages >10 kg S/ha/yr. Numerous other stations in the East reported S 19 deposition >5 kg S/ha/yr. Total S deposition in the U.S. west of the 100th meridian was 20 relatively low, with all recording stations reporting <2 kg S/ha/yr and many reporting <1 kg 21 S/ha/yr. S was primarily deposited in the form of wet SC>4 2 followed in decreasing order by a 22 smaller proportion of dry SC>2 and a much smaller proportion of deposition as dry SC>42 . 23 New scientific evidence exists to address each of the areas of uncertainty raised in the 24 previous reviews (summarized above). Based on the new evidence, the current ISA concludes 25 that: 26 (1) The evidence is sufficient to infer a causal relationship between acidifying deposition 27 (to which both NOX and SOX contribute) and effects on biogeochemistry related to 28 terrestrial and aquatic ecosystems; and biota in terrestrial and aquatic ecosystems. 29 (2) The evidence is sufficient to infer a causal relationship between N deposition, to 30 which NOX and NHX contribute, and the alteration of A) biogeochemical cycling of N 31 and carbon in terrestrial, wetland, freshwater aquatic, and coastal marine ecosystems; March 2010 131 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 B) biogenic flux of methane (CH4), and N2O in terrestrial and wetland ecosystems; 2 and C) species richness, species composition, and biodiversity in terrestrial, wetland, 3 freshwater aquatic and coastal marine ecosystems. 4 (3) The evidence is sufficient to infer a causal relationship between S deposition and 5 increased Hg methylation in wetlands and aquatic environments. 6 Subsequent to the previous review of the NOX secondary standard, a great deal of 7 information on the contribution of atmospheric deposition associated with ambient NOX has 8 become available. Chapter 3 of the REA provides a thorough assessment of the contribution of 9 NOX to nitrogen deposition throughout the U. S., and the relative contributions of ambient NOX 10 and reduced forms of nitrogen. Staff concludes that based on that analysis, ambient NOX is a 11 significant component of atmospheric nitrogen deposition, even in areas with relatively high 12 rates of deposition of reduced nitrogen. In addition, staff initially concludes that atmospheric 13 deposition of oxidized nitrogen contributes significantly to total nitrogen loadings in nitrogen 14 sensitive ecosystems. 15 As discussed throughout the risk and exposure assessment document, there are several 16 key areas of risk that are associated with ambient concentrations of NOX and SOX. In previous 17 reviews of the NOX and SOX secondary standards, the standards were designed to protect against 18 direct exposure of plants to ambient concentrations of the pollutants. A significant shift in 19 understanding of the effects of NOX and SOX has occurred since the last reviews, reflecting the 20 large amount of research that has been conducted on the effects of deposition of nitrogen and 21 sulfur to ecosystems. The most significant risks of adverse effects to public welfare are those 22 related to deposition of NOX and SOX to both terrestrial and aquatic ecosystems. These risks fall 23 into two categories: acidification and nutrient enrichment. These made up the emphasis of the 24 REA, and are most relevant to evaluating the adequacy of the existing standards in protecting 25 public welfare from adverse ecological effects. March 2010 132 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 4 5 6 7 8 9 10 11 12 13 14 15 4.5.1 To what extent do the current NOX and SOX secondary standards provide protection from adverse effects associated with deposition of atmospheric NOX, and SOX which results in acidification in sensitive aquatic and terrestrial ecosystems? The focus of the REA case studies was on determining whether deposition of sulfur and oxidized nitrogen in locations where ambient NOX and SOX was at or below the current standards was resulting in acidification and related effects. This review has focused on identifying ecological indicators that can link atmospheric deposition to ecological effects associated with acidification. NOX and SOX contribute to acidification in both aquatic and terrestrial ecosystems, although the indicators of effects differ. While there are some geographic areas with both terrestrial and aquatic ecosystems that are vulnerable to acidification, the case study areas do not fully overlap. Figure 4-20 shows the locations of the case studies evaluated in the REA. 0 250 500 750 1,000 —^^™ Kilometers Figure 4-20. National map highlighting the 9 case study areas evaluated in the REA. March 2010 133 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 4.5.1.1 Aquatic Acidification 2 Based on the case studies conducted for lakes in the Adirondacks and streams in 3 Shenandoah National Park, staff concludes that there is significant risk to acid sensitive aquatic 4 ecosystems at atmospheric concentrations of NOX and SOX at or below the current standards. 5 This conclusion is based on application of the MAGIC model to estimate the effects of 6 deposition at levels consistent with atmospheric NOX and SOX concentrations that are at or below 7 the current standards. An important ecological indicator for aquatic acidification effects is ANC, 8 measuring the acid buffering capacity of a waterbody, and the case study focused on evaluating 9 whether locations were likely to be below critical values of ANC given deposition levels 10 associated with NOX and SOX concentrations that meet the current standards. In addition, the case 11 studies assessed the ecological effects and some of the known ecosystem services that are 12 associated with different levels of ANC in order to associate the ecological indicator with 13 measures of public welfare that may be adversely affected by deposition levels consistent with 14 concentrations of NOX and SOX that meet the current standards. 15 Staff concludes that the evidence and risk assessment support strongly a relationship 16 between atmospheric deposition of NOX and SOX and ANC, and that ANC is an excellent 17 indicator of aquatic acidification. Staff also concludes that at levels of deposition associated with 18 NOX and SOX concentrations at or below the current standards, ANC levels are expected to be 19 below benchmark values that are associated with significant losses in fish species richness (REA 20 Section 4) 21 Many locations in sensitive areas of the U.S. have ANC levels below benchmark levels 22 for ANC classified as severe, elevated, or moderate concern (see Figure 2-1). The average 23 current ANC levels across 44 lakes in the Adirondack case study area is 62.1 (moderate 24 concern), however, 44 percent of lakes had deposition levels exceeding the critical load for an 25 ANC of 50, and 28 percent of lakes had deposition levels exceeding the critical load for an ANC 26 of 20 |ieq/L (REA Section 4.2.4.2). This indicates that almost half of the 44 lakes in the 27 Adirondacks case study area are at an elevated concern levels, and almost a third are at a severe 28 concern level. These levels are associated with greatly reduced fish species diversity, and losses 29 in the health and reproductive capacity of remaining populations. Based on assessments of the 30 relationship between number offish species and ANC level in both the Adirondacks and 31 Shenandoah areas, the number of fish species is decreased by over half at an ANC level of 20 March 2010 134 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 neq/L relative to an ANC level at 100 |ieq/L (REA Figure 4.2-1). At levels below 20 |ieq/L, 2 populations of sensitive species, such as brook trout, may decline significantly during episodic 3 acidification events. When extrapolated to the full population of lakes in the Adirondacks area 4 using weights based on the EMAP probability survey (REA 4.2.6.1), 36 percent of lakes 5 exceeded the critical load for an ANC of 50 jieq/L and 13 percent of lakes exceeded the critical 6 load for an ANC of 20 jieq/L. 7 Many streams in the Shenandoah case study area also have levels of deposition that are 8 associated with ANC levels classified as severe, elevated, or moderate concern. The average 9 ANC under recent conditions for the 60 streams evaluated in the Shenandoah case study area is 10 57.9, indicating moderate concern. However, 85 percent of streams had recent deposition 11 exceeding the critical load for an ANC of 50 |ieq/L, and 72 percent exceeded the critical load for 12 an ANC of 20 |ieq/L. As with the Adirondacks area, this suggests that significant numbers of 13 sensitive streams in the Shenandoah area are at risk of adverse impacts on fish populations under 14 recent conditions. Many other streams in the Shenandoah area are likely to experience conditions 15 of elevated to severe concern based on the prevalence in the area of bedrock geology associated 16 with increased sensitivity to acidification suggesting that effects due to stream acidification could 17 be widespread in the Shenandoah area (REA 4.2.6.2). 18 The ISA notes that large portions of the Eastern U.S. are acid sensitive, and that current 19 deposition levels exceed those that would allow recovery of the most acid sensitive lakes in the 20 Adirondacks (ISA ES). In addition, because of past loadings, areas of the Shenandoah are 21 sensitive to current deposition levels (ISA ES). Much of the West is naturally less sensitive to 22 acidification, and as such, less focus is placed on the adequacy of the existing standards in these 23 areas, with the exception of the mountainous areas of the West, which experience episodic 24 acidification due to deposition. 25 While most (99 percent) of stream kilometers in the U.S. are not chronically acidified 26 under current conditions, a recent survey found sensitive streams in many locations in the U.S., 27 including the Appalachian mountains, the Coastal Plain, and the Mountainous West (ISA 28 Section 4.2.2.3). In these sensitive areas, between 1 and 6 percent of stream kilometers are 29 chronically acidified. 30 The ISA notes that "consideration of episodic acidification greatly increases the extent 31 and degree of estimated effects for acidifying deposition on surface waters." (ISA Section March 2010 135 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 3.2.1.6) Some studies show that the number of lakes that could be classified as acidified based on 2 episodic acidification is 2 to 3 times the number of lakes classified as acidified based on chronic 3 ANC. These episodic acidification events can have long term effects on fish populations (ISA 4 Section 3.2.1.6). Under recent conditions, episodic acidification has been observed in locations 5 in the Eastern U.S. and in the Mountainous Western U.S. (ISA Section 3.2.1.6). 6 It can therefore be concluded that recent levels of NOX and SOX are associated with 7 deposition that leads to ANC values below benchmark values known to cause ecological harm in 8 sensitive aquatic systems, including lakes and streams in multiple areas of the U.S. These 9 changes are known to have impacts on ecosystem services such as reductions in recreational 10 fishing. While other ecosystem services (e.g. habitat provisioning, subsistence fishing, and 11 biological control as well as many others) are potentially affected by reductions in ANC, 12 confidence in the specific translation of ANC values to these additional ecosystem services is 13 much lower. 14 4.5.1.2 Terrestrial Acidification 15 Based on the case studies on sugar maple and red spruce habitat, staff concludes that 16 there is significant risk to terrestrial ecosystems from acidification at atmospheric concentrations 17 of NOX and SOX at or below the current standards. This conclusion is based on application of the 18 simple mass balance model to deposition levels associated with NOX and SOX concentrations at 19 or below the current standards. The ecological indicator selected for terrestrial acidification is the 20 base cation to aluminum ratio (BC:A1), which has been linked to tree health and growth. The 21 results of the REA strongly support a relationship between atmospheric deposition of NOX and 22 SOX and BC:A1, and that BC:A1 is a good indicator of terrestrial acidification. At levels of 23 deposition associated with NOX and SOX concentrations at or below the current standards, BC: Al 24 levels are expected to be below benchmark values that are associated with significant losses in 25 tree health and growth. Such degradation of terrestrial ecosystems could affect ecosystem 26 services such as habitat provisioning, endangered species, goods production (timber, syrup, etc.) 27 and many others. 28 Many locations in sensitive areas of the U.S. have Bc/Al levels below benchmark levels 29 classified as providing low to intermediate levels of protection to tree health. At a Bc/Al ratio of 30 1.2 (intermediate level of protection), red spruce growth can be reduced by 20 percent. At a March 2010 136 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Bc/Al ratio of 0.6 (low level of protection), sugar maple growth can be reduced by 20 percent. 2 The REA did not evaluate broad sensitive regions. However, in the sugar maple case study area 3 (Kane Experimental Forest), recent deposition levels are associated with a Bc/Al ratio below 1.2, 4 indicating between intermediate and low level of protection, which would indicate the potential 5 for a greater than 20 percent reduction in growth. In the red spruce case study area (Hubbard 6 Brook Experimental Forest), recent deposition levels are associated with a Bc/Al ratio slightly 7 above 1.2, indicating slightly better than an intermediate level of protection (REA Section 8 4.3.5.1) 9 Over the full range of sugar maple, 12 percent of evaluated forest plots exceeded the 10 critical load for a Bc/AL ratio of 1.2, and 3 percent exceeded the critical load for a Bc/Al ratio of 11 0.6. However, there was large variability across states. In New Jersey, 67 percent of plots 12 exceeded the critical load for a Bc/Al ratio of 1.2, while in several states on the outskirts of the 13 range for sugar maple, e.g. Arkansas, Illinois, no plots exceeded the critical load for a Bc/Al ratio 14 of 1.2. For red spruce, overall 5 percent of plots exceeded the critical load for a Bc/Al ratio of 15 1.2, and 3 percent exceeded the critical load for a Bc/Al ratio of 0.6. In the major red spruce 16 producing states (Maine, New Hampshire, and Vermont), critical loads for a Bc/AL ratio of 1.2 17 were exceeded in 0.5, 38, and 6 percent of plots. 18 The ISA reported one study that estimated 15 percent of U.S. forest ecosystems exceeded 19 the critical loads for acidity for N and S deposition by >250 eq/ha/year under current conditions 20 (ISA Section 4.2.1.3). Staff believes that this represents a significant portion of sensitive 21 terrestrial ecosystems. 22 It can therefore be concluded that recent levels of NOX and SOX are associated with 23 deposition that leads to BC:A1 values below benchmark values that cause ecological harm in 24 some sensitive terrestrial ecosystems. While effects are more widespread for sugar maple, there 25 are locations with low to intermediate levels of protection from effects on both sugar maple and 26 red spruce. While there are many other ecosystem services, including timber production, natural 27 habitat provision, and regulation of water, climate, and erosion, potentially affected by 28 reductions in BC:A1, linkages of BC:A1 values to these additional ecosystem services is on the 29 whole not well understood. March 2010 137 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 4.5.2 To what extent does the current NOX secondary standard provide protection 2 from adverse effects associated with deposition of atmospheric NOX, which 3 results in nutrient enrichment effects in sensitive aquatic and terrestrial 4 ecosystems? 5 Nutrient enrichment effects are due to nitrogen loadings from both atmospheric and non- 6 atmospheric sources. Evaluation of nutrient enrichment effects requires an understanding that 7 nutrient inputs are essential to ecosystem health. The specific long term levels of nutrients in a 8 system affect the types of species that occur over long periods of time. Short term additions of 9 nutrients can affect species competition, and even small additions of nitrogen in areas that are 10 traditionally nutrient poor can have significant impacts. In certain limited situations, additions of 11 nitrogen can increase rates of growth, and these increases can have short term benefits in certain 12 managed ecosystems. As noted earlier, this review of the standards is focused on unmanaged 13 ecosystems. As a result, in assessing adequacy of the current standards, we are focusing on the 14 adverse effects of nutrient enrichment in unmanaged ecosystems. However, the following 15 discussion provides a brief assessment of effects in managed ecosystems. 16 Impacts of nutrient enrichment in managed ecosystems may be positive or negative 17 depending on the levels of nutrients from other sources in those areas. Positive effects can occur 18 when crops or commercial forests are not receiving enough nitrogen nutrients. Nutrients 19 deposited on crops from atmospheric sources are often referred to as passive fertilization. 20 Nitrogen is a fundamental nutrient for primary production in both managed and unmanaged 21 ecosystems. Most productive agricultural systems require external sources of nitrogen in order to 22 satisfy nutrient requirements. Nitrogen uptake by crops varies, but typical requirements for wheat 23 and corn are approximately 150 kg/ha/yr and 300 kg/ha/yr, respectively (NAPAP, 1990). These 24 rates compare to estimated rates of passive nitrogen fertilization in the range of 0 to 5.5 kg/ha/yr 25 (NAPAP, 1991). 26 Information on the effects of changes in passive nitrogen deposition on forestlands and 27 other terrestrial ecosystems is very limited. The multiplicity of factors affecting forests, including 28 other potential stressors such as ozone, and limiting factors such as moisture and other nutrients, 29 confound assessments of marginal changes in any one stressor or nutrient in forest ecosystems. 30 The ISA notes that only a fraction of the deposited nitrogen is taken up by the forests, most of 31 the nitrogen is retained in the soils (ISA 3.3.2.1). In addition, the ISA indicates that forest March 2010 138 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 management practices can significantly affect the nitrogen cycling within a forest ecosystem, and 2 as such, the response of managed forests to NOX deposition will be variable depending on the 3 forest management practices employed in a given forest ecosystem (ISA Annex C C.6.3) 4 Increases in the availability of nitrogen in N-limited forests via atmospheric deposition could 5 increase forest production over large non-managed areas, but the evidence is mixed, with some 6 studies showing increased production and other showing little effect on wood production (ISA 7 3.3.9). Because leaching of nitrate can promote cation losses, which in some cases create nutrient 8 imbalances, slower growth and lessened disease and freezing tolerances for forest trees, the net 9 effect of increased N on forests in the U.S. is uncertain (ISA 3.3.9). 10 In managed agricultural ecosystems, nitrogen inputs from atmospheric NOX comprise a 11 small fraction (less than 3 percent) of total nitrogen inputs, which include commercially applied 12 fertilizers as well as applications of composted manure. And because of the temporal and spatial 13 variability in atmospheric deposition of NOX, it is unlikely that farmers would alter their 14 fertilization decisions based on expected nitrogen inputs from NOX. And, in some locations, 15 farmers need less nitrogen inputs due to production of excess nitrogen through livestock. In some 16 locations, nitrogen production through livestock waste exceeds the absorptive capacity of the 17 surrounding land, and as such, excess nitrogen from deposition of NOX in those locations reduces 18 the capacity of the system to dispose of excess nitrogen, potentially increasing the costs of waste 19 management from livestock operations (Letson and Gollehon, 1996). A USD A Economic 20 Research Service report found that in 1997, 68 counties with high levels of confined livestock 21 production had manure nitrogen levels that exceed the assimilative capacity of all the county's 22 crop and pasture land (Gollehon et al, 2001). In those locations, additional nitrogen inputs from 23 NOX deposition will result in excess nitrogen, leading to nitrogen leaching and associated effects. 24 4.5.3 Aquatic Nutrient Enrichment 25 The REA case studies focused on coastal estuaries and revealed that while current 26 ambient loadings of atmospheric NOX are contributing to the overall deposit!onal loading of 27 coastal estuaries, other non-atmospheric sources are contributing in far greater amounts in total, 28 although atmospheric contributions are as large as some other individual source types. The 29 ability of current data and models to characterize the incremental adverse impacts of nitrogen 30 deposition is limited, both by the available ecological indicators, and by the inability to attribute March 2010 139 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 specific effects to atmospheric sources of nitrogen. The REA case studies used as the ecological 2 indicator for aquatic nutrient enrichment, an index of eutrophication known as the Assessment of 3 Estuarine Trophic Status Eutrophication Index (ASSETS El). This index is a six level index 4 characterizing overall eutrophication risk in a waterbody. This indictor is not sensitive to 5 relatively large changes in nitrogen deposition. In addition, this type of indicator does not reflect 6 the impact of nitrogen deposition in conjunction with other sources of nitrogen. 7 For example, if NOX deposition is contributing nine tenths of the nitrogen loading 8 required to move a waterbody from an ASSETS El category of "moderate" to a category of 9 "poor", zeroing out NOX deposition will have no impact on the ASSETS El value. However, if 10 an area were to decide to put in place decreases in nitrogen loadings to move that waterbody 11 from "poor" to "moderate," the area would have to reduce the full amount of the loadings 12 through other sources if atmospheric deposition were not considered. Thus, the adverse impact of 13 atmospheric nitrogen is in its contribution to the overall loading, and reductions in NOX will 14 decrease the amount of reductions from other sources of nitrogen loadings that would be required 15 to move from a lower ASSETS El category to a higher category. NOX deposition can also be 16 characterized as reducing the risk of a waterbody moving from a higher ASSETS El category to 17 a lower category, by reducing the vulnerability of that waterbody to increased loadings from 18 non-atmospheric sources. 19 Based on the above considerations, staff preliminarily concludes that the ASSETS El is 20 not an appropriate ecological indicator for estuarine aquatic eutrophication. Staff further 21 concludes that additional analysis is required to develop an appropriate indicator for determining 22 the appropriate levels of protection from N nutrient enrichment effects in estuaries related to 23 deposition of NOX. As a result, staff is unable to make a determination as to the adequacy of the 24 existing secondary NOX standard in protecting public welfare from N nutrient enrichment effects 25 in estuarine aquatic ecosystems. 26 Additionally, nitrogen deposition can alter species composition and cause eutrophication 27 in freshwater systems. In the Rocky Mountains, for example, deposition loads of 1.5 to 2 28 kg/ha/yr which are well within current ambient levels are known to cause changes in species 29 composition in diatom communities indicating impaired water quality (ISA Section 3.3.5.3). It 30 then seems apparent then that the existing secondary standard for NOX does not protect such 31 ecosystems and their resulting services from impairment. March 2010 140 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 4.5.4 Terrestrial Nutrient Enrichment 2 The scientific literature has many examples of the deleterious effects caused by excessive 3 nitrogen loadings to terrestrial systems. Several studies have set benchmark values for levels of 4 N deposition at which scientifically adverse effects are known to occur. These benchmarks are 5 discussed more thoroughly in Chapter 5 of the REA. Large areas of the country appear to be 6 experiencing deposition above these benchmarks for example, Fenn et al. (2008) found that at 7 3.1 kg N/ha/yr, the community of lichens begins to change from acidophytic to tolerant species; 8 at 5.2 kg N/ha/yr, the typical dominance by acidophytic species no longer occurs; and at 10.2 kg 9 N/ha/yr, acidophytic lichens are totally lost from the community. Additional studies in the 10 Colorado Front Range of the Rocky Mountain National Park support these findings and are 11 summarized in Chapter 6.0 of the Risk and Exposure Assessment. These three values (3.1, 5.2, 12 and 10.2 kg/ha/yr) are one set of ecologically meaningful benchmarks for the mixed conifer 13 forest (MCF) of the pacific coast regions. Nearly all of the known sensitive communities receive 14 total nitrogen deposition levels above the 3.1 N kg/ha/yr ecological benchmark according to 15 the 12 km, 2002 CMAQ/NADP data, with the exception of the easternmost Sierra Nevadas. 16 MCFs in the southern portion of the Sierra Nevada forests and nearly all MCF communities in 17 the San Bernardino forests receive total nitrogen deposition levels above the 5.2 N kg/ha/yr 18 ecological benchmark. 19 Coastal Sage Scrub communities (CSS) are also known to be sensitive to community 20 shifts caused by excess nitrogen loadings. Wood et al. (2006) investigated the amount of nitrogen 21 utilized by healthy and degraded CSS systems. In healthy stands, the authors estimated that 3.3 22 kg N/ha/yr was used for CSS plant growth (Wood et al., 2006). It is assumed that 3.3 kg N/ha/yr 23 is near the point where nitrogen is no longer limiting in the CSS community. Therefore, this 24 amount can be considered an ecological benchmark for the CSS community. The majority of the 25 known CSS range is currently receiving deposition in excess of this benchmark. Thus, staff 26 concludes that recent conditions where NOX ambient concentrations are at or below the current 27 NOX secondary standards are not adequate to protect against anticipated adverse impacts from N 28 nutrient enrichment in sensitive ecosystems (systems where N is not limiting). March 2010 141 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 4.6 TO WHAT EXTENT DO THE CURRENT NOX AND/OR SOX 2 SECONDARY STANDARDS PROVIDE PROTECTION FROM 3 OTHER ECOLOGICAL EFFECTS (E.G., MERCURY 4 METHYLATION) ASSOCIATED WITH THE DEPOSITION OF 5 ATMOSPHERIC NOX, AND/OR SOX? 6 It is stated in the ISA (ISA Sections 3.4.1 and 4.5) that mercury is a highly neurotoxic 7 contaminant that enters the food web as a methylated compound, methylmercury. Mercury is 8 principally methylated by sulfur-reducing bacteria and can be taken up by microorganisms, 9 zooplankton and macroinvertebrates. The contaminant is concentrated in higher trophic levels, 10 including fish eaten by humans. Experimental evidence has established that only inconsequential 11 amounts of methylmercury can be produced in the absence of sulfate. Once methylmercury is 12 present, other variables influence how much accumulates in fish, but elevated mercury levels in 13 fish can only occur where substantial amounts of methylmercury are present. Current evidence 14 indicates that in watersheds where mercury is present, increased SOX deposition very likely 15 results in additional production of methylmercury which leads to greater accumulation of MeHg 16 concentrations in fish (Munthe et al, 2007; Drevnick et al., 2007). 17 The production of meaningful amounts of methylmercury (MeHg) requires the presence 18 of SO42" and mercury, and where mercury is present, increased availability of SO42" results in 19 increased production of MeHg. There is increasing evidence on the relationship between sulfur 20 deposition and increased methylation of mercury in aquatic environments; this effect occurs only 21 where other factors are present at levels within a range to allow methylation. The production of 22 methylmercury requires the presence of sulfate and mercury, but the amount of methylmercury 23 produced varies with oxygen content, temperature, pH, and supply of labile organic carbon (ISA 24 Section 3.4). In watersheds where changes in sulfate deposition did not produce an effect, one or 25 several of those interacting factors were not in the range required for meaningful methylation to 26 occur (ISA Section 3.4). Watersheds with conditions known to be conducive to mercury 27 methylation can be found in the northeastern United States and southeastern Canada. The 28 relationship between sulfur and methylmercury production is addressed qualitatively in Chapter 29 6 of the Risk and Exposure Assessment. March 2010 142 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 With respect to sulfur deposition and mercury methylation, the final ISA determined: The 2 evidence is sufficient to infer a causal relationship between sulfur deposition and increased 3 mercury methylation in wetlands and aquatic environments. However, staff did not conduct a 4 quantitative assessment of the risks associated with increased mercury methylation under current 5 conditions. As such, staff are unable to make a determination as to the adequacy of the existing 6 SC>2 standards in protecting against welfare effects associated with increased mercury 7 methylation. 8 4.7 REFERENCES 9 BJ. Finlayson-Pitts and J.N. Pitts, 2000, Chemistry of the Upper and Lower Troposhere, 10 Academic Press, San Diego, CA 11 Drevnick, P.E., D.E. Canfield, P.R. Gorski, A.L.C. Shinneman, D.R. Engstrom, D.C.G. Muir, 12 G.R. Smith, PJ. Garrison, L.B. Cleckner, J.P. Hurley, R.B. Noble, R.R. Otter, and J.T. 13 Oris. 2007. Deposition and cycling of sulfur controls mercury accumulation in Isle 14 Royale fish. Environmental Science and Technology ¥7(21):7266-7272. 15 Fenn, M.E., S. Jovan, F. Yuan, L. Geiser, T. Meixner, and B.S. Gimeno. 2008. Empirical and 16 simulated critical loads for nitrogen deposition in California mixed conifer forests. 17 Environmental Pollution 755(3 ): 492-511. 18 Munthe, 1, R.A. Bodaly, B.A. Branfireun, C.T. Driscoll, C.C. Gilmour, R. Harris, M. Horvat, M. 19 Lucotte, and O. Malm. 2007. Recovery of mercury-contaminated fisheries. Ambio 36:33- 20 44. 21 Scheffe, R.D., P. A. Solomon, R. Husar, T. Hanley, M. Schmidt, M. Koerber, M. Gilroy, J. 22 Hemby, N. Watkins, M. Papp, J. Rice, J. Tikvart andR. Valentinetti, The National 23 Ambient Air Monitoring Strategy: Rethinking the Role of National Networks, JAWMA, 24 ISSN: 1047-3289 J. Air & Waste Manage. Assoc. 2009, 59:579-590 DOI: 10.3155/1047- 25 3289.59.5.579 26 U.S. EPA (Environmental Protection Agency). 1982. Review of the National Ambient Air Quality 27 Standards for Sulfur Oxides: Assessment of Scientific and Technical Information. March 2010 143 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 OAQPS Staff Paper. EPA-450/5-82-007. U.S. Environmental Protection Agency, Office 2 of Air Quality Planning and Standards, Research Triangle Park, NC. 3 U.S. EPA (Environmental Protection Agency). 1995. Review of the National Ambient Air Quality 4 Standards for Nitrogen Dioxide: Assessment of Scientific and Technical Information. 5 OAQPS Staff Paper. EPA-452/R-95-005. U.S. Environmental Protection Agency, Office 6 of Air Quality Planning and Standards, Research Triangle Park, NC. September. 7 U.S. EPA (Environmental Protection Agency). 2008. Integrated Science Assessment (ISA) for 8 Oxides of Nitrogen and Sulfur-Ecological Criteria (Final Report). EPA/600/R- 9 08/082F. U.S. Environmental Protection Agency, National Center for Environmental 10 Assessment-RTF Division, Office of Research and Development, Research Triangle 11 Park, NC. Available at http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=201485. 12 Wood, Y., T. Meixner, PJ. Shouse, and E.B. Allen. 2006. Altered Ecohydrologic response 13 drives native shrub loss under conditions of elevated N-deposition. Journal of 14 Environmental Quality 35:76-92. 15 March 2010 144 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 5. CONCEPTUAL DESIGN OF AN ECOLOGICALLY RELEVANT 2 MULTI-POLLUTANT STANDARD 3 The objective of this chapter is to describe the conceptual design for a national ambient 4 air quality standard that links ecological indicators of concern to ambient air indicators of NOX 5 and SOX. In Chapter 4 of this policy assessment, the limitations of the design of the current 6 secondary standards are described as they apply to protection of sensitive ecosystems. The 7 conceptual design described in this chapter addresses those limitations. The overall concept for 8 the standards starts by recognizing that the fundamental welfare effects associated with ambient 9 NOX and SOX occur through the process of deposition to sensitive ecosystems. As detailed in 10 Chapter 4, previous NOX and SOX NAAQS reviews only considered effects to vegetation via 11 stomatal exposure. There is now sufficient data to link atmospheric concentrations to adverse 12 effects in ecosystems that are caused by exposure via deposition to soils and surface waters. 13 Deposition is a direct consequence of atmospheric concentration; however it is also modified by 14 factors that vary across the landscape (e.g. elevation and groundcover). Likewise, ecological 15 response to deposition can vary according to ecosystem sensitivity and the ecological indicator 16 of concern. This is the first time a secondary standard for deposition effects related to NOX and 17 SOX has been developed; therefore the conceptual design of a potential standard is described here 18 prior to the specific details on the indicator, level, form and averaging time for such a potential 19 standard that are presented in chapter 6. 20 5.1 COMPONENTS OF THE DESIGN 21 There are four main components to the conceptual design of the standard: atmospheric 22 and ecological indicators, deposition metrics, functions that relate indicators to deposition 23 metrics and factors that modify the functions. These components of the design are illustrated in 24 Figure 5-1. The squares represent indicators. Ecological indicators are chemical or biological 25 components of the ecosystem that can be linked to N and S deposition based on scientific 26 evidence. Air quality indicators are the chemical species of the criteria air pollutants that best 27 represent the atmospheric pollutants that cause ecological harm in the criteria pollutant 28 categories of oxides of nitrogen and oxides of sulfur. Triangles indicate functions in which two 29 variables are related. The ecological effect function is the relationship between the ecological March 2010 145 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 4 5 6 9 10 11 12 13 14 15 16 17 18 19 20 21 indicator and deposition over a range of values. The atmospheric deposition transformation function is the relationship between deposition and the atmospheric concentration of an air quality indicator. The circles represent factors which will modify the functions. Modifying factors can vary across the landscape. The spatial heterogeneity of modifying factors can be challenging to characterize, and therefore in some cases we present multiple options for how to incorporate them into the design. Ecological Indicator Variable/Fixed Modifying Factors Variable/Fixed Modifying Factors Deposition Metric Atmospheric Deposition Transformation Function Ecological Response to Deposition Function Air Quality Indicator(s) Fig 5-1. Schematic diagram of the conceptual design of the standard. 5.1.1 For which effects is there sufficient information to support setting standards? After review of the ISA and REA, CAS AC concluded that aquatic acidification should be the focus for developing a multi-pollutant standard, based on the quantity and quality of data. CASAC also recommended that, in addition to aquatic acidification, the EPA should consider multiple ecological indicators and made the following statement in their letter to the EPA on August 28, 2009: ".. .the Panel finds the information in the current REA sufficient to inform setting separate standards for terrestrial acidification, eutrophication of western alpine lakes and terrestrial nutrient enrichment. However, the Panel believes that setting a standard for coastal nutrient enrichment would be difficult because of the substantial inputs of non-atmospheric sources of N to these systems." The following sections describe the conceptual design for standards based on aquatic acidification, terrestrial acidification, eutrophication of high elevation western lakes and March 2010 146 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 terrestrial nutrient enrichment. The focus of the first draft will be on aquatic acidification, but 2 this general conceptual framework will apply to a broader set of potential endpoints. 3 5.2 ECOLOGICAL COMPONENTS OF THE STANDARD: AQUATIC 4 ACIDIFICATION 5 Details of the conceptual design of the NOX and SOX NAAQS based on aquatic 6 acidification effects are presented in this section. A summary of our over all approach is given 7 here to help provide context and support for the more detailed discussions that follow. 8 At the catchment scale, ambient NOy and SOX add to the total deposition of N and S that 9 lead to aquatic acidification. NHX is often another big component of the total N deposition. The 10 load of deposition that causes a desired level of ANC will vary depending on the characteristics 11 of the ecosystem. The level of ANC is tied to the degree of biological harm to the system from 12 aquatic acidification. 13 The components of the standard are modified for application to aquatic acidification and 14 presented in Fig 5-2. The bidirectional arrows emphasize that the order in which one considers 15 the links between ANC and atmospheric concentrations of NOX and SOX is conceptually 16 important to the standard design. Moreover, different questions may be answered by working 17 through Fig 5-2 from the left to the right versus the right to the left. For example, working from 18 left to right, when a level of ANC is specified the deposition loadings of N and S that would 19 cause the specified level of ANC can be calculated; in essence this would be a critical load for a 20 specified ANC limit. A comparison between the total amount of deposited N and S to the critical 21 load would determine whether the specified level of ANC is achieved for a catchment. Let's now 22 work through the equation from right to left. If the amount of N and S deposited to a given 23 catchment is known, you could calculate the level of ANC that would result. The calculated 24 ANC could then be compared to a benchmark value of ANC. In both of these approaches the 25 amount of reduced N would be subtracted from the total N deposition to calculate deposition 26 from NOy. The atmospheric concentrations of NOy and SOX would be calculated from the 27 deposition of NOy and S according to the methods presented in section 5.4. To determine the 28 appropriate conceptual design from the ecological components of the standard, the analysis from 29 the REA is evaluated in which critical loads were calculated for a target value of ANC, thereby 30 working from left to right on Fig 5-2. March 2010 147 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 4 5 6 9 10 11 12 13 14 15 16 17 18 Level of ANC related to biological effects Acidification Model that relates ANC to deposition at catchment- scale Deposition Loading of N+S which represent national scale landscape categories <=>/ \<=>| Atmospheric Deposition Transformation Function Concentration of the Air Quality Indicator(s) Relationship between the amount of deposition and the effect on the selected ecological indicator, ANC (described in 5.2) Relationship between the amount of deposition and the concentration ofNOx and SOx (described in 5.4) Fig 5-2. Schematic diagram of the conceptual design of the standard based on aquatic acidification. From left to right, if a desired level of ANC is known, then the concentration of the atmospheric indicators that will cause that level may be calculated. From right to left, if the if the concentration of the air quality indicators are known than the ANC that will be caused may be calculated. The secondary NAAQS would apply to all areas of the country. It is not practical to evaluate each catchment individually, and that is not the appropriate approach for a national standard. Here, EPA staff proposes to categorize landscape features nationally, such that within a category there are generally similar characteristics as far as the relationship of total deposited N and S to the ANC. Every part of the country would be assigned into one of these bins/ landscape categories. The secondary NAAQS would be based on a judgment as to a specified level of ANC. For each national acid-sensitivity bin/ landscape category there would be a range of critical loads for a specified ANC limit from the individual catchments within the total population aggregated to an acid-sensitivity category. Given that, the EPA would develop a deposition metric and associated tradeoff curve that represented the percentage of the catchments that would achieve the ANC (DLo/oECo). Therefore a judgment would also need to be made to determine the March 2010 148 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 percentage of ecosystems that would be targeted to achieve a specified ANC level that applies to 2 a bin/category. 3 The following discussions in this section focus on the ecological components of the 4 standard (ecological indicator, the deposition metric, the ecological response function and its 5 modifiers). Questions that are relevant to the design of the standard are used to organize these 6 discussions. The first series of questions (section 5.2.1) considers information presented in the 7 ISA and REA relevant to the conceptual design, while the second series of questions (section 8 5.2.2) presents the proposed conceptual design in more detail with an example calculation based 9 on the Adirondacks case study presented in section 5.5. 10 5.2.1 Conceptual design considerations from the ISA and REA 11 This section presents discussion of the ecological components of the design based on 12 information in the ISA and REA. The information presented here is considered in the 13 development of the design options that are proposed (section 5.2.2). 14 5.2.1.1 Does the available information provide support for the use of ecological 15 indicators to characterize the responses of aquatic ecosystems to nitrogen and 16 sulfur deposition ? 17 Ecological indicators of acidification in aquatic ecosystems can be chemical or 18 biological components of the ecosystem that are demonstrated to be altered by the acidifying 19 effects of N and S deposition based on scientific evidence. A desirable ecological indicator for 20 aquatic acidification will be one that is measurable or estimable, linked causally to deposition of 21 N and S, and linked causally to ecological effects known or anticipated to adversely affect public 22 welfare. 23 As summarized in Chapter 2, aquatic acidification is indicated by changes in the surface 24 water chemistry of ecosystems. In turn, the alteration of surface water chemistry has been linked 25 to negative effects on the biotic integrity of freshwater ecosystems. There are a suite of chemical 26 indicators that can be used to assess the effects of acidifying deposition on lake or stream acid- 27 base chemistry. These indicators include acid neutralizing capacity (ANC), surface water pH and 28 concentrations of SC>42", NCV, Al, and Ca2+; the sum of base cations; and the recently developed 29 base cation surplus. ANC is the most widely used chemical indicator of acid sensitivity and has 30 been found in various studies to be the best single indicator of the biological response and health March 2010 149 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 of aquatic communities in acid-sensitive systems (Lien et al., 1992; Sullivan et al., 2006). The 2 utility of the ANC criterion lies in the association between ANC and the surface water 3 constituents that directly contribute to or ameliorate acidity-related stress, in particular pH, Ca2+, 4 and Al. ANC is also used because it integrates overall acid status (ISA 3.2.3 and REA 5.2.1) and 5 the acid-related stress for biota that occupies the water that can be directly related to biological 6 impairment, specifically the number offish species (ISA 3.2.3). 7 EPA staff thus concludes that the available information provides support for the use of 8 ecological indicators to characterize the responses of aquatic ecosystems to nitrogen and sulfur 9 deposition, and that ANC is the most supportable indicator. 10 5.2.1.2 Does the available information provide support for the development of a 11 function that relates total nitrogen and sulfur deposition to ecological 12 indicators? 13 There is evidence to support the link between deposition of N and S, water chemistry and 14 biota. Atmospheric deposition of NOX and SOX causes aquatic acidification through the input of 15 acid anions (e.g. N(V and SC>42") The anions are deposited either directly to the aquatic 16 ecosystem, or indirectly via terrestrial ecosystems. In other words, when the anions are mobile in 17 the terrestrial soil, they can leach into adjacent waterbodies. Acidification of ecosystems is 18 reflected in a robust relationship between ANC of water and the deposition of NOX and SOX. 19 In the REA, the relationship between deposition and ANC was investigated using models 20 of ecosystem acidification (REA Chapter 4 and REA Appendix 4). These models characterize 21 the relationship between deposition N and S and the ability of an ecosystem to counterbalance or 22 buffer the deposition. The utility of the ecosystem acidification models is for simulating a variety 23 of water and soil acidification responses at the laboratory, plot, hillslope, and catchment scales. 24 For example, the ANC value caused by the current amount of deposition could be calculated, or, 25 the level of deposition that causes a specified level of an ecosystem endpoint could be calculated 26 (i.e. a critical load for ANC=50) (ISA appendix A). 27 The models used in the REA were the Steady State Water Chemistry model (SSWC), the 28 First-order Acid Balance model (FAB) and the Model of Acidification of Groundwater in 29 Catchment (MAGIC). The SSWC and FAB models were used to calculate critical loads for 30 specified ANC levels in the case study areas. MAGIC was used to develop weathering rates that March 2010 150 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 4 5 6 7 were needed for the Shenandoah critical loads calculation and the F-factor was used for weathering rates in the in the Adirondacks. MAGIC was also used to show long-term trends between anthropogenic N and S deposition on ANC dating back to pre-industrial times. It is important to note that acidification models are data intensive. Water chemistry data from the TIME and LTM programs, which are part of the Environmental Monitoring and Assessment Program (EMAP), were input to the acidifcation models. An abbreviated summary of acidification models and data inputs is given in Table 5.2-1, a complete list is in Appendix A. Table 5-1. Illustration of how selected models and water chemistry data were used to calculate critical loads in the REA. Adirondack Shenandoah Weathering rate as input to CL model F-factor MAGIC Water chemistry data input to CL model EMAP EMAP CL calculation: single value sswc sswc CL calculation: critical load function FAB FAB 9 In summary, the EPA staff concludes that the available information supports using the 10 acidification models to characterize the relationship between total nitrogen and sulfur deposition 11 and the ANC ecological indicator. 12 5.2.1.3 Does a quantified relationship exist between the level of a relevant ecological 13 indicator to an amount of nitrogen and sulfur deposition? 14 A quantified relationship exists between the level of ANC and nitrogen and sulfur 15 deposition. This relationship was analyzed to determine current risk for two case study areas, the 16 Adirondacks and Shenandoahs, in the PvEA using a time series analysis and a critical load 17 approach. The time series analysis was conducted using MAGIC and recent monitoring data. The 18 critical loads analysis used water chemistry data from the Temporally Integrated Monitoring of 19 Ecosystems (TIME) program and Long-term Monitoring (LTM) to calculate critical loads with the 20 SSWC and FAB models. 21 Long-term trends in surface water nitrate, sulfate and ANC were modeled using MAGIC 22 for the two case study areas. This data was used to compare current surface water conditions 23 (2006) with preindustrial conditions (i.e. preacidification or 1860). The results showed a March 2010 151 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 dramatic increase in the number of acidified lakes, characterized as a decrease in ANC levels, 2 since the onset of anthropogenic N and S deposition (REA Appendix 4 Section 5) 3 More recent trends in ANC, over the time period from 1990 to 2006, were assessed using 4 monitoring data collected at the two case study areas. In both case study areas, nitrate and sulfate 5 deposition decreased over this time period. In the Adirondacks, this corresponded to a decreased 6 concentration of nitrate and sulfate in the surface waters and an increase in ANC (REA 4.2.4.2). 7 In the Shenandoahs, there was a slight decrease in nitrate and sulfate concentration in surface 8 waters corresponding to modest increase in ANC from 50 ueq/L in 1990 to 67 ueq/L in 2006 9 (REA 4.2.4.3 and REA Appendix 4 Section 3.4) 10 A critical load for ANC is the amount (or load per year) of N and S deposition above 11 which a selected level of ANC will be exceeded for individual water bodies. In the REA case 12 study analyses, critical loads and their exceedances were calculated for four values of ANC (i.e., 13 ANC of 0, 20, 50, and 100 ueq/L) for 169 lakes in the Adirondacks and 60 streams in the 14 Shenandoahs. Those four ANC values correspond to important points along the ANC response 15 curve that are associated with levels of ecosystem impairment. The case studies used steady-state 16 critical loads models and focus on the combined load of sulfur and nitrogen deposition, below 17 which the ANC level would still support healthy aquatic ecosystems. For each waterbody, the 18 total deposition in the year 2002 was compared with the estimated critical loads for the four 19 critical limit thresholds to determine which sites exceed their critical limit of deposition and 20 biological protection level. Estimates of deposition were based on the sum of measured wet 21 deposition values from the year 2002 NADP network and modeled dry deposition values based 22 on the year 2002 emissions and meteorology using the Community Multiscale Air Quality 23 (CMAQ) model, respectively (REA 4.2). It is important to note that a single level of ANC may 24 be caused by a range of deposition values due to heterogeneous sensitivity among watersheds. 25 In summary, EPA staff concludes that a quantified relationship exists between the level 26 of surface water ANC and an amount of nitrogen and sulfur deposition. This relationship is 27 demonstrated by long-term trends going back to preindustrial conditions in the 1860s, recent 28 trends since the 1990s and critical loads modeling based on 2002 deposition data. Models are the 29 best way to evaluate how multiple environmental factors alter the relationship ANC and 30 deposition. March 2010 152 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 5.2.1.4 What are the important variables in the ecological response to deposition 2 relationship^)? 3 There are numerous variables that modify the ANC to deposition relationship. The effects 4 of these modifiers are described by models that parameterize ecosystems to simulate the process 5 of acidification. The steady-state models used for critical loads analysis in the REA required 6 input data for between 17 and 20 environmental parameters. 7 The basic principle of the steady-state approach is to determine the maximum acid input 8 that will balance the system at a biogeochemical safe-limit. Safe-limit is a subjective term that 9 relates to a particular benchmark (e.g. ANC = 20, 50, 100), representing protection against 10 specific types and magnitudes of aquatic ecosystem response. The steady-state models that were 11 used in the REA relate an aquatic ecosystem's critical load to the weathering rate of its drainage 12 basin expressed in terms of the base cation flux. Weathering rate of geologic parent material is 13 the main source of base cations to an ecosystem. It is considered one of the governing factors to 14 ecosystem critical loads, and therefore an important variable in the ecological response to 15 deposition relationship. Landscape features that are correlated to ecosystem acid-sensitivity 16 include lithology, elevation, percent forested watershed, and watershed area (Sullivan et al. 17 2007). A more detailed summary of the models and the environmental variables incorporated 18 into the models that were used in the REA is presented in Appendix A. 19 Numerous environmental variables affect the acidification process. Therefore the input 20 data required to run acidification models is rather extensive. For example, MAGIC, a dynamic 21 process based model of acidification, requires atmospheric deposition fluxes for the base cations 22 and strong acid anions as inputs to the model. The volume discharge for the catchment must also 23 be provided to the model. Values for soil and surface water temperature, partial pressure of 24 carbon dioxide and organic acid concentrations must also be provided at the appropriate 25 temporal resolution. The aggregated nature of the model requires that it be calibrated to 26 observational data from a system before it can be used to examine potential system response. The 27 calibration procedure requires that stream water quality, soil chemical and physical 28 characteristics, and atmospheric deposition data be available for each catchment. The water 29 quality data needed for calibration are the concentrations of the individual base cations (Ca, Mg, 30 Na, and K) and acid anions (Cl, SC>42", and NCV) and the pH. The soil data used in the model 31 include soil depth and bulk density, soil pH, soil cation-exchange capacity, and exchangeable March 2010 153 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 bases in the soil (Ca, Mg, Na, and K). The atmospheric deposition inputs to the model must be 2 estimates of total deposition (wet and dry). 3 In summary, the EPA staff concludes there are numerous variables which modify the 4 ANC to deposition relationship. The relationships between environmental factors are described 5 by models that parameterize ecosystems to simulate the process of acidification. Weathering rate 6 of geologic parent material is the main source of base cations to an ecosystem, and it is therefore 7 considered one of the governing factors of ecosystem critical loads. Landscape features that are 8 correlated to ecosystem acid-sensitivity include lithology, elevation, percent forested watershed, 9 and watershed area. Consideration of the effects of environmental variables on the relationship 10 between environmental variables is extensive in ecosystem acidification models. The calibration 11 procedure requires that stream water quality, soil chemical and physical characteristics, and 12 atmospheric deposition data be available for each catchment. 13 5.2.1.5 Are these relationships applicable regionally ? 14 The relationship between ANC and N + S deposition based on catchment- scale modeling 15 is applicable regionally. Response to N and S deposition will vary catchment by catchment. 16 However, modeling every catchment in a region (i.e. a spatial area that includes a large 17 population of individual catchments) is implausible due to the extensive data requirements to 18 inform the simulations. A method to extrapolate watershed-scale analysis to a region was 19 developed in the REA. In that method, the critical loads (combined N+S load) developed for the 20 case study sites were applied over a region using water quality data. Critical load exceedance 21 (i.e., the amount of actual deposition above the critical load, if any) was calculated for each 22 waterbody in the region to quantify the number of lakes or streams that receive harmful levels of 23 deposition. Lakes and streams with positive exceedance values, where actual deposition was 24 above its critical load, were not protected at that critical limit (e.g. ANC= 20, 50, 100; REA 25 appendix 4). 26 In the Adirondack case study conducted in the REA, critical load exceedances were 27 extrapolated to lakes defined by the New England EMAP probability survey. The EMAP 28 probability survey was designed to estimate, with known confidence, the status, extent, change, 29 and trends in condition of the nation's ecological resources, such as surface water quality. For 30 the Adirondack Case Study Area, the regional EMAP probability survey of 117 lakes were used March 2010 154 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 to infer the number of lakes and percentage of lakes that receive acidifying deposition above 2 their critical load of a population of 1,842 lakes. ANC limits of 20, 50, and 100 ueq/L were 3 examined. 4 In the Shenandoah case study, critical load exceedances were extrapolated using the 5 SWAS-VTSSS LTM quarterly monitored sites to the population of brook trout streams that do 6 not lie on limestone bedrock and/or are not significantly affected by human activity within the 7 watershed. The total number of brook trout streams represented by the SWAS-VTSSS LTM 8 quarterly monitored sites is approximately 310 streams out of 440 mountain headwater streams 9 known to support reproducing brook trout. ANC limits of 20, 50, and 100 ueq/L were examined. 10 (REA Appendix 4.3.1). 11 In summary, approaches were developed in the REA to extrapolate the ANC-deposition 12 relationship across a region. The data requirements for these approaches include (1) calculation 13 of critical loads of ANC using a catchment-scale model (2) stream chemistry data across the 14 region of concern, and (3) deposition loads across the region. With this information the 15 deposition load that would cause the stream to exceed the critical limit of ANC was calculated as 16 the critical load exceedance. 17 5.2.1.6 Are these relationships applicable nationally ? 18 The relationship between ANC and N + S deposition is applicable nationally. Areas that 19 have similar geologic underpinnings and weathering rates should show similar sensitivity to NOX 20 and SOX deposition. The critical load modeling that was used in the REA case studies requires 21 parameterization to each catchment. The spatial scale is small (e.g. catchment level) and the data 22 requirements are great (17+ environmental variables for each catchment) to use this method to 23 determine critical loads across all sensitive regions of the U.S. at this time. It is important to note 24 that acid-sensitivity often varies from catchment to catchment. Even if we did calculate critical 25 loads data for each catchment, aggregation of the catchment-scale data is appropriate for a 26 national standard. 27 The technique developed in the REA for extrapolating catchment-specific results to a 28 regional area determines the number of streams in a given area that show critical load (CL) 29 exceedances based on a selected value of ANC and deposition values for 2002. The approach 30 developed in the case study for extrapolating catchment-specific results to a regional area is not March 2010 155 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 immediately applicable across the U.S. because data for surface water chemistry and data for 2 other input parameters is not available at a national scale. 3 To summarize, the relationship between ANC and N + S deposition is applicable 4 nationally. However the data required for critical loads analysis and extrapolation that is 5 available on the regional scale is not available at the national scale. Considering this current data 6 limitation the utility of the extrapolation approach developed in the REA to the national-scale is 7 limited. Additional national-scale approaches are discussed in section 5.2.3. 8 5.2.1.7 Summary 9 In summary, EPA staff concludes that the available information from the ISA and REA 10 supports the following characterization of aquatic acidification. First, there is sufficient support 11 for the use of ecological indicators to characterize the responses of aquatic ecosystems to 12 nitrogen and sulfur deposition, and that ANC is the most supportable indicator. The available 13 information supports using the acidification models to characterize the ecological response, using 14 ANC as the indicator, to nitrogen and sulfur deposition. Models are the best way to evaluate how 15 multiple environmental factors alter the relationship ANC and deposition. 16 Heterogeneous sensitivity among watersheds is due in part to landscape features. 17 Weathering rate of geologic parent material is the main source of base cations to an ecosystem, 18 and is therefore considered one of the governing factors of ecosystem critical loads. Landscape 19 features that are correlated to ecosystem acid-sensitivity include lithology, elevation, percent 20 forested watershed, and watershed area. 21 Modeling every catchment in a region is implausible due to the extensive data 22 requirements. The relationship between ANC and N + S deposition is applicable regionally. A 23 method to extrapolate watershed-scale analysis to a region was developed in the REA. In that 24 method, the critical loads (combined N+S load) developed for the case study sites were applied 25 over a region using water quality data. The data requirements for the regional extrapolation 26 include (1) calculation of critical loads for ANC using a catchment-scale model (2) stream 27 chemistry data across the region of concern, and (3) deposition loads across the region. The 28 approach developed in the case study areas is not immediately applicable across the U.S. because 29 data for critical loads modeling and surface water chemistry is not available at a national scale. 30 However, it is important to note that the relationship between ANC and N+S deposition is March 2010 156 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 applicable nationally. Areas that have similar geologic underpinnings should show similar 2 sensitivity to NOX and SOX deposition. 3 5.2.2 Design options for aquatic acidification 4 The following design options describe the conceptual approach to integrating the 5 ecological components of the standard outlined in section 5.1: ecological indicator, modifying 6 factors, ecological response function and deposition metric. The goal is to illustrate how levels of 7 NOX and SOX can be set to protect areas of the U.S. from acidic deposition. 8 5.2.2.1 Is it appropriate to use ANC as the ecological indicator for the conceptual 9 design of the NOX and SOX standard based on aquatic acidification ? 10 There is strong evidence supporting that ANC is an appropriate ecological indicator for 11 aquatic acidification as discussed in Chapter 2 and Section 5.1.1 (as well as ISA 3.2.3 and REA 12 5.2.1). Options for the level of the indicator are discussed in Chapter 6. The options for the levels 13 are derived from experimental and observed evidence in the scientific literature showing the 14 biological effects over a range of ANC values. 15 5.2.2.2 What is the appropriate ecosystem acidification model(s) to represent the 16 ecological response function ? 17 In the REA, critical loads were calculated for specified ANC levels using the SSWC and 18 FAB models, these are referred to as acidification models, acid balance models or critical loads 19 models. The different assumptions of each modeling approach have implications that should be 20 considered in the conceptual design of a deposition-based NOX and SOX standard. Most notably, 21 biogeochemical pathways of N and S deposition are considered differently in the two models. In the 22 SSWC model, sulfate is assumed to be a mobile anion (i.e. S leaching = S deposition), while nitrogen is 23 retained in the catchment by various processes. This assumption that all N is retained by the ecosystem 24 and does not contribute to acidification is incorrect in many instances because nitrate leaching is 25 observed. If nitrogen is leaching out of an ecosystem, obviously it has not been retained. Nitrate leaching 26 is determined from the sum of the measured concentrations of nitrate and ammonia in the runoff. The 27 critical load for sulfur that is calculated by SSWC can be corrected for the amount of nitrogen that 28 contributes to acidification. When an exceedence value for the critical load is calculated, the critical load 29 is subtracted from S deposition plus the amount of nitrate leaching, as it represents the difference between March 2010 157 Draft - Do Not Quote or Cite ------- 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX N deposition and N retention by the ecosystem. N leaching data used in this calculation is considered robust. In contrast to the SSWC approach, the FAB model includes more explicit modeling of N processes including soil immobilization, denitrification, and wood removal, in-lake retention of N and S, as well as lake size. Although N cycling is more detailed in the FAB model, there is greater uncertainty in the input data needed to characterize the components of the N cycle. The FAB model yields a deposition load function for a specified level of an endpoint. This function is characterized by three nodes that are illustrated on Figure 5-3: 1) the maximum of amount of N deposition when S deposition equals zero (DLmax (N)), 2) the amount of N deposition that will be captured by the ecosystem before it leaches (DLmin(N)) and 3) the maximum amount of S sulfur deposition considering the N captured by the ecosystem (DLmax (S)). The function represents many unique pairs of N and S deposition that will equal the critical load for acidifying deposition. The slope portion of the function will vary according to attributes of the water body that is modeled, including lake size and in-lake retention. H exposition Figure 5-3. The depositional load function. A third modeling approach, which synthesizes components of each model used in the REA, is suggested by staff for catchment scale modeling in developing the NAAQS. The foundation of the proposed approach is the SSWC model because there is high confidence in the input data required. The SSWC model for aquatic acidification is expressed as equation 1. (1) 22 where, March 2010 158 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 DLANciim(N+S) = depositional load of S and N that does not cause the ecosystems to exceed a 2 given ANCiim 3 [BC]0* = the preindustrial concentration of base cations (equ/L) 4 ANCumit = a "target" ANC level (equ/L) 5 Q= surface water runoff (m/yr) (this is typically equal to precipitation -evapotranspiration 6 7 This model could be further constrained by a quantity of N which would which would be 8 taken up, immobilized or denitrified by ecosystems and adjust the quantity of deposition required 9 to meet a specified critical load. This term is represented as DLmin(N) in the FAB model and 10 illustrated in Fig. 5-3. For application in the NAAQS and in the following discussion, the 1 1 parameter is designated with the abbreviation NEco- The acid-base model constrained by NEco is 12 expressed by equation 2. 13 DLANClimN + S=(BC -ANC^ + N^ (2) 14 1 5 where, 16 Neco= nitrogen retention and denitrification by terrestrial catchment and nitrogen retention in the 17 lake 18 19 The term Neco could be derived multiple ways, each yielding different ultimate results. 20 The first is by taking the mean value calculated to represent the long-term amount of N an 21 ecosystem can immobilize and denitrify before leaching (i.e. N saturation) that is derived from 22 the FAB model [denoted as DLmin(N) in the FAB model]. This approach requires the input of 23 multiple ecosystem parameters. Its components are expressed by eq 3. 24 Neco = fNupt + Nret + (l - r \Nmm + Nden ) (3) 25 where, 26 Nupt= nitrogen uptake by the catchment 27 Nimm= nitrogen immobilization by the catchment 28 Nden=denitrification of nitrogen in the catchment, March 2010 159 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 N-et = in-lake retention of nitrogen 2 f =forest cover in the catchment (dimensionless parameter) 3 r = fraction lake/catchment ratio (dimensionless parameter) 4 5 The second approach for estimating Neco is to take the difference between N deposition 6 and measured N leaching in a catchment as expressed by eq 4. 7 Neco=DL(N)-Nleach (4) 8 It is unclear which approach for calculating NECO should be used in developing the 9 NAAQS. The two equations can result in quite different values (See section 5.4 for an example 10 calculation). 11 To summarize, the SSWC model assumes N deposited to the ecosystem is retained by the 12 ecosystem, while also assuming that all S deposition is leached and contributes to aquatic 13 acidification. The critical load is calculated for S deposition, and the N that contributes to 14 acidification is incorporated into the exceedance calculation. The FAB model considers a 15 detailed accounting of the N cycle; however confidence in the input data to the model is more 16 uncertain. The FAB approach yields a function which may be solved by many unique pairs of N 17 and S deposition. A minimum amount of N deposition that will be captured by the ecosystem 18 before it leaches is included in the calculation of the maximum amount of S deposition. A third 19 approach is suggested by staff as the most appropriate approach for informing the structure of the 20 NOX and SOX secondary standard. This approach constrains the critical load calculated from a 21 SSWC method by a value of NEco [previously defined as DLmin(N)] which accounts for the 22 amount of N deposition that would be taken up by the ecosystem and, therefore, would not 23 contribute to acidification. 24 5.2.2.3 How are results of acidification models aggregated to adequately represent a 25 larger spatial area and inform a deposition metric? 26 So far in this section, the ecological indicator would be established as ANC. Acidification 27 models are considered the best way to describe the relationship between ANC and deposition and 28 to describe how this relationship is altered by modifying factors. If deposition is known the 29 model may be run to calculate the resultant ANC. If a target ANC level is desired the model may March 2010 160 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 be run to calculate the corresponding deposition load that should not be exceeded (i.e. the critical 2 load). The following discussion will focus on the critical load application of the acidification 3 model. It is important to emphasize that the acidification models are only applied at the spatial 4 scale of the catchment. Spatial aggregation of critical loads are necessary to inform the 5 discussion of appropriate design and levels of a national standard. 6 Acidification models are parameterized for catchments. The critical loads that they 7 calculate for N and S deposition based on a specified ANC limit vary at the small spatial scale of 8 the catchment to the degree that acid-balancing properties of the catchments vary. Despite this 9 variation, the goal of aggregating critical loads from multiple catchments is to develop an 10 appropriately representative deposition value, which is adequately protective of ecosystems and 11 could be applied over larger spatial areas. 12 Staff proposes evaluating the critical loads for a specified ANC limit of a population of 13 waterbodies to calculate a benchmark deposit!onal load in which a specified percentage of the 14 population does not exceed their critical load. This approach uses the distribution of critical loads 15 from a population to derive a value that is intended to provide protection over a spatial area that 16 is larger than the individual catchment for which a single critical load may be calculated. An 17 example of this technique is calculated in section 5.5. The ecological indicator would be a single 18 value of ANC, and the acidification models would calculate the critical loads for the specified 19 ANC level for individual catchments across a spatial area. The deposition metric would be an 20 amount of deposition such that a specified percentage of a population of water bodies does not 21 exceed a critical load for the specified value of ANC. The deposition metrics could be calculated 22 for populations of catchments that are categorized according to acid-sensitivity, as described in 23 the next section. 24 5.2.2.4 How are modifying factors of the ecological response to deposition function 25 considered at the national-scale? 26 As previously noted, critical loads for ANC vary at a small spatial scale, catchment by 27 catchment. As it is implausible to model the acidification status of every catchment in the U.S, 28 an alternative is to develop a deposition metric for a population of catchments, assuggested in the 29 previous section. The following design options focus on relating acid-sensitivity, based on ANC, 30 to a feature(s) of the landscape at a national-scale by creating acid-sensitivity categories. A March 2010 161 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 population of catchments could then be defined to represent these categories and a representative 2 deposition metric chosen. 3 Acid sensitivity classes based on bed rock geology 4 Here an approach is presented in which ecosystem sensitivity to acidification is 5 categorized into classes based on bedrock geology/ lithology. The approach is supported by 6 conclusions from the ISA in which geologic bedrock is determined to be the governing factor 7 that drives ecosystem sensitivity to acidification (ISA 3.2.4.1). Specifically, geologic bedrock 8 with a low base cation supply leads to ecosystems that are sensitive to acidifying deposition. A 9 method to develop a deposition metric, based on the distribution of critical loads of a 10 representative population, for each category of acid-sensitivity is presented here. 11 A map was developed to capture the heterogeneity of geologic bedrock that occurs across 12 the eastern U.S. and link it to ecosystem acid-sensitivity (Fig5-4). The method is based on 13 Sullivan et al.(2007) in which 70+ primary lithologies are grouped into 5 categories of acid- 14 sensitivity, using ANC as the ecosystem indicator upon which acid-sensitivity is based. Sullivan 15 et al. (2007) evaluated multiple features of the landscape and found that geology is the landscape 16 parameter that governs ecosystem sensitivity to acidic deposition. The analysis in Sullivan et al. 17 2007 was conducted in the Southern Appalachian Mountains region, which included sites from 18 the states of GA, TN, NC, KT, VA and WV. EPA is conducting additional analyses to further 19 test the concept that lithology correlates to acid sensitivity in case study areas and in the western 20 U.S. EPA staff intends that some of these additional analyses will be available at for review in 21 the second draft of the policy assessment. 22 As previously stated, acidification often varies catchment by catchment. Therefore there 23 will be variation in terms of acid-sensitivity among catchments within each acid-sensitivity class 24 designated by the map. Despite this variation, lithology is a nationally applicable landscape 25 feature which is known to govern acid-sensitivity. Ultimate detail and rigor would be provided 26 by modeling deposition and consequential acidification of each catchment in the U.S., an 27 approach which would require knowledge of 17+ environmental parameters for each catchment. 28 However classification of the landscape into categories based on geology provides a national - 29 scale landscape feature to extrapolate the results of catchment-scale modeling. March 2010 162 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX Acid-Sensitive Areas of the Eastern United States A Classification based on Bedrock Geology L'S Environmental Protection Agency Projection Alter*EquwAr F*]t4_e*iMig 0 OOOOCO FUM fMXTiH-g 000003? C*ntr« M About the Classification Tn« nwp !* MM on • 250.000 ttMtoek Q.MW uu from OM US O»=**jC8i Su-v»y. rl Mt M*n ti*i lifted Mwd on « mwrnod .n Sunvin *l M ..2006,1. --im vw [ | Slate Boundaries Acid Sensitivity ^^| Cartxinale - Least Sensitive Silaceous - Less Sensilive Argillaceous • Sensitive ~'j Felsic - Water 1 2 Fig 5-4. A map of acid sensitive areas of the Eastern U.S. developed from a 3 lithology-based five-unit geologic classification system after methods in Sullivan 4 etal. (2007). 5 Acid sensitivity based on multiple landscape features 6 Although bedrock geology is a governing factor of acid sensitivity, multiple factors have 7 been shown to contribute to sensitivity. Topography is a characteristic of the landscape that is 8 often shown to correlate with acid-sensitivity, specifically low elevations, which generally 9 receive some cations from higher elevation sites, are less sensitive that higher elevation sites 10 (ISA 3.2.4.1). Could both topography and bedrock geology be included a national map of acid- 11 sensitivity? A map of high elevation could be layered over the map of bedrock categories. If all 12 high elevation areas were within the sensitive geologic categories, then the additional parameter 13 would further refine the spatial resolution of sensitivity within the bedrock categorization. 14 Moreover, the approach will provide more spatial detail on the sensitivity within areas already 15 considered sensitive based on bedrock geology. It's unclear if elevation alone would help March 2010 163 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 identify new sensitive areas. It's also unclear if greater spatial resolution of sensitivity within 2 areas already identified as sensitive would be helpful in terms of relating the national-scale 3 landscape features to critical loads. Should additional multiple features be considered when 4 categorizing the landscape according to acid-sensitivity? We are providing this design option to 5 elicit comment; it is presented as a conceptual idea. 6 5.2.2.5 How is a deposition metric developed so that critical loads for catchments are 1 aggregated to adequately represent classes of acid sensitivity based on 8 geology? 9 The values that represent a deposition metric for the acid-sensitivity categories could be 10 derived from the critical loads calculated for the case study analysis in the REA. The case study 11 sites (Adirondack and Shenandoah areas) occur in areas that are predominately composed of the 12 two most acid-sensitive types of bedrock geology. Therefore the case study sites would represent 13 those sensitivity categories. The deposition and atmospheric concentration tradeoff curves for a 14 specified level of ANC for each bedrock geology site would be based on a deposition metric 15 derived from the distribution of critical loads within the case study areas. It could be a central 16 value such as the mean or median value or a value representing a percentile of the distribution, 17 such as the 95th percentile. Central estimates, such as the mean, would likely not be projected to 18 achieve the target ANC of the majority of acid-sensitive ecosystems; therefore it may be 19 preferable to calculate the spatially aggregated value for some percentage of catchments to 20 project achieving the ANC for the more sensitive ecosystem types. For example, if projecting 21 85%, 90% or 95% of the aquatic ecosystems achieving the ANC is selected, then the deposition 22 metric that represents the critical load for the 85th, 90th or 95th percentile of the population would 23 be selected. An example calculation for the Adirondacks is presented in section 5.5. 24 5.2.2.6 How is reduced nitrogen appropriately considered in the deposition metric? 25 Reduced forms of nitrogen deposition are quickly converted to nitrate in the environment 26 and use up the assimilative capacity of ANC at the same rate as oxidized forms of nitrogen 27 deposition; therefore, reduced nitrogen deposition must be accounted for in the watershed. There 28 are two basic approaches to accounting for the use of this assimilative capacity. 29 The suggested approach is to subtract the loadings of reduced forms of nitrogen derived 30 for a given spatial area from the deposition metric that represents selected percentage of critical March 2010 164 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 loads for a given population, such that the resultant deposition metric is for sulfur and oxidized 2 nitrogen only. This approach assumes that the reduced forms of nitrogen deposition are relatively 3 constant over time. This assumption could lead to over or under protection for an area depending 4 on whether the actual concentrations of reduced forms of nitrogen increase or decrease over 5 time. An example for how to subtract reduced nitrogen from the deposition metric based on 6 nitrogen and sulfur is given in section 5.5. 7 5.2.2.7 Summary 8 In summary, the ecological components of the conceptual design for a standard base on 9 aquatic acidification include the ecological indicator, ecological response function and its 10 modifiers and the deposition metric. A summary how each component is considered in the 11 conceptual design is given in Table 5-2. Using ANC as the ecological indicator, an approach is 12 suggested for using an acidification model constrained by a parameter for ecosystem N retention 13 to represent the ecological response function. The best way to calculate ecosystem N retention is 14 as of yet unclear. It is proposed that the national landscape is categorized in terms of criteria that 15 denote acid-sensitivity. It is well known that bedrock geology is a governing factor of acid- 16 sensitivity, in other words ecosystem response is modified across the landscape due in part to 17 bedrock geology. It is unclear if landscape categorization based on geology is the best approach 18 or other criteria/combination of criteria should be used. 19 The distribution of critical loads for a specified target ANC from a population of 20 catchments representing an acid-sensitivity category, based on geology or some combination of 21 factors, can be calculated From this a deposition metric, an amount of deposition, could be 22 calculated such that a specified target percentage of the population of water bodies in the acid- 23 sensitivity category does not exceed a critical load for the specified value of ANC. Moreover, the 24 deposition metric would reflect both the selected level of ANC and the percentage of catchments 25 in the representative population that do not exceed their critical load. Reduced nitrogen 26 deposition, average over a determined spatial scale, would be subtracted from the deposition 27 metric yielding a value for allowable deposition from NOy and SOX. The deposition from NOy 28 and SOX would be converted to atmospheric concentrations of NOy and SOX by the methods 29 described in section 5.4. March 2010 165 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX Table 5-2. Summary of the ecological components of design option 1. Values given for illustrative purposes only. Levels are discussed in Chapter 6 and ultimately selected by the administrator. Modifying Factor Geology 5 categories of sensitivity Variable/Fixed Modifying Factors Deposition Metric Ecological Response to Deposition Function Ecological Indicator Atmospheric Deposition Transformation Function See Section 5.4 Determined by the % of Acidifcation mode ecosystems represented Proposed levels based on biological effects Concentration of Air Quality Indicator(s) Ecological Indicator Ecological Response Function Modifying Factor Deposition Metric ANC; level reflects degree of Effects on aquatic biota in the ecosystem Acidification model constrained by a parameter for N retention Acid-sensitivity categories, based on geologic bed rock or a combination of factors, that may be applied at a national scale Determined from the distribution of critical loads from a population that can be related to an acid- sensitivity category. Reduced nitrogen subtracted from the deposition metric to yield allowable deposition from NOX and SOX. March 2010 166 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 5.3 ECOLOGICAL COMPONENTS OF THE STANDARD: 2 TERRESTRIAL ACIDIFICATION, TERRESTRIAL NUTRIENT 3 ENRICHMENT AND SURFACE WATER NUTRIENT 4 ENRICHMENT 5 These effects were not included in the conceptual design for the first draft of the PA, 6 however a brief summary of our approach for developing standards that are protective of these 7 ecological effects follows. 8 5.3.1 Terrestrial Acidification 9 The deleterious effects of terrestrial acidification on tree species is indicated by base 10 cation to aluminum ratio (Be: Al) of soils. Critical load functions were developed in the REA that 11 relate Bc:Al threshold values (0.6, 1.2 and 10) to values of N+S deposition using the simple mass 12 balance (8MB) model. The exceedance of these critical loads were calculated at the two study 13 sites and then extrapolated over 24 states. Like aquatic acidification, sensitivity of terrestrial 14 ecosystems to acidification is linked to the geologic bedrock. Moreover, areas that are sensitive 15 to aquatic acidification should also be sensitive to terrestrial acidification. Therefore, an 16 approach similar to that described for aquatic acidification could be developed. This would mean 17 that a critical load based on Bc:Al at either 1.2 or 10 would be calculated to protect a percentage 18 of the terrestrial landscape. This value would then be assigned to categories of acid sensitivity 19 based on geology. 20 This could result in two standards, one for aquatic ecosystems and one for terrestrial 21 ecosystems. This leads to the question, are aquatic or terrestrial ecosystem more sensitive? To 22 answer this question, an analysis was conducted in which critical loads for the Adirondacks and 23 Shenandoah case study areas were calculated based on the terrestrial ecosystem indicator, Be: Al, 24 at the level of 1.2 and 10. The terrestrial critical loads were compared to the critical loads for 25 aquatic ecosystems. A full description of this analysis and results is available in Chapter 7, the 26 results are briefly summarized here. In the Adirondacks case study area, 7 of the 16 watersheds 27 had terrestrial critical acid loads (based on a Bc:Al of 10.0) that were lower and therefore more 28 sensitive to acidification than all the lakes in the watershed. However, when the terrestrial critical 29 loads were calculated with a Bc:Al soil solution ratio of 1.2, only 5 of the 16 watersheds were 30 protected by a terrestrial critical load that was lower than the aquatic critical loads of the lakes. In March 2010 167 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 the Shenandoah case study area terrestrial critical loads offered a higher level of protection than 2 aquatic critical loads in only one watershed. If two standards were proposed, the one that allows 3 lower ambient levels of NOX and SOX would be controlling in a given area. 4 5.3.2 Terrestrial and surface water nutrient enrichment 5 NOX and NHX are the main contributors to nitrogen deposition. The effects of nitrogen 6 deposition on terrestrial ecosystems and surface waters are many. Most notable are the effects on 7 ecosystem biodiversity found across the U.S and affecting multiple taxonomic groups including 8 vascular plants, algae, mycorrhiza and lichens (ISA 3.3). Unlike terrestrial and aquatic 9 acidification, there is no one, well-supported chemical or biological indicator of ecosystem 10 effects that occurs across the nation. In order to develop a NAAQS based on nitrogen enrichment 11 effects there needs to be one indicator that can be applied across the nation. It is possible that we 12 could develop an index in which information on different ecological indicators could be input 13 and the output would be an index score that could be consistently applied across the U.S. It is not 14 clear how to develop such an index. 15 Nitrogen critical loads are known for many ecosystem endpoints in the U.S. and are 16 published in the scientific literature. Additionally, critical loads for ecosystems in Europe, many 17 of which are similar to U.S. ecosystems, have been reported for over a decade, they are 18 continually refined through periodic assessments of the scientific literature, and they are 19 currently supported by a strong weight of peer-reviewed scientific information (ISA 3.3). 20 Additional critical load modeling was not conducted in the REA because of two factors. There 21 are numerous reports in the peer-reviewed scientific literature and there is no model available to 22 conduct such analysis for multiple endpoints and ecosystems. However, based on nitrogen 23 critical loads published in the literature, the REA evaluated the extent of the landscape 24 represented by those critical loads and their exceedances (REA 5.0). 25 A standard that integrates acidification and nutrient effects could conceptually be quite 26 simple. The total nitrogen deposition allowed for a deposition metric based on acidification could 27 be constrained so that it does not exceed a value based on a deposition metric for a nutrient 28 related effect. March 2010 168 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 5.3.3 Summary 2 Conceptual design of NOX and SOX NAAQS were not developed for terrestrial 3 acidification and terrestrial/surface water nitrogen enrichment in the first draft PA, however a 4 brief summary of a potential structure for these ecological effects is presented. The ecological 5 indicator for terrestrial acidification would be Bc:Al because it relates to both atmospheric 6 deposition of N+S and deleterious effects on tree growth. Critical loads would be related to acid- 7 sensitivity categories and calculated according to similar methods presented for aquatic 8 acidification effects. This could result in two standards, one for aquatic ecosystems and one for 9 terrestrial ecosystems. If two standards were proposed, the one that allows lower ambient levels 10 of NOX and SOX would be controlling in a given area. Unlike terrestrial and aquatic acidification, 11 there is no one, well-supported ecological indicator of nitrogen deposition effects that occurs 12 across the nation. In order to develop a NAAQS based on nitrogen enrichment effects there 13 needs to be one indicator that can be applied across the nation. Although, the specifics of an 14 approach are unclear, it may be possible that we could develop an index in which information on 15 different ecological indicators could be input and the output would be an index score that could 16 be consistently applied across the U.S. A standard that integrates acidification and nutrient 17 effects could conceptually be quite simple. The total nitrogen deposition allowed for a deposition 18 metric based on acidification could be constrained so that it does not exceed a value based on a 19 deposition metric for a nutrient related effect. 20 5.4 LINKING DEPOSITION TO ATMOSPHERIC CONCENTRATION 21 5.4.1 Background 22 Atmospheric pollutants deposit onto land and water surfaces through at least two major 23 mechanisms: direct contact with the surface (dry deposition), and transfer into liquid 24 precipitation (wet deposition). The magnitude of each deposition process is related to the 25 ambient concentration through the time-, location-, process- and species-specific deposition 26 velocity (Seinfeld and Pandis, 1998) and can be conceptualized as: 27 DePlD>y =v1Dry-C,Amb (1) 28 Dep,Wet=v,wret-CiAmb (2) March 2010 169 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Dry . Wet , , , , . . , . . ^ Dry , ^ Wet . . . 1 where vt and v, are the dry and wet deposition velocities, Dept and Dept are the dry and 2 wet deposition fluxes, Cf is the ambient concentration, and the /' subscript indicates the 3 pollutant species under study. The wet deposition velocity term is a conceptualized term and not 4 a state variable that allows for the grouping of wet and dry deposition. The total deposition of 5 each pollutant is .- r^ Tot r^ Dry 7-. Wet ,~\ 6 Dept =Depi +Dept (3) 7 Substituting Equations 1 and 2 into Equation 3 yields 87-. Tot Dry /^< Amb Wet ,~t Amb / A\ DePl =vt v -C, +v,. -Ct (4) 9 The total deposition of sulfur or nitrogen would therefore be: t r\ r-\ Tot ^~^ / Dry Wet \ /-i Amb /c\ 10 Dep^ = ^ (v,. 'y+vi ) • mi • C, (5) ; 11 where m is the molar ratio of the atom (sulfur or nitrogen) of interest to the /'th pollutant. 12 Ambient sulfur- and nitrogen-containing pollutants include gases such as sulfur dioxide (SO2), 13 ammonia (NH3), various nitrogen oxides (NO, NO2, HONO, N2O5), nitric acid (HNO3), and 14 organic nitrates such as peroxyacetyl nitrates (PAN); as well as particulate species such as sulfate 15 (SC>42"), nitrate (N(V), and ammonium (NH4+). As discussed in chapter 4, the definitions of NOy 16 and SOX species for the purposes of this review include the sulfur-containing species above and 17 the above oxidized forms of nitrogen (NOy); ammonia and ammonium are not currently included 18 as listed pollutants (see Chapter 8 for an expanded discussion of the role of NHX). 19 5.4.2 Aggregation Issues 20 Equation 5 provides a relationship for converting sulfur or nitrogen deposition to 21 "equivalent" ambient concentrations,. A major issue to consider during such conversion is the 22 treatment of spatial, temporal and chemical resolutions of the deposition data and the resulting 23 standards. Since the objective is to set an ambient air quality standard for total oxidized sulfur 24 and nitrogen, and this is also the chemical resolution provided by the ecosystem models, it is 25 convenient to use a relationship with the following form: March 2010 170 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX i r\ Tot T T- .-"r A mb / s- \ 1 DePs\N = VSIN • CsiN (6) 2 where VS/N can be considered an aggregateddeposition to ambient air transformation ratio, 3 referred to herein as the deposition transformation ratio, that relates total deposition of sulfur or 4 nitrogen to the total ambient concentration, and represents an average of the species specific v,Tot 5 ( = v,Dry + v,Wet) values in Equation 5. The sulfur and nitrogen concentrations are the result of 6 applying the ni; values to the C;^11 values in Equation 5. 7 Since the deposition critical loads are expressed in terms of annual total deposition, the 8 most relevant averaging time for equivalent ambient concentrations is the annual average. Data 9 used to derive annual VS/N values will need to have the same spatial representativeness as the 10 depositonal loads. To be clear, the deposition transformation ratio is not a state variable, but 11 simply is a calculated term that facilitates the linkage between deposition and concentrations 12 which is a necessary step in developing ambient air indicators that are used to assess compliance 13 with a NAAQS. There will be a tendency that is not scientifically defensible to compare 14 deposition ratios with deposition velocities that are uniquely determined on a species by species 15 basis influenced by numerous factors as discussed earlier. 16 5.4.3 Air Quality Simulation Models 17 Ideally, VS/N values would be derived for each area of interest from concurrently collected 18 sulfur and nitrogen deposition and concentration measurements. However, no monitoring 19 network currently exists that can provide such information. We therefore propose using output of 20 the CMAQ model for initial calculation of VS/N values. 21 CMAQ provides both concentrations and depositions for a large suite of pollutant species 22 on an hourly basis for 12 km grids across the continental U.S. Its comprehensive structure is 23 ideal for providing VS/N values that appropriately address the chemical and temporal aggregation 24 issues discussed above, and weighted spatial averages of the gridded data can be used for areas 25 that span multiple grid cells. Potential concerns with using CMAQ-predicted concentrations and 26 depositions for this purpose stem from the various, but unquantifiable uncertainties in model 27 formation and input data, which will be discussed in the next draft of this PAD. 28 CMAQ does not directly calculate or use VS/N values; instead the following procedures 29 are used in the code to model deposition: March 2010 171 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 1) vdry values of gaseous pollutants are calculated in the CMAQ weather module called 2 the Meteorology-Chemistry Interface Processor (MCIP) through a complex function of 3 meteorological parameters (e.g. temperature, relative humidity) and properties of the geographic 4 surface (e.g. leaf area index, surface wetness) 5 2) vdry values for paniculate pollutants are calculated in the aerosol module of CMAQ, 6 which, in addition to the parameters needed for the gaseous calculations, also accounts for 7 properties of the aerosol size distribution 8 3) vwet values are not explicitly calculated. Wet deposition is derived from the cloud 9 processing module of CMAQ, which performs simulations of mass transfer into cloud droplets 10 and aqueous chemistry to incorporate pollutants into rainwater, all of which is conceptually 11 contained in the vwet parameter in Equation 2. 12 Due to lack of direct measurements, no performance evaluations of CMAQ's dry 13 deposition calculations can be found; however, the current state of MCIP is the product of 14 research that has been based on peer-reviewed literature from the past two decades (EPA, 1999) 15 and is considered to be EPA's best estimate of dry deposition velocities. Some bias has been 16 found between CMAQ's wet deposition predictions and measured values (Morris et al., 2005); 17 recent analyses suggest that poor simulation of precipitation could be responsible for this (Davis 18 and Swall, 2006), which can potentially be dealt with by recalculating wet deposition using 19 precipitation measurements. Although the model is continually undergoing improvement, 20 CMAQ is EPA's state-of-the-science computational framework for calculating deposition 21 velocities, and was therefore the logical first choice as a source for VS/N values. 22 5.4.4 Oxidized Sulfur and Nitrogen Pollutant Species 23 Ideally, all possible air pollutant species that contribute to ecological adversity would be 24 considered for VS/N values. The pollutant list is constrained by the source of VS/N values, which is 25 currently CMAQ output. Table 1 lists the oxidized sulfur and nitrogen species currently available 26 in CMAQ whose data will be used for VS/N values. 27 One issue that needs explicit consideration is the contributions of particles larger than 28 PM2.5 to sulfur and nitrogen deposition. A recent review of particle deposition measurements 29 (Grantz, Garner, and Johnson, 2003) showed that coarse particles generally deposit far more 30 sulfate and nitrate in forest ecosystems than fine particles. However, CMAQ does not currently March 2010 172 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 provide simulationsof coarse paniculate sulfate and nitrate. This is an issue that needs to be 2 addressed by developers of either the model or the future SOX/NOX measurement network to set 3 scientifically sound standards. 4 5.4.5 Example Calculations 5 Figure 5-5 shows annual inverse VS/N values16 calculated for each 12 km grid in the 6 eastern and western domains for a 2002 CMAQ v4.6 simulation, which is the quantity that would 7 be used for conversion of deposition load tradeoff curves which illustrate (see Section 6) the 8 combinations of NOy and SOX conventartions that would correspond to an established critical 9 load. Figure 5-6 shows an example application of these ratios for a lake in the Adirondacks. 10 Deposition load tradeoff curves for this lake (see Section 6for their calculation) are multiplied by 11 the inverse VS/N value from the appropriate grid cell in Figure 1 to convert those depositions to 12 ambient concentrations of sulfur and nitrogen. 13 A CMAQ v4.7 simulation for multiple years (2002-2005) recently became available, 14 which was used to examine the inter-annual variability of inverse VS/N values. The grid-specific 15 coefficients of variation (CV) are shown in Figure 3. Figure 5-7 shows that CV values are 16 relatively small (< 25%) in the Adirondacks and Shenandoah case study areas. This suggests that 17 a 3-year average of the ratios may be a sufficiently stable representation of deposition velocities 18 for converting the deposition load curves to ambient concentrations in future applications. 16 Inverse VS/N values represent the multiplier needed to convert deposition levels into atmospheric concentrations of NOx and SOx. March 2010 173 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 Table 5-3. Oxidized sulfur and nitrogen species currently available in CMAQ simulations. Note that PNA concentrations are not available in current CMAQ extractions. :"'.:illlimit C.'IfL-i-, •inlthr Oxides Nirr.igsn Oxides Clinn:i: :L| C.M. u; • S|::::'i-r-^ Hvn.l..)! SO, so*- .\O NOj NOJ -NA MONO PAN PA.NX NTH PNA Hpsris-i Ntn.:" Hiilliir Dioxide Hill rare Nirrogsn Oxide \irrogen Dioxide Nitrata Dinit-mgen peiitoxide Nirric Acid Itroxvacetvl nitrate Hijjher oi'der peroxyflcetyl nitrates Orfiaiiic Nitrates BMCkt >'..,!:-,- Predoiuiiiautly pftrticulat* PredomiDantly paniculate 4 5 6 Figure 5-5. VS/N values for each grid cell in the eastern (right) and western (left) U.S. domains. The top maps are for sulfur and the bottom are for nitrogen. March 2010 174 Draft - Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX Deposition Concentration 1 2 3 4 7 13 20 N deposition (kg/hay) -ANC100 ANC50 - ANC20 Current • Conditions (CMAQl Figure 5-6. Schematic Diagram illustrating the procedure for converting deposition tradeoff curves of sulfur and nitrogen to atmospheric concentrations of SOX and NOX. March 2010 175 Draft -Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX Coeflicent ol Variation of N Cone Dep ralio, 2002-2005 3 4 a) Coellicenl of Variation ol S Cone.Dep ralio. 2002-2005 b) ' CV (%) rSO.O 37.5 25.0 \ 12.5 0.0 CV (%) — 50.0 37.5 25.0 [12.5 0.0 Figure 5-7. Inter-annual coefficients of variation (CV) of a) nitrogen and b) sulfur VS/N values, based on a series of 2002-2005 CMAQ v4.7 simulation. March 2010 176 Draft -Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 5.5 EXAMPLE CALCULATION FOR THE CONCEPTUAL DESIGN 2 AND DERIVATION OF AAPI 3 Section 5.2 describes a proposed conceptual design for a NOX and SOX NAAQS based on 4 aquatic acidification. To summarize the process of acidification, atmospheric deposition of NOX 5 and SOX contributes to acidification in aquatic ecosystems through the input of acid anions, such 6 as NO3" and SO42". The acid-base balance of headwater lakes and streams is controlled by the 7 level of this acidifying deposition of NOs" and SC>42" and a series of biogeochemical processes 8 that produce and consume acidity in the watershed. The biotic integrity of freshwater ecosystems 9 is then a function of the, acid-base balance and the resulting acidity-related stress on the biota 10 that occupy the water. Given some "benchmark level" of ANC [ANClimit]) that appropriately 1 1 protects biological integrity, the depositional load of acidity DL(N+S) is simply the input flux of 12 acid anions from atmospheric deposition that result in a surface water ANC level equal to the 13 [ANClimit] when balanced by the sustainable flux of base cations input and the sinks of nitrogen 14 and sulfur in the watershed catchment. 15 5.5.1 Example calculation for the conceptual design 16 This section summarizes and provides an example calculation of the approach proposed 17 by EPA staff to calculate (1) the acid-base balance of a catchment for a specified ANC level, (2) 18 the N and S deposition tradeoff curves for a deposition metric, which represents a specified 19 percentage of the total population of water bodies that do not exceed their critical load at a 20 specified ANC level and (3) the conversion from tradeoff curves for N and S deposition to those 21 for atmospheric concentrations of NOy and SOX. The equations representing deposition loads and 22 associated tradeoff curves for a specified level of ANC are the basis for deriving the form of the 23 standard discussed above in section (5.5.2). 24 Equation (1) expresses the model that we suggest using to determine the amount of N and 25 S that may be deposited onto a catchment to yield a specified level of ANC. 26 DL^ (N + S)= ([BCl - [ANC^ ])Q + Neco (1) 27 where, March 2010 177 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 DLANciim(N+S) = depositional load of S and N that does not cause the ecosystems to exceed a 2 given ANCiim 3 [BC]0* = the preindustrial concentration of base cations (equ/L) 4 ANCumit = a "target" ANC level (equ/L) 5 Q= surface water runoff (m/yr) (this is typically equal to precipitation -evapotranspiration 6 Neco= nitrogen retention and denitrification by terrestrial catchment and nitrogen retention in the 7 lake 8 The term Neco could be derived multiple ways. The first is by taking the mean value 9 calculated to represent the long-term amount of N an ecosystem can immobilize and denitrify 10 before leaching (i.e. N saturation) that is derived from the FAB model. This approach requires 1 1 the input of multiple ecosystem parameters. Its components are expressed by eq 2. 12 Neco = JNupt + Nret + (l - rlNmm + Nden ) (2) 13 where, 14 Nupt= nitirogen uptake by the catchment 15 Nimm= nitrogen immobilization by the catchment soil 16 Nden=denitrification of nitrogen in the catchment, 17 Nret = in-lake retention of nitrogen 18 f =forest cover in the catchment (dimensionless parameter) 19 r = fraction lake/catchment ratio (dimensionless parameter) 20 21 The second approach for estimating Neco is to take the difference between N deposition 22 and measured N leaching in a catchment as expressed by eq 3. 23 Neco=DL(N)-Nleach (3) 24 N deposition is composed of NHX deposition (NHxdep) and NOy deposition. It is known that 25 NHxdep contributes to acidification, however the definition of NOX in the CAA does not include 26 NHX, and as such is not defined to provide protection from the acidifying effects of NHX. 27 Therefore, DLANciim(N) is separated into NHX and NOy. 28 DL^^Ncy + S = pL^^(N)-DL^^(mx) + DL^^(SOx) (4) March 2010 178 Draft -Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Equation 1 and 4 will differ catchment by catchment because the acid-base balance of a 2 catchment is a function of site-specific characteristics. However, for the standard it is desirable to 3 calculate a deposition load for a specified ANC not for an individual catchment, but a larger 4 population of catchments. The site specific values from equation 1 can be used to derive such a 5 deposition loading, here called the deposition metric, which represents a group or percentage of 6 water bodies that reach a specified ANC (or higher). For example, if it is desired that all water 7 bodies reach a specified ANC, the allowable amount of deposition for all water bodies is equal to 8 the lowest value calculated from equation 1 for the population of water bodies. Because the 9 deposition metric represents a percentage of individual catchments from a population of water 10 bodies, and not an individual catchment like DLANciim(S+N), the deposition metric is noted by 11 the follow abbreviation DLo/oEC0. 12 As an example of the above approach, we evaluate the population of 169 waterbodies in 13 the Adirondacks used in the REA analysis. For each individual waterbody in the population 14 DLANciim(S+N) at ANCum = 50 was calculated using the two equations for deriving the Neco 15 term (eq 2 and 3). The distribution of deposition loads for the population was assessed and Table 16 5-5 shows the a few selected values for DLo/oECo. The mean value for DLo/oECo for the 169 water 17 bodies is presented, as well as the values for which 50, 75, 85, 95 and 100% of the water bodies 18 in the population will not exceed their critical load at ANC=50. Note, only 32% of water bodies 19 would not exceed their critical load at ANC=50 for the mean value DLo/oECo because variability is 20 high in the data set. The deposition and atmospheric concentration tradeoff curves for DLo/oECO 21 equal to 32% and 50% are plotted in the subsequent figures. The Administrator will choose 22 which % of water bodies are projected to reach a targeted level of ANC as part of the overall 23 decision on the elements of the standard; this selection may be higher or lower than the examples 24 given here. March 2010 179 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX Table 5-4. Example Calculations for Determining the Percent of Water Bodies Achieving Target ANC Levels This example is based the population of DLANciimfor and ANC=50 for 169 catchments in the Adirondacks. These catchments occur across on three categories of geologic sensitivity. We could separate the DLANciim values into sensitivity categories (if info is available) and do the analysis for each category or calculate one DLANciim for combined geologic categories. Units are in meq/m2/yr. Mean Stdev Ster Rank %tile 50% 75% 85% 95% 100% NHX dep 20.40 3.22 0.25 Neco (eq2) 19.19 3.03 0.23 DLo/oECO(S+N) using Neco eq 2 162.36 162.92 13.04 99.33 65.62 54.89 45.12 30.22 Neco (eq3) 63.95 11.15 0.86 DLo/oECO(S+N) using Neco eq 3 207.55 165.42 13.24 139.22 110.37 95.53 83.99 59.07 % of lakes within the population that have ANC > 50 31.7% 50% 75% 85% 95% 100% 4 5 6 The deposition tradeoff curves for N and S based on DLo/oECO at ANC=50 using the two approaches for Neco and protective of 32 and 50% of the population of water bodies, are plotted on Fig 5-8 and 5-9. The values for the maximum deposition values for N and S are given in Table 5-5. Table 5-5. Values for N and S deposition tradeoff curves for ANC = 50, protecting 32 and 50% of the population, in Adirondacks case study area as illustrated on Fig 5-8 and Fig 5-9. Units are in meq/m2/yr unless noted otherwise. % protection 32 50 32 50 Eq2 Eq2 Eq3 Eq3 NHxdep 20.4 20.4 20.4 20.4 Neco 19.19 19.19 63.75 63.75 DLo/oECo (max N) 162.36 99.33 207.5 139.22 DLo/oECo (max S) 143.97 80.14 143.6 75.27 DLo/oECo (max NOY) 141.96 78.9.3 187.15 118.82 March 2010 180 Draft -Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 200 ANC=50 & 32% lakes protected ANC=50 & 50% lakes protected Neco NHx Deposition Neco=19.19 0 NHxdep=20.40 5Q 100 150 N (meq/m2/yr) 200 Figure 5-8. Tradeoff curve for S and N deposition to protect from aquatic acidification in the Adirondacks using Neco equation 2. 4 5 6 200 -£ 150 ^•» c\i |-100 ANC=50 & 32% lakes protected ANC=50 & 50% lakes protected Neco NHx Deposition Max(S) =143.6 Max(N) =207.5 0 100 N (meq/m2/yr) 150 200 Figure 5-9. Tradeoff curve for S and N deposition to protect from aquatic acidification in the Adirondacks using Neco equation 3. March 2010 181 Draft -Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 4 5 6 7 As previously stated, it is known that NHX deposition (NHxdep) contributes to acidification. However, the criteria pollutant listed by EPA pursuant to section 108 (a) of the Act is oxides of nitrogen does not include NHX, and as such is not defined to provide protection from the acidifying effects of NHX. Therefore, in order to represent the role of NHxdep as a component of acidification it is subtracted from DLo/oECO(S+N). The difference is the total allowable deposition from NOy and SOX to protect a selected % of catchments in the population at a selected level of ANC [DLo/oECO (S + NOy)] as expressed in equation 5. DL%ECO (NOY +S) = DL% %ECO - NHX DEP (5) 9 The NOy and S deposition tradeoff curves for ANC =50, protecting 32 and 50% of the 10 water bodies, are presented in Table 5-6 and plotted on Fig 5-10 and 5-11. If NHX deposition is 11 greater than Neco, then Neco disappears from the tradeoff curve (i.e. Fig 5-11). Table 5-6. Values for NOy and S deposition tradeoff curves for ANC = 50, protecting 32 and 50% of the population in Adirondacks case study area as illustrated on Fig 5.10 and Fig 5.11. Units are in meq/m2/yr unless noted otherwise. % protection 32 50 32 50 Eq2 Eq2 Eq3 Eq3 NHxdep 20.4 20.4 20.4 20.4 Neco (Noy) Neco < NHxdep Neco < NHxdep 43.35 43.35 DLmax(S) 141.96 78.93 143.6 75.27 DLmax(Noy) 141.96 78.93 187.15 118.82 12 March 2010 182 Draft -Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 200 c\i o-100 C/3 0 ANC=50 & 32% lakes protected ANC=50 & 50% lakes protected 50 100 150 Noy (meq/m2/yr) 200 Figure 5-10. Tradeoff curve for S and NOy deposition to protect from aquatic acidification in the Adirondacks using Neco equation 2. 4 5 6 200 150 100 0 ANC=50 & 32% lakes protected ANC=50 & 50% lakes protected Neco Max(S)=143.6 50 100 150 NOy (meq/m2/yr) Max(Noy)=187.15 200 Figure 5-11. Tradeoff curve for S and NOy deposition to protect from aquatic acidification in the Adirondacks using Neco equation 3. The tradeoff curves for the atmospheric concentration of NOy and SOX are presented in Fig 5-12 and 5-13. Deposition values for NOy and S (from Table 5-6, Fig 5-10 and 5-11) were multiplied by the ratio of concentrations to depositions (previously referred to as aggregate March 2010 183 Draft -Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 effective deposition velocitiesl?) for NOX and SOX (VSOX = 0.03824755 ng/m3/meq/m2 and 2 VNOX= 0.04386373 |j,g/m3/meq/m2). This is expressed in equation 5. These velocities were 3 calculated by taking the median value of the concentration of oxidized N to deposition of 4 oxidized N ratio in CMAQ for all grid cells over the Adirondack case study area. [DL%ECO (N o J- Vnoy\+ [DL%ECO (s) • Vsox] = DL%ECO (N+S)- NHX VDEP (6) 6 7 CO E X o CO ANC=50 & 32% lakes protected ANC=50 & 50% lakes protected 0.00 10.00 Noy (ug/m3) Figure 5-12. Tradeoff curve for atmospheric concentration of SOX and NOy to protect from aquatic acidification in the Adirondacks using Neco equation 2. 17 Note to reviewers: in previous drafts we have referred to the ratios of deposition to concentration for NOy and SOx as "aggregate effective velocities." We are revisiting this choice of terms, as it is not as accurate a reflection of the parameter as we might prefer. The concern with continuing to use the term "velocity" in this context is that it will be misinterpreted by the scientific community, and in order to avoid confusion, we will likely replace the term with "deposition ratio" or some other term that more accurately describes the parameter. March 2010 184 Draft -Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 10.00 ANC=50 & 32% lakes protected ANC=50 & 50% lakes protected IE ^) X O C/5 0.00 0.00 10.00 Noy (ug/rrr) 1 2 Figure 5-13. Tradeoff curve for atmospheric concentration of SOX and NOy to 3 protect from aquatic acidification in the Adirondacks using Neco equation 3. 4 5.5.2 Derivation of the Atmospheric Acidification Potential Index (AAPI): 5 While the conceptual framework above provides a means for calculating tradeoff curves 6 associated with a specific level of protection (indicated by a target ANC level) and a specific 7 percentage of ecosystems protected within an overall sensitive area, it does not provide a clearly 8 integrated statement that can be expressed as a level such as would be needed for the secondary 9 standard. The goal of this development of the AAPI is to create an index which can be applied 10 across the nation to convey the potential of an ecosystem to become acidified from atmospheric 11 deposition. 12 The definition of the AAPI form considered here is: 13 Annual Average AAPI: Natural background ANC minus the contribution to 14 acidifying deposition from NHX, minus the acidifying contribution of NOy and 15 SOX. This term is essentially a calculated ANC value that represents a percentage 16 of catchments in a population. 17 In order to derive the AAPI, we start with the basic framework of critical loads discussed 18 in the example above. 19 The approach used to calculate N and S deposition values for a specified ANC at a 20 catchment-scale is expressed in Equation 1. The deposition value for a specified ANC will vary March 2010 185 Draft -Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 from catchment to catchment based on how the properties that counterbalance the acidifying 2 deposition vary among catchments. Equation 5 expresses how to calculate a deposition metric for 3 a specified ANC for a population of waterbodies that could represent a national acid-sensitivity 4 category. Moreover, the quantity of deposition equal to a specified ANC limit (i.e. critical load) 5 will vary in eq 1 and 5 depending on the characteristics of the catchment or population of 6 catchments, respectively. The goal for a secondary NOX and SOX NAAQS is to develop a form 7 for the standards that allows us to set a single value for the standard across the U.S. To 8 accomplish this, we rearrange equation (1) to solve for ANC (place ANC on the left hand side of 9 the equation): 10 Q. ANC^ = Neco + [BC]0-Q- [Dl(N] + DL(s)] (7) (8) 12 In order to develop a form for the standard in which the level can be expressed as a single 13 national value related to protection against effects that occur at specific values of ANC, a 14 simplified version of equation (8) is: 15 ANC]im=g(-)-DL(N + S) (9) 16 where, g()= sustainable flux of base cations from the ecosystem + ecological sinks of N. This 17 term is equivalent to the pre-industrial ANC level, or the natural background ANC, expressed as: 18 g() = ~Neco+[BCl (10) 19 Building from equation 9, total nitrogen deposition is split into oxidized and reduced 20 nitrogen because we need to be able to specify the standards in terms of oxides of nitrogen, and 21 so the contribution of reduced nitrogen has to be separated. 22 ANC^=g()~[DL(NOT) + DL(S)]~-D^NHX) (11) March 2010 186 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 where, 2 DL(Nox)= the deposit!onal load of oxidized nitrogen 3 DL(NHX)= the deposit onal load of reduced nitrogen, NHX. 4 5 In order to judge whether an ecosystem or group of ecosystems meets the ANCimut given 6 observed NOy and SOX levels, the associated depositional loadings of NOy and S can be 7 compared directly against calculated deposition tradeoff curves (eq 4), atmospheric 8 concentrations of NOy and SOX can be compared against the atmospheric concentration tradeoff 9 curves (eq 5) or, loadings of NOX and SOX can be input into the following equations to obtain the 10 calculated value of ANC, equal to ANC*: 11 ANC* = g(-) - [L(Noy) + l(SOx)] - L(NHx) (11) 12 where, 13 ANC*= the calculated value of ANC given loadings of N and S for comparison against an 14 ANCHmit. 15 L(NOX+S)= the load of NOX+S anions based on observed atmospheric concentrations of NOy and 16 SOX 17 L(NHX) = the load of reduced nitrogen deposition 18 [Note that L(N) = L(NOX+NHX)] 19 20 In equation 11, the ANC* will vary based on the deposition load inputs of Nox, NHX and 21 S at the site of interest. The deposition loads caused by NOy and S and NHX are inputs, leading to 22 ANC* = g~[L(Nax) + L(S)]~-DltNHx) (12) 23 If ANC* < ANCiim, then the deposition of N and S exceeds the deposition load to maintain 24 ANCiimit. ANC* is still representative of the calculated ANC based on specific catchment level 25 estimates of g, Q and NHX. 26 AAPI is equivalent to the equation for calculating ANC* when the catchment specific 27 values for g in equation (9) in Section 5.5.1. are replaced by representative values for acid 28 sensitive areas (based on a percentile of water bodies targeted for an ANC level selected by the March 2010 187 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Administrator), Q and NHX are replaced by average values for aggregate ecosystem areas, and 2 L(Nox) and L(S) are replaced by terms translating atmospheric NOy and SOX into deposition: (13) 4 where NOy and SOX are concentrations of NOy and SOX, respectively, VNOY and VSOX are the 5 ratios of deposition to concentrations (deposition transformation ratios) for NOy and SOX, 6 respectively. 7 5.6 REFERENCES 8 Davis, J. M., Swall, J. L., 2006. An examination of the CMAQ simulations of the wet deposition 9 of ammonium from a Bayesian perspective. Atmospheric Environment 40, 4562-4573. 10 EPA, 1999. Science Algorithms of the EPA Models-3 Community Multiscale Air Quality 1 1 (CMAQ) Modeling System. Tech. Rep. EPA/600/R-99/030, U.S. Environmental 12 Protection Agency, Washington DC. 13 Grantz, D., Garner, J., Johnson, D., 2003. Ecological effects of particulate matter. Environment 14 International 29, 213-239. 15 Lien L; Raddum GG; Fjellheim A. (1992). Critical loads for surface waters: invertebrates and 16 fish. (Acid rain research report no 21). Oslo, Norway: Norwegian Institute for Water 17 Research 18 Morris, R. E., McNally, D. E., Tesche, T. W., Tonnesen, G., Boylan, J. W., Brewer, P., 2005. 19 Preliminary evaluation of the community multiscale air quality model for 2002 over the 20 Southeastern United States. Journal of the Air and Waste Management Association 55, 21 1694-1708. 22 Seinfeld, J., Pandis, S., 1998. Atmospheric Chemistry and Physics. John Wiley and Sons, Inc., 23 New York. March 2010 188 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Sullivan TJ; Webb JR; Snyder KU; Herlihy AT; Cosby BJ. (2007). Spatial distribution of acid- 2 sensitive and acid-impacted streams in relation to watershed features in the southern 3 Appalachian mountains. Water Air Soil Pollut, 182, 57-71. 4 Sullivan TJ; Fernandez U; Herlihy AT; Driscoll CT; McDonnell TC; Nowicki NA; Snyder KU; 5 Sutherland JW. (2006). Acid-base characteristics of soils in the Adirondack Mountains, 6 New York. Soil Sci Soc Am J, 70, 141-152. 7 March 2010 189 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 6. OPTIONS FOR ELEMENTS OF THE STANDARD 2 The elements of the standard include the ambient air indicator, the form, the level and the 3 averaging time. The "indicator" of a standard defines the chemical species or mixture of the 4 criteria air pollutant that is to be measured in determining whether an area attains the standard. 5 The "form" of a standard defines the air quality statistic that is to be compared to the level of the 6 standard in determining whether an area attains the standard. The "averaging time" defines the 7 period of time over which the air quality indicator is averaged, e.g. annual average. The "level" 8 is the specific quantity to which the air quality statistic will be compared. 9 EPA has historically established NAAQS so that the locally-monitored ambient 10 concentration of an air pollutant indicator is compared against a specified numerical level of 11 atmospheric concentration, using a specified averaging time and statistical form. For example, 12 the current secondary standard for oxides of nitrogen uses ambient concentrations of NC>2 as the 13 indicator. Attainment is determined by comparing the annual arithmetic mean of the measured 14 maximum daily 1-hour NO2 concentrations, for a calendar year, against the level of 0.053 ppm. 15 As discussed in Chapters 4 and 5, a standard using this kind of approach for defining indicator, 16 averaging time, form, and level is not the most appropriate way to protect sensitive ecosystems 17 from effects associated with ambient concentrations of NOX and SOX. Moreover, the inherently 18 complex and variable linkages between ambient concentrations of NOX and SOX, their deposited 19 forms of nitrogen and sulfur, and the ecological responses that are associated with public welfare 20 effects call for consideration of a more complex and ecologically relevant design of the standard 21 that reflects these linkages. 22 Chapter 5 provided a conceptual framework for a secondary standard that is designed to 23 provide protection of ecosystems against the effects associated with deposition of ambient 24 concentrations of NOX and SOX. This conceptual framework takes into account variable factors, 25 such as atmospheric and ecosystem conditions that modify the amounts of deposited NOX and 26 SOX, and the associated effects of deposited N and S on ecosystems. Based on the conceptual 27 framework described in Chapter 5, this chapter provides a set of potential options for specifying 28 the elements of the framework to define a secondary standard for NOX and SOX. Our 29 development of options for the standards recognizes the need for a nationally applicable standard 30 for protection against adverse effects to public welfare, while recognizing the complex and 31 heterogeneous interactions between atmospheric concentrations of NOX and SOX, deposition, and March 2010 190 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 ecological response. These options will include elements of the framework related to the air 2 quality indicator, the averaging time, the form, and the level, which are based on the ecological 3 indicator, the ecological response to deposition function, the deposition metric, and the 4 atmospheric deposition transformation function. 5 To make the transition from the conceptual framework in Chapter 5, which is developed 6 largely around the concept of critical loads, to elements of the standard, we propose to focus on 7 developing a form of the standard that is based on the concepts of critical loads of NOX and SOX 8 deposition linked to target ANC values, recognizing the limitations in available data and related 9 uncertainties. Our goal in developing the form of the standard is to create an index, directly 10 expressed in terms of atmospheric concentrations of NOy and SOX, that can be applied across the 11 nation to convey the potential of an ecosystem to become acidified from atmospheric deposition. 12 This chapter is structured around questions related to the various elements of a standard. 13 The chapter begins in section 6.1 with a discussion of atmospheric indicators. Section 6.2 then 14 discusses averaging times for the atmospheric indicators. Section 6.3 suggests a possible 15 ecologically relevant form of the standard. Section 6.4 provides a discussion of issues regarding 16 the spatial area over which a standard might be evaluated, and related issues regarding spatial 17 averaging within areas. Section 6.5 discusses options for specifying target levels for the 18 ecological indicator for aquatic acidification. Section 6.6 addresses issues relating to monitoring 19 of the atmospheric indicators. Section 6.7 concludes with a discussion of potential ranges of 20 levels for the standard. 21 6.1 WHAT ATMOSPHERIC INDICATORS OF OXIDIZED NITROGEN 22 AND SULFUR ARE APPROPRIATE FOR USE IN A SECONDARY 23 NAAQS THAT PROVIDES PROTECTION FOR PUBLIC WELFARE 24 FROM EXPOSURE RELATED TO DEPOSITION OF N AND S? 25 WHAT AVERAGING TIMES AND STATISTICS FOR SUCH 26 INDICATORS ARE APPROPRIATE TO CONSIDER? 27 Staff concludes that indicators other than NC>2 and SC>2 should be considered as the 28 appropriate pollutant indicators for protection against the acidification effects associated with 29 deposition of NOX and SOX. This conclusion is based on the recognition that all forms of March 2010 191 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 oxidized nitrogen and sulfur in the atmosphere contribute to deposition and resulting 2 acidification, and as such NO2 and 862 are incomplete indicators. Furthermore, staff concludes 3 that NOy (total oxidized nitrogen) should be considered as an appropriate indicator for oxides of 4 nitrogen. NOy is defined as NOX (NO and NO2) and all oxidized NOX products: including NO, 5 NO2, and all other oxidized N-containing compounds transformed from NO and NO2 (Finlayson- 6 Pitts and Pitts, 2000). As described in Chapter 4, this set of compounds includes NO2 + NO + 7 HNO3 + PAN +2N2O5 + HONO+ NO3 + organic nitrates + paniculate NO3. Staff concludes that 8 SOX should be considered as an appropriate indicator for oxides of sulfur. SOX includes sulfur 9 monoxide (SO), sulfur dioxide, sulfur trioxide (SO3), and disulfur monoxide (S2O), and 10 particulate-phase S compounds that result from gas-phase sulfur oxides interacting with particles. 11 In principle, measured NOy based on catalytic conversion of all oxidized species to NO 12 followed by chemiluminescence NO detection is consistent with this definition. We recognize 13 the caveats associated with instrument conversion efficiency and possible inlet losses which are 14 discussed in Section 5.6. The development of the function that converts atmospheric 15 concentrations of NOy and SOX to N and S deposition which incorporates NOy estimates is based 16 on the Community Multi-scale Air Quality (CMAQ) model (EPA, 1999). CMAQ treats the 17 dominant NOy species as explicit species while the minor contributing non-PAN organic 18 nitrogen compounds are aggregated. Total oxidized sulfur, SOX, requires independent 19 measurements of particle bound sulfate and gaseous sulfur dioxide; methodology and network 20 considerations are discussed in Section 5.6. The CMAQ treatment of SOX is the simple addition 21 of both species which are treated explicitly in the model formulation. All particle size fractions 22 are included in the CMAQ SOX estimates. At this time, we consider the contribution of coarse 23 fraction (aerodynamic diameters between 2.5 and 10 microns) particle bound sulfate to be 24 insignificant from a measurement perspective. Consequently, the routinely measured sulfate 25 from IMPROVE and EPA speciation networks, as well as CASTNET, are viable candidates for 26 measurement consideration. Consistent with units and the charge balance relationships applied in 27 ecosystem acidification models, only mass as sulfur or nitrogen is considered requiring 28 conversion of reported particle bound sulfate and nitrate. Precipitation mass is not included 29 explicitly as part of an atmospheric NAAQS indicator. March 2010 192 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 6.2 WHAT IS THE APPROPRIATE AVERAGING TIME FOR THE AIR 2 QUALITY INDICATORS NOY AND SOX TO PROVIDE 3 PROTECTION OF PUBLIC WELFARE FROM ADVERSE EFFECTS 4 FROM ACIDIFICATION? 5 Based on the review of the scientific evidence, welfare effects associated with 6 acidification result from annual cumulative deposition of nitrogen and sulfur, reflected in effects 7 on the chronic ANC level (measured as annual ANC). It is important to note that chemical 8 changes can occur over both long- and short-term timescales. Short-term (i.e., hours or days) 9 episodic changes in water chemistry can also have significant biological effects. Episodic 10 chemistry refers to conditions during precipitation or snowmelt events when proportionately 11 more drainage water is routed through upper soil horizons that tend to provide less acid 12 neutralizing than was passing through deeper soil horizons. Surface water chemistry has lower 13 pH and acid neutralizing capacity (ANC) during events than during baseflow conditions. One of 14 the most important effects of acidifying deposition on surface water chemistry is the short-term 15 change in chemistry that is termed "episodic acidification." Some streams may have chronic or 16 base flow chemistry that is suitable for aquatic biota, but may be subject to occasional acidic 17 episodes with lethal consequences. Episodic declines in pH and ANC are nearly ubiquitous in 18 drainage waters throughout the eastern United States and are caused partly by acidifying 19 deposition and partly by natural processes. As noted in Chapter 3 of the ISA, while ecosystems 20 are also affected by episodic increases in acidity due to pulses of acidity during high rainfall 21 periods and snowmelts, protection against these episodic acidity events can be achieved by 22 establishing a higher chronic ANC level. Episodic acidification can result from either shorter 23 term deposition episodes, or from longer term deposition on snowpack. Snowmelt can release 24 stored N deposited throughout the winter, leading to episodic acidification in the absence of 25 increased deposition during the actual episodic acidification event. Protection against a low 26 chronic ANC level is provided by reducing overall annual average deposition levels for nitrogen 27 and sulfur. This supports the conclusion that long term NOX and SOX concentrations are 28 appropriate to provide protection against low chronic ANC levels, which protects against both 29 long term acidification and acute acidic episodes. March 2010 193 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Long term concentrations are often measured using annual averages. However, given the 2 multi-year nature of responses to chronic acidification, multi-year averages of concentrations of 3 NOy and SOX may also be appropriate. In the second draft policy assessment, we will provide an 4 expanded discussion of the support for different options for the averaging time to best represent 5 long-term concentrations of NOy and SOX related to chronic acidification. 6 6.3 WHAT FORM(S) OF THE STANDARD ARE MOST APPROPRIATE 7 TO PROVIDE PROTECTION OF SENSITIVE ECOSYSTEMS 8 FROM THE EFFECTS OF ACIDIFYING DEPOSITION RELATED 9 TO AMBIENT NOX AND SOX CONCENTRATIONS? 10 Based on the evidence for joint effects of NOX and SOX through acidifying deposition, 11 staff concludes that it is appropriate to consider changes to the form of the existing NOX and SOX 12 secondary standards to provide protection to ecosystems. Staff notes that in recent reviews of the 13 secondary ozone standards, EPA has considered use of a form of the standard that reflects 14 ecologically relevant exposures, by using a cumulative index which weights exposures at higher 15 concentrations greater than those at lower concentrations based on scientific literature 16 demonstrating the cumulative nature of (Vinduced plant effects and the need to give greater 17 weight to higher concentrations (EPA, 2007). See 75 FR 2938, 2999 (Janaury 19, 2010) In order 18 to recognize the roles that NOX and SOX play in acidification based on their acidifying potentials, 19 and to incorporate the important roles that reduced nitrogen and non-atmospheric variables play 20 in determining the acidifying potentials of NOX and SOX, staff suggests using an Atmospheric 21 Acidification Potential Index (AAPI) that is a more ecologically relevant form relative to the 22 current ambient concentration based forms, based on the derivations in Section 5.5.1. The intent 23 of the AAPI is in effect to weight atmospheric concentrations of NOX and SOX by their 24 propensity to contribute to acidification through deposition, given the fundamental acidifying 25 potential of each pollutant, and the ecological factors that govern acid sensitivity in different 26 ecosystems. Thus the APPI is more relevant to protecting ecosystems from acidifying deposition 27 compared to simple ambient concentration forms which do not reflect factors that affect 28 acidifying potential. 29 The AAPI is closely tied to the ecological indicator of acidification, ANC, so that the 30 form of AAPI is intended to identify the atmospheric concentrations of NOX and SOX that will March 2010 194 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 result in an equivalent level of a target ANC for a percentage of aquatic ecosystems within a 2 particular acid sensitive area. Thus, this form is ecologically relevant as it is tied directly to the 3 ecological indicator that is most directly linked with known ecological effects. 4 The AAPI incorporates the processes which modify both rates of deposition and 5 ecological response to deposition caused by NOX and SOX. There is strong evidence in the 6 scientific literature demonstrating that the amount of deposition caused by NOX and SOX is 7 modified by atmospheric and landscape factors. Within the ecosystem there are factors, such as 8 bedrock geology and topography, which modify the acidifying potential of the nitrogen and 9 sulfur deposition resulting from ambient NOX and SOX concentrations. In addition reduced 10 nitrogen contributes to total nitrogen loading. In this review, reduced nitrogen is treated as an 11 additional modifying factor within the ecosystem, which reduces the buffering capacity of the 12 ecosystem, and therefore it increases the impact or sensitivity to additional loading from oxidized 13 forms of nitrogen. In effect this leaves less allowable deposition loading from NOX and SOX 14 before the ecosystem fails to achieve a target ANC level. Based on this evidence staff concludes 15 that the form should include landscape and atmospheric factors, including reduced nitrogen, 16 which modify the acidifying potential of ambient NOX and SOX concentrations. This form is 17 consistent with the language of the CAA as discussed in Section 1.5. 18 Selecting a more ecologically-relevant secondary standard form would also be directly 19 responsive to the recommendation of the 2004 National Research Council's report titled Air 20 Quality Management in the United States (NRC, 2004) which encourages the Agency to evaluate 21 its historic practice of setting the secondary NAAQS equal to the primary. 22 In theory, the AAPI could address acidification potential related to both terrestrial 23 acidification and aquatic acidification. For this first draft policy assessment, as discussed in 24 Chapter 5, we define the AAPI for protection against aquatic acidification. In the second draft 25 policy assessment, we will explore the potential to include protection against terrestrial 26 acidification in the AAPI or a related index. 27 The definition of the AAPI form considered here is: 28 Annual Average AAPI: Natural background ANC minus the contribution to 29 acidifying deposition from NHX, minus the acidifying contribution of deposition 30 from NOy and SOX. 31 Building from the derivation of ANC* provided in Section 5.5.2, the AAPI is equivalent 32 to the equation for calculating ANC* when the catchment specific values for g in equation (9) in March 2010 195 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Section 5.5.2. are replaced by representative values for acid sensitive areas (based on a percentile 2 of water bodies targeted for an ANC level selected by the Administrator), Q and NHX are 3 replaced by average values for aggregate ecosystem areas, and L(N0x) and L(S) are replaced by 4 terms translating atmospheric NOy and SOX into deposition: (1) 6 where NOy and SOX are concentrations of NOy and SOX, respectively, VNOY and VSOX are the 7 ratios of deposition to concentrations (deposition transformation ratios) for NOy and SOX, 8 respectively. Deposition transformation ratios are the estimated relationships between 9 atmospheric concentrations of NOy and SOX and the collocated deposition of Nox and S. See 10 Chapter 5.4.4 and 5.4.5 for further description of calculation of ratios of deposition to 1 1 concentrations. 12 Note that while equation (1) is used to calculate the value of AAPI for any observed 13 values of NOy and SOX, the level of the standard for AAPI selected by the administrator should 14 reflect a wide number of factors, including desired level of protection indicated by a target 15 ANCumit, the specified percentile of waterbodies projected to achieve the target ANC, and the 16 various factors and uncertainties involved in specifying all of the other aspects of the standard, 17 such as the classification of landscape areas, the specification of reduced nitrogen deposition, the 18 methodology to determine deposition of NOy and SOX, and the averaging time. As such the 19 administrator may choose an AAPI level higher or lower than the target ANCumit to reflect the 20 combined effect of the all of the components of the standard and their related uncertainty, such 21 that the chosen AAPI, in the context of the overall standard, reflects her informed judgment as to 22 a standard that is sufficient but not more than necessary to protect against adverse public welfare 23 effects. 24 How are AAPI parameters determined? 25 Other than ambient levels of NOX and SOX, which would be measured values, EPA would 26 determine and specify all of the values for the AAPI parameters, as discussed below. 27 The natural background ANC, g, is a calculated value and is determined by two 28 components: [BC]o* which is closely associated with underlying bedrock which strongly 29 influences the contribution of base cations due to weathering, for which a representative value March 2010 196 Draft -Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 could be determined for a limited set of geologic acid sensitivity classes, and Neco, which 2 represents the amount of deposited nitrogen that is available for acidification due to uptake, 3 denirification and immobilization. Neco is estimated using two different approaches: (1) the 4 individual terms are estimated through available data and modeling or (2) Neco is calculated as 5 nitrogen deposited minus nitrogen leached, using streamwater measurements of nitrate for 6 leaching and estimates of nitrogen deposition based on model results and measurements. The 7 details of these procedures are addressed in chapter 4 and Appendix 4 of the REA. 8 The runoff parameter Q for each acid sensitive area is determined based on USGS 9 mapping of runoff values (REFERENCE NEEDED). 10 VNO?, VSOX are calculated from CMAQ by dividing the annual average NOy, SOX 11 concentration by the total NOy or SOX deposition, respectively, for each grid cell and then 12 aggregating all grid cells in the acid sensitive area. 13 L(NHX) is calculated using the same procedures applied to CMAQ results for deposited 14 NHX. 15 The VNOY and VSOX are spatially variable, and for the purposes of setting the standard, 16 are determined based on the ratios of total sulfur and nitrogen depositions to concentrations from 17 CMAQ model outputs (see Chapter 5 for details of calculation of deposition ratios). V^oy, VSOX 18 are calculated from CMAQ by dividing the annual average NOy or SOX concentration by the total 19 NOy or SOX deposition, respectively, for each grid cell and then computing the mean or median 20 of all grid cells in the acid sensitive area (the decision on whether the median or mean value 21 should be used is an option for discussion; the mean will give more weight to outlier values 22 relative to the median). 23 NHX is spatially variabe and determined based on monitored and/or CMAQ modeled 24 outputs. The average NHX deposition across grid cells within an acid sensitive region will be 25 used to represent the deposit!onal load of NHX. 26 There will be multiple combinations of concentrations of NOX and SOX that result in a 27 specific value of the AAPI. There will be no single combination of NOX and SOX that solves for a 28 particular value of AAPI in all locations, easured concentrations of annual average NOX and SOX 29 necessary to meet the standards are thus expressed conditionally by the equality in (1), and not 30 by fixed quantities. March 2010 197 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 In order to provide a set of values for elements of the form, e.g. to develop a specific set 2 of parameter values for g, VNOY, Vs, and NX, we propose to classify locations in the U.S. into a 3 set of areas based on sensitivity to acidification. Each area would be assigned a classification for 4 the g parameter; for example, as described in Section 5.2.2.4, a set of classes of acidification 5 sensitivity might be able to be developed based on underlying bedrock geology, or bedrock 6 geology plus other ecosystem variables. The g parameter (natural background, or preindustrial, 7 ANC) would then be estimated for each of those sensitivity classes, based on the critical load 8 modeling available for each class. Each acid sensitive area would then be assigned a value of g 9 based on the geology class in which it falls. In the case of VNOY, Vs, and NHX, values for specific 10 areas would be estimated based on the best available monitoring and/or modeling data. Given the 11 limited availability of measured deposition velocities, staff concludes that the calculated 12 deposition ratios based on the CMAQ modeling from 2005 provides the best available source of 13 estimates of VNOY and Vs. Evaluation of the stability of these estimates of deposition ratios over 14 time (see Chapter 5) suggests that in most acid sensiive areas, deposition ratios are quite stable, 15 with a coefficient of variation less than 25 percent across a four year period. While there are a 16 limited number of sites that directly measure deposition of reduced nitrogen, staff concludes that 17 the most widely available and defensible estimates of reduced nitrogen deposition (NHX) are the 18 estimates obtained from the CMAQ modeling from 2005.18 19 It is important to note for this form of the standard that the same AAPI can be obtained 20 with differentcombinations of ambient NOX and SOX concentrations. The implication of the form 21 of the standard expressed in equation (1) is that there will be a tradeoff curve that reflects the 22 combinations of NOX and SOX that satisfy equation (1) for any specific value of the standard. The 23 shape of the tradeoff curve will depend on the specific values of G, VNOY, Vs, and NHX for a 24 limited number of specific areas classified based on acid-sensitivity. As discussed in Chapter 5, 25 all parts of the U.S. would be classified into areas based on acid-sensitivity. Within each such 26 area, EPA would specify the parameter values of APPI, leading to a specific tradeoff curve for 27 each area. The levels of NOy and SOX that meet an AAPI standard expressed for a given 28 g() [preindustrial ANC], Q, L(NHX) and VNOy and VSOx: 18 Note to readers: Maps of CMAQ 2005 estimates of NHx deposition will be included in the second draft policy assessment, along with an evaluation of the representativeness of the 2005 NHx deposition for characterizing conditions over a multiyear period. March 2010 198 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 12 1 VNOy • NOy + V^-ST = g() • Q - L(NHx) - AAPI • Q (2) 2 Note that \^NOy • NOy + VST • ST\ is essentially the critical load of NOy and S, expressed in terms 3 of atmospheric concentrations. As such, equation (2) can also be expressed in a form similar to a 4 typical critical load equation as discussed in Chapter 5, e.g. 5 VNOy • NOy + VST-ST = (BC*0 - APPIjQ - Neco - L(NHx) (3) 6 This expression is based on 7 g() = (BC* +)- L(NHx) - AAPI • Q (4) 8 The pairs of NOy and SOX that will meet a given AAPI limit are related through the following 9 equations 10 NO;=Cmm(NOy) (5) 11 SO* =C (SOx)\/NO < C (NO ) (6) x max \ / y min \ y / \ / (Cmm(NOy)-CmK(NOy)) •NO*yVNOy>Cmm(NOy) (7) 13 where, 14 NO*y is the coordinate point for NOy 15 SO*X is the coordinate point for SOX 16 Cmax (SOx) is the concentration of SOX in the atmosphere consistent with DLmax (S) 17 Cmax (NOy) is the concentration of NOy in the atmosphere consistent with DL max (N) 18 Cmm (NO ) is the concentration of NOy in the atmosphere consistent with DL min (N) 19 Cmax (SOx) = — DLmax (S) (8) March 2010 199 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX Cmm (NOY ) = -t—DL^(N - NHx)\/NHx < DLmm (N) (9) wo. =OVNHX>DLmm(N) (10) 4 where DLmax(S), DLmax (N), and DLmin(N).are based on the critical load within a sensitive 5 areas that protects a specified percentile (e.g. 95 °) of water bodies in the area. 6 Note that Cmin(NOy) is a conditional function determined by the relationship between 7 total nitrogen buffering capacity in an ecosystem and the amount of reduced nitrogen deposition. 8 When reduced nitrogen deposition exceeds the buffering capacity of an ecosystem, then all 9 atmospheric oxidized nitrogen contributes to acidification. When reduced nitrogen deposition is 10 less than the buffering capacity of an ecosystem, then some amount of NOy is buffered (i.e. is 1 1 reflected in Cmin(NOy) but that amount reflects the contribution of NHX to total nitrogen (the 12 amount of buffering capacity used up by reduced nitrogen). In this case, some fraction of the 13 atmospheric oxidized nitrogen may not contribute to acidification. 14 Recall that these three variables are conditional on the chosen level of APPI, and reflect 15 the deposit! onal loadings that are associated with an equivalent level of ANC, e.g. for an APPI of 16 50, the DLmax(S), DLmax(N), and DLm;n(N) are associated with an ANC of 50. Also recall than 17 DLmax(S) for a given ANC is a function of the "natural" flux of base cations to a watershed, 18 runoff, and the amount of sulfur retention within a waterbody; DLm;n(N) is the minimum amount 19 of deposition of total nitrogen (NHX + NOX) that catchment processes can effectively remove 20 without contributing to the acidic balance; and DLmax(N) for a given ANC is a function of 21 DLm;n(N) and the "natural" flux of base cations to a watershed, runoff, and the amount of 22 nitrogen retention within a waterbody, assuming S is zero. In our framework, DLm;n(N) is 23 calculated from the FAB critical load modeling (equation 5 from Attachment A of the REA) or 24 estimated through measured or modeled values of total nitrogen deposition and nitrate leaching. 25 As discussed in Chapter 5, the specific estimation of G, VNOY, Vs, and NHxin a specific 26 sensitive area will depend on the spatial scale of the sensitive area. Sensitivity can be assessed at 27 the level of individual catchments, however, this presents practical limitations for establishing March 2010 200 Draft -Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 meaningful standards, as there are thousands of catchments within the U.S. Binning classes of 2 sensitivity within larger spatial areas, e.g. the sensitive ecosystem areas displayed in Figure 6-1 3 (reproducing Figure 4.2-2 in the REA), can provide a more manageable set of values of G, VNoy, 4 Vs, and NHX. These parameters can be estimated in several ways for the larger spatial areas. 5 Mean or median values can be generated across catchments, however, this would lead to 6 parameter estimates that do not reflect conditions in the more sensitive lakes in the region. 7 Alternatively, in order to provide a desired level of protection in these larger defined spatial 8 areas, estimates based on higher percentiles of the distributions of parameters across catchments 9 can be generated, e.g. the 75th or 95th percentile values of G, VNOY, Vs, and NHxr could be used to 10 provide protection for the more vulnerable aquatic ecosystems, however this would potentially 11 lead to over-protection for less vulnerable ecosystems in the area. The Administrator may 12 consider the balance between protection of particularly sensitive ecosystems and the overall 13 protection for ecosystems in an area as an important element to consider in making decisions 14 about the target level of ANC and the percent of aquatic ecosystems within an area targeted to 15 achieve the specified ANC level. One potentially important modification to this process would 16 be to first remove water bodies that are naturally acidic (e.g. that will not benefit from reductions 17 in atmospheric NOX and SOX deposition) from the distribution of water bodies in the area prior to 18 determining the mean or 95th percentile. This will increase the likelihood that the estimated g 19 parameter will be representative of ecosystems within an area that are sensitive to NOX and SOX 20 deposition. The second draft policy assessment will explore the implications of alternative 21 combinations of target ANC and percent of aquatic ecosystems protected at the target ANC in 22 areas of different sizes. The second draft policy assessment will also explore methods for 23 determining values of g for areas that are clearly not sensitive to acidification from deposition of 24 NOX and SOX. These areas may be areas that have very high levels of natural buffering, or may 25 also be areas that are naturally acidified, such that the value of g is less than the target value of 26 ANC. In these naturally acidified areas, reducing deposition from NOX and SOX will not be 27 beneficial, because the areas are adapted to high levels of acidity. March 2010 201 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2 3 4 5 6 /lAppalacfiian Plateau Figure 6-1. Ecosystems sensitive to acidifying deposition in the Eastern U.S. (Note that Florida represents a special case where high levels of natural acidification exist unrelated to deposition) This map does not include all sensitive areas in the U.S. Certain mountainous areas of the Western U.S. are also sensitive to acidifying deposition. March 2010 202 Draft -Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 6.4 WHAT ARE THE APPROPRIATE SPATIAL EXTENTS OF THE 2 BOUNDARIES FOR EVALUATING AAPI? WITHIN THOSE 3 BOUNDARIES, WHAT ARE THE APPROPRIATE STATISTICS TO 4 USE IN CALCULATING THE PARAMETERS OF THE AAPI, E.G. 5 G, VNOY, Vs, AND NHX? WITHIN THOSE BOUNDARIES, WHAT S 6 THE APPROPRIATE SPATIAL AVERAGING FOR THE AIR 7 QUALITY INDICATORS NOY AND SOX TO PROVIDE 8 PROTECTION OF PUBLIC WELFARE FROM ADVERSE EFFECTS 9 FROM ACIDIFICATION? 10 [Note to reviewers: This section will be added in the second draft policy assessment. In 11 the second draft we plan to provide initial sets ofparameteir values for acid sensitive areas of 12 the U.S., and include an exploration of how the standard might be specified for areas of the U.S. 13 that are not sensitive to deposition o/"NOx and SOX. In addition, we plan to discuss the 14 correlation between the extent of a spatial area and the importance of evaluating alternative 15 percentiles of critical loads to protect a percentage of water bodies in an area, and to discuss 16 how averaging of the VNOY, FSOX, andNHx should be conducted to best represent the 17 parameters for an area.] 18 6.5 WHAT ARE THE OPTIONS FOR SPECIFYING THE TARGETS 19 FOR THE ECOLOGICAL INDICATOR FOR AQUATIC 20 ACIDIFICATION? 21 Chapter 5 discusses the rationale for use of ANC as the ecological indicator best suited to 22 reflect the sensitivity of aquatic ecosystems to acidification. ANC as an indicator of acidification 23 is causally linked to a number of measures of adversity to ecosystems, including declines in fish 24 populations and diversity of aquatic species. ANC is also causally linked with deposition of 25 nitrogen and sulfur. ANC is thus ideally suited to serve as the bridge between deposition and 26 ecological effects. As such, staff concludes that ANC is the best available choice as the 27 ecological indicator. CASAC has agreed that ANC represents a suitable ecological indicator for 28 aquatic acidification (EPA-CASAC-09-013). Results from the REA confirm that ANC may be March 2010 203 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 used to establish impacts from current depositional loadings (REA 4.2.6). As explained above, 2 ANC is an indicator of the effects expected to occur given the natural buffering capacity of an 3 ecosystem and the loadings of nitrogen and sulfur resulting from atmospheric deposition. A 4 target ANC limit based on a desired level of protection is an important input to the decisions of 5 the level of AAPI and the percent of ecosystems to be protected. 6 6.5.1 What levels of impairment are related to alternative levels of ANC? 7 As discussed in Chapters 2, 3, and 4, specific levels of ANC are associated with differing 8 levels of ecosystem impairment, with higher levels of ANC resulting in fewer ecosystem 9 impacts, and lower levels resulting in both higher intensity of impacts and a broader set of 10 impacts. Logistic regression of species presence/absence data against ANC provides a 11 quantitative dose-response function, which indicates the probability of occurrence of an 12 organism for a given value of ANC. For example, the number offish species present in a 13 waterbody has been shown to be positively correlated with the ANC level in the water, with 14 higher values supporting a greater richness and diversity offish species (Figure 6-2). The 15 diversity and distribution of phyto-zooplankton communities are also positively correlated with 16 ANC. 17 The relationship between ANC and ecosystem impacts is non-linear, with a sigmoidal 18 shape. For freshwater systems, ANC levels can be grouped into five major classes: <0, 0-20, 20- 19 50, 50-100, and >100 microequivalents per liter (ueq/L), with each range representing a 20 probability of ecological damage to the community. The five categories of ANC and expected 21 ecological effects are described Table 2-1 in Chapter 2 and are supported by a large body of 22 research completed throughout the eastern United States (Sullivan et al., 2006). 23 Biota are generally not harmed when ANC values are >100 microequivalents per liter 24 (ueq/L). The number offish species also peaks at ANC values >100 ueq/L. This suggests that at 25 ANC greater than 100, little risk from acidification exists in most aquatic ecosystems. At ANC 26 levels below 100 ueq/L, overall health of an aquatic community can be maintained; however, 27 fish fitness and community diversity begin to decline. At ANC levels between 100 and 50 ueq/L, 28 the fitness of sensitive species (e.g., brook trout, zooplankton) also begins to decline. When ANC 29 concentrations are <50 ueq/L, negative effects on aquatic biota are observed, including large 30 reductions in diversity offish species, and changes in health offish populations, affecting March 2010 204 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 reproductive ability and fitness. ANC levels below 50 are generally associated with death or loss 2 of fitness of biota that are sensitive to acidification. (ISA 5.2.2.1 and REA 5.2.1.2). 3 Based on the field data from the Adirondacks and Shenendoah case study areas, ANC 4 levels less than 50 are clearly adverse to ecosystem health, and are likely to lead to reductions in 5 ecosystem services related to recreational fishing. ANC levels between 50 and 100 are 6 potentially adverse to ecosystem health, and may result in losses in ecosystem services, but the 7 effects are less severe and greater uncertainty exists as to the magnitude of ecosystem service 8 impacts. A more comprehensive discussion of uncertainties related to ecological effects at 9 different ANC levels and related ecosystem services will be included in the second draft policy 10 assessment. 11 The implications of the data from the Adirondacks and Shenendoah case study areas for 12 relating ANC to adverse ecological impacts is transferable to other acid sensitive areas of the 13 U.S. The relationship between species diversity and ANC is quite similar between the two case 14 study areas (see REA Figure 4.2-1), which have different water body types and different 15 geological and topographical features. While the species composition and thereby relative 16 sensitivities of species are likely to vary across the landscape, the rate of impact is likely to be 17 similar. The plot in Figure 6-2 shows a rapid decrease in fish species between an ANC of 100 18 and an ANC of 0. This trend is what would be expected in many systems given similar changes 19 in ANC. 20 Consideration of the appropriate levels of ANC to target in the standard to reduce the 21 likelihood of effects from aquatic acidification can be based upon the above presented categories 22 of aquatic status in Table 2-1. Using this information as well as information provided by both the 23 ISA and REA, the lowest two categories (0 and 0<20) would appear inadequate to protect against 24 catastrophic loss of ecosystem function. While ecological effects occur at ANC levels below 50, 25 the degree and nature of those effects is less significant than at levels below 20. Therefore, three 26 levels of ANC - 20, 50, and 100 - would provide the Administrator with reasonable range of 27 options in designing an AAPI for protecting public welfare. 28 Given the level of ecosystem impairment occurring at ANC levels below 50, staff suggest 29 that the greatest support is for the Administrator to consider a range for the target ANC between 30 50 and 100 as a basis for the design of the standard. Selection of target ANC values closer to 50 31 places less weight on the vulnerability of sensitive aquatic ecosystems, while selection of target March 2010 205 Draft-Do Not Quote or Cite ------- 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX ANC values closer to 100 places more weight on sensitive species within acid sensitive ecosystems. Staff conclude that while target ANC values between 20 and 50 will not result in complete impairment of aquatic systems, the level of damages due to ANC as you get lower in this range are highly likely to result in adverse impacts to public welfare in many locations, due to the significant reductions in the number offish species in affected waterbodies, and the reductions in health and reproductive fitness offish populations and other aquatic organisms. Severe Elevated Moderate 14 . ....... . -200 -100 0 100 200 300 400 ANC(ueq/L) 500 Figure 6-2. Number offish species per lake or stream versus ANC level and 19 aquatic status category (colored regions) for lakes in the Adirondack Case Study Area (Sullivan et al., 2006). The target ANC level specified in designing the standard is only one part in determining the overall protectiveness of the standard. The degree of protectiveness is based on all elements of the standard, including the target ANC, the size of the spatial areas over which the standard is applied, the percent of aquatic ecosystems targeted within a spatial area that is selected by the Administrator to achieve the selected ANC level, the atmospheric indicator, the method for calculating g, the calculated values for the deposition transformation ratios (VNOX and VSOX), 19 The aquatic status categories are based on the literature and are discussed in detail in the REA (REA Appendix 4- 20) March 2010 206 Draft -Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 and the calculated value for reduced nitrogen deposition (NHX). There are widely varying 2 degrees of uncertainty associated with all of these elements, some being much more certain and 3 others being much less certain. The specified target ANC level is a crucial part of developing a 4 standard that is requisite to protect, but it is the overall design and content of the standard that 5 must be considered in judging the adequacy of protection it provides. 6 Consideration of the target ANC should also reflect that an adequate level of ANC should 7 protect against episodic as well as long term effects. Selecting a higher chronic ANC level can 8 provide greater protection against short term peaks in acidification. In addition, selection of ANC 9 values in the range of 20 to 50 provides less protection against these short term episodic effects. 10 Selection of target ANC values in the range from 50 to 100 provides additional protection 11 against episodic peaks in acidification. 12 When considering the appropriate level of a standard to protect against aquatic 13 acidification, it is necessary to take into account both the time period desired for recovery as well 14 as the potential of recovery. Ecosystems become adversely impacted by acidifying deposition 15 over long periods of time and have variable time frames and abilities to recover from such 16 perturbations. Modeling presented in the REA (REA Section 4.2.4) shows the estimated ANC 17 values for Adirondack lakes and Shenandoah streams under pre-acidification conditions and 18 indicates that for a small percentage of lakes and streams, natural ANC levels would have been 19 below 50. Therefore, for these waterbodies, no reduction in input is likely to achieve an ANC of 20 50 or greater. Conversely, for some lakes and streams the level of perturbation from long periods 21 of acidifying deposition has resulted in very low ANC values compared to estimated natural 22 conditions. For such waterbodies, the time to recovery would be largely dependent on future 23 inputs of acidifying deposition. These concepts become important in the consideration of the 24 desired level of protection of a standard and will be discussed further in the next draft of this 25 document. March 2010 207 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 6.6 WHAT ARE THE APPROPRIATE AMBIENT AIR MONITORING 2 METHODS TO CONSIDER IN DEVELOPING THE STANDARDS? 3 6.6.1 What measurements would be used to characterize NOy and SOX ambient air 4 concentrations for the purposes of the AAPI based standard? 5 Ambient NOy, gaseous 862 and particulate sulfate concentrations would be used in 6 determining compliance with the AAPI. This would require measurements of NOy, sulfate and 7 sulfur dioxide, all which are conducted as part of current routine monitoring networks (section 8 3.2). There are issues requiring resolution associated with Federal Reference or Equivalency 9 Measurement (FRM/FEM) status of measurement techniques, that to date have served as 10 supplemental information, which will require resolution. A FRM for SC>2 exists, but not for NOy 11 or sulfate. Only recently have NOy measurements, which historically were viewed as research 12 venue measurements, been incorporated as "routine" observations, partly as a result of the NCore 13 program. Acquiring FRM status may require better characterization of the conversion 14 efficiencies, mass loss and clear guidance on operating and siting procedures. Particulate sulfate 15 has been measured for several years in the IMPROVE, CASTNET and EPA CSN networks. The 16 nation has over 500 24-hour average, every third day sulfate measurements produced by the 17 PM2.5 speciation networks (IMPROVE and EPA CSN) and nearly 80 CASTNET sites that 18 provide continuous weekly average samples of sulfate with an open inlet accommodating all 19 particle sizes. However, with minor exceptions, the PM2.5 fraction accounts for nearly all sulfate 20 mass. The sample collection period is not an issue for gaseous measurements of NOy and SO2 21 that operate continuously. Some concerns have been raised about the possibility of exclusion of 22 coarse particles from NOy samplers operating at low flow conditions as well as potential 23 difficulties of reducing organically bound and mineralized nitrate. These conversion efficiency 24 and particle size fraction issues are viewed by EPA as relatively minor mass accounting issues 25 that require more clarification but not necessarily technical resolution. 26 6.6.2 What sampling frequency would be required? 27 The averaging time for the standard is likely to be an annual average. Conceptually, 28 extended sampling periods as long as one year would be adequate for the specific purposes of 29 comparing to a standard. However, future assessments that characterize acidification and form March 2010 208 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 the scientific basis for subsequent standards reviews and allow for systematic checking of 2 progress through accountability procedures benefit from more highly resolved data, especially 3 the evaluation of air quality models that are key components of N/S deposition assessments. In 4 addition, many of the monitoring approaches that are used throughout the nation sample (or at 5 least report out) on daily (PM2.5 chemical speciation), weekly (CASTNET) and hourly (all 6 inorganic gases) periods. There is a tradeoff to consider in sampling period design. For example, 7 the weekly CASTNET collection scheme covers all time periods throughout a year, but only 8 provides weekly resolution that misses key temporal and episodic features valuable for 9 diagnosing model behavior. The every third day, 24-hour sampling scheme used in IMPROVE 10 and EPA speciation monitoring does provide more information for a specific day of interest yet 11 misses 2/3 of all sampling periods. The missing sampling period generally is not a concern when 12 aggregating upward to a longer term average value as the sample number adequately represents 13 an aggregated mean value. Additionally, there is a benefit to leveraging existing networks which 14 should be considered in sampling frequency recommendations. A possible starting point would 15 be to assume gaseous oxidized species, NOy and SC>2, are run continually all year reporting 16 values every hour, consistent with current routine network operations. Sulfate sampling periods 17 should coincide with either the chemical speciation network schedules or CASTNET. There are 18 advantages to coordinating with either network. Ammonia gas and ammonium ion present 19 challenges in that they are not routinely sampled and analyzed for, and the combined quantity, 20 NHX is of interest. Because NHX is of interest, some of the problems of volatile ammonia loss 21 from filters may be mitigated. However, for model diagnostic purposes, delineation of both 22 species at the highest temporal resolution is preferred. While levels of deposited reduced 23 nitrogen would be specified by EPA for purposes of the APPI, monitoring of reduced nitrogen 24 would be important but would not be used in the APPI itself. 25 6.6.3 What are the spatial scale issues associated with monitoring for compliance, 26 and how should these be addressed? 27 The observation network for NOy, NHX and SOX is very modest and includes a 28 monitoring network infrastructure that is largely population oriented. While there is platform and 29 access infrastructure support provided by CASTNET, NADP and IMPROVE, those locations by 30 themselves are not likely to provide the needed spatial coverage to address acid sensitive March 2010 209 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 watersheds across the United States. Ambient monitoring at every watershed may not be required 2 due to the nature of the ambient air quality in acid sensitive areas. An understanding of the 3 spatial variability of NOy, NHX, sulfate and SO2 will help inform monitoring. Critical load 4 models are based on annual averages, which effectively serves to dampen much of the spatial 5 variability. Furthermore, the development of an area-wide depositional load tradeoff curve 6 implies focus on region wide characterization. Toward that end, CMAQ concentration fields will 7 provide insight into the likely spatial representativeness of monitors leading to efficient 8 application of monitoring resources. For example, the CMAQ based spatial coefficient of 9 variation (standard deviation/mean) of oxidized nitrogen in the Adirondacks was 1.46%. 10 Improved dry deposition estimates will result from enhancements of ambient monitoring 11 addressing the N/S secondary standards as each additional location could serves a similar role 12 that existing CASTNET sites provide in estimating dry deposition. 13 6.7 TAKING INTO CONSIDERATION INFORMATION ABOUT 14 ECOSYSTEM SERVICES AND OTHER FACTORS RELATED TO 15 CHARACTERIZING ADVERSITY FOR THE ECOLOGICAL 16 EFFECTS BEING ASSESSED IN THIS REVIEW, WHAT IS AN 17 APPROPRIATE RANGE OF ALTERNATIVE STANDARDS FOR 18 THE AGENCY TO CONSIDER? 19 The secondary NAAQS will reflect the public welfare policy judgments of the 20 Administrator, based on the science, as to the level of air quality which is requisite to protect the 21 public welfare from any known or anticipated adverse effects associated with the pollutant in the 22 ambient air. The exposure and risk assessment provide information regarding the effects 23 associated with a number of different welfare endpoints at different levels of air quality, 24 expressed in terms of the joint annual mean concentrations of NOX and SOX determined such that 25 specific levels of ecosystem protection (for example, ANC greater than 50) are met. Staff also 26 recognizes that in certain naturally acidic ecosystems, even though the ecological benchmarks 27 are exceeded, e.g. ANC may be quite low; NOX and SOX are not contributing to effects because 28 those systems have chronic natural acidity and will not benefit from reductions in atmospheric 29 deposition. The secondary NAAQS are not intended to provide protection in these types of March 2010 210 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 naturally acidic systems. As noted earlier, we will be exploring methods to address the design of 2 the standard relative to these naturally acidic systems in the second draft policy assessment. The 3 secondary NAAQS are focused on providing protection in areas where ambient NOX and SOX are 4 resulting in effects in ecosystems with low natural levels of acidification that are highly sensitive 5 to additional inputs of acid deposition. 6 Staff believes that ecosystem effects of NOX and SOX deposition in aquatic ecosystems 7 are an important public welfare effect of concern. There are several sources of benchmark values 8 for ANC that can help to inform a determination of adversity. [Additional information on 9 benchmark values will be provided in the second draft policy assessment] Staff concludes that 10 achieving ANC in the range of 50 to 100 would be likely to provide adequate protection against 11 the effects of acidification on ecosystems. 12 Based on our analyses of risks of impacts on aquatic species diversity and fitness and on 13 the basis of the scientific effects literature, we anticipate that achieving the upper end of this 14 ANC range would substantially decrease the effects of acidification due to NOX and SOX on 15 aquatic ecosystems. Additionally, it is anticipated that achieving the upper end of this range 16 would provide increased protection from NOX and SOX in areas with higher levels of variability 17 in ecosystem sensitivity due to variability in meteorology, bedrock geology, topography, land 18 use characteristics, or reduced nitrogen deposition. 19 These ANC levels are estimated to protect sensitive aquatic ecosystems from significant 20 negative effects of NOX and SOX deposition on aquatic biota, including large reductions in 21 diversity offish species, and changes in health offish populations, affecting reproductive ability 22 and fitness. It is recognized, however, that a standard set within this range would not protect the 23 most sensitive aquatic ecosystems or species within those ecosystems from the effects of NOX 24 and SOX. At ANC levels below 100, while overall health of an aquatic community can be 25 maintained, ANC levels are expected to be such that fish fitness and community diversity begin 26 to decline. At ANC levels between 100 and 50, ANC levels are expected to be such that the 27 fitness of sensitive species (e.g., brook trout, zooplankton) also begins to decline. Staff notes that 28 at levels of ANC above 100, biota are generally not harmed. As such, achieving an ANC of 29 greater than 100 would be expected to result in little damage from NOX and SOX deposition to 30 aquatic ecosystems. March 2010 211 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 Specifying an appropriate range of levels for an AAPI standard that is designed and 2 specified as discussed above involves consideration of the degree to which any specific AAPI 3 would lead to achieving the desired ANC level, and a judgment as to the degree of protection of 4 public welfare that is warranted. In general, staff initially conclude that it would be appropriate 5 for the Administrator to consider an AAPI in the range of 50 to 100. Selection of a range of 6 AAPI and selection of a specific level of AAPI within that range should incorporate a wide 7 number of considerations, including the percent of water bodies within acid sensitive areas that 8 the Administrator determines should be protected at the targeted ANC level. 9 The Administrator should consider the uncertainties in the ecological effects observed in 10 the literature and the adversity to public welfare associated with those effects. In determining the 11 requisite level of protection for the public welfare from effects on aquatic ecosystems, the 12 Administrator will need to weigh the importance of the predicted risks of these effects in the 13 overall context of public welfare protection, along with a determination as to the appropriate 14 weight to place on the associated uncertainties and limitations of this information. 15 In addition, selection of a specific level of AAPI should consider uncertainties in the 16 design and calculation of the parameters included in the AAPI, including uncertainties in the 17 characterization of natural background ANC (indicated by g in the AAPI equation), spatial and 18 temporal averaging of aggregate effective deposition velocities (indicated by VNOY and VSOX in 19 the AAPI equation), and spatial and temporal averaging of NHX deposition (indicated by NHX in 20 the AAPI equation). 21 March 2010 212 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX i 7. CO-PROTECTION FOR OTHER EFFECTS USING 2 STANDARDS TO PROTECT AGAINST ACIDIFICATION 3 To this point, the standard for NOX and SOX centers on ecosystem protection against 4 aquatic acidification. This chapter focuses on the level of co-protection that this standard would 5 provide for other ecological effects, including terrestrial acidification, terrestrial nutrient 6 enrichment, and estuarine eutrophication. 7 7.1 TO WHAT EXTENT WOULD A STANDARD SPECIFICALLY 8 DEFINED TO PROTECT AGAINST AQUATIC ACIDIFICATION 9 LIKELY PROVIDE PROTECTION FROM TERRESTRIAL 10 ACIDIFICATION? 11 In order to understand the level of protection provided by a NOX/SOX standard based on 12 aquatic acidification to protect against terrestrial acidification effects, an analysis was conducted 13 comparing the critical loads for lakes and streams that would be developed to protect for an 14 aquatic ANC of 50 to the critical loads to protect for either a terrestrial Be: Al ratio of 1.2 or 10 15 averaged across a watershed area. See Appendix B for full analysis results. The analysis selected 16 16 watersheds with 29 lakes in the Adirondacks case study area, 4 watersheds randomly selected 17 from each of 4 categories of sensitivity reported in the REA: highly sensitive, moderately 18 sensitive, low sensitivity, and not sensitive. In the Shenandoah case study area, there were a 19 limited number of watersheds in the low sensitivity and not sensitive range, so 18 of the 20 20 streams in 16 watersheds selected were located in highly and moderately sensitive categories. 21 Results for the Adirondacks showed that critical loads for 29 lakes at an ANC of 50 were 22 lower for 13 lakes than the critical load for the terrestrial watershed areas at a Bc:Al ratio of 10 23 and for 21 lakes at a Bc:Al ratio of 1.2. Perhaps more significant was the result that 13 of the 16 24 lakes in the highly and moderately sensitive areas had a lower critical load than the Bc:Al 10 25 areas and 16 of 16 lakes in the highly and moderately sensitive areas had lower critical loads 26 than the Bc:Al 1.2 areas. The Shenandoah region reflected similar results. See table 7.1 below 27 for tabulated results. 28 March 2010 213 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX Table 7-1. Results of comparing aquatic ANC50 critical loads to average terrestrial watershed area Bc:Al ratios. Left numbers in each column are the number of lakes or streams that had a lower critical load than the terrestrial calculated critical load. Right numbers in each column are the number of lakes that had a higher critical load than the watershed calculated terrestrial critical loads. Adirondack Be: Al 10 Adirondack Be Al 1.2 Shenandoh Bc:Al 10 Shenandoh Bc:Al 1.2 Highly Sensitive 7-0 7-0 13-0 13-0 Moderately Sensitive 6-3 9-0 5-0 5-0 Low Sensitivity 0-7 5-2 0-1 0-1 Not Sensitive 0-6 0-6 0-1 0-1 2 In summary, a comparison of the terrestrial and aquatic critical acid loads for watersheds 3 in the Adirondacks and Shenandoah Case Study Areas indicated that, in general, the aquatic 4 critical acid loads offered greater protection to the watersheds than did the terrestrial critical 5 loads. Generally in situations where the terrestrial critical loads were more protective, the lakes 6 or streams in the watershed were rated as having "Low Sensitivity" or "Not Sensitive" to 7 acidifying nitrogen and sulfur deposition. Conversely, when the water bodies were more 8 sensitive to deposition ("Highly Sensitive" or "Moderately Sensitive"), the aquatic critical acid 9 loads generally provided a greater level of protection against acidifying nitrogen and sulfur 10 deposition in the watershed. In the next draft of the Policy Assessment Document, we intend to 11 expand this analysis by comparing more levels of ANC to other Bc:Al ratios. 12 7.2 TO WHAT EXTENT WOULD A STANDARD SPECIFICALLY 13 DEFINED TO PROTECT AGAINST AQUATIC ACIDIFICATION 14 LIKELY PROVIDE PROTECTION FROM TERRESTRIAL 15 NUTRIENT ENRICHMENT? 16 This question will be answered in the next draft of the Policy Assessment Document. 17 Once maximum depositonal loads are calculated for broad areas, we can compare the derived 18 maximum NOy limits to nutrient enrichment benchmarks found in the REA. Benchmarks for 19 lichens, grasses, mychorrizae, and diatoms will be compared to the aquatic acidification limits 20 for nitrogen. March 2010 214 Draft -Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 7.3 TO WHAT EXTENT WOULD A STANDARD SPECIFICALLY 2 DEFINED TO PROTECT AGAINST AQUATIC ACIDIFICATION 3 LIKELY PROVIDE PROTECTION FROM AQUATIC NUTRIENT 4 ENRICHMENT? 5 The REA found that deposition of reactive nitrogen contributed to eutrophication of 6 estuaries; however, it was also noted that atmospheric deposition of nitrogen is only part of the 7 total nitrogen load to the estuaries. Due to the complications of separating out the effects of 8 atmospheric deposition from the effects of other nitrogen loads, CASAC did not recommend that 9 a secondary NAAQS be set to specifically protect against estuarine eutrophication. In the next 10 draft of the Policy Assessment Document, we will attempt to analyze the benefit to the 11 Chesapeake Bay that attaining an aquatic acidification standard would provide by decreasing 12 nitrogen deposition to the watershed. 13 14 March 2010 215 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX i 8. CONSIDERATION OF ISSUES REGARDING REDUCED 2 AND OXIDIZED FORMS OF NITROGEN 3 [To be added in the second draft Policy Assessment] 4 5 March 2010 216 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 9. INITIAL CONCLUSIONS 2 Staff initial conclusions on the elements of the secondary NOX and SOX standards for the 3 Administrator's consideration in making decisions on the secondary NOX and SOX standards are 4 summarized below, together with supporting conclusions from previous chapters. We recognize 5 that selecting from among alternative policy options will necessarily reflect consideration of 6 qualitative and quantitative uncertainties inherent in the relevant evidence and in the assumptions 7 of the quantitative exposure and risk assessments. Any such standard should protect public 8 welfare from any known or anticipated adverse effects associated with the presence of the 9 pollutant(s) in the ambient air. In providing these options for consideration, we are mindful that 10 the Act requires standards that, in the judgment of the Administrator, are requisite to protect 11 public welfare. The standards are to be neither more nor less stringent than necessary. 12 To evaluate whether the current secondary NAAQS is adequate or whether consideration 13 of revisions is appropriate, the conclusions and options for the Administrator to consider in this 14 review are based on effects-, exposure- and risk-based considerations. The exposure and risk 15 assessments reflect the availability of new tools, assessment methods, and a larger and more 16 diverse body of evidence than was available in the last reviews. We have taken a weight of 17 evidence approach that evaluates information across the variety of research areas described in the 18 ISA and in addition includes assessments of air quality, exposures, and qualitative and 19 quantitative risks associated with alternative air quality scenarios. 20 Staff notes that since the last review, additional policy-relevant developments have 21 occurred that may also warrant consideration by the Administrator when making decisions about 22 what is requisite to protect public welfare. The NRC report (described in Chapter 6) states: 23 "Whatever the reason that led EPA to use identical primary and secondary NAAQS in the past, it 24 is becoming increasingly evident that a new approach will be needed in the future. There is 25 growing evidence that the current forms of the NAAQS are not providing adequate protection to 26 sensitive ecosystems and crops" (NRC, 2004). 27 The last review raised the following key issues as a rationale for not setting a separate 28 standard for NOX to protect against acidification and nutrient enrichment effects in sensitive 29 ecosystems: 30 1) Lack of enough consistent information to support a revision of the current secondary 31 standard to protect these aquatic systems. March 2010 217 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 2) Lack of adequate quantitative evidence on the relationship between deposition rates 2 and environmental impacts 3 3) Significant uncertainties with regard to the long-term role of nitrogen deposition in 4 surface water acidity and with regard to the quantification of the magnitude and 5 timing of the relationship between atmospheric deposition and the appearance of 6 nitrogen in surface water. 7 In this current review, staff concludes that important new information has become 8 available since the last review that supports revising the current NOX and SOX standards. 9 Specifically, the ISA has concluded that there are causal relationships between NOX and SOX 10 acidifying deposition and effects on aquatic and terrestrial ecosystems, and the ISA and REA 11 provide substantial quantitative evidence of effects occurring in locations that meet the current 12 NC>2 and SC>2 standards. In addition, substantial new information, based on observational data 13 and rigorous atmospheric modeling, has become available regarding the role of both nitrogen and 14 sulfur deposition in acidification of sensitive water bodies. This information is sufficient to 15 inform the development of revised secondary standards for NOX and SOX to protect against the 16 effects of acidification20. While there is also new information available on the role of nitrogen 17 deposition on nutrient enrichment effects in terrestrial and aquatic ecosystems, and the ISA 18 concludes there is a causal relationship between NOX and nutrient enrichment effects, for this 19 first draft policy assessment, staff have focused on acidification effects due to the substantially 20 greater amount of information available to inform the development of secondary standards. 21 Staff highlights the progress made in considering the joint nature of ecosystem responses 22 to acidifying deposition of NOX and SOX, and notes that the ability to consider revisions to the 23 NOX and SOX secondary standards has been enhanced by our ability to consider a joint standard 24 for NOX and SOX to protect against acidification effects. The development of an appropriate form 25 of the standard linked to a common indicator of aquatic acidification, ANC, is also a significant 26 step forward, as it allows for development of a standard for aquatic acidification designed to 27 provide generally the sme degree of protection across the country, while still reflecting the 28 underlying variability in ecosystem sensitivity to acidifying NOX and SOX deposition. 29 20 As we have note earlier in the document, in this draft we have focused on aquatic acidification. However, in the second draft policy assessment we plan to more fully explore the possibility of expanding the conceptual model to address terrestrial acidification. March 2010 218 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 9.1 CONCLUSIONS 2 As noted throughout this document, because of the complex interactions between NOX 3 and SOX in the atmosphere and their impacts once deposited in ecosystems, the consideration of 4 indicators, averaging times, forms, and levels for the two pollutants is being conducted jointly. In 5 addition, as discussed in Chapters 5 and 6, we are considering structures for the standards that 6 reflect a more scientifically derived understanding of the relationships between atmospheric 7 concentrations of NOX and SOX and the primary indicators of ecosystem impacts. 8 With respect to soil and water effects information, we have evaluated the conclusions 9 drawn at the end of the last review in light of more recent evidence from studies for a variety of 10 ecological effects endpoints. We place greater weight on U.S. studies due to the species-, site-, 11 and climate-specific nature of ecological responses. With respect to quantitative exposure- and 12 risk-based considerations, we have relied on both monitored and modeled NOX and SOX ambient 13 concentrations and related deposition, as described in Chapter 3 of the REA. 14 Uncertainties associated with the exposure and risk assessments are also discussed, 15 including, where possible, some sense of the direction and/or magnitude of the uncertainties that 16 should be taken into account as one considers these estimates. As with any analysis that relies on 17 complex scientific models, there are a number of unknown and unquantifiable sources of 18 uncertainty. However, each model that has been applied in the risk and exposure assessment 19 represents the best available science and the models have all been subject to substantial levels of 20 peer-review. 21 The following secondary NAAQS conclusions encompass the breadth of policy-relevant 22 considerations described in this policy assessment: 23 (1) Based on the policy-relevant findings from the ISA described in Chapter 2, and while 24 recognizing that important uncertainties and research questions remain, staff conclude 25 that great progress has been made since the last reviews of the secondary standards 26 for NOX and SOX. We generally find support in the available effects-based evidence 27 for consideration of NOX and SOX standards that are at least as protective as the 28 current standard and do not find support for consideration of NOX and SOX standards 29 that are less protective than the current standard. The staff also concludes that 30 consideration of joint standards for NOX and SOX is appropriate given the common 31 atmospheric processes governing the deposition of NOX and SOX to sensitive March 2010 219 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 ecosystems, and given the combined effects of N and S deposition on acidification of 2 soil and water. 3 (2) Staff concludes that ambient NOX is a significant component of atmospheric nitrogen 4 deposition, even in areas with relatively high rates of deposition of reduced nitrogen. 5 Staff make this conclusion based on the analysis in Chapter 3 of the REA, which 6 provides a thorough assessment of the contribution of NOX to nitrogen deposition 7 throughout the U.S., and the relative contributions of ambient NOX and reduced forms 8 of nitrogen. 9 (3) Staff concludes based on the case study results provided in the REA, that current 10 levels of NOX and SOX are associated with deposition that leads to ANC values below 11 benchmark values that cause ecological harm and losses in ecosystem services. Staff 12 concludes that the evidence and risk assessment support strongly a relationship 13 between atmospheric deposition of NOX and SOX and ANC, and that ANC is an 14 excellent indicator of aquatic acidification. Staff also concludes that at levels of 15 deposition associated with NOX and SOX concentrations at or below the current 16 standards, ANC levels are expected to be below benchmark values that are associated 17 with significant losses in fish species richness, which is associated with reductions in 18 recreational fishing services. While there are many other ecosystem services 19 potentially affected by reductions in ANC, including subsistence fishing, natural 20 habitat provision, and biological control, confidence in the specific translation of 21 ANC values to these additional ecosystem services is much lower. 22 (4) Losses in aquatic resources associated with ANC levels below 50 are clearly 23 associated with significant losses in economic value. Based on the best available data, 24 just in the northeastern U.S., current acidification levels are resulting in $4 million to 25 $300 million in damages annually from lost recreational fishing. This estimate 26 represents only a fraction of the total economic value of ecosystem damages as many 27 impacted resources are not amenable to economic valuation methods. In addition, 28 economic damages are also likely to occur in other areas affected by acidification, 29 including New England, the Appalachian Mountains (northern Appalachian Plateau 30 and Ridge/Blue Ridge region), and the Upper Midwest. Staff concludes that reducing 31 acidifying deposition of NOX and SOX will result in improvements in public welfare March 2010 220 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 by increasing the quantity and quality of ecosystem services, including recreational 2 fishing and other services associated with improved water quality. 3 (5) Staff initially concludes based on the case study results that current levels of ambient 4 NOX and SOX are associated with deposition that leads to BC:A1 values below 5 benchmark values that cause ecological harm and losses in ecosystem services. Staff 6 concludes that the evidence and risk assessment support strongly a relationship 7 between atmospheric deposition of NOX and SOX and BC:A1, and that BC:A1 is a 8 good indicator of terrestrial acidification. Staff also concludes that at levels of 9 deposition associated with NOX and SOX concentrations at or below the current 10 standards, BC:A1 levels are expected to be below benchmark values that are 11 associated with significant losses in tree health and growth, which are associated with 12 reductions in timber production. While there are many other ecosystem services, 13 including maple syrup production, natural habitat provision, and regulation of water, 14 climate, and erosion, potentially affected by reductions in BC:A1, confidence in the 15 specific translation of BC:A1 values to these additional ecosystem services is much 16 lower. 17 (6) On the basis of the acidification and nutrient enrichment effects that have been 18 observed to still occur under current ambient conditions and those predicted to occur 19 under the scenario of just meeting the current secondary NAAQS, staff concludes that 20 the current secondary NAAQS are inadequate to protect the public welfare from 21 known and anticipated adverse welfare effects from aquatic and terrestrial 22 acidification associated with deposition of NOX and SOX.. As discussed above, this 23 conclusion derives from several lines of evidence. 24 (7) Staff has concluded, based on the completeness of the available evidence and 25 quantitative risk information, that effects due to aquatic and terrestrial acidification 26 are most suitable for defining secondary standards for NOX and SOX. Staff notes that 27 in developing a standard designed to protect against the effects of acidification due to 28 deposition of NOX and SOX, the resulting standards may not provide protection 29 against known effects associated with nutrient enrichment in aquatic and terrestrial 30 ecosystems. March 2010 221 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 (8) It is appropriate to consider using indicators other than NC>2 and 862 as the indicators 2 for a standard that is intended to address the ecological effects associated with 3 deposition of NOX and SOX to sensitive ecosystems. Given the reasons discussed in 4 Chapters 2, 4, and 5 of this policy assessment, staff concludes thatNOx, as defined in 5 the CAA, is best represented by the atmospheric indicator NOy, defined as NO2 + NO 6 + HNO3 + PAN +2N2O5 + HONO+ NO3 + organic nitrates + paniculate NO3 is the 7 more appropriate indicator of oxides of nitrogen, and that SOX, defined to include 8 sulfur monoxide (SO), sulfur dioxide, sulfur trioxide (SO3), and disulfur monoxide 9 (S2O), and particulate-phase S compounds, is the more appropriate indicator of 10 oxides of sulfur. 11 (9) It is appropriate to use the annual average of concentrations of NOy and SOX as the 12 averaging time for the secondary standards, based on the chronic nature of 13 acidification, and the protection against episodic acidification provided by a standard 14 based on annual average concentrations. 15 (10) It is appropriate to consider changing the form of the secondary standards for NOX 16 and SOX as the current form does not take into account the linkages between NOX and 17 SOX in the causation of effects associated with acidification of aquatic ecosystems. 18 Based on the causal linkages between NOX and SOX, deposition of N and S, and the 19 indicator of acidification, ANC, staff concludes that the current forms should be 20 replaced with an atmospheric acidification potential index (AAPI), which reflects the 21 important roles of underlying ecosystem characteristics, determinants of deposition, 22 and reduced nitrogen deposition in determining the potential effects from deposition 23 of NOX and SOX. 24 (11) Staff initial conclusions regarding the elements of the standard, e.g. the target ANC, 25 spatial extent of areas in which the standard will be evaluated, percentiles of aquatic 26 ecosystems within sensitivity classes to be protected for alternative target ANC 27 values, calculated values of deposition transformation ratios, natural buffering 28 capacity, and reduced nitrogen deposition will be provided in the second draft of the 29 policy assessment. In addition, staff initial conclusions regarding consideration of 30 uncertainty and variability in elements of the standard will be developed in the second 31 draft. March 2010 222 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 9.2 SUMMARY OF KEY UNCERTAINTIES AND RESEARCH 2 RECOMMENDATIONS RELATED TO SETTING A SECONDARY 3 STANDARD FOR NOX AND SOX 4 [This section is still under development. Summary of key uncertainties to be added in 5 second draft policy assessment. Research and data needs are partial lists that will be more 6 completely developed in subsequent versions.] 7 9.2.1 Research Needs to Reduce Uncertainty in the Next Review (focused on 8 aquatic acidification) 9 Based on the information presented in this policy assessment, several information gaps 10 arise that suggest further research is needed in the following areas: 11 • Developing relationships between aquatic acidity as measured by ANC, and effects on 12 ecological effects and ecosystem services, especially due to incremental changes 13 • Developing nationwide weathering rates, or weathering rates for aquatic ecosystems 14 sensitive to acidification 15 • Developing a better understanding of the uncertainty in critical loads for acidity 16 • Developing methods for calculating critical loads for surface water acidity when data are 17 absent or of poor quality 18 • Evaluating ways to combine multiple critical load estimates for surface waters and soils on 19 a national scale 20 • Estimating ways to determine critical load parameters across different media (e.g., surface 21 waters, soils). 22 9.2.2 Data Needs to Reduce Uncertainty in the Next Review (focused on aquatic 23 acidification) 24 Improved measurements of reduced nitrogen: Nitrification processes within watershed 25 soil, sediment and vegetation systems effectively convert ammonia gas and ammonium ions to 26 nitrates, which contribute to the overall acidifying loads in ecosystems; consequently, the 27 atmospheric contributions of reduced nitrogen must be accounted for in acidification 28 assessments. We would expect that all or a subset of ambient monitoring platforms supporting March 2010 223 Draft-Do Not Quote or Cite ------- Policy Assessment for the Review of the Secondary National Ambient Air Quality Standards for NOX and SOX 1 the N/S secondary standard will measure both ammonia gas and ammonium ion along with 2 oxidized sulfur and nitrogen species. 3 Extended modeling of air quality and deposition to inform monitoring network design: In 4 addition to providing deposition inputs for watershed models and critical loads analysis, the 5 spatial and temporal flexibility afforded by air quality modeling can support monitoring network 6 design and in inform the averaging time period (one or more years) to more appropriately 7 account for inter-annual variability in NOX and SOX concentrations. 8 Development of data fusion approaches to combine model results with observational 9 data: Consideration also will be given to fusing model results with observation fields to improve 10 spatial resolution by taking advantage of the landscape, emissions and meteorological 11 information that affect spatial gradients while relying on observations to reduce the influence of 12 model uncertainties. March 2010 224 Draft-Do Not Quote or Cite ------- 3 APPENDIX A 4 CHAPTER 5: CONCEPTUAL DESIGN OF THE 5 STANDARD 6 7 8 First External Review Draft 9 10 11 12 Prepared by: 13 14 U.S. Environmental Protection Agency 15 Office of Air Quality Planning and Standards 16 Research Triangle Park, NC 27711 17 18 19 ------- ------- Appendix A 1 TABLE OF CONTENTS 2 Appendix A Chapter 5: Conceptual Design of the Standard 1 3 A. 1 Technical summary of methods used in the REA Aquatic Acidification analysis 1 4 A.2 Technical summary of critical loads modeling in the REA 2 5 1.2.1 Preindustrial Base Cation Concentration 5 6 1.2.2 F-factor 6 7 9 LIST OF TABLES 10 Table A. 1. Brief summary of objects and methods used in the REA Aquatic Acidification 11 analysis 1 12 Table A.2 Illustrates SSWC Approach -Environmental Variables 7 13 Table A.3 FAB Approach -Environmental Variables 8 14 15 March 2010 i Draft - Do Not Quote or Cite ------- Appendix A 1 This page intentionally left blank. 2 March 2010 ii Draft - Do Not Quote or Cite ------- Appendix A i APPENDIX A 2 CHAPTERS: CONCEPTUAL DESIGN OF THE STANDARD 3 This is supplemental information to support the discussion of the conceptual design of the 4 standard that is presented in Chapter 5 of the Policy Assessment Document. The aquatic 5 acidification analyses developed in the REA used a number of different models and calculation 6 techniques that are important for the development of the standard. The goal of this Appendix is 7 to summarize information from the REA analysis that is most relevant to the Policy Assessment. 8 A brief summary of the REA analyses are presented in section 1. In section 2 there is a general 9 summary and technical discussion of the critical loads modeling approaches that were used in the 10 REA, followed by a brief description of MAGIC model data requirements. 11 A. 1 TECHNICAL SUMMARY OF METHODS USED IN THE REA 12 AQUATIC ACIDIFICATION ANALYSIS 13 The aquatic acidification analysis is presented in Chapter 4 and Appendix 4 of the REA. 14 The analysis uses multiple techniques to show the relationship between ANC and NOX and SOX 15 deposition, as well as determine the current level of risk to water bodies that occur in sensitive 16 areas. A brief summary of the techniques and objectives of the REA analysis is given in Table 1. 17 Table A.I. Brief summary of objects and methods used in the REA Aquatic Acidification 18 analysis. Technique Time-series graphs of current conditions MAGIC Objectives 1 1 2 3 Data from monitoring networks collected from 1990 to 2006 were plotted to show trends in concentrations of pollutants, deposition and acidification for each case study site. The data included surface water concentration of nitrate, sulfate and ANC; deposition of sulfate and nitrate; as well as air concentration of SOX, NOX and NH4 Used to estimate the relationship between ANC values and anthropogenic NOX and SOX emission from the past (preacidification -I860), present (2002 and 2006) and projected into the future (2020 and 2050). Analysis included 44 lakes from Adirondacks and 60 streams from Shenandoah. Used to develop input parameters for critical loads modeling (i.e. weathering rates) Used for uncertainty analysis March 2010 Draft - Do Not Quote or Cite ------- Appendix A Technique Critical Loads modeling Regional Extrapolation Objectives 1 2 3 1 2 SSWC and FAB models used to calculate critical loads for critical limits of ANC = 0, 20, 50, 100 Critical loads for ANC critical limits calculated for 169 lakes in the Adirondacks and 60 streams in the Shenandoah using water quality data from monitoring sites collected in 2006 Critical loads exceedences calculated by comparing the critical loads that were calculated by SSWC with deposition data from NADP for wet deposition and CMAQ for dry deposition, both for the year 2002 117 of the critical loads calculated for the Adirondacks were extrapolated to lakes defined by the New England EMAP probability survey, representing 1842 lakes, to infer the # of lakes that exceeded their critical load 69 of the critical loads calculated for the Shenandoah were extrapolated to 330 streams based on bed rock geology classification. 1 2 A.2 TECHNICAL SUMMARY OF CRITICAL LOADS MODELING IN 3 THE REA 4 The critical load of acidity for lakes or streams was derived from present-day water 5 chemistry using a combination of steady-state models. Both the Steady-State Water Chemistry 6 (SSWC) model and First-order Acidity Balance model (FAB) is based on the principle that 7 excess base-cation production within a catchment area should be equal to or greater than the acid 8 anion input, thereby maintaining the ANC above a preselected level (Reynolds and Norris, 2001; 9 Posch et al. 1997). These models assume steady-state conditions and assume that all SC>42 in 10 runoff originates from sea salt spray and anthropogenic deposition. Given a critical ANC 11 protection level, the critical load of acidity is simply the input flux of acid anions from 12 atmospheric deposition (i.e., natural and anthropogenic) subtracted from the natural (i.e., 13 preindustrial) inputs of base cations in the surface water. Final Risk and Exposure Assessment 14 September 2009 Appendix 4, Attachment A - 15 Aquatic Acidification Case Study Atmospheric 15 deposition of NOX and SOX contributes to acidification in aquatic ecosystems through the input of 16 acid anions, such as NO3- and SC>42 The acid balance of headwater lakes and streams is 17 controlled by the level of this acidifying deposition of NO3- and SC>42 and a series of 18 biogeochemical processes that produce and consume acidity in watersheds. The biotic integrity 19 of freshwater ecosystems is then a function of the acid-base balance, and the resulting acidity- March 2010 Draft - Do Not Quote or Cite ------- Appendix A 1 related stress on the biota that occupy the water. The calculated ANC of the surface waters is a 2 measure of the acid-base balance: 3 ANC = [BC]* - [AN]* (1) 4 where [BC]* and [AN]* are the sum of base cations and acid anions (NO3- and SC>42), 5 respectively. Equation (1) forms the basis of the linkage between deposition and surface water 6 acidic condition and the modeling approach used. Given some "target" ANC concentration 7 [ANClimit]) that protects biological integrity, the amount of deposition of acid anions (AN) or 8 depositional load of acidity CL(A) is simply the input flux of acid anions from atmospheric 9 deposition that result in a surface water ANC concentration equal to the [ANClimit] when 10 balanced by the sustainable flux of base cations input and the sinks of nitrogen and sulfur in the 11 lake and watershed catchment. 12 Critical loads for nitrogen and sulfur (CL(N) + CL(S) ) or critical load of acidity CL(A) 13 were calculated for each waterbody from the principle that the acid load should not exceed the 14 nonmarine, nonanthropogenic base cation input and sources and sinks in the catchment minus a 15 neutralizing to protect selected biota from being damaged: 16 CL(N) + CL(S) or CL(A) = BC*dep + BCw - Ecu - AN - ANClimit (2) 17 Where, 18 BC*dep = (BC*=Ca*+Mg*+K*+Na*), nonanthropogenic deposition flux of base cations BCw = 19 the average weathering flux, producing base cations 20 Ecu (Bc=Ca*+Mg*+K*) = the net long-term average uptake flux of base cations in the biomass 21 (i.e., the annual average removal of base cations due to harvesting) 22 AN = the net long-term average uptake, denitrification, and immobilization of nitrogen anions 23 (e.g. NO3-) and uptake of SO42 24 ANClimit = the lowest ANC-flux that protects the biological communities. 25 Since the average flux of base cations weathered in a catchment and reaching the lake or 26 streams is difficult to measure or compute from available information, the average flux of base 27 cations and the resulting critical load estimation were derived from water quality data (Henriksen 28 and Posch, 2001; Henriksen et al., 1992; Sverdrup et al., 1990). Weighted annual mean water 29 chemistry values were used to estimate average base cation fluxes, which were calculated from 30 water chemistry data collected from the Temporally Integrated Monitoring of Ecosystems March 2010 3 Draft - Do Not Quote or Cite ------- Appendix A 1 (TIME)/Long-Term Monitoring (LTM) monitoring networks, that include Adirondack Longterm 2 Monitoring (ALTM), Virginia Trout Stream Sensitivity Study (VTSSS), and the Shenandoah 3 Watershed Study (SWAS), and Environmental Monitoring and Assessment Program (EMAP) 4 (see Section 4.1.2.1 of Chapter 4). 5 The preacidification nonmarine flux of base cations for each lake or stream, BC*0, is 6 BC*0 = BC*dep + BCw - Ecu (3) 7 Thus, critical load for acidity can be rewritten as 8 CL(N) + CL(S) = BC*0 - AN - ANClimit = Q.([BC*]0 - [AN] - [ANC]limit), (4) 9 where the second identity expresses the critical load for acidity in terms of catchment runoff (Q) 10 m/yr and concentration ([x] = X/Q). The sink of nitrogen in the watershed is equal to the uptake 11 (Nupt), immobilization (Nimm), and denitrification (Nden) of nitrogen in the catchment. Thus, 12 critical load for acidity can be rewritten as 13 CL(N) + CL(S) = (fNupt + (1 - r)(Nimm + Nden)} + ( [BC]0* - [ANClimit])Q (5) 14 where f and r are dimensionless parameters that define the fraction of forest cover in the 15 catchment and the lake/catchment ratio. The in-lake retention of nitrogen and sulfur was assumed 16 to be negligible. Equation 5 described the FAB model that was applied when sufficient data was 17 available to estimate the uptake, immobilization, and denitrification of nitrogen and the 18 neutralization of acid anions (e.g. NO3-) in the catchment. In the case were data was not 19 available, the contribution of nitrogen anions to acidification was assumed to be equal to the 20 nitrogen leaching rate (Nleach) into the surface water. The flux of acid anions in the surface 21 water is assumed to represent the amount of nitrogen that is not retained by the catchment, which 22 is determined from the sum of measured concentration of NO3- and ammonia in the stream 23 chemistry. This case describes the SSWC model and the critical load for acidity is 24 CL(A) = Q.([BC*]0 - [ANC]limit) (6) 25 where the contribution of acid anions is considered as part of the exceedances calculation (see 26 Section 1.2.5, below). For the assessment of current condition in both case study areas, the 27 critical load calculation described in Equation 6 was used for most lakes and streams. The lack of 28 sufficient data for quantifying nitrogen denitrification and immobilization prohibited the wide 29 use of the FAB model. In addition, given the uncertainty in quantifying nitrogen denitrification March 2010 4 Draft - Do Not Quote or Cite ------- Appendix A 1 and immobilization, the flux of nitrogen anions in the surface water was assumed to more 2 accurately reflect the contribution of NO3- to acidification. Several major assumptions are made: 3 (1) steady-state conditions exist, (2) the effect of nutrient cycling between plants and soil is 4 negligible, (3) there are no significant nitrogen inputs from sources other than atmospheric 5 deposition, (4) ammonium leaching is negligible because any inputs are either taken up by biota 6 or adsorbed onto soils or nitrate compounds, and (5) longterm sinks of sulfate in the catchment 7 soils are negligible. 8 1.2.1 Preindustrial Base Cation Concentration 9 Present-day surface water concentrations of base cations are elevated above their 10 steady state preindustrial concentrations because of base cation leaching through ion exchange in 11 the soil due to anthropogenic inputs of SC>42 to the watershed. For this reason, present-day 12 surface water base cation concentrations are higher than natural or preindustrial levels, which, if 13 not corrected for, would result in critical load values not in steady-state condition. To estimate 14 the preacidification flux of base cations, the present flux of base cations was estimated, 15 BC*t, given by BC*t = BC*dep + BCw-Ecu +BCexc, (7) 16 Where BCexc = the release of base cations due to ion-exchange processes. Assuming that 17 deposition, weathering rate, and net uptake have not changed over time, BCexc can be obtained 18 by subtracting Equation 5 from Equation 7: 19 BCexc = BC*t-BC*0 (8) 20 This present-day excess production of base cations in the catchment was related to the long-term 21 changes in inputs of nonmarine acid anions (ASO*2 + ANO3) by the F-factor (see below): 22 BCexc = F (ASO*2 + ANO3) (9) 23 For the preacidification base cation flux, solving Equation 5 for BC*0 and then substituting 24 Equation 8 for BCexc and explicitly describing the long-term changes in nonmarine acid ion 25 inputs: 26 BC*0 = BC*t-F(SO*4,t-SO*4,0+NO*3,t-NO*3,0) (10) 27 The preacidification NO3- concentration, NO*3,0, was assumed to be zero. March 2010 5 Draft - Do Not Quote or Cite ------- Appendix A 1 1.2.2 F-factor 2 An F-factor was used to correct the concentrations and estimate preindustrial base concentrations 3 for lakes in the Adirondack Case Study Area. In the case of streams in the 4 Shenandoah Case Study Area, the preindustrial base concentrations were derived from the 5 MAGIC model as the base cation supply in 1860 (hindcast) because the F-factor approach is 6 untested in this region. An F-factor is a ratio of the change in nonmarine base cation 7 concentration due to changes in strong acid anion concentrations (Henriksen, 1984; Brakke et al., 8 1990): 9 F=([BC*]t-[BC*]0)/([SO4*]t-[SO4*]0 + [NO3*]t-[NO3*]0), (12) 10 where the subscripts t and 0 refer to present and preacidification conditions, respectively. If F=l, 11 all incoming protons are neutralized in the catchment (only soil acidification); at F=0, none of 12 the incoming protons are neutralized in the catchment (only water acidification). The F-factor 13 was estimated empirically to be in the range 0.2 to 0.4, based on the analysis of historical data 14 from Norway, Sweden, the United States, and Canada (Henriksen, 1984). Brakke et al. (1990) 15 later suggested that the F-factor should be a function of the base cation concentration: 16 F = sin (Ti/2 Q[BC*]t/[S]) (13) 17 where 18 Q = the annual runoff (m/yr). [S] = the base cation concentration at which F=l; and for 19 [BC*]t>[S] F is set to 1. For Norway [S] has been set to 400 milliequivalents per cubic meter 20 (meq/m3)(circa.8 mg Ca/L) (Brakke et al., 1990). The preacidification SO42- concentration in 21 lakes, [SO4*]0, is assumed to consist of a constant atmospheric contribution and a geologic 22 contribution proportional to the concentration of base cations (Brakke et al., 1989). The 23 preacidification SO42- concentration in lakes, [SO4*]0 was estimated from the relationship 24 between [SO42-]o* and [BC]t* based on work completed by Henriksen et al., 2002 as described 25 by the following equation: 26 [SO42-]o* = 15 + 0.16 * [BC]t* (14) March 2010 6 Draft - Do Not Quote or Cite ------- Appendix A 1 2 Table A.2 Illustrates SSWC Approach - Environmental Variables CL(A) = BC*dep + BCW - Bcu - ANClimit CL(A) = Q ([BC*]0 - [ANC]limit) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Variable Code BC dep BCW Bcu ANClimit Ca* Mg* Na* K* S04* CL S04* NO3* Q [BC*]0 [S04*]0 [N03*]0 F Description Sum (Ca*+Mg*+K*+Na*), nonanthropogenic deposition flux of base cations Average weathering flux of base cations Sum (Ca+Mg+K), the net long-term average uptake flux of base cations in the biomass Lowest ANC-flux that protects the biological communities Sea Salt corrected Surface water concentration (ueq/L) growing season average. (Ca - (CL x 0.0213)) Sea Salt corrected Surface water concentration (ueq/L) growing season average. (Mg - (CL x 0.0669)) Sea Salt corrected Surface water concentration (ueq/L) growing season average. (Na - (CL x 0.557)) Sea Salt corrected Surface water concentration (ueq/L) growing season average. (K - (CL x 0.0.0206)) Sea Salt corrected Surface water concentration (ueq/L) growing season average. (SO4 - (CL x 0.14)) Surface water concentration (ueq/L) growing season average. Surface water concentration (ueq/L) growing season average. Surface water concentration (ueq/L) growing season average. The annual runoff (m/yr) Preindustrial flux of base cations in surface water, corrected for sea salts Preindustrial flux of sulfate in surface water, corrected for sea salts Preindustrial flux of nitrate, corrected for sea salts Calculated factor Source WetNADPandDry CASTNET Calculated (5-17) USFS-FIA data Set Water quality data Water quality data Water quality data Water quality data Water quality data Water quality data Water quality data Water quality data USGS Calculated from water quality data Estimated Equal to 0 Fix values 4 5 March 2010 Draft - Do Not Quote or Cite ------- Appendix A 1 Table A.3 FAB 2 DL(N) + DL(S) = Approach - Environmental Variables {fNupt + (1 - r)(Nimm + Nden) + (Nret + Sret)} [BC]0* - [ANClimit])Q 1 2 3 4 5 6 7 8 9 10 11 12 13 14 14 15 16 17 18 19 20 Variable Code Ndepo ANClimit [BC*]0 Ca* Mg* Na* K* SO4* CL S04* NO3* Q f r Nret ^ret Nupt -L Mmm Nden Lake Size WSH Description Total N deposition Lowest ANC-flux that protects the biological communities Preindustrial flux of base cations in surface water, corrected for sea salt Sea Salt corrected Surface water concentration (ueq/L) growing season average. (Ca - (CL x 0.0213)) Sea Salt corrected Surface water concentration (ueq/L) growing season average. (Mg - (CL x 0.0669)) Sea Salt corrected Surface water concentration (ueq/L) growing season average. (Na - (CL x 0.557)) Sea Salt corrected Surface water concentration (ueq/L) growing season average. (K - (CL x 0.0.0206)) Sea Salt corrected Surface water concentration (ueq/L) growing season average. (SO4 - (CL x 0.14)) Surface water concentration (ueq/L) growing season average. Surface water concentration (ueq/L) growing season average. Surface water concentration (ueq/L) growing season average. The annual runoff (m/yr) f is a dimensionless parameter that define the fraction of forest cover in the catchment r is a dimensionless parameter that define the lake/catchment ratio The in-lake retention of nitrogen The in-lake retention of sulfur The net long-term average uptake flux of N in the biomass Immobilization of N in the soils Denitrification Lake size (ha) Watershed area (ha) Source NADP/CMAQ Set Calculated from water quality data Water quality data Water quality data Water quality data Water quality data Water quality data Water quality data Water quality data Water quality data USGS Estimated Estimated USFS-FIA data Estimated fix value Estimated fix value DLMs Calculated March 2010 Draft - Do Not Quote or Cite ------- Appendix A 1 Data requirements for MA GIC 2 The MAGIC model (Cosby et al., 1985a; 1985b; 1985c) is a mathematical model (a 3 lumped-parameter model) of soil and surface water acidification in response to atmospheric 4 deposition based on process-level information about acidification. A process model, such as 5 MAGIC, characterizes acidification into (l)a section in which the concentrations of major ions 6 are assumed to be governed by simultaneous reactions involving SC>42" adsorption, cation 7 exchange, dissolution-precipitation- speciation of aluminum, and dissolution-speciation of 8 inorganic carbon; and (2) a mass balance section in which the flux of major ions to and from the 9 soil is assumed to be controlled by atmospheric inputs, chemical weathering, net uptake and loss 10 in biomass and losses to runoff. At the heart of MAGIC is the size of the pool of exchangeable 11 base cations in the soil. As the fluxes to and from this pool change over time owing to changes in 12 atmospheric deposition, the chemical equilibria between soil and soil solution shift to give 13 changes in surface water chemistry. The degree and rate of change of surface water acidity thus 14 depend both on flux factors and the inherent characteristics of the affected soils. 15 There are numerous input data required to run MAGIC making it rather data intensive. 16 Atmospheric deposition fluxes for the base cations and strong acid anions are required as inputs 17 to the model. These inputs are generally assumed to be uniform over the catchment. The volume 18 discharge for the catchment must also be provided to the model. In general, the model is 19 implemented using average hydrologic conditions and meteorological conditions in annual 20 simulations, i.e., mean annual deposition, precipitation and lake discharge are used to drive the 21 model. Values for soil and surface water temperature, partial pressure of carbon dioxide and 22 organic acid concentrations must also be provided at the appropriate temporal resolution. 23 The aggregated nature of the model requires that it be calibrated to observed data from a 24 system before it can be used to examine potential system response. Calibrations are based on 25 volume weighted mean annual or seasonal fluxes for a given period of observation. The length of 26 the period of observation used for calibration is not arbitrary. Model output will be more reliable 27 if the annual flux estimates used in calibration are based on a number of years rather than just 28 one year. There is a lot of year-to-year variability in atmospheric deposition and catchment 29 runoff. Averaging over a number of years reduces the likelihood that an "outlier" year (very dry, 30 etc.) is used to specify the primary data on which model forecasts are based. On the other hand, March 2010 9 Draft - Do Not Quote or Cite ------- Appendix A 1 averaging over too long a period may remove important trends in the data that need to be 2 simulated by the model. 3 The calibration procedure requires that stream water quality, soil chemical and physical 4 characteristics, and atmospheric deposition data be available for each catchment. The water 5 quality data needed for calibration are the concentrations of the individual base cations (Ca, Mg, 6 Na, and K) and acid anions (Cl, SC>42", and N(V) and the pH. The soil data used in the model 7 include soil depth and bulk density, soil pH, soil cation-exchange capacity, and exchangeable 8 bases in the soil (Ca, Mg, Na, and K). The atmospheric deposition inputs to the model must be 9 estimates of total deposition, not just wet deposition. In some instances, direct measurements of 10 either atmospheric deposition or soil properties may not be available for a given site with stream 11 water data. In these cases, the required data can often be estimated by: (a) assigning soil 12 properties based on some landscape classification of the catchment; and (b) assigning deposition 13 using model extrapolations from some national or regional atmospheric deposition monitoring 14 network. Soil data for model calibration are usually derived as aerially averaged values of soil 15 parameters within a catchment. If soils data for a given location are vertically stratified, the soils 16 data for the individual soil horizons at that sampling site can be aggregated based on horizon, 17 depth, and bulk density to obtain single vertically aggregated values for the site, or the stratified 18 data can be used directly in the model. 19 March 2010 10 Draft - Do Not Quote or Cite ------- 2 3 4 Methodologies for National Terrestrial and Aquatic 5 Acidification Maximum Depositional Load e Approaches: Determining Weathering Rates 7 8 9 First External Review Draft 10 11 12 13 Prepared for: 14 15 U.S. Environmental Protection Agency 16 Office of Air Quality Planning and Standards 17 Research Triangle Park, NC 27711 18 19 20 Prepared by: 21 22 RTI International 23 P.O. Box 12194 24 Research Triangle Park, NC 27709 25 26 EPA Contract Number EP-D-06-003 27 RTI Project Number 0209897.004.080 28 29 30 31 January 11, 2009 32 33 HRTI INTERNATIONAL ------- ------- Appendix B 1 Table of Contents 2 3 1. PURPOSE 1 4 2. OVERVIEW OF ACIDIFICATION 1 5 2.1 EPA's Integrated Science Assessment and Risk and Exposure Assessment 1 6 2.2 Aquatic Acidification and Critical Acid Loads 4 7 2.1.2 Terrestrial Acidification and Critical Acid Loads 9 8 3. AQUATIC BASE CATION WEATHERING METHODOLOGY 14 9 3.1 Aquatic Base Cation Weathering 14 10 3.2 Methodologies for Determining Base Cation Weathering Values in the United States 16 11 3.2.1 Difficulties in estimating base cation weathering 16 12 3.2.2 Approaches to estimating BCW for Aquatic Acidification 17 13 3.3 Proposed Methodology for Estimating and Mapping Base Cation Weathering for Aquatic 14 Critical Acid Load Calculations 27 15 3.3.1 Potential limitations of proposed methodology 33 16 3.3.2 Uncertainty analyses 34 17 4. TERRESTRIAL BASE CATION WEATHERING METHODOLOGY 35 18 4.1 Introduction 35 19 4.2 Terrestrial Base Cation Weathering 36 20 4.3 Methodologies for Determining Base Cation Weathering Values in the United States 39 21 4.3.1 Difficulties in estimating base cation weathering 39 22 4.3.2 Approaches to estimating BCw: 40 23 4.3.3 Proposed methodology for estimating and mapping base cation weathering for 24 terrestrial critical acid load calculations 50 25 4.3.5 Potential limitations of proposed methodology 78 26 4.3.6 "Field Tests" of model and uncertainty analyses 79 27 5. CONCLUSIONS AND RECOMMENDATIONS 81 28 6. REFERENCES 82 29 APPENDIX 1 Potentially Applicable National-Scale Geochemical Data 97 30 APPENDIX 2 References for Table 3-2: Applications of the MAGIC Model 102 31 32 March 2010 i Draft - Do Not Quote or Cite ------- Appendix B List of Figures 2 2-1. (a) Number offish species per lake or stream versus acidity, expressed as acid neutralizing 3 capacity for Adirondack Case Study Area lakes (Sullivan et al., 2006). (b) Number offish 4 species among 13 streams in Shenandoah National Park. Values of acid neutralizing 5 capacity are means based on quarterly measurements from 1987 to 1994. The regression 6 analysis shows a highly significant relationship (p < .0001) between mean stream acid 7 neutralizing capacity and the number offish species 5 8 2-2. The relationship between the Bc/Al ratio in soil solution and the percentage of tree species 9 (found growing in North America - native and introduced species) exhibiting a 20% 10 reduction in growth relative to controls (after Sverdrup and Warfvinge, 1993) 10 11 3-1. Process steps for estimating BCW using the MAGIC model with regional extrapolation 28 12 4-1. Areas of continental U.S. that were covered during the last glacial event (Reed and Bush, 13 2005) 39 14 4-2. Process Steps for Estimating BCW Using the PROFILE Regional Model 52 15 4-3. Map Showing the Distribution and Status of SSURGO Data 64 16 4-4. Soil Sampling Locations Included in the USGS Shacklette Dataset 66 17 4-5. Sample Density of USGS National Geochemical Survey 68 18 4-6. NRCS Soil Pedon Sample Pit Locations (30,000 total) 70 19 4-7. NRCS Soil Pedon Pit Sample Locations with Geochemical and Mineralogy Data 71 20 21 22 List of Tables 23 2-1. Aquatic Status Categories 6 24 2-2. Summary of Linkages between Acidifying Deposition, Biogeochemical Processes That 25 Affect Ca2+, Physiological Processes That Are Influenced by Ca2+, and Effect on Forest 26 Function 11 27 2-3. The Three Indicator (Bc/Al)cnt Soil Solution Ratios and Corresponding Levels of Protection 28 to Tree Health and Critical Loads 14 29 3-1. Review of Modeling Approaches (and models) to Estimate Base Cation Weathering for 30 Aquatic Critical Acid Load Determinations 23 31 3-2. Locations of Previous MAGIC Applications within the U.S. and Canada1 29 32 3-3. Input Data Requirements of MAGIC Model 31 33 4-1. Review of modeling approaches (and models) to estimate base cation weathering for 34 terrestrial critical acid load determinations 46 35 4-2. The fourteen dominant minerals modeled within PROFILE 50 36 4-3a. Data required to estimate BCW with the regional PROFILE model (version 5.0). The data in 37 this table must be input by the user and are currently available as a continuous coverage 38 layers for at least a portion of the conterminous United States 53 39 4-3b. Data required to estimate BCW with the regional PROFILE model (version 5.0). The data in 40 this table must be input by the user and are not currently available as a continuous 41 coverage layers for at least a portion of the conterminous United States (will require 42 development of national coverage layer) 53 March 2010 ii Draft - Do Not Quote or Cite ------- Appendix B 1 4-3c. Data required to estimate BCW with the regional PROFILE model (version 5.0). The data in 2 this table are used to support calculations within the model and should be reviewed by the 3 user 54 4 4-4. Available datasets and databases for the conterminous United States that could be used to 5 estimate BCW with the regional application of the PROFILE model (version 5.0) 55 6 4-5. Nitrogen and base cation uptake by forest type (from McNulty et al., 2007) 58 7 4-6 Datasets with Geochemical and Mineralogy Data for US Soils 61 8 4-7. Long-Term Ecological Research (LTER) sites that could potentially be suitable as "field 9 test" sites to validate BCW estimates generated with the regional application of the 10 PROFILE model (version 5.0) 74 11 12 March 2010 iii Draft - Do Not Quote or Cite ------- Appendix B 1 2 March 2010 iv Draft - Do Not Quote or Cite ------- Appendix B 1 1. PURPOSE 2 The purpose of this Work Assignment Task is to develop methodologies for estimating 3 national terrestrial and aquatic acidification maximum depositional loads. Separate approaches 4 are developed for terrestrial and aquatic acidification because biogeochemical processes in 5 aquatic and terrestrial ecosystems for nitrogen and sulfur are not identical. Information about the 6 key physical, chemical, and biological parameters needed to predict acidification potential in 7 ecosystems is not always available. For example, weathering rates are key to acidification but are 8 not available in all parts of the U.S. Knowledge of an ecosystem's weathering characteristics 9 enables a more accurate assessment of whether acidifying deposition can be neutralized or 10 exceeds an ecosystem's critical load beyond which negative effects in aquatic and terrestrial 11 health may occur. 12 This report presents an introduction to aquatic and terrestrial acidification, followed by 13 reviews of different approaches to estimating base cation weathering and detailed methodologies 14 that could be used to estimate base cation weathering for aquatic and terrestrial critical load 15 calculations. 16 2. OVERVIEW OF ACIDIFICATION 17 2.1 EPA's Integrated Science Assessment and Risk and Exposure Assessment 18 Deposition of SOX, NOX, and NHX can lead to ecosystem exposure to acidification. The 19 Integrated Science Assessment (ISA) for Oxides of Nitrogen and Sulfur-Ecological Criteria 20 (FinalReport) (ISA) (U.S. EPA, 2008) reports that acidifying deposition has altered major 21 biogeochemical processes in the United States by increasing the sulfur and nitrogen content of 22 soils, accelerating sulfate (SC>42 ) and nitrate (NOs ) leaching from soil to drainage water, 23 depleting soil exchangeable base cations (especially calcium [Ca2+] and magnesium [Mg2+]) 24 from soils, and increasing the mobility of aluminum (Al) within the soil (U.S. EPA, 2008, 25 Section 3.2.1) 26 The extent of soil acidification is a critical factor that regulates virtually all acidification- 27 related ecosystem effects from sulfur and nitrogen deposition. Soil acidification occurs in 28 response to both natural factors and acidifying deposition (U.S. EPA, 2008, Section 3.2.1). 29 Under conditions of low atmospheric deposition of nitrogen and sulfur, the naturally produced March 2010 1 Draft - Do Not Quote or Cite ------- Appendix B 1 bicarbonate anion is often the dominant mobile anion, with SC>42" and N(V playing a limited role 2 with respect to cation leaching. Increased atmospheric deposition of sulfur and nitrogen can 3 result in marked increases in SC>42" and NCV soil fluxes resulting in the concomitant leaching of 4 base cations (Ca2+, Mg2+) and toxic cations (Aln+ and H+). 5 Acidification can impact the health of terrestrial and aquatic ecosystems. One of the 6 effects of soil acidification is the increased mobility of dissolved inorganic Al, which is toxic to 7 tree roots, fish, algae, and aquatic invertebrates (U.S. EPA, 2008, Sections 3.2.1.5, 3.2.2.1, and 8 3.2.3). 9 The changes in major biogeochemical processes and soil conditions caused by acidifying 10 deposition have significant ramifications for the water chemistry and biological functioning of 11 associated surface waters. Surface water chemistry indicates the negative effects of acidification 12 on the biotic integrity of freshwater ecosystems. Surface water chemistry integrates the sum of 13 terrestrial and aquatic processes that occur upstream within a watershed. Important terrestrial 14 processes include nitrogen saturation, forest decline, and soil acidification (Stoddard et al., 15 2003). Thus, water chemistry integrates and reflects changes in soil and vegetative properties and 16 biogeochemical processes (U.S. EPA, 2008, Section 3.2.3.1). 17 Ecological effects occur at four levels of biological organization: (1) the individual; (2) 18 the population, which is composed of a single species of individuals; (3) the biological 19 community, which is composed of many species; and (4) the ecosystem. Low ANC 20 concentrations are linked with negative effects on aquatic systems at all four of these biological 21 levels. For the individual level, impacts are assessed in terms of fitness (i.e., growth, 22 development, and reproduction) or sublethal effects on condition. Surface water with low ANC 23 concentrations can directly influence aquatic organism fitness or mortality by disrupting ion 24 regulation and can mobilize dissolved inorganic aluminum, which is highly toxic to fish under 25 acidic conditions (i.e., pH <6 and ANC <50 ueq/L). For example, research showed that as the pH 26 of surface waters decreased to <6, many aquatic species, including fish, invertebrates, 27 zooplankton, and diatoms, tended to decline sharply causing species richness to decline 28 (Schindler, 1988). Van Sickle and colleagues (1996) also found that blacknose dace (Rhinichthy 29 spp.) were highly sensitive to low pH and could not tolerate inorganic Al concentrations greater 30 than about 3.7 micromolar (uM) for extended periods of time. For example, they found that after March 2010 2 Draft - Do Not Quote or Cite ------- Appendix B 1 6 days of exposure to high inorganic Al, blacknose dace mortality increased rapidly to nearly 2 100%. 3 At the community level, species richness and community structure can be used to 4 evaluate the effects of acidification. Species composition refers to the mix of species that are 5 represented in a particular ecosystem, whereas species richness refers to the total number of 6 species in a stream or lake. Acidification alters species composition and richness in aquatic 7 ecosystems. There are a number of species common to many oligotrophic waterbodies that are 8 sensitive to acidification and cannot survive, compete, or reproduce in acidic waters. In response 9 to small to moderate changes in acidity, acid-sensitive species are often replaced by other more 10 acid-tolerant species, resulting in changes in community composition and richness, but with little 11 or no change in total community biomass. The effects of acidification are continuous, with more 12 species being affected at higher degrees of acidification. At a point, typically a pH <4.5 and an 13 ANC <0 ueq/L, complete to near-complete loss of many classes of organisms occur, including 14 fish and aquatic insect populations, whereas others are reduced to only a few acidophilic forms. 15 These changes in species integrity are because energy cost in maintaining physiological 16 homeostasis, growth, and reproduction is high at low ANC levels (Schreck, 1981, 1982; 17 Wedemeyer et al., 1990). 18 In EPA's Risk and Exposure Assessment for Review of the Secondary National Ambient 19 Air Quality Standards for Oxides of Nitrogen and Sulfur (U.S. EPA, 2009), the negative impacts 20 of acidifying deposition were assessed by conducting case studies of 1) aquatic acidification in 21 Adirondack Mountains lakes and Shenandoah Mountains streams, and 2) terrestrial acidification 22 in red spruce and sugar maple forests in the White Mountains of New Hampshire and in 23 Pennsylvania, respectively. The results of these case studies revealed the significance of base 24 cation weathering in predicting aquatic and terrestrial acidification impacts. The results further 25 highlighted the need to select weathering methodologies that can be applied across geologically 26 diverse ecosystems in the United States. This report uses the information from the Risk and 27 Exposure Assessment as a starting point to identify and evaluate approaches to predicting 28 weathering at other locations and larger scales in the United States. In this report, RTI 29 recommends methodologies (including computer models) for application in the United States, 30 assesses the availability of input data for those methodologies, identifies potential remedies to March 2010 3 Draft - Do Not Quote or Cite ------- Appendix B 1 limited data availability, and describes uncertainties with the methodologies in predicting 2 acidification impacts. 3 2.2 Aquatic Acidification and Critical Acid Loads 4 Surface water chemistry is a primary indicator of acidification and the resulting negative 5 effects on the biotic integrity of freshwater ecosystems. Chemical parameters can be used to 6 assess effects of acidifying deposition on lake or stream acid-base chemistry. These receptors 7 include surface water pH and concentrations of SC>42", NCV, Al, and Ca2+; the sum of base 8 cations; and the recently developed base cation surplus. Another widely used water chemistry 9 indicator for both atmospheric deposition sensitivity and effects is acid neutralizing capacity 10 (ANC). The utility of the ANC criterion lies in the association between ANC and the surface 11 water constituents that directly contribute to or ameliorate acidity-related stress, in particular pH, 12 Ca2+, and Al. ANC is also used because it integrates overall acid status and because surface 13 water acidification models do a better job projecting ANC than they do for projecting pH and 14 dissolved inorganic Al concentrations. 15 For the purpose of this study, ANC of surface waters is simply measured as the total 16 amount of strong base ions minus the total amount of strong acid anions: 17 ANC = (Ca2++ Mg2++ K++ Na++ NH4) - (SO42" + NCV+CO (2-1) 18 The unit of ANC is usually microequivalents per liter (ueq/L). If the sum of the 19 equivalent concentrations of the base cations exceeds those of the strong acid anions, then the 20 ANC of a waterbody will be positive. To the extent that the base cation sum exceeds the strong 21 acid anion sum, the ANC will be higher. Higher ANC is generally associated with high pH and 22 Ca2+ concentrations; lower ANC is generally associated with low pH and A13+ concentrations and 23 a greater likelihood of toxicity to biota. 24 Low ANC coincides with effects on aquatic systems (e.g., individual species fitness loss 25 or death, reduced species richness, altered community structure). At the community level, 26 species richness is positively correlated with pH and ANC (Kretser et al., 1989; Rago and 27 Wiener, 1986) because energy cost in maintaining physiological homeostasis, growth, and 28 reproduction is high at low ANC levels (Schreck, 1981, 1982; Wedemeyer et al., 1990). For 29 example, Sullivan and colleagues (2006) found a logistic relationship between fish species 30 richness and ANC class for Adirondack Case Study Area lakes (Figure 2-la) that indicates the March 2010 4 Draft - Do Not Quote or Cite ------- Appendix B 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 probability of occurrence of an organism for a given value of ANC. In the Shenandoah Case Study Area, a statistically robust relationship between acid-base status of streams and fish species richness was also documented (Figure 2-lb). In fact, ANC has been found in various studies to be the best single indicator of the biological response and health of aquatic communities in acid-sensitive systems (Lien et al., 1992; Sullivan et al., 2006). Biota are generally not harmed when ANC values are >100 microequivalents per liter (ueq/L). The number offish species also peaks at ANC values >100 ueq/L (Bulger et al., 1999; Driscoll et al., 2001; Kretser et al., 1989; Sullivan et al., 2006). Below 100 ueq/L, ANC fish fitness and community diversity begin to decline (Figure 2-1). At ANC levels between 100 and 50 ueq/L, the fitness of sensitive species (e.g., brook trout, zooplankton) also begins to decline. When ANC concentrations are <50 ueq/L, they are generally associated with death or loss of fitness of biota that are sensitive to acidification (Kretser et al., 1989; Dennis and Bulger, 1995). l-,\ Number of Fish Specicsj k PO Q1 N) i^ £Ti £U ft pyj 1 J 1 1 1 1 L 1 1 Jp0 --..— J i •^wfi^^ " f 1 — — i 1 1 1 1 — -200 -100 0 100 200 300 400 5< ANC(ueq/U (b) 3 3 . . M 5 Jl & 1 - u. 1 2 i 2 z 2 1 - in n • ! ~^_ !-"' \ \.'\ i £ \x U* 1 ^1 t | \ M |1 flf! \ !!< t i : | § 1 i t i •25 0 25 50 TZ ICO 115 150 17S 200 225 250 27 Average AUC (jieq/L] Figure 2-1. (a) Number offish species per lake or stream versus acidity, expressed as acid neutralizing capacity for Adirondack Case Study Area lakes (Sullivan et al., 2006). (b) Number of fish species among 13 streams in Shenandoah National Park. Values of acid neutralizing capacity are means based on quarterly measurements from 1987 to 1994. The regression analysis shows a highly significant relationship (p < .0001) between mean stream acid neutralizing capacity and the number of fish species. When ANC levels drop to <20 ueq/L, all biota exhibit some level of negative effects. Fish and plankton diversity and the structure of the communities continue to decline sharply to levels where acid-tolerant species begin to outnumber all other species (Matuszek and Beggs, 1988; Driscoll et al., 2001). Stoddard and colleagues (2003) showed that to protect biota from episodic acidification in the springtime, base flow ANC levels need to have an ANC of at least March 2010 Draft - Do Not Quote or Cite ------- Appendix B 1 30-40 ueq/L. Complete loss offish populations and extremely low diversity of planktonic 2 communities occur when ANC levels stay <0 ueq/L. Only acidophilic species are present, but 3 their population numbers are sharply reduced (Sullivan et al., 2006). 4 The critical load approach can be used to connect current deposition of nitrogen and 5 sulfur to the acid-base condition and biological risk to biota of lakes and streams in the study 6 through the defined ANC thresholds. Calculating critical load exceedances (i.e., the amount of 7 deposition above the critical load) allows the determination of whether current deposition poses a 8 risk of acidification to a given group of waterbodies. Low critical load values (i.e., less than 50 9 meq/m2 yr) mean that the watershed has a limited ability to neutralize the addition of acidic 10 anions, and hence, it is susceptible to acidification. The greater the critical load value, the greater 11 the ability of the watershed to neutralize the additional acidic anions and protect aquatic life. 12 This approach also allows for the comparison of different levels of ANC thresholds (e.g., 0 13 ueq/L (acidic), 20 ueq/L (minimal protection), 50 ueq/L (moderate protection), and 100 ueq/L 14 (full protection)) and their associated risk to the biological community. Table 2-1 provides a 15 summary of the biological effects experienced at each of these limits. Table 2-1. Aquatic Status Categories Category Label ANC Levels* Expected Ecological Effects Acute Concern <0 micro equivalent per Liter (ueq/L) Near complete loss offish populations is expected. Planktonic communities have extremely low diversity and are dominated by acidophilic forms. The number of individuals in plankton species that are present is greatly reduced. Severe Concern 0-20 ueq/L Highly sensitive to episodic acidification. During episodes of high acidifying deposition, brook trout populations may experience lethal effects. Diversity and distribution of zooplankton communities decline sharply. Elevated Concern 20-50 ueq/L Fish species richness is greatly reduced (i.e., more than half of expected species can be missing). On average, brook trout populations experience sublethal effects, including loss of health, reproduction capacity, and fitness. Diversity and distribution of zooplankton communities decline. Moderate Concern 50-100 ueq/L Fish species richness begins to decline (i.e., sensitive species are lost from lakes). Brook trout populations are sensitive and variable, with possible sublethal effects. Diversity and distribution of zooplankton communities also begin to decline as species that are sensitive to acidifying deposition are affected. March 2010 Draft - Do Not Quote or Cite ------- Appendix B Category Label ANC Levels* Expected Ecological Effects Low Concern >100 |jeq/L Fish species richness may be unaffected. Reproducing brook trout populations are expected where habitat is suitable. Zooplankton communities are unaffected and exhibit expected diversity and distribution. 1 There are numerous methods and models that can be used to calculate critical loads for 2 acidity. Drawing on the peer-reviewed scientific literature (Dupont et al., 2005), this study will 3 use a steady-state critical load model that uses surface water chemistry as the base for calculating 4 the critical load. A combination of the Steady-State Surface Water Chemistry (SSWC) and First- 5 Order Acidity Balance (FAB) models were used to calculate the critical load. Both the SSWC 6 model and FAB are based on the principle that excess base-cation production within a catchment 7 area should be equal to or greater than the acid anion input, thereby maintaining the ANC above 8 a preselected level (Reynolds and Norris, 2001; Posch et al., 1997). These models assume 9 steady-state conditions and assume that all SO42 in runoff originates from sea salt spray and 10 anthropogenic deposition. Given a critical ANC protection level, the critical load of acidity is 11 simply the input flux of acid anions from atmospheric deposition (i.e., natural and 12 anthropogenic) subtracted from the natural (i.e., preindustrial) inputs of base cations in the 13 surface water. 14 Critical loads for nitrogen and sulfur (CL(N) + CL(S)) or critical load of acidity CL(A) 15 are calculated for each waterbody from the principle that the acid load should not exceed the 16 nonmarine, nonanthropogenic base cation input and sources and sinks in the catchment minus a 17 neutralizing to protect selected biota from being damaged: 18 CL(N) + CL(S) or CL(A) = BC*dep + BCW - Bcu - AN - ANCiimit (2-2) 19 where 20 BC dep = nonanthropogenic deposition flux of base cations 21 (BC*=Ca*+Mg*+K*+Na*) 22 BCW = the average weathering flux, producing base cations 23 Bcu = the net long-term average uptake flux of base cations (Bc=Ca*+Mg*+K*) in 24 the biomass (i.e., the annual average removal of base cations due to 25 harvesting) March 2010 7 Draft - Do Not Quote or Cite ------- Appendix B 1 AN = the net long-term average uptake, denitrification, and immobilization of 2 nitrogen anions (e.g. N(V) and uptake of S(V 3 ANCumit = the lowest ANC-flux that protects the biological communities. 4 In order to estimate a critical load from water quality data alone, a relation to the 5 preacidification nonmarine flux of base cations for each lake or stream, BC „, is used. 6 BC*0 = BC*dep + BCW - Bcu (2-3) 7 Thus, the critical load for acidity can be rewritten as 8 CL(N) + CL(S) = BC*0 - AN - ANC,,mt = Q.([ BC*]0 - [AN] - [ANC],,mt) (2-4) 9 where the second identity expresses the critical load for acidity in terms of catchment runoff (Q) 10 m/yr and concentration ([x] = X/Q). In cases where data are available, the FAB model is applied 11 to quantify the [AN] term of the critical load calculation (derivation provided in Appendix 4, 12 Attachment A of U.S. EPA, 2009). Where data are not available the contribution of nitrogen 13 anions to acidification was assumed to be equal to the nitrogen leaching rate into the surface 14 water. The flux of acid anions in the surface water is assumed to represent the amount of 15 nitrogen that is not retained by the catchment, which is determined from the sum of measured 16 concentration of N(V and ammonia in the stream chemistry. This case describes the SSWC 17 model and the critical load for acidity is 18 CL(A) = Q.([BO]o-[ANC]iimit) (2-5) 19 where the contribution of acid anions is considered as part of the exceedances calculation. With 20 this approach several major assumptions are made: (1) steady-state conditions exist, (2) the effect 21 of nutrient cycling between plants and soil is negligible, (3) there are no significant nitrogen 22 inputs from sources other than atmospheric deposition, (4) ammonium leaching is negligible 23 because any inputs are either taken up by biota or adsorbed onto soils or nitrate compounds, and 24 (5) long-term sinks of sulfate in the catchment soils are negligible. 25 To determine a value for BC*0 with the SSWC method, estimates of BCdep are available 26 from previous works including the recent REA (U.S. EPA, 2009). Assumptions or estimates for 27 BCU and AN can be made based on attributes of the area of study, including vegetation March 2010 8 Draft - Do Not Quote or Cite ------- Appendix B 1 characteristics. But the average flux of base cations weathered in a catchment and reaching the 2 lake or streams (BCW) is difficult to measure or compute from available information (Henriksen 3 and Posch, 2001; Henriksen et al., 2002; Langan et al., 2001). In the previous work for the Risk 4 and Exposure Assessment case studies (U.S. EPA, 2009) the average flux of base cations and the 5 resulting critical load estimation were derived from water quality data (Henriksen and Posch, 6 2001; Henriksen et al., 1992; Sverdrup et al., 1990). Weighted annual mean water chemistry 7 values were used to estimate average base cation fluxes, which were calculated from water 8 chemistry data collected from several national and regional monitoring programs. For a national 9 assessment, however, new methods must be developed to estimate the BCW flux, which is critical 10 to the critical load calculation, through consistent, nationally-applicable means. 11 2.1.2 Terrestrial Acidification and Critical Acid Loads 12 Due to the impact of acidifying nitrogen and sulfur deposition on soil solution base cation 13 (Be) and aluminum concentrations, the Bc/Al ratio in the soil solution is often used as the 14 chemical or critical indicator of terrestrial acidification. It was recently used as an indicator in the 15 U.S. EPA's Risk and Exposure Assessment for oxides of nitrogen and oxides of sulfur (U.S. 16 EPA, 2009). This Bc/Al ratio links acidifying deposition to biological responses or end points, 17 such as reduced plant or tree growth, within an ecosystem. In a meta-analysis of studies that 18 explored the relationship between Bc/Al ratio in soil solution and tree growth, Sverdrup and 19 Warfvinge (1993a) reported the Bc/Al ratios at which growth was reduced by 20% relative to 20 control trees. This 20% reduction in tree growth was selected as the critical value because it was 21 thought to represent a significant reduction in growth (H. Sverdrup personal communication, 22 2009b) and approximates the Bc/Al value that would result in a 10% reduction in normal tree 23 growth under field conditions (Sverdrup and Warfvinge, 1993a). Figure 2-2 presents the 24 findings of Sverdrup and Warfvinge (1993 a) based on 46 of the tree species (native and 25 introduced) that grow in North America. This summary indicates that there is a 50% chance of 26 negative tree response (i.e., >20% reduced growth) at a soil solution Bc/Al ratio of 1.2 and a 27 75% chance at a Bc/Al ratio of 0.6. These findings clearly demonstrate a relationship between 28 Bc/Al ratio and tree health; as the Bc/Al is reduced, there is a greater likelihood of a negative 29 impact on tree health. March 2010 9 Draft - Do Not Quote or Cite ------- Appendix B 10 a o "o 2 1 - o E/3 a 0 ta o PQ .01 ' ( ^\ V, ,,,,,.. (Bc/AlJrit = 1.2 %-^_^_^ ^^* ***** *~*^^ i UP/ Ah = n h 'v,. \J— '^' ^^^irit W.LJ ^^- ^^^\ \ V ) 25 50 75 1( Cumulative Percentage of Species Exhibiting Reduced Growth Response DO 2 Figure 2-2. The relationship between the Bc/AI ratio in soil solution and the percentage of 3 tree species (found growing in North America - native and introduced species) exhibiting 4 a 20% reduction in growth relative to controls (after Sverdrup and Warfvinge, 1993). 5 The tree species most commonly studied in North America to assess the impacts of 6 acidification due to total nitrogen and sulfur deposition are red spruce (i.e., Picea Rubens) and 7 sugar maple (i.e., Acer saccharum). Both species are found in the eastern United States, and soil 8 acidification is widespread throughout this area (Warby et al., 2009). Based on the results from a 9 compilation of laboratory studies, red spruce growth can be reduced by 20% at a Bc/AI soil 10 solution ratio of approximately 1.2, and a similar reduction in growth may be experienced by 11 sugar maple at a Bc/AI ratio of 0.6 (Sverdrup and Warfvinge 1993a). 12 Red spruce is found scattered throughout high-elevation sites in the Appalachian 13 Mountains, including the southern peaks. Noticeable fractions of the canopy red spruce died 14 within the Adirondack, Green, and White mountains in the 1970s and 1980s. Although a variety 15 of conditions, such as changes in climate and exposure to ozone, may impact the growth of red 16 spruce (Fincher et al., 1989; Johnson et al., 1988), acidifying deposition has been implicated as 17 one of the main factors causing this decline. Based on the research conducted to date, acidifying 18 deposition can cause a depletion of base cations in upper soil horizons, Al toxicity to tree roots, 19 and accelerated leaching of base cations from foliage (U.S. EPA, 2008, Section 3.2.2.3). Such 20 nutrient imbalances and deficiencies can reduce the ability of trees to respond to stresses, such as 21 insect defoliation, drought, and cold weather damage (DeHayes et al., 1999; Driscoll et al., March 2010 10 Draft - Do Not Quote or Cite ------- Appendix B 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 2001), thereby decreasing tree health and increasing mortality. Additional linkages between acidifying deposition and red spruce physiological responses are indicated in Table 4.3-1. Within the southeastern United States, periods of red spruce decline slowed after the 1980s, when a corresponding decrease in SC>2 emissions, and therefore acidic deposition, was recorded (Webster et al., 2004). Sugar maple is found throughout the northeastern United States and the central Appalachian Mountain region. This species has been declining in the eastern United States since the 1950s. Studies on sugar maple have found that one source of this decline in growth is related to both acidifying deposition and base-poor soils on geologies dominated by sandstone or other base-poor substrates (Bailey et al., 2004; Horsley et al., 2000). These site conditions are representative of the conditions expected to be most susceptible to impacts of acidifying deposition because of probable low initial base cation pools and high base cation leaching losses (U.S. EPA, 2008, Section 3.2.2.3). The probability of a decrease in crown vigor or an increase in tree mortality has been noted to increase at sites with low Ca2+ and Mg2+ as a result of leaching caused by acidifying deposition (Drohan and Sharpe, 1997). Low levels of Ca2+ in leaves and soils have been shown to be related to lower rates of photosynthesis and higher antioxidant enzyme activity in sugar maple stands in Pennsylvania (St. Clair et al., 2005). In addition, plots of sugar maples in decline were found to have Ca2+/Al ratios less than 1, as well as lower base cation concentrations and pH values compared with plots of healthy sugar maples (Drohan et al., 2002). Sugar maple regeneration has also been noted to be restricted under conditions of low soil Ca2+ levels (Juice et al., 2006). These indicators have all been shown to be related to the deposition of atmospheric nitrogen and sulfur. Additional linkages between acidifying deposition and sugar maple physiological responses are indicated in Table 2-2. Table 2-2. Summary of Linkages between Acidifying Deposition, Biogeochemical Processes That Affect Ca2+, Physiological Processes That Are Influenced by Ca2+, and Effect on Forest Function Biogeochemical Response to Acidifying deposition Leach Ca2+ from leaf membrane Reduce the ratio of Ca2+/AI in soil and soil solutions Physiological Response Decrease the cold tolerance of needles in red spruce Dysfunction in fine roots of red spruce blocks uptake of Ca2+ Effect on Forest Function Loss of current-year needles in red spruce Decreased growth and increased susceptibility to stress in red spruce March 2010 11 Draft - Do Not Quote or Cite ------- Appendix B Biogeochemical Response to Acidifying deposition Reduce the ratio of Ca2+/AI in soil and soil solutions Reduce the availability of nutrient cations in marginal soils Physiological Response More energy is used to acquire Ca2+ in soils with low Ca2+/AI ratios Sugar maples on drought-prone or nutrient-poor soils are less able to withstand stresses Effect on Forest Function Decreased growth and increased photosynthetic allocation to red spruce roots Episodic dieback and growth impairment in sugar maple Source: Fenn and colleagues, 2006. 1 Although the main focus of the Terrestrial Acidification Case Study outlined in the Risk 2 and Exposure Assessment for Review of the Secondary National Ambient Air Quality Standards 3 for Oxides of Nitrogen and Sulfur (U.S. EPA, 2009) was an evaluation of the negative impacts of 4 nitrogen and sulfur deposition on soil acidification and tree health, it should be recognized that 5 under certain conditions, nitrogen and sulfur deposition can have a positive impact on tree health. 6 Nitrogen limits the growth of many forests (Chapin et al., 1993; Killam, 1994; Miller, 1988), and 7 therefore, in such forests, nitrogen deposition may act as a fertilizer and stimulate growth. 8 Forests where critical acid loads are not exceeded by nitrogen and sulfur deposition could 9 potentially be included within this group of forests that respond positively to deposition. These 10 potential positive growth impacts of nitrogen and sulfur deposition are discussed further, and the 11 results of case study analyses are presented in Attachment A of Appendix 5 of the Risk and 12 Exposure Assessment (U.S. EPA, 2009). 13 In summary, among potential influencing factors, including elevated ozone levels and 14 changes in climate, the acidification of soils is one of the factors that can negatively impact the 15 health of red spruce and sugar maple. Mortality and susceptibility to disease and injury can be 16 increased and growth decreased with acidifying deposition. Therefore, the health of sugar maple 17 and red spruce was used as the endpoints (ecological responses) to evaluate acidification in 18 terrestrial systems. "Health" in the context of the Risk and Exposure Assessment terrestrial 19 acidification case study was defined as the physiological condition of a tree that impacts growth 20 and/or mortality. 21 The Simple Mass Balance (8MB) model was used to estimate critical loads of acidity in 22 the Risk and Exposure Assessment case study (Equation 2-1). The full derivation of this equation 23 is detailed in the TCP Mapping and Modeling Manual (UNECE, 2004). March 2010 12 Draft - Do Not Quote or Cite ------- Appendix B 1 CL(S + N) = BCdep - Cldep + BCW -Bcu + N, + Nu + Nde - ANCle,cnt (2-1) 2 where 3 CL(S+N) = forest soil critical load for combined nitrogen and sulfur acidifying 4 deposition ((N+S)comb) 5 BCdep = base cation (Ca2+ + K+ + Mg2+ + Na+) deposition1 6 Cldep = chloride deposition 7 BCW = base cation (Ca2+ + K+ + Mg2+ + Na+) weathering 8 Bcu = uptake of base cations (Ca2+ + K+ + Mg2+) by trees 9 N; = nitrogen immobilization 10 Nu = uptake of nitrogen by trees 11 Nde = denitrification 12 ANCie,crit = forest soil acid neutralizing capacity of critical load leaching 13 Some of these parameters had defined or selected input values (BCdep, Cldep, N;, Nu and Nde), 14 while four of these parameters, including BCW, Bcu, Nu and ANCie,Crit, required calculation. 15 For the Risk and Exposure Assessment's terrestrial acidification case study, three values 16 of the indicator of critical load, expressed as (Bc/Al)crit soil solution ratio, were selected to 17 represent different levels of tree protection associated with total nitrogen and sulfur deposition: 18 0.6, 1.2, and 10 (Table 2-3). The (Bc/Al)crit ratio of 0.6 represents the highest level of impact 19 (lowest level of protection) to tree health and growth and was selected because 75% of species 20 found growing in North America experience reduced growth at this Bc/Al ratio. In addition, a 21 soil solution Bc/Al ratio of 0.6 has been linked to a 20% and 35% reduction in sugar maple and 22 red spruce growth, respectively. The (Bc/Al)crit ratio of 1.2 is considered to represent a moderate 23 level of impact, as the growth of 50% of tree species (found growing in North America) was 24 negatively impacted at this soil solution ratio. The (Bc/Al)crit ratio of 10.0 represents the lowest 25 level of impact (greatest level of protection) to tree growth; it is the most conservative value used 26 in studies that have calculated critical loads in the United States and Canada (Canada (McNulty 27 et al., 2007; NEG/ECP, 2001; Watmough et al., 2004). 1 The ICP Mapping and Modeling Manual (UNECE, 2004) recommends that wet deposition be corrected for sea salt on sites within 70 km of the coast. Both the HBEF and KEF case study areas are greater than 70 km from the coast, so this correction was not used. March 2010 13 Draft - Do Not Quote or Cite ------- Appendix B Table 2-3. The Three Indicator (Bc/AI)crit Soil Solution Ratios and Corresponding Levels of Protection to Tree Health and Critical Loads Indicator (Bc/AI)crit Soil Solution Ratio 0.6 1.2 10.0 Level of Protection to Tree Health Low Intermediate High Critical Load High Intermediate Low 1 The prediction of tree protection achieved using each of these three indicator ratios of 2 0.6, 1.2, and 10.0 includes an important estimation of base cation weathering as shown in 3 Equation 2-1, above. The purpose of this report is to describe the methodologies, data 4 requirements, data availability, and uncertainties associated with estimating base cation 5 weathering. 6 3. AQUATIC BASE CATION WEATHERING METHODOLOGY 7 The ISA (US EPA, 2008) reports that the principal factor governing the sensitivity of 8 terrestrial and aquatic ecosystems to acidification from sulfur and nitrogen deposition is geology 9 (particularly surficial geology). Geologic formations having low base cation supply generally 10 underlie the watersheds of acid-sensitive lakes and streams. Other factors that contribute to the 11 sensitivity of soils and surface waters to acidifying deposition include topography, soil 12 chemistry, land use, and hydrologic flowpath. Surface waters in the same setting can have 13 different sensitivities to acidification, depending on the relative contributions of near-surface 14 drainage water and deeper groundwater (Chen et al., 1984; Driscoll et al., 1991; Eilers et al., 15 1983). Lakes and streams in the United States that are sensitive to episodic and chronic 16 acidification in response to SOX, and to a lesser extent NOX, deposition tend to occur at relatively 17 high elevation in areas that have base-poor bedrock, high relief, and shallow soils (U.S. EPA, 18 2008, Section 3.2.4.1). 19 3.1 Aquatic Base Cation Weathering 20 Base cation weathering for aquatic acidification critical loads must be representative of 21 the catchment around the waterbody of interest. This aspect of quantification of the weathering 22 rate provides the difference when calculating weathering rates for aquatic versus terrestrial 23 analysis purposes. The process of weathering itself provides the only natural in-soil source of 24 alkalinity that is available to neutralize acidity inputs to the system over the long term. Chemical 25 weathering of the mineral matrix within soils supplies base cations that are removed from soil March 2010 14 Draft - Do Not Quote or Cite ------- Appendix B 1 due to acid inputs. Therefore, the rate of weathering of the soils within a catchment is dependent 2 on the chemical and physical properties of the soil (Sverdrup et al., 1992; Whitfield et al., 2006). 3 As indicated in Section 2.2, the average flux of base cations weathered in a catchment and 4 reaching the lake or streams (BCW) is difficult to measure or compute from available information 5 (Henriksen and Posch, 2001; Henriksen et al., 2002; Langan et al., 2001). Approaches also differ 6 based on whether the weathering rate needs to account for only in-soil processes (profile 7 measurements and models) or whether it needs to account for the flux of base cations to surface 8 water (spatially integrated catchment data and models) (Langan et al., 2001). 9 In the Aquatic Acidification case study in the REA Report (U.S. EPA, 2009), BCW rates 10 were not directly calculated. Instead, the F-factor approach was used to calculate the pre- 11 acidification, non-marine flux of base cations (BC*0) for each lake or stream. An F-factor 12 (explained in Section 3.2.2) is a ratio of the change in non-marine base cation concentration due 13 to changes in strong acid anion concentrations (Henriksen, 1984; Brakke et al., 1990), as shown 14 in the following equations: 15 BC*0 = BC*t - F (SO*4,t - SO*4,o + NO*3,t - NO*3)0) (3-1) 16 F = ([BC*]t - [BC*]0)/([S04*]t - [S04*]o + [NO3*]t - [NO3*]o) (3-2) 17 where the subscripts t and 0 refer to present and pre-acidification conditions, respectively. The 18 pre-acidification N(V concentration, NO*3j0, was assumed to be zero. Several attempts have 19 been made to create empirical relations for the F-factor and the pre-acidification SO/ 20 concentration. Although the Aquatic Acidification case study relied on two of these relations, it 21 must be noted that they were developed for areas outside of the U.S. and, therefore, cannot be 22 applied to the conditions found within U.S. soils and climates without introducing a source of 23 uncertainty (Henriksen and Posch, 2001; Henriksen et al., 2002; Brakke et al., 1989; Posch et al., 24 1997). Notwithstanding the lack of U.S.-based empirical relations, the F-factor can be used to 25 derive BCW estimates. Assuming that all atmospheric deposition of base cations that falls within 26 a catchment passes through to the surface water and that one can accurately estimate the uptake 27 of base cations within the catchment, the BCW could ultimately be backed out of these 28 relationships. However, both of these assumptions are likely to introduce an additional amount of 29 uncertainty into the BCW estimates. March 2010 15 Draft - Do Not Quote or Cite ------- Appendix B 1 For a national aquatic acidification assessment, different methods must be employed to 2 estimate BCW rates. In some studies, simple assumptions for the BCW are utilized. For instance, in 3 a study by Dupont and colleagues (2005) using the SSWC, the authors assumed that weathering 4 rates were time-independent and did not affect critical load estimates. In more advanced process 5 modeling applications, such as ones using the Model of Acidification of Groundwater in 6 Catchments (MAGIC), weathering rates can be adjusted during calibration and allowed to vary 7 over ranges like 0 and 5 times the observed watershed base cation export for base cation 8 weathering (Sullivan et al., 2004). There are several different approaches to estimating the 9 weathering rate of a soil or a catchment, ranging from empirical relations to mass balance 10 methods to calibrated process models. According to Whitfield and colleagues (2006) "to date no 11 method has proven to be superior in application to different soil types and differing levels of soil 12 acidification." The remainder of this section is intended to examine the BCW estimation methods 13 that would be applicable to a national aquatic acidification critical loads analysis giving 14 consideration to the limitations of the method and the possible data and processing requirements 15 for the analysis. 16 3.2 Methodologies for Determining Base Cation Weathering Values in the United States 17 3.2.1 Difficulties in estimating base cation weathering 18 Consideration must be given to several factors in the estimation of base cation weathering 19 fluxes for aquatic acidification (Sverdrup et al., 1992; Whitfield et al., 2006; Rapp and Bishop, 20 2009; Henriksen and Posch, 2001; Henriksen et al., 2002): 21 1. The weathering contribution of the entire catchment must be understood and not 22 simply the weathering contribution of certain soil profiles within the catchment. 23 Additionally, the various types of land use (e.g. agriculture or forest) within a 24 catchment may all affect weathering rates differently. 25 2. When utilizing soil profile weathering methods, the characteristics of the entire soil 26 profile must be considered and weighted according to catchment composition as 27 opposed to only the rooting zone in individual profiles as used in determining 28 weathering for terrestrial acidification purposes. 29 3. Based on the critical load method chosen, it is often necessary to assume that the BCW 30 remains constant over the length of the analysis. While this simplifies the estimation 31 of BCW, it introduces uncertainty into any analysis. The length of the analysis March 2010 16 Draft - Do Not Quote or Cite ------- Appendix B 1 scenario must be sufficiently long and have supporting data in order to provide a 2 long-term average, which is not subjected to short-term variations. 3 4. The data requirements for a national assessment necessitate using similar data sources 4 for all applications so that assumptions and methods can remain constant across the 5 nation. 6 5. The application of any empirical relations for calculation of BCW or intermediate 7 component of BCW (e.g., the F-factor) must be validated against the geographic region 8 in which they will be applied. Given that most empirical relations developed to date 9 were based on data from European nations, these relations need to be recalibrated to 10 data from the U.S. 11 Given all of these factors, estimation of BCW for a national application poses a significant 12 challenge. The methods detailed in the following section seek to balance the limitations and 13 benefits of each approach to estimation of BCW. 14 3.2.2 Approaches to estimating BCW for Aquatic Acidification 15 Work presented in the scientific literature over the last two decades provides several 16 different approaches researchers have taken to estimate the BCW rates for aquatic effects. These 17 approaches do not always differentiate between the actual weathering processes in-soil and the 18 other ion exchange processes taking place (Langan et al., 2001). Approaches to estimating BCW 19 also vary between terrestrial and aquatic studies. Aquatic studies of acidification must capture 20 the weathering rates of all soil horizons which contribute base cations and not solely the rooting 21 zone as specified in terrestrial acidification studies (Whitfield et al., 2006). 22 Four general categories of approaches are outlined for determining BCW for aquatic 23 acidification critical loads calculations using the SSWC. 24 1. Budgets studies of catchments or watersheds; 25 2. Historical weathering rate determinations; 26 3. Empirical relations; and 27 4. Process-based models. 28 In the case of empirical data relations and process-based models, specific methods are 29 provided. The strengths and weaknesses of either the general category or specific approach in 30 terms of both utilization in aquatic acidification critical loads calculations and estimation of BCW 31 are examined in the following paragraphs and Table 3-1. March 2010 17 Draft - Do Not Quote or Cite ------- Appendix B 1 Budget Studies - Budget studies are simple means of determining fluxes within a system 2 by balancing the masses coming into and going out of a system. In determining the BCW, a mass 3 balance would be performed around the base cations fluxes within a watershed, where 4 atmospheric deposition constitutes the main source input and streamflow the main output. Within 5 the balance, base cation retention is also accounted for through uptake by biomass and 6 immobilization in the soil. The BCW developed from budget studies represent integrated values 7 for the whole watershed as desired for aquatic acidification estimates as opposed to only 8 weathering from the rooting zone as desired for terrestrial acidification (Sverdrup and Warfvinge 9 1988; Miller, 2001). Depending on how the balance is set up, the balance can be a single 10 equation around the total base cation flux or a series of equations for each individual cation. The 11 setup of the equations leads to the primary limitation of the method in that while it is a relatively 12 simple concept, the individual fluxes within the balance are not easily measured or known 13 (Bricker et al., 2003). 14 Most mass balance calculations require an assumption of steady-state behavior. This 15 assumption is easily justifiable over long periods of record. Additional limitations of the method 16 evolve from the number of unknown fluxes (e.g. weathering rate of individual minerals) within 17 the equations defining the balance. Researchers have utilized a variety of techniques to overcome 18 this limitation, including applying simplifying assumptions or adding additional equations. 19 However, with each assumption or additional equation, a greater amount of uncertainty that must 20 be quantified is added into the analysis. Data sources for a mass balance can also be variable 21 depending on the complexity of the relationships defined within the balance. While databases 22 and studies may exist for major elements at a variety of sites, comparable data for trace or more 23 complex elements may be lacking (Velbel and Price, 2007). 24 Historical Rate Determinations - This approach is detailed in Section 4 for terrestrial 25 acidification approaches. Because the BCW flux required for aquatic acidification approaches 26 requires characterization of the whole soil profile averaged across a catchment or watershed, this 27 approach can become computationally intensive for aquatic purposes. While it is possible to 28 conduct such an approach on a small scale for an aquatic acidification assessment, it is more 29 likely suited to terrestrial applications and so explanation is provided in those sections of the 30 document. March 2010 18 Draft - Do Not Quote or Cite ------- Appendix B 1 Empirical Relations - A number of empirical relationships have been developed to 2 calculate BCW, or related factors, from water quality data alone. Empirical relationships are only 3 as strong as the data on which they are based and are only applicable to the geographic regions 4 from which the calibration data is obtained. 5 F-Factor: The F-factor is defined as the ratio of change in non-marine base cation concentrations 6 due to changes in strong acid anion concentrations (Henriksen, 1984; Brakke et al., 1990). (See 7 Section 3.1.) A situation where F = 1 indicates that only soil acidification occurs within the 8 catchment, i.e. all incoming protons are neutralized in the catchment. When F = 0, then only 9 water acidification is occurring and none of the incoming protons are neutralized in the 10 catchment. Using historical data from Norway, Sweden, U.S.A. and Canada, the F-factor was 11 estimated empirically to be in the range 0.2-0.4 (Henriksen, 1984). Several empirical 12 relationships have been developed in order to calculate the F-factor based on current base cation 13 concentrations using data from Norway (Brakke et al., 1990) or on pre-acidification base cation 14 concentration using data from Finland (Posch et al., 1993). 15 There are several limitations to using the F-factor. While it is simple to apply anywhere the data 16 is available to satisfy the empirical relations, these relations are really only valid in Norway, 17 Finland, or wherever the specific relation was derived. In several instances, researchers have 18 applied the Norway- or Finland-based relations to Canadian (Watmough et al., 2005) and U.S. 19 study locations (Henriksen et al., 2002; Dupont et al., 2005; U.S. EPA, 2009) with the assumption 20 that the empirical equations provide adequate characterization of the relationship between base 21 cation concentrations and the F-factor. 22 A second major limitation in utilizing the F-factor is that this derived factor does not specifically 23 quantify the BCW flux. Instead it provides calculation of the base cations leached from the soil, 24 which includes BCw and base cations derived through deposition inputs to the system, or 25 removed by harvesting (Henriksen et al., 2002; Rapp and Bishop, 2009). Although the SSWC is 26 most often used with the F-factor, in a national application where we seek to specifically quantify 27 the BCW, an alternative method should be used. 28 Indicator element in conjunction with weathering ratios: Chen et al., 2004: "Weathering rates at 29 Arbutus watershed could also be obtained using sodium as indicator element, as described by 30 Gbondo-Tugbawa et al. (2001). The weathering inputs of the indicator element (sodium) could be 31 derived using a mass balance approach, and the derived sodium weathering rate was used in 32 conjunction with base cation weathering ratios reported by Johnson and Lindberg (1992) for the 33 HF to derive weathering rates of other base cations. Using this method, the weathering rates of 34 sodium and calcium derived for Arbutus watershed are very similar to values derived through 35 calibration, whereas rates of magnesium and potassium derived using these two methods showed 36 some discrepancies (Table II)." 37 Weathering rates vs. Stream chemistry or landscape variables: This approach begins with a set of 38 BCW for a specific set of water bodies. The values of BCW are then regressed against the stream 39 chemistry parameters, such as ANC, in order to find a correlation relationship. These regression 40 relationships are then applied to stream chemistry of other water bodies within a defined region of 41 interest to find the BCW for the water bodies. In areas where stream chemistry is not available, 42 landscape variables can be used in place to find correlations with the BCW. 43 The limitation with this approach is that a statistically significant number of BCW values must be 44 available from which to create a regression relationship. Also, the region in which the 45 extrapolation is valid must be defined. In work by Sullivan and colleagues (2004), extrapolation March 2010 19 Draft - Do Not Quote or Cite ------- Appendix B 1 of modeled BCW values was completed using groupings of physiographic region and ANC class. 2 The value of this approach is that it can specifically be applied to BCW and can be based on 3 modeled, monitored, or estimated BCW values as long as there are a sufficient number of values 4 for extrapolation. Other works have extrapolated ANC values based on chemistry and landscape 5 variables in a similar manner with a high level of success (Sullivan et al., 2007b, Nanus et al., 6 2009). 7 Process-based Models - Mineral weathering terms within modeling simulations can be a 8 large source of uncertainty as the weathering term utilized in most process models, in attempts to 9 represent reality, impacts the loss of base cations to surface waters. Therefore, when little is 10 known about the true weathering rate or the constraints on its values, models must utilize 11 calibration procedures against in-stream water chemistry data to arrive at a likely weathering rate 12 (Chen et al., 2004; Sullivan et al., 2004). 13 Process-based models vary greatly in their range of processes represented, complexity of 14 representations, time step, and required data inputs. Overall, there is no perfect model but the 15 best candidate for a task can be chosen provided the available data, the area of concern, and the 16 goals of the analysis. In this case, we would seek to use all available data resources in order to 17 derive a range of spatially-explicit BCW values across the nation. 18 Descriptions of the four candidate process models available for use across the country for 19 determining BCW are provided below. In order to provide as concise a description as possible, 20 these model summaries are taken directly from the scientific literature. Summation of the 21 strengths and weaknesses of each model is provided after the model description. 22 DavCent-Chem: "DayCent-Chem links together two widely accepted and tested models—(1) a 23 daily time-step nutrient cycling and soil hydrology model, version 5 of the DayCent model 24 [Parton et al., 1998], and (2) PHPvEEQC, an aqueous geochemical equilibrium model [Parkhurst 25 and Appelo, 1999]—to form a model that simulates N, P, S, and carbon (C) ecosystem dynamics 26 and soil and stream water acid-base chemistry (fig. 1.2). DayCent-Chem computes atmospheric 27 deposition, soil water fluxes, snowpack and stream dynamics, plant production and uptake, soil 28 organic matter decomposition, mineralization, nitrification, and denitrification (left side of fig. 29 1.2) while utilizing PHREEQC's low-temperature aqueous geochemical equilibrium calculations, 30 including CO2 dissolution, mineral denudation, and cation exchange, to compute soil water and 31 stream chemistry (right side of figure). DayCent-Chem's daily soil solution and stream water 32 chemistry calculations make it possible to use the model to investigate the potential for episodic 33 acidification" (Hartman et al., 2009). 34 DayCent-Chem was recently applied to eight different mountain watersheds from the west to the 35 east with success in certain capacities, therefore, making it a suitable candidate for a national 36 analysis. These applications did highlight difficulties in determining realistic weathering rates in 37 certain areas. With DayCent-Chem, a user must specify an initial value for weathering, which 38 may be adjusted during calibration. In several instances, this value was first set to measured or 39 estimated values for the area of interest and then modified largely during calibration (Hartman et 40 al., 2009). While the daily time step and biotic processes represented by the model provide a March 2010 20 Draft - Do Not Quote or Cite ------- Appendix B 1 more complex view of the environment, they also add a complexity to the model that appears to 2 greatly impact the estimation of the key parameter for this analysis. 3 MAGIC: "MAGIC is a lumped-parameter model of intermediate complexity, developed to 4 predict the long-term effects of acidic deposition on surface water chemistry [Cosby et al., 1985a, 5 1985b]. The model simulates soil solution chemistry and surface water chemistry to predict the 6 monthly and annual average concentrations of the major ions in these waters. MAGIC consists of 7 (1) a section in which the concentrations of major ions are assumed to be governed by 8 simultaneous reactions involving SO42" adsorption, cation exchange, dissolution-precipitation- 9 speciation of Al and dissolution-speciation of inorganic C and (2) a mass balance section in 10 which the flux of major ions to and from the soil is assumed to be controlled by atmospheric 11 inputs, chemical weathering, net uptake and loss in biomass and losses to runoff. At the heart of 12 MAGIC is the size of the pool of exchangeable base cations in the soil. As the fluxes to and from 13 this pool change over time owing to changes in atmospheric deposition, the chemical equilibria 14 between soil and soil solution shift to give changes in surface water chemistry. The degree and 15 rate of change of surface water acidity thus depend both on flux factors and the inherent 16 characteristics of the affected soils" (Sullivan et al., 2004). 17 The strengths of MAGIC lie in its simplicity and ability to be applied for a large number of 18 lakes/streams in batch processes. MAGIC has been in use since the 1980s, has been widely 19 applied within the eastern portions of the country with more limited applications in the West. (See 20 Section 3.3 for further discussion.) The simplicity of MAGIC's mass balances approach also 21 counts as one of its limitations because it may not account for all of the biotic processes that 22 affect the weathering rate. MAGIC determines the BCW through calibration to water chemistry 23 data. The "fuzzy optimization" procedures now built into MAGIC allow for an optimized value 24 of the BCW to be determined from a series of calibrations at each modeling location (Sullivan et 25 al., 2004). 26 PnET-BGC: "PnET-BGC is an integrated forest-soil-water model that has been used to assess the 27 effects of air pollution and land disturbances on forest and aquatic ecosystems [Gbondo-Tugbawa 28 et al., 2001]. The model was developed by linking two submodels: PnET-CN (PnET-carbon and 29 nitrogen) [Aber et al., 1997] and BGC [Gbondo-Tugbawa et al., 2001]. The main processes in the 30 model include tree photosynthesis, growth and productivity, litter production and decay, 31 mineralization of organic matter, immobilization of nitrogen, nitrification [Aber et al., 1997], 32 vegetation and organic matter interactions of major elements, abiotic soil processes, solution 33 speciation, and surface water processes [Gbondo-Tugbawa et al., 2001].... For lake simulations, 34 it is assumed that the water column is completely mixed. The model predicts monthly 35 concentrations and fluxes of major solutes in lake water, monthly concentrations and pools of 36 exchangeable cations and adsorbed sulfate in soil, and monthly fluxes of major solutes from soil 37 and forest vegetation" (Zhai et al., 2008). 38 Chen and others (2004) nicely summarize the tradeoffs associated with utilizing the PnET-BGC 39 model: "A strength of PnET-BGC over other acidification models is its ability to simulate 40 [vegetation and microbial processes]. However, this representation can also be a limitation. The 41 model depicts large element pools in soil and large fluxes through biotic processes. Any change 42 in these pools and fluxes will greatly influence the element budgets. If these simulated fluxes are 43 not accurate, then model predictions will misrepresent element dynamics." Additionally, almost 44 all of the PnET-BGC applications to date have been completed within the eastern portions of the 45 country mostly focusing in the Adirondacks and Hubbard Brook Experimental Forest (Gbondo- 46 Tugbawa et al., 2001; Chen et al., 2004; Zhai et al., 2008). Expanding this model to western or 47 southern areas would necessitate large amounts of data gathering and processing as well as testing 48 of the representation of the biotic processes found in these differing ecosystems. March 2010 21 Draft - Do Not Quote or Cite ------- Appendix B 1 PROFILE: "PROFILE [WarfVinge and Sverdrup, 1992] is a steady-state soil chemical model 2 with a weathering rate sub-model that calculates weathering rates (for each base cation) explicitly 3 using independent soil properties. Mineral dissolution reactions governing the rate of weathering 4 involve many components in the liquid phase including H2O, FT, OH", CO2 and organic acids. 5 These serve as the principle method for cataloguing the contribution of chemical reactions 6 between soil solution and silicate minerals to base cation release. Inhibition of the reactions 7 through increased concentrations of the products is accounted for by rate reduction factors. 8 Precipitation of secondary minerals is subtracted from the total base cation release rate. Climate 9 data, soil properties and detailed soil mineralogy are used as inputs to the model [WarfVinge and 10 Sverdrup, 1992]" (Whitfield et al., 2006). 11 The PROFILE model is more fully explained in Section 4 for the terrestrial BCW approaches. 12 While the PROFILE model provides a highly deterministic, process-based representation of 13 mineral weathering, trying to utilize this model to determine the base cations weathering and 14 reaching surface water bodies requires the representation of all soil horizons that may contribute 15 to weathering and the summarization of BCW calculations by catchments surrounding each water 16 body of interest. These two qualifications on top of the basic PROFILE application introduce a 17 large amount of complexity into the modeling analysis. 18 19 March 2010 22 Draft - Do Not Quote or Cite ------- Appendix B Table 3-1. Review of Modeling Approaches (and models) to Estimate Base Cation Weathering for Aquatic Critical Acid Load Determinations Model Approach Budget Studies Historical Rate Determinations Empirical Data Relations F-Factor Description of Method mass balance of inputs and outputs of base cations within catchment or watershed loss of base cations in soil profile relative to stable element (Zr, Ti, quartz or rutile) modeled relationships between surface water characteristics and site conditions or atmospheric deposition measures a factor that combines the effects of deposition and weathering Data Require- ments medium low low- high low- medium Model Complexity Medium Low low- high Low Suitability for Estimating BCW for Aquatic Critical Acid Load Determinations in The United States low; BCW estimate is often an integrated value for whole catchment or watershed medium; restricted to sites with young soils of known age (eg., soils that have formed since the most recent glacial event, -20,000 years ago) low-medium low; most accurately applied to sites similar to those where the model was derived; if new derivations can be completed for the U.S. the suitability of this method would increase Suitability tor Mapping BCW Over Large Regions in The United States low - medium (based on data availability); may require Sr isotope ratio of stream chemistry to separate exchangeable versus weathered base cation sources low; restricted to sites with young soils and sites where historical rate determinations have been conducted low-medium low; most accurately applied to sites similar to those where the model was derived; if new derivations can be completed for the U.S. the suitability of this method would increase References Brickeret al., 1993; Velbel and Price, 2007 Sverdrup et al., 1998; Sverdrup et al., 1990 Brakke et al., 1990; Henriksen and Posch, 2001; Henriksen et al., 2002; Rapp and Bishop, 2009 March 2010 23 Draft - Do Not Quote or Cite ------- Appendix B Model Approach Indicator element in conjunction with weathering ratios Weathering rates vs. Stream chemistry or landscape variables Process-Based Models Description of Method determine weathering rate through mass balance methods for element such as sodium (Na) then apply defined ratios to determine weathering rates of additional elements utilize weathering rates determined by other methods and extrapolate to additional areas based on site characteristics Steady-state and dynamic models that rely on mathematical relationships representing soil and surface water processes Data Require- ments low-medium low- medium medium - high Model Complexity low low medium - high Suitability for Estimating BCW for Aquatic Critical Acid Load Determinations in The United States low; most accurately applied to sites similar to those where the model was derived medium; suitability will depend on the ease at which derived weathering rates can be obtained and how strong the regressions between BCW and site characteristics are Suitability tor Mapping BCW Over Large Regions in The United States low; most accurately applied to sites similar to those where the model was derived medium; relatively good success has been had at extrapolating BCwto additional sites based on stream chemistry; suitability will depend on data availability References Gbondo-Tugbawa et al., 2001,2002; Chen etal.,2004 Sullivan et al., 2004, 2007a, 2007b; Webb et al., 1994; Nanus et al., 2009 March 2010 24 Draft - Do Not Quote or Cite ------- Appendix B Model Approach DayCent-Chem MAGIC PnET-BGC Description of Method Mineral weathering rates are set and then calibrated within the process model; Rates are specified by mineral and not necessarily base cations alone BCW determined through calibration to fulfill the requirements of a catchment mass balance by optimizing simulated soil and surface water chemistry to monitored values BCW determined through calibration and held constant throughout dynamic modeling simulations Data Require- ments high medium - high high Model Complexity high medium- high medium - high Suitability for Estimating BCW for Aquatic Critical Acid Load Determinations in The United States medium - high; provides daily time step results which can be used to estimate time to recovery or time to damage; uncertainty on how well model can simulate BCW in some areas will impact confidence of results in these areas medium - high; numerous applications in the east with some, but fewer in number, applications in the west; little coverage in the Midwest but these areas are less of a concern for aquatic acidification effects low-medium; model applications mostly completed only within northeastern U.S. vegetation and other biotic processes represented by the model would need validation to other regions of the country Suitability tor Mapping BCW Over Large Regions in The United States medium; DayCent-Chem has had trouble in estimating mineral weathering rates in some areas of the country medium- high; will be restricted in areas where soils data are lacking (some western areas); otherwise, highly applicable in any areas where MAGIC applications have been completed low-medium; because BCwis found through calibration alone for this model, the other model processes and input data must be validated for any application before calibration can be used for BCW References Hartman et al., 2007; Hartman et al., 2009 Cosby et al., 1985a, 1985b, 1989a; Sullivan et al., 2004; Sullivan et al., 2008 Gbondo-Tugbawa et al., 2001; Chen et al., 2004; Zhaietal., 2008 March 2010 25 Draft - Do Not Quote or Cite ------- Appendix B Model Approach PROFILE Description of Method BCW determined as a function of weathering of individual soil minerals and field- based soil and biotic conditions Data Require- ments high Model Complexity high Suitability for Estimating BCW for Aquatic Critical Acid Load Determinations in The United States medium - high; may have restrictions in desert regions and areas that are lacking necessary data; also must be able to characterize catchment summary values and not solely individual profiles Suitability tor Mapping BCW Over Large Regions in The United States medium - high; may have restrictions in desert regions and areas that are lacking necessary data References Warfvinge and Sverdrup, 1992 and 1995; Sverdrup, 1990 March 2010 26 Draft - Do Not Quote or Cite ------- Appendix B 1 3.3 Proposed Methodology for Estimating and Mapping Base Cation Weathering for 2 Aquatic Critical Acid Load Calculations 3 In determining the proposed methodology for a national assessment, the identified 4 strengths and weaknesses of each approach in the previous section had to be weighed against one 5 another. Because every method required a large environmental data component, the largest 6 deciding factor in the proposed approach became the number and spatial representation of 7 previous applications of an approach within the United States. This decision factor immediately 8 ruled out applying any of the empirical relationships (e.g. F-factor, relation of BCW to stream 9 chemistry) derived primarily with data from other countries, although it did not rule out deriving 10 new relationships using the same methods. Ultimately, the F-factor approach was not chosen 11 because it did not directly provide a BCW rate. Additionally, application of an empirical relation 12 alone provided little information on the long-term versus current state of the ecosystem. 13 Therefore, a combination of a process-based model determination of BCW rates with regional 14 expansion of these rates through empirical relations is proposed at the methodology for a 15 national assessment. 16 Utilizing a process-based model, which can calibrate BCW rates to stream or lake 17 chemistry across any number of years, provides a credible long-term estimate of the BCW rate 18 that can be input into the SSWC in order to obtain the system critical load. The process-based 19 model most widely applied throughout the U.S. to date is the MAGIC model. The intermediate 20 complexity of this model provides a balance between data inputs required to run the model and 21 the processes involving base cations, nitrogen, and sulfur within a watershed, which is 22 considered a requirement of providing a national assessment. Finally, because of its wide 23 application, MAGIC has been extensively tested against independent databases providing the 24 opportunity for iterative model testing and refinement (Sullivan, 2000). 25 The following steps outline the main processes of the method: 26 Step 1. Definition of MAGIC study sites and the regions to which each grouping of 27 study sites may be extrapolated. 28 Step 2. Data gathering and processing for population of the MAGIC model for 29 each study site with additional regional data gathering of available 30 stream/lake chemistry and landscape parameters. 31 Step 3. MAGIC modeling application on selected stream/lake study sites where 32 BCW is arrived at through calibration against water chemistry data. March 2010 27 Draft - Do Not Quote or Cite ------- Appendix B 1 2 3 Step 4. Extrapolation of BCW for modeled streams/lakes to other waterbodies within the region through correlation analysis using stream chemistry data, where available, and landscape parameters in its absence. Figure 3-1 provides a flow chart of these steps and their components. Identification MAGIC study sites and applicable regions for extrapolation Identification of input data for the MAGIC model and of stream chemistry and landscape parameters Input Data Classes Atmospheric Deposition (including annual precipitation) Construct Model Input Tables MAGIC Application at Study Sites with Calibration to Water Chemistry for BCw Correlation Analysis between Site BCw Values and Stream Chemistry or Landscape Parameters Application of Regional Regression Relationships Developed by Site Grouping to Un-Modeled Sites 5 6 7 CMapping of BCw Values Across \ Regions of the Nation J Figure 3-1. Process steps for estimating BCW using the MAGIC model with regional extrapolation March 2010 28 Draft - Do Not Quote or Cite ------- Appendix B 1 Step 1. Definition of MAGIC study sites and the regions to which each grouping of 2 study sites may be extrapolated. 3 MAGIC has been used to assess acidification impacts in a large number of areas across 4 North America (Table 3-2). These previous applications should be utilized where possible to 5 provide a starting point for the national analysis. Sites within the eastern United States likely 6 provide a wide range of coverage from which initial extrapolations can begin. Within the mid- 7 west and western areas of the country, additional sites will need to be investigated. Authors of 8 these studies should be contacted to obtain data sources and model results. Previous model 9 applications should be compared for the years and objectives of the analysis and input data to 10 determine if the results already created could be utilized in an extrapolation analysis without 11 rerunning the model. Table 3-2. Locations of Previous MAGIC Applications within the U.S. and Canada1 Location(s) 25 lakes in south-central Ontario, Canada 2 catchments located in Nova Scotia, Canada Maryland 36 lake catchments in the Adirondack Mountains of New York 40 to 50 sites within each of three physiographic provinces in the eight-state southern Appalachian Mountains region 33 representative watersheds in the Adirondacks Shenandoah National Park 60 Southern Appalachian streams Joyce Kilmer And Shining Rock Wilderness Areas (North Carolina/Tennessee) Monongahela National Forest, West Virginia Shasta Lake, Idaho Libby Lake, Montana Popo Agie Wilderness, WY, and Weminuche Wilderness, CO Rocky Mountain, Grand Teton, Sequoia, and Mount Rainier National Parks The Loch, a subalpine lake in Rocky Mountain National Park in Colorado 2 locations in the Sierra Nevadas Reference Aherne, J, P.J. Dillon, and B.J. Cosby. 2003. Dennis, I.F., T.A. Clair, and B.J. Cosby. 2005 Ellis, H., and M. Bowman. 1994. Church, M.R. and J. Van Sickle. 1999. Sullivan, T.J., B.J. Cosby, AT. Herlihy, J.R. Webb, A.J. Bulger, K.U. Snyder, P.P. Brewer, E.H. Gilbert, and D.L. Moore. 2004. Sinha, R., M.J. Small, P.P. Ryan, T.J. Sullivan, and B.J. Cosby. 1998. Bulger, A. J; Dolloff, C. A.; Cosby, B. J.; Eshleman, K. N.; Webb, J. R., and Galloway, J. N. 1995 Bulger AJ, Cosby BJ, Webb JR. 2000. Sullivan, T.J. and B.J. Cosby. 2002 Sullivan, T.J. and B.J. Cosby. 2004 Eilers J.M., B.J. Cosby, J.A. Bernet, T.A. Sullivan, 1998. Bernett, J.A., Eilers J.M., B.J. Cosby. 1997. Sullivan, T.J., Cosby, B.J., Bernert, J.A., and Eilers, J.M. 1998. Cosby and Sullivan. 2001 Sullivan, T.J., B.J. Cosby, K.A. Tonnessen, and D.W. Clow. 2005. Sullivan and Eilers, 1996 References from this table are presented in Appendix 2. March 2010 29 Draft - Do Not Quote or Cite ------- Appendix B 1 When selecting sites for MAGIC analyses that will later be used in an extrapolation 2 analysis, Sullivan and colleagues (2004) outlined two key considerations: 3 1. Do not select too many watersheds for modeling that occurred in the same general 4 area in order to avoid skewing the results too heavily to one portion of the region for 5 the following extrapolation step 6 2. Screen sites to remove those in which the water chemistry data were not internally 7 consistent or for which available data suggested the possibility of significant 8 influence from road salt, geological sulfur, land use, or insect defoliation. 9 Step 2. Data gathering and processing for population of the MAGIC model for 10 each study site with additional regional data gathering of available 11 stream/lake chemistry and landscape parameters. 12 The data requirements of the MAGIC model are summarized in Table 3-3. The table 13 includes both data inputs derived from monitoring data and constant parameters that the user 14 must set based on available data and methods suggested by previous MAGIC applications. 15 Additional information on data inputs can be found in: Cosby and colleagues, 1985a; Cosby et 16 al., 1985b; Sullivan and Cosby, 2004; Sullivan et al., 2007c. 17 Due to the wide range of water quality monitoring assessments conducted within the 18 United States, a large amount of water quality data is typically available to work from. Similarly, 19 in recent times advances and expansions of atmospheric modeling have been conducted 20 providing a large amount of deposition estimates from which to pull model input data. The area 21 of data most lacking, especially in the western United States, is the composition of soils. Sources 22 of soil data are discussed in Section 4.3.3. Given that there may be areas in which soils data are 23 not available, work by Sullivan and others used a tiered assessment of MAGIC applications to 24 overcome this obstacle. The tiers consisted of: (1) chemistry data were available from within the 25 watershed to be modeled with multiple soil sampling sites in an individual watershed aggregated 26 on an area-weighted basis; (2) soils data within the catchment were missing but were available 27 from a nearby watershed underlain by similar geology; and (3) soils data were neither available 28 from within the watershed nor from nearby watersheds on similar geology. In order to populate 29 soil characteristics for tier 2 and 3 watersheds, a surrogate approach was used meaning that these 30 watersheds were paired with a watershed for which all input data were available. In order to be 31 paired, watersheds had to have similar streamwater characteristics (ANC, sulfate, and base cation March 2010 30 Draft - Do Not Quote or Cite ------- Appendix B 1 concentrations), physical characterization (location, elevation), and bedrock geology data 2 (Sullivan et al., 2004). Table 3-3. Input Data Requirements of MAGIC Model Data Class Catchment Stream Chemistry Aqueous Phase - Equilibrium Constants Solid Phase - Weathering and Exchange Constants Soil Composition Atmospheric Deposition Data Element Area Relative area of lake/stream PH ANC Ca'+ Mg^+ K+ Na+ SO/" N03- cr Aluminum solubility constant Slope of pH-pAl relationship Organic acid Organic aluminum Inorganic aluminum speciation Inorganic carbon speciation and dissociation of water Cation exchange selectivity coefficients Weathering rates (Ca"+, Mg"+, K+, NH4, S042", Cr, N03", F) Thickness Total cation exchange capacity Exchangeable bases (Ca^+, Mg^+, K+, and Na+) Bulk Density Porosity PH Sulfate adsorption half saturation Aluminum solubility constant Slope of pH-pAl relationship Annual precipitation Ca^+ Mg'+ K+ Na+ SO/" NH4 NO3" cr Measure fraction unitless eq/L eq/L eq/L eq/L eq/L eq/L eq/L eq/L logio unitless logio logio eq/m^/yr Depth (m) eq/kg mg/kg kg/mj fraction unitless eq/nr3 logio unitless Volume (m/yr) Total annual deposition (eq/ha/yr) March 2010 31 Draft - Do Not Quote or Cite ------- Appendix B 1 Step 3. MAGIC modeling application on selected stream/lake study sites where 2 BCW is arrived at through calibration against water chemistry data. 3 As was completed with the REA (U.S. EPA, 2009), batch processing of MAGIC models 4 at a range of sites can be completed. Calibration of those sites with available data (streamwater 5 chemistry, soil chemical and physical characteristics, and atmospheric deposition) is completed 6 by setting values of the "fixed parameters" within the system and comparing the output of the 7 model run to the observed values of such characteristics as stream ANC. There are eight 8 parameters optimized through this method including the BCW rate. The eight observations used to 9 drive the calibration procedure include the current soil exchangeable pool size and current output 10 flux of each of the four base cations. The model is iteratively run adjusting the "fixed 11 parameters" from a specified range of values (representing uncertainty in knowledge of these 12 parameters), so that the outputs match the observed parameters within an acceptable margin of 13 error. The set of "fixed parameters" that are obtained that allow the model to meet this 14 acceptable of margin of error become the range of calibrated parameters from which the median 15 is chosen to represent the parameter value for the watershed. "The use of median values assures 16 that the simulated responses approximate the most likely behavior of each watershed, given the 17 assumptions inherent in the model and the data used to constrain and calibrate the model" 18 (Sullivan et al., 2004). This "fuzzy optimization" procedure has been developed for use with 19 MAGIC modeling to help quantify the uncertainties within the modeled parameters (Sullivan et 20 al., 2004). Using these calibration procedures of each site MAGIC run will provide not only an 21 estimate of BCW but an expected range of values in which BCW falls, thereby providing bounds 22 and certainty limits for the following extrapolation step. 23 Step 4. Extrapolation of BCW for modeled streams/lakes to other waterbodies 24 within the region through correlation analysis using stream chemistry data, 25 where available, and landscape parameters in its absence. 26 Regionalization of MAGIC modeling results can be completed through either "binning" 27 sites based on characteristics like physiographic region and ANC concentration (Sullivan et al., 28 2004) or creating regional regressions to relate site characteristics (chemistry or landscape) to a 29 parameter of interest (e.g., ANC; Sullivan et al., 2007a). In order to provide some measure of 30 "goodness of fit" to the extrapolations, we have chosen to proceed with creating regression 31 relationships between the BCW determined through calibration of the MAGIC model and either March 2010 32 Draft - Do Not Quote or Cite ------- Appendix B 1 water chemistry or landscape parameters. In previous studies, the landscape variables considered 2 for regression relationships with ANC have included elevation, watershed area, ecoregion, 3 lithology, forest type and geological sensitivity class (Sullivan et al., 2007b). We expect to 4 follow similar methods to create the relations with BCW (i.e., the response variable). Within each 5 region of extrapolation landscape variables appropriate to the region will be selected. For 6 example, the types of forest selected for inclusion may vary between an extrapolation in the 7 Southern Appalachians as opposed to the Rocky Mountains. In all instances, there must be 8 adequate representation of the variable within all modeled and non-modeled watersheds or it will 9 be eliminated from the pool of candidate variables available for regression analysis. 10 Sullivan and others (2007b) relied on the corrected Aikake's Information Criteria (AIC) 11 to evaluate all possible correlation relations. The corrected version of the evaluation criteria was 12 used because of the relatively small sample sizes available from which to build the regressions. 13 Additional evaluation criteria can easily be applied for choosing the best-fitting and most 14 meaningful regressions for extrapolation from a set of individual modeled sites to a larger set of 15 regional sites. Potential criteria for evaluating individual variables within correlation models 16 include partial F tests, t-values, and variance inflation factors. To evaluate the model as a whole 17 statistics such as PRESS, coefficient of determination, adjusted coefficient of determination, 18 Mallow's Cp, and root mean square error of the model can all be utilized (Helsel and Hirsch, 19 1992). If a commercial statistical package, such as SAS, is chosen to complete this portion of the 20 analysis the predefined routines and groupings of evaluation statistics can be employed with 21 relative ease. 22 3.3.1 Potential limitations of proposed methodology 23 The limitations with the proposed methodology can be divided into five distinct 24 categories: 25 1. MAGIC is an intermediate level model that does not take into account biotic 26 processes which may affect the calculation of BCW rates within a watershed. 27 2. While MAGIC is the most widely applied acidification model within the U.S. it still 28 faces the challenge of having limited applications in the Midwest and western states. 29 3. As an extension of the bias in eastern applications, processing and organization of the 30 data required for input into the MAGIC model in the East far exceeds that of the 31 West. Additionally, there is an indication that soils data are more incomplete or hard March 2010 33 Draft - Do Not Quote or Cite ------- Appendix B 1 to obtain in the West. Note that the terrestrial acidification national assessment faces 2 even greater demands in terms of soil composition data needs. As such, there can be a 3 combined effort in obtaining new data that will benefit both assessments. 4 4. The population and calibration of specific site applications of MAGIC across the 5 country constitutes a major modeling effort. However, it may be possible to leverage 6 previous applications. 7 5. The proposed approach calls for the creation of several different regional 8 extrapolations of BCW rates based on sets of individual MAGIC applications. The 9 success of these extrapolations remains to be seen and will depend upon the 10 limitations mentioned above in even applying the MAGIC model at a multitude of 11 locations and upon the availability of a statistically significant number of model 12 outcomes on which to base the regressions for each regional analysis. 13 While these limitations may seem extensive, there are many possibilities for overcoming 14 the limitations. For example, criteria on model application years can be relaxed to include more 15 of the previously completed MAGIC applications in lieu of updating and rerunning models at the 16 same sites. And, joint data collection between the aquatic and terrestrial acidification 17 assessments can allow the most efficient use of resources and demands on other agencies. 18 3.3.2 Uncertainty analyses 19 As typical with any process based model, the major uncertainties in MAGIC include 20 input data variability, model calibration uncertainty, and the ability of the mathematical model 21 processes to represent reality. Within this national analysis, there will also be uncertainty 22 associated with regional extrapolation of modeling results from individual watersheds to the 23 region. However, Sullivan and colleagues (2004) state that these "errors and uncertainties are not 24 additive, but rather would be expected to some extent to cancel each other out." 25 Several research projects have undertaken attempts to quantify the relative magnitude of 26 the effects of sources of uncertainty for regional, long-term MAGIC simulations using Monte 27 Carlo methods (Cosby et al., 1989b, 1989c, 1990; Hornberger et al., 1989, 1990). While the 28 results of these studies indicated that the different sources of uncertainty can have varying levels 29 of impacts on the outputs of the MAGIC model, the development of the "fuzzy optimization" 30 technique for calibration was designed to reduce these impacts of uncertainty. With "fuzzy 31 optimization" there is an explicit accounting within different uncertainty categories and a March 2010 34 Draft - Do Not Quote or Cite ------- Appendix B 1 resulting time-variable measure of overall simulation uncertainty for each state variable. One 2 way the optimization procedure reduces uncertainty is in its selection of parameter and variable 3 values for the "fixed parameters" from distributions of possible values rather than having a user 4 select a single value during a single calibration (Sullivan et al., 2004). 5 Outside of the operation and calibration of the MAGIC model, the parameterization of 6 the input data provides another source of uncertainty. As identified in previous sections, the soils 7 composition data are expected to be the greatest source of uncertainty. If a tiered approach to 8 populating soils data for watersheds lacking in data is used, uncertainty with the method can be 9 examined by calibrating selected tier 1 watersheds twice, once using the appropriate site-specific 10 soils data, and a second time using borrowed soils data from an alternate site, using either tier 2 11 or tier 3 protocols. A comparison between the results from each of the scenarios can then be 12 made to determine the magnitude of difference in output parameters. If this analysis can be done 13 at multiple sites, than a sensitivity analysis can be performed over the results to determine if 14 there is a consistent bias in results from modeling analyses utilizing tier 2 or 3 procedures 15 (Sullivan et al., 2004). 16 4. TERRESTRIAL BASE CATION WEATHERING METHODOLOGY 17 4.1 Introduction 18 Geology is one of the most important factors in determining the potential sensitivity of an 19 area to terrestrial acidification (U.S. EPA, 2008, Section 3.2.4). In particular, the characteristics 20 of the soils and the upper portion of the bedrock can impact the acid-neutralizing ability of the 21 soils in a particular area. Acid-sensitive soils are those which contain low levels of exchangeable 22 base cations and low base saturation (U.S. EPA, 2008, Section 3.2.4). Bedrock composition and 23 soil pH are two characteristics that are directly related to the ability of a system to neutralize 24 acid. Soils overlying bedrock, such as calcium carbonate (e.g., limestone), which is reactive with 25 acid, are more likely to successfully neutralize acidifying deposition than soils overlying 26 nonreactive bedrock. In addition, soils with higher pH (i.e., more alkaline) have a greater 27 capacity to neutralize acidifying deposition. 28 This section reviews the effect of acidification known as base cation weathering, 29 describes its significance in estimating critical acid loads, and identifies methodologies for 30 estimating base cation weathering. Further, this report recommends a methodology for potential March 2010 35 Draft - Do Not Quote or Cite ------- Appendix B 1 use in the review of the NOX and SOX secondary National Ambient Air Quality Standards and 2 describes the steps and information resources needed to apply that methodology across the 3 United States. 4 4.2 Terrestrial Base Cation Weathering 5 In the calculation of terrestrial critical acid loads using the simple mass balance (8MB) 6 methodology, base cation weathering (BCW)2 is defined as "the release of base cations from 7 minerals in the soil matrix due to chemical dissolution" (UNECE, 2004), and this weathering 8 occurs in the rooting zone of the soil profile and consists of the release of calcium (Ca2+), 9 magnesium (Mg2+), potassium (K+) and sodium (Na+). It does not include the removal of base 10 cations from soil ion exchange complexes (cation exchange sites) or the degradation of soil 11 organic matter. Base cations from these sources have already been released through the 12 weathering process. Base cation weathering is often a dominant source of base cations in soils, 13 replacing Ca2+, Mg2+, K+ and Na+ that are lost through leaching and uptake by plant (Langan et 14 al., 1995; Langan et al., 1996; Ouimet, 2008). Therefore, BCW plays an important role in 15 determining the sensitivity of a site to acidifying nitrogen and sulfur deposition (Hodson and 16 Langan 1999a). The BCW term is also one of the most influential parameters in the 8MB 17 calculations of terrestrial critical acid loads. Li and McNulty (2007) determined that 49% of the 18 variability in critical load estimates was due to this term. Sverdrup and colleagues (1995) 19 (reference in Langan et al., 1996) determined that BCW can account for 90% of the variation in 20 critical loads. 21 For the Terrestrial Acidification case study in the Risk and Exposure Assessment (U.S. 22 EPA, 2009), BCW rates were calculated using the clay-substrate method (McNulty et al., 2007). 23 This method was selected for the Risk and Exposure Assessment because it is one of the most 24 commonly used methods to estimate BCW for critical load analyses in North America (Ouimet et 25 al., 2006; Watmough et al., 2006; McNulty et al., 2007; Pardo and Duarte, 2007), and has been 26 used to map critical loads across the United States (McNulty et al., 2007). However, the 27 applicability of the clay-substrate method is most likely limited because it is an empirical model 28 that appears to be based on a modification of the soil type - texture approximation method that 29 was developed on a restricted number sites in northern Europe that were glaciated during the last 2 Within the 8MB equation, Bcw refers to the weathering of Ca2+, Mg2+ and K+. March 2010 36 Draft - Do Not Quote or Cite ------- Appendix B 1 glacial advance (CLAD, 2009; H. Sverdrup personal communication, 2009a, UNECE, 2004). It 2 relies on a classification of the acidity of soil parent material and soil clay content and consists of 3 three equations (equations 4-1 - 4-34). 4 Acid Substrate: BCe = (56.7 x %clay)- (p.32 x (%clay)2) (4-1) 5 Intermediate Substrate: BCe = 500 + (53.6 x %clay)- (o. 18 x (%clay)2) (4-2) 6 Basic Substrate: BCe = 500 + (59.2 x %clay) (4-3) 7 where 8 BCe = empirical soil base cation (Ca2+ + K+ + Mg2+ + Na+) weathering rate 9 (eq/ha/yr) 10 % clay = the percentage of clay (determined by particle size) within the rooting zone 11 of soil profile. 12 Critical load experts from both the United States and Canada have commented that the clay- 13 substrate model, in general, appears to perform well in young soils that have formed since the 14 last glaciations (approximately 20,000 years before present). However, the model may not be 15 suitable or provide accurate estimates on older, more weathered soils that were not impacted by 16 the last glaciation (P. Arppersonal communication, 2009). These soils have undergone 17 weathering for a longer period of time and the relationships between clay particle size and base 18 cation release may not be as strong as in younger soils (H. Sverdrup personal communication, 19 2009a). To our knowledge, however, there have been no published studies that have tested this 20 hypothesis and compared BCW estimates generated with the clay-substrate model and other 21 methods on sites underlain by old, more weathered and recently glaciated soils. At least one 22 study has compared the clay-substrate BCW method with estimates from other models on 23 glaciated soils in Canada and found that the rate estimates were similar within the area of 24 assessment (Whitfield et al., 2006). 25 Results from the Risk and Exposure Assessment (U.S. EPA, 2009) appear to support the 26 distinction between the suitability of applying the clay-substrate model to glaciated versus older, 27 non-glaciated soil environments. As outlined in Appendix 5 of the Risk and Exposure March 2010 37 Draft - Do Not Quote or Cite ------- Appendix B 1 Assessment, the regression analysis assessing the relationship between the growth of sugar 2 maple (Acer saccharum) and critical acid load exceedance was not significant (p=0.38) when all 3 plots were included in the analysis. However, when the analyses were restricted to sites located 4 on younger, glaciated soils, which resulted in the removal of 25% of the data from the analyses, 5 the linear regression relationship was significant at the p=0.10 level. Improvements in the 6 significance of the relationship may, in part, have been due to the greater accuracy of the BCW 7 estimates in the critical load calculations for the plots north of the glaciations line. 8 The majority of the conterminous United States was not directly impacted by the most 9 recent glacial advance (Figure 4-1) and some of the soils in these areas have not been influenced 10 by glaciations in at least 700,000 years (Sverdrup et al., 1992). Only ten states had their full land 11 area impacted by glaciers during the glacial advance 20,000 years ago. Therefore, if the concerns 12 and supportive results regarding the suitability of the clay-substrate model for the estimation of 13 BCW on older, non-recently-glaciated soils are correct, the model may not be an appropriate 14 method to estimate BCW for a large portion of the United States. Given that the BCW parameter is 15 one of the most influential variables within the 8MB calculations to estimate critical acid loads, 16 it is particularly important to use a method that provides accurate and defendable estimates of 17 BCW. Therefore, any and all future work focused on estimating and mapping terrestrial critical 18 acid loads in the United States, should acknowledge the potential limitations of the clay-substrate 19 model and consider the adoption of a BCW modeling approach that is transferable and can be 20 applied to multiple locations and different soil conditions and soil ages. March 2010 38 Draft - Do Not Quote or Cite ------- Appendix B I Approximate area affected by most recent glaciation 1 2 Figure 4-1. Areas of continental U.S. that were covered during the last glacial event 3 (Reed and Bush, 2005). 4 4.3 Methodologies for Determining Base Cation Weathering Values in the United States 5 4.3.1 Difficulties in estimating base cation weathering 6 Base cation weathering is one of the most difficult parameters to estimate (Sverdrup et 7 al., 1990; Ouimet and Duchesne, 2005; Langan et al., 1996), as it is a function of a time, soil 8 mineralogy, and a variety of other environmental biotic and abiotic factors. Weathering occurs 9 over centuries and millenia and results in the chemical and physical alterations of parent material 10 and minerals. Minerals that are present in the soil may no longer resemble the original bedrock 11 parent material (C. Smith personal communication, 2009). In addition, the soil may be derived 12 from parent material that was transported to its current location and does not resemble the 13 underlying bedrock. Abiotic factors including temperature and moisture and location on the March 2010 39 Draft - Do Not Quote or Cite ------- Appendix B 1 landscape and biotic factors including vegetation and soil microbes can also impact base cation 2 weathering through removal of base cations and chemical weathering of minerals (Brady and 3 Weil, 2002). Combined, these factors pose many challenges to determining BCW in the soil 4 profile for terrestrial critical acid load estimations. As a result, a variety of BCW methods and 5 approaches have been developed (Sverdrup et al., 1990). 6 4.3.2 Approaches to estimating BCw: 7 Methods and models that have been developed to estimate BCW for critical acid load 8 determinations differ significantly in the approaches used to generate weathering estimates 9 (Langan et al., 1995; Sverdrup et al., 1990; UNECE, 2004). For the purposes of this work 10 assignment, BCW methods and models are grouped into three main approaches: 11 1. budget studies of catchments or watersheds; 12 2. historical weathering rate determinations; and 13 3. empirical and mathematical models. 14 Each of these approaches vary in complexity, data intensity and scalability, thereby offering 15 different strengths and weaknesses to estimating BCW. In addition, these approaches differ in 16 their abilities to map BCW over regional and larger land areas. Table 4-1 provides a summary of 17 the approaches to BCW, critical load models that use the approaches, the strengths and 18 weaknesses of the different models, and the suitability of the approaches and models to map BCW 19 and therefore critical loads over large areas. For a model to be suitable for large-scale mapping, it 20 must be quick to apply, supported by existing databases, not require extensive and costly 21 analyses, and be transferable to sites with varying conditions and geological histories (Sverdrup 22 etal., 1990). 23 Budget Studies - The budget study approach, also referred to as input-output balances 24 (Kolka et al., 1996; Langan et al., 1996; Starr et al., 1998), estimates BCW as a component of the 25 mass balance input and output of cations within a catchment or watershed (Langan et al., 1996; 26 Sverdrup et al., 1990; Sverdrup et al., 1998). In most catchments, the main source of input is 27 atmospheric deposition and output is streamflow, and base cation retention is accounted for 28 through uptake by biomass and immobilization in the soil. Base cation weathering is therefore 29 determined through mass balance differences between these different input, output and storage 30 pools. The main strengths of this method are that it only requires a moderate amount of input 31 data and relies on data collected from the catchment or watershed. In addition, it offers the March 2010 40 Draft - Do Not Quote or Cite ------- Appendix B 1 potential for mapping multiple catchments, if the necessary input data is available. However, 2 several of the drawbacks to this method include an assumption that the catchment is in a steady - 3 state condition and the cation exchange capacity does not change over time (Langan et al., 1996; 4 Miller, 2001; Sverdrup et al., 1998). In addition, it is often difficult to determine BCW within the 5 rooting zone of individual soil profiles because the BCW estimates from budget studies represent 6 integrated values for the whole watershed, the full soil profile, bedrock weathering and all soil 7 processes (Sverdrup and Warfvinge 1988; Miller, 2001). It is also difficult to separate 8 contributions of base cations from exchange sites versus mineral weathering and chemical 9 dissolution (Sverdrup et al., 1990; Miller, 2001). Therefore, it is challenging, and potentially 10 erroneous, to use budget studies of catchments to estimate BCW for terrestrial critical acid loads. 11 It may be possible to modify the budget study approach and evaluate base cation input and 12 output in the rooting zone of individual soil profiles (Kolka et al., 1996), and to separate base 13 cations from exchanges soil pool versus weathering sources using techniques such as the analysis 14 of strontium (Sr) isotope ratios (Miller et al., 1993). However, the soil profile approach would 15 require lysimeter measurements of soil solution chemistry at each site and the soil solution would 16 also need to be analyzed for Sr isotope ratios (87Sr/86Sr). Both of these analyses would be very 17 time intensive and would not be practical over large areas. 18 Historical Rate Determinations - The historical weathering rate approach, also 19 sometimes referred to as element depletion (Langan et al., 1996; Miller, 2001) or pedological 20 mass balance (PMB) (Ouimet and Duchesne, 2005; Ouimet, 2008), estimates BCW by 21 determining the relative depletion of base cations to the depletion of a stable element as a 22 function of the age of the soil profile (Langan et al., 1996). Zirconium (Zr), titanium (Ti), rutile 23 and sometimes quartz are typically selected as the stable soil elements for this method (Langan et 24 al., 1996; Sverdrup et al., 1998) because they are very resistant to weathering (Starr et al., 1998). 25 This technique is commonly applied to soils that were formed since the last glaciation, and 26 characterizes the ratio of base cations to the stable element in the upper weathered soil horizons 27 and the unweathered C horizon. It assumes that post glaciation, the mineral matrix of the soil 28 consisted of freshly ground material that was not previously exposed to weathering (Sverdrup et 29 al., 1998), the lowermost soil is representative of the parent material and the stable element is 30 found in a constant proportion throughout the soil profile (Langan et al., 1996; Starr et al., 1998). 31 Over time, base cations are weathered and lost from the profile through uptake and leaching, but March 2010 41 Draft - Do Not Quote or Cite ------- Appendix B 1 the concentration of the stable element remains constant due to resistance to weathering (Starr et 2 al., 1998). Main strengths of this approach are that it is a good technique to estimate weathering 3 in young soils, does not require a large amount of data and BCW is relatively easy to calculate. 4 However, this approach also presents some major weaknesses. It estimates the historic BCW rate 5 which may be differ from the current weathering rate. Historic weathering rates may 6 underestimate current BCW because the historical weathering occurred under more neutral 7 conditions with less acidifying deposition (Sverdrup et al., 1990). Conversely, historic 8 weathering rates may be higher than current BCW if the original post glaciations soil contained a 9 significant proportion of easily weathered material that have since been depleted (Miller, 2001). 10 Studies have indicated that the initial phase of weathering lasts a few hundred to several 11 thousand years and can deplete a maximum of 25% of the mass during this period (Sverdrup et 12 al., 1998). In addition, the historic rate approach is not suitable for older, more weathered soils. 13 In such soils, it is often difficult to determine the amount of time since the last glacial or mass 14 disturbance event that caused the formation of newly ground material (H. Sverdrup personal 15 communication, 2009b). Therefore, this method cannot be applied to all locations and it is often 16 difficult or impossible to extrapolate results to larger geographical areas. Since a large proportion 17 of the soils in the United States were not influenced by the most recent glaciation, the historic 18 rate approach to estimate BCW for terrestrial critical acid load estimates could only be applied to 19 a fraction of the land area. 20 Empirical and Mathematical Models - Empirical and mathematical models estimate BCW 21 based on laboratory- and field-based relationships between soil, abiotic and biotic factors. Over 22 the past several decades, a large number of BCW models have been developed for terrestrial 23 critical acid load determinations. Initial models were developed from a limited number of sites 24 and data. More recent models incorporate a larger number of factors and are more complex and 25 data intensive. One of the first BCW models was the Skokloster Assignment which is a semi- 26 empirical method that was devised during the Critical Load Workshop in Skokloster, Sweden in 27 1988 (UNECE, 2004). It divides minerals into 5-6 mineral classes based on the dominant 28 weatherable soil minerals and assigns a range of critical acid loads to each. This method was 29 later expanded to include a larger range of minerals and to estimate BCW based on the relative 30 abundance of fast versus slow weathering minerals. The Skokloster Assignment was originally 31 based on soils with density, moisture content, clay content and pH conditions similar to the soils March 2010 42 Draft - Do Not Quote or Cite ------- Appendix B 1 in the three Gardsjon catchments in Sweden, and was validated against a preliminary version of 2 the PROFILE model, described further below (Hodson and Langan 1999b; H. Sverdruppersonal 3 communication, 2009b). 4 A second model, the Soil Type - Texture Approximation assigns weathering rate classes 5 to soils based on soil texture and parent material acidity classes. It was developed for European 6 forest soils (UNECE, 2004). As described earlier, it is believed that the clay-substrate model that 7 is used extensively throughout North America was derived from the Soil Type - Texture 8 Approximation. 9 A third model, the Total Base Cation Content Correlation was developed using Zr(SiO4) 10 and historical rate approach applied to eleven sites in Sweden (UNECE, 2004). Correlation 11 between historical BCW rates and the total content of base cations in the undisturbed bottom soil, 12 corrected for temperature, were used to develop equations to estimate the weathering of Ca2+, K+ 13 and Mg+2 (Olsson et al., 1993). For a more complete review and description of these three 14 empirical models see Hodson and Langan (1999) and UNECE (2004). In general, the main 15 benefits of these models are the minimal data requirements, transferability, and the potential to 16 be applied to multiple sites. Therefore, such models offer good options for mapping of BCW for 17 critical acid load determinations. However, these models also have several key weaknesses 18 which limit their utility for estimating BCW in many locations, including a large proportion of the 19 United States. All of the models were determined using data from a limited number of sites 20 within Sweden and other regions of Europe and are based on average or generalized 21 relationships. Therefore, similar to the clay-substrate model, these models may do a reasonable 22 job of estimating BCW on sites that were recently glaciated and/or have similar conditions to the 23 Swedish sites. However, they should not be applied everywhere, as they may poorly estimate 24 BCW on sites with older, more weathered soils. As stated for Total Base Cation Content 25 Correlation, the method is only applicable to granitic soils (Hodson and Langan, 1999) and 26 should be used with caution because the relationships are based on Nordic geological history 27 (UNECE, 2004). 28 A fourth model that supported the creation of some of the aforementioned empirical BCW 29 models, and is currently in its 5th version is PROFILE (Warfvinge and Sverdrup 1992 and 1995; 30 Sverdrup, 1990). PROFILE (version 5.0) is a mechanistic, mathematical, steady-state, kinetics 31 model that calculates the weathering of Ca2+, Mg2+, and K+ in each horizon of a soil profile March 2010 43 Draft - Do Not Quote or Cite ------- Appendix B 1 (Akselsson et al., 2005). It is unique and differs from other empirical models in that it calculates 2 BCW rates for soil from independently measured geochemistry and soil conditions (Jonsson et al., 3 1995; Ouimet and Duschesne 2005). It combines laboratory-based evaluations of mineral- 4 specific chemical dissolution with field-based conditions and other soil measurements to 5 estimate individual weathering rates of Ca2+, Mg2+, and K+ (Langan et al., 1996). The model 6 includes 14 of the most common primary and secondary soil minerals (Table 4-2)3, and their 7 release of base cations in five separate reactions (Sverdrup et al., 1990; Hodson et al., 1997; 8 Sverdrup etal., 1998): 9 i) Reaction with hydrogen ion (H+) and dissolved aluminum (Al) 10 ii) Reaction with water and dissolved Al 11 iii) Reaction with hydroxyl ion (OH") and dissolved Al 12 iv) Reaction with carbon dioxide (CO?) 13 v) Reaction with strongly complexing polydentate organic acids 14 The reaction rates are calculated using constants contained within the model and data input by 15 the user, and the total base cation release rate by chemical weathering is calculated as the sum of 16 all parallel simultaneous process rates minus the rate of precipitation of secondary solid phases 17 (Sverdrup et al., 1998). The rate equation (Equation 4-4) for the weathering of all minerals 18 within the rooting zone of the soil profile is defined as: 19 RW = 2(horizonS)2(minerals)/ T> ' AexP ' X' '® ' Z 20 where 21 rt = dissolution rate of mineral /' (kmolc/m2/s) - sum of the 5 separate reactions 22 Aexp = exposed surface of mineral matrix (m2/m3) 23 9 = soil moisture saturation (m3/m3) 3 Thirteen additional minerals can be added to PROFILE, as necessary (H. Sverdrup personal communication). March 2010 44 Draft - Do Not Quote or Cite ------- Appendix B 1 Xj = fraction of mineral /' in the mineral matrix of the soil horizon 2 z = soil layer thickness (m) 3 The weathering rate is either increased or reduced by different soil and biotic and abiotic 4 conditions, many of which are entered as input data by the user. Input data includes site climatic 5 and deposition attributes, soil physical and chemical characteristics and biological components 6 that influence the soil chemistry and BCW (input data required by PROFILE discussed further in 7 Section 4.3.3). As summarized by Jonsson and colleagues (1995), "The weathering rate is 8 increased by a high H+ concentration, a high soil moisture (water) content, and a high CC>2 9 pressure. Weathering reactions are product inhibited, i.e., decrease by high concentrations of 10 reaction products in the soil solution such as inorganic aluminum and base cations. The surface 11 activity is calculated as dependent on the mineral surface area, temperature and soil moisture 12 saturation. The soil temperature impact on the weathering rate is expressed as an Arrhenius 13 equation, as dependent on the activation energy. The soil moisture saturation is important for the 14 reaction rate as the reactions will only take place on wetted surfaces. The degree of surface 15 wetting, and thus surface activity, is considered to be a function of the soil moisture saturation. 16 This is calculated from soil bulk density, the solid particle density and the volumetric water 17 content." For a more detailed description of the theory and calculations behind PROFILE, see 18 Sverdrup (1990), Sverdrup and Warfvinge (1992, 1993a, 1995). 19 A main weakness of the PROFILE model is that it is data intensive and complex, and can 20 be difficult to parameterize. However, PROFILE does offer some significant benefits that set it 21 apart from the other models. As described, it determines current BCW rates from laboratory- 22 derived weathering rates of individual minerals and therefore is not bound to data from a specific 23 location or region. Therefore, it can be used to determine and map BCW over large areas (Miller 24 et al., 1993). Although it was developed in Sweden, it has been successfully applied to the 25 mapping of BCW and critical loads in the Northeastern United States, Maryland, Minnesota, 26 Pennsylvania, Thailand, China, Argentina, and Greece (Duan et al., 2002; Miller, 2001; Sverdrup 27 et al., 1992; H. Sverdrup personal communication, 2009b). March 2010 45 Draft - Do Not Quote or Cite ------- Appendix B Table 4-1. Review of modeling approaches (and models) to estimate base cation weathering for terrestrial critical acid load determinations. Model Approach Budget Studies Historical Rate Determinations Empirical Models Description of Method mass balance of inputs and outputs of base cations within catchment, watershed or soil profile loss of base cations in soil profile relative to stable element (Zr, Ti, quartz or rutile) modeled relationships between soil attributes and abiotic and biotic site conditions Data Requirements medium low low- high Model Complexity medium low low- high Suitability for Estimating BCW for Terrestrial Critical Acid Load Determinations in The United States low; BCw estimate is often an integrated value for whole catchment or watershed medium; restricted to sites with young soils of known age (e.g., soils that have formed since the most recent glacial event, -20,000 years ago) low- high Suitability for Mapping BCw Over Large Regions in The United States low- medium (based on data availability); may require Sr isotope ratio of stream chemistry to separate exchangeable versus weathered base cation sources low; restricted to sites with young soils and sites where historical rate determinations have been conducted low- high References Sverdrup et al., 1998; Sverdrup et al., 1990 Sverdrup et al., 1998; Sverdrup et al., 1990 March 2010 46 Draft - Do Not Quote or Cite ------- Appendix B Model Approach Skokloster Assignment Soil Type - Texture Approximation Description of Method BCW rate categorically determined by relative abundance of minerals grouped into 5-6 weathering rate classes; originally developed for soils similar to those found in the 3 Gardsjon catchments in Sweden BCW categorically determined as a function of parent material acidity and soil texture, modified by temperature; developed from data from European forest soils Data Requirements low low Model Complexity low low Suitability for Estimating BCW for Terrestrial Critical Acid Load Determinations in The United States low; most accurately applied to sites similar to those where the model was derived low- medium; most accurately applied to sites similar to those where the model was derived Suitability for Mapping BCw Over Large Regions in The United States low; most accurately applied to sites similar to those where the model was derived low- medium; most accurately applied to sites similar to those where the model was derived References UNECE, 2004; Hodson and Langan, 1999 UNECE, 2004; Hodson and Langan, 1999 March 2010 47 Draft - Do Not Quote or Cite ------- Appendix B Model Approach Total Base Cation Content Correlation Clay-Substrate Model Description of Method BCW determined by correlations between historical rate determinations (Zr) and total content of base cations in the undisturbed bottom soil, corrected for temperature; based on data from eleven sites in Sweden BCW determined by one of three equations based on parent material acidity and % clay content; most likely a modification of the Soil Type - Texture Approximation Data Requirements low low Model Complexity low low Suitability for Estimating BCW for Terrestrial Critical Acid Load Determinations in The United States low; restricted to sites with granitic soils and Nordic geological histories low- medium; most accurately applied to sites similar to those where the model was derived (most likely young soils formed since the last glaciation) Suitability for Mapping BCw Over Large Regions in The United States low; restricted to sites with granitic soils and Nordic geological histories low- medium; most accurately applied to sites similar to those where the model was derived (most likely young soils formed since the last glaciation) References UNECE, 2004; Hodson and Langan, 1999 original source unknown; Ouimet et al., 2006; Watmough et al., 2006; McNulty et al., 2007; Pardo and Duarte, 2007 March 2010 48 Draft - Do Not Quote or Cite ------- Appendix B Model Approach PROFILE Description of Method BCW determined as a function of weathering of individual soil minerals and field- based soil and biotic conditions Data Requirements high Model Complexity high Suitability for Estimating BCW for Terrestrial Critical Acid Load Determinations in The United States medium - high; may have restrictions in desert regions and areas that are lacking necessary data Suitability for Mapping BCw Over Large Regions in The United States medium - high; may have restrictions in desert regions and areas that are lacking necessary data References Warfvinge and Sverdrup, 1992 and 1995; Sverdrup, 1990 March 2010 49 Draft - Do Not Quote or Cite ------- Appendix B Table 4-2. The fourteen dominant minerals modeled within PROFILE. Dominant Minerals K-Feldspar Plagioclase Albite Hornblende Pyroxene Epidote Garnet Biotite Muscovite Fe-Chlorite Mg-Vermiculite Apatite Kaolinite Calcite 1 2 4.3.3 Proposed methodology for estimating and mapping base cation weathering for 3 terrestrial critical acid load calculations 4 As has been outlined in the above review, there are multiple approaches to estimate BCW 5 for terrestrial critical acid loads. However, not all are suitable for both calculating and mapping 6 terrestrial critical acid loads throughout the United States. Such an approach has to be quick and 7 easy to apply, be supported by available data and be easily and accurately transferable to sites 8 within the United States that differ in soil, biotic and abiotic properties and conditions. In 9 addition, as stated by Miller (2001), "the most promising approach for a logically consistent 10 estimation of the present-day weathering rate over broad regions is the application of model(s) 11 that predict the weathering rate from first principles, given detailed measurements of the soil 12 environment and laboratory-derived rate constants for specific mineral weathering reactions." 13 Therefore, an approach that is based on soil mineralogy and weathering of individual minerals is 14 preferable. Of all the models that are currently available for determining BCW for terrestrial 15 critical acid load determinations, PROFILE meets these requirements and appears to be the most 16 suitable. Methodologically, it has few location restrictions and models BCW based on site- 17 specific mineralogy and soil and site conditions. In addition, it has already been successfully 18 applied in both glaciated and non-glaciated regions of the United States to estimate and map BCW March 2010 50 Draft - Do Not Quote or Cite ------- Appendix B 1 and critical acid loads (Miller et al., 1993; Sverdrup et al., 1992; H. Sverdruppersonal 2 communication, 2009b/ Although, as with all models, PROFILE does have some weaknesses 3 and limitations (discussed further in Section 4.3.5) that need to be acknowledged and addressed 4 prior to application, critical load experts, in general, agree that PROFILE is the best model to 5 date for estimating and mapping BCW rates for terrestrial critical acid load determinations in the 6 United States (J. Aherne personal communication, 2009, J. Cosby personal communication, 7 2009, J. Lynch personal communication, 2009, R. Ouimet personal communication, 2009, H. 8 Sverdrup personal communication, 2009b). 9 There are two forms of PROFILE (version 5.0) that can be used for estimating BCW: the 10 single site application which estimates BCW for a single location or soil profile, and the regional 11 application which PROFILE can be run for a region or conterminous areas (C. Akselsson 12 personal communication, 2009). For mapping BCW in the conterminous United States, the 13 regional application of the model would be applied, and the estimation and mapping of BCW 14 would involve two main steps: 15 Step 1. Identification of input data required by PROFILE and development of 16 spatial data layers, national databases and default values for each data 17 element within the model 18 Step 2. Determination of polygon layer to spatially define the BCW rates and 19 development of continuous coverage map of calculated BCW values. 20 These process steps are further illustrated in the flowchart presented in Figure 4-2. March 2010 51 Draft - Do Not Quote or Cite ------- Appendix B Identification of input data required by PROFILE to estimate BCV In Existing National-Level GIS Coverages (see table 3a) P > u t Dat a Class Requires Development and Delineation of National-Level GIS Datalayers (see table 3b) 1 es PROFILE Default Values Requires review by user (see table 3c) F Construct National GIS Data Layers and convert to raster format Determine and Delineate BCw polygons Calculate mean values for National GIS Data Layers for each BCw Polygon Organize and Format BCw Polygon data as PROFILE model input file i 2 Calculate BCw For BCw polygons Figure 4-2. Process Steps for Estimating BCW Using the PROFILE Regional Model March 2010 52 Draft - Do Not Quote or Cite ------- Appendix B 1 2 3 4 5 6 7 8 9 10 11 12 Step 1. Identification of input data required by PROFILE and development of spatial data layers, national databases and default values for each data element within the model PROFILE (version 5.0) is data intensive and requires the user to provide or review a total of 26 soil, climatic and biological input data (Table 4-3a-c). In addition, to run the regional application of PROFILE, it would be necessary to have each of these data parameters available as continuous coverages. A large proportion of these variables are already included in existing databases in the United States and could be easily converted into continuous coverage data layers for the conterminous United States, if not currently available in continuous coverage format. Others, such as soil mineralogy, would need be modeled or constructed from other data. Still others may need to be represented by default values from the literature, until more unique, spatially site-specific values are determined. Table 4-3a. Data required to estimate BCW with the regional PROFILE model (version 5.0). The data in this table must be input by the user and are currently available as a continuous coverage layers for at least a portion of the conterminous United States. PARAMETER precipitation cation deposition anion deposition number of soil layers soil layer height temperature dry soil bulk density run-off UNITS m/yr kEq/ha/yr kEq/ha/yr # m °C kg/mj m/yr DESCRIPTION 30-year long-term average ammonium (NH4+), Ca^+, Mg^+, K+, Na+, Al - wet and dry deposition sulphate (SO/"), chloride (Cl~) and nitrate (NO3~) - wet and dry deposition up to 5 layers (with the forest floor/organic layer being the first horizon) by layer mean annual soil temperature by layer by layer number between 0 and the precipitation rates. If there is no lateral flow, runoff rate should equal the precipitation rate times the % of precipitation leaving the last soil layer. 13 Table 4-3b. Data required to estimate BCW with the regional PROFILE model (version 5.0). The data in this table must be input by the user and are not currently available as a continuous coverage layers for at least a portion of the conterminous United States (will require development of national coverage layer). PARAMETER net uptake cation uptake nitrogen uptake litterfall soil water content surface area UNITS kEq/ha/yr % % kEq/ha/yr mj/mj nf/mj DESCRIPTION nitrogen (N), Caz+, Mg^+, K+ - only applied if biomass through harvesting or fire % of total soil profile (all soil layers combined should Can be estimated using root distribution % of total soil profile (all soil layers combined should Can be estimated using root distribution removed sum to 100%). sum to 100%). N, Ca^+, Mg^+ and K+ - input to forest floor by layer soil surface area by layer March 2010 53 Draft - Do Not Quote or Cite ------- Appendix B PARAMETER logKgibbsite dissolved organic carbon (DOC) mineralogy UNITS - mg/L % DESCRIPTION by layer by layer % abundance of 14 dominant mineral groups (K-Feldspar, Plagioclase, Albite, Hornblende, Pyroxene, Epidote, Garnet, Biotite, Muscovite, Fe-Chlorite, Mg-Vermiculite, Apatite, Kaolinte, Calcite) Table 4-3c. Data required to estimate BCW with the regional PROFILE model (version 5.0). The data in this table are used to support calculations within the model and should be reviewed by the user. PARAMETER forest canopy net mineralization precipitation entering soil horizon precipitation leaving soil horizon CO2 pressure immobilization nitrification denitrifi cation nutrient uptake kinetics UNITS kEq/ha/yr kEq/ha/yr % % xatm - - - - DESCRIPTION N, Ca^+, Mg^+, K+ - nutrients removed by or leached from canopy N, Ca , Mg^+, K+ - net accumulation of soil organic matter expressed as % of precipitation. If no lateral flow, % leaving top layer should be same as % entering underlying layer expressed as % of precipitation. If no lateral flow, % leaving top layer should be same as % entering underlying layer entered as multiple of atmospheric pressure; typically ranges from 5 in the organic horizons to 40 in the mineral soil layers nitrogen immobilization - constant constant constant coupled vs. uncoupled uptake of N and base cations / uptake mechanism (unspecific, vanselow and none) 2 A total of eight parameters including climate, deposition, run-off and many of the soil 3 variables have data available as continuous coverages for the conterminous the United States 4 (Table 4-4), and for most of these variables, data exist for all 48 states. However, some of these 5 databases are missing variables and/or data or may need to be modified. Currently, there is no 6 data that describes wet and dry Al deposition. This data, however, is not available in most 7 locations where PROFILE is applied, and this parameter is typically left blank within model (H. 8 Sverdrup personal communication, 2009b). Therefore, the absence of this datalayer in the United 9 States should not pose a problem for the BCW estimates. The soil temperature parameter within 10 the SSURGO database is poorly populated and data only exists for seventeen states. However, 11 mean annual air temperature is often used as a surrogate for soil temperature within PROFILE 12 because the two temperature measures are similar in some of regions (Miller, 2001). In some 13 cases, models describing the relationship between air and soil temperature are also available 14 (e.g., Yin and Arp, 1993). Therefore, the use of air temperature instead of soil temperature or 15 modeled soil temperature could be explored with the application of PROFILE in the United 16 States, if necessary. March 2010 54 Draft - Do Not Quote or Cite ------- Appendix B Table 4-4. Available datasets and databases for the conterminous United States that could be used to estimate BCW with the regional application of the PROFILE model (version 5.0). DATA Total annual precipitation Average maximum air temperature Average minimum air Temperature Run-off Dry cation deposition (NH4+, Ca2+, Mg2+, K+, Na+) Wet cation deposition (NH4+, Ca2+, Mg2+, K+, Na+) Dry anion deposition (S042-, Cr, N03') Wet anion deposition (SO42", CI", NO3") NH4+ and NO3- wet and dry deposition Soil horizon depth Soil bulk density SOURCE URL or REFERENCE http://prism.oregonstate.edu/products/matrix.pht ml?vartype=ppt&view=maps http://prism.oregonstate.edu/products/matrix.pht ml?vartype=ppt&view=maps http://prism.oregonstate.edu/products/matrix.pht ml?vartype=ppt&view=maps http://pubs.er.usgs. gov/djvu/HA/ha_710_plt.djvu http://www.epa.gov/castnet/data.html http://www.epa.gov/castnet/data.html, http://nadp.sws.uiuc.edu/maplib/grids/2008/ http://www.epa.gov/castnet/data.html http://www.epa.gov/castnet/data.html, http://nadp.sws.uiuc.edu/maplib/grids/2008/ Community Multiscale Air Quality (CMAQ) - http://www.epa.gov/AMD/CMAQ/ HZDEPT_R field of chorizon table (httpV/soildatamart.nrcs.usda.gov/documents/S SURGOMetadataTableColumnDescriptions.pdf) DB3BAR_R field of CHORIZON table (http://soildatamart.nrcs.usda.gOv/documents/S SURGOMetadataTableColumnDescriptions.pdf) DATE(S) OF AVAILABLE DATA 1971-2000 1971-2000 1971-2000 1951-1980 1987-2008 1987-2008, 1994-2006 1987-2008 1987-2008, 1994-2006 2002 1987-2008, 1994-2006 N/A UNITS in/yr °F °F in/yr kg/ha kg/ha kg/ha kg/ha kg/ha cm g/cm3 RESOLUTION 0.64 km2 0.64 km2 0.64 km2 1:2,000,000 86 stations in the 48 conterminous states 86 stations in the 48 conterminous states, 6.25 km2 86 stations in the 48 conterminous states 86 stations in the 48 conterminous states, 6.25 km2 12km2 1:12,000- 1:63,360 1:12,000- 1:63,360 STATES WITH COVERAGE all all all all all (except: ID, SD, NE, NM, and TX) Extrapolated 400 km from each station all (except ID, SD, NE, NM, and TX) Extrapolated 400 km from each station all (except ID, SD, NE, NM, and TX) Extrapolated 400 km from each station all (except ID, SD, NE, NM, and TX) Extrapolated 400 km from each station all all all March 2010 55 Draft - Do Not Quote or Cite ------- Appendix B DATA Soil texture (% sand, silt, and clay) Soil stoniness (% of soil with particles >2mm) Soil temperature SOURCE URL or REFERENCE SANDTOT R, SILTTOT R, and CLAYTOT R fields of CHORIZON table (httpV/soildatamart.nrcs.usda.gov/documents/S SURGOMetadataTableColumnDescriptions.pdf) FRAGVOL_R in CHFRAGS table (http://soildatamart.nrcs.usda.gOv/documents/S SURGOMetadataTableColumnDescriptions.pdf) SOITEMPMM field of the COSOILTEMP table (http://soildatamart.nrcs.usda.gOv/documents/S SURGOMetadataTableColumnDescriptions.pdf) DATE(S) OF AVAILABLE DATA N/A N/A N/A UNITS % % °C (average by month) RESOLUTION 1:12,000- 1:63,360 1:12,000- 1:63,360 1:12,000- 1:63,360 STATES WITH COVERAGE all all AK,CA,CO,GA,ID,KS,MI,MN, MO,MT,NC,NE,NM,OR,PR, TX,VA March 2010 56 Draft - Do Not Quote or Cite ------- Appendix B 1 Although a portion of the input data to estimate BCW with PROFILE are already available 2 as national data coverages, there are nine additional input parameters that are not currently 3 described by nationwide datasets, and nine parameters that are built into the regional application 4 of the model and may require review and adjustment prior to applying PROFILE in the United 5 States. The nine input parameters that would require the development of national GIS coverages 6 or datasets that could be applied throughout the Unites States include: net uptake, % base cation 7 and nitrogen uptake, litterfall, soil water content, surface area, logKgibbsite, mineralogy, and 8 dissolved organic carbon (DOC). 9 Net Uptake 10 A national dataset of net uptake of nutrients by forest systems could be developed using 11 the approach outlined by McNulty and colleagues (1997). Briefly, the United States Forest 12 Service (USFS) and United States Geological Survey (USGS) dataset describing the 21 different 13 forest types would be used to map forest cover in the 48 states, and nitrogen and base cation 14 (Ca2+, Mg2+, K+) uptake by each forest type would be determined using the average values 15 presented in Table 4-5. These values were calculated by McNulty and colleagues (2007) and 16 incorporate annual volume growth by region from the USFS Forest Inventory and Analysis 17 (FIA) database and nitrogen and base cation contents by tree species and tree component from 18 the Tree Chemistry Database (Pardo et al., 2004). Net uptake would only be necessary for sites 19 that are actively managed and experience removal of biomass through logging and/or fire. 20 Therefore, based on the assumption that only wilderness and conservation areas are not harvested 21 or managed, these nitrogen and base cation uptake estimates would only be applied to forest 22 areas that are not designated as wilderness by the National Wilderness Preservation System of 23 the United States (McNulty et al., 2007). 24 March 2010 57 Draft - Do Not Quote or Cite ------- Appendix B Table 4-5. Nitrogen and base cation uptake by forest type (from McNultyetal., 2007). FOREST COVER TYPE white-red-jack pine spruce fir longleaf slash pine loblolly shortleaf pine oak pine oak hickory oak-gum-cypress elm-ash-cottonwood maple-beech-birch aspen-birch douglas-fir hemlock-sitka-spruce ponderosa pine western white pine lodgepole pine Larch fir-spruce Redwood Chaparral pinyon-juniper western hardwoods NITROGEN UPTAKE (eq/ha/yr) 59.07 54.27 154.74 140.41 129.71 102.56 124.18 79.74 101.76 81.69 109.89 98.88 75.29 40.69 40.19 65.1 94.65 100.92 106.6 40.87 135.21 BASE CATION UPTAKE (eq/ha/yr) 77.14 83.72 227.22 208.58 213.75 254.87 235.68 156.3 190.51 125.46 179.03 161.12 174.39 37.11 61.25 77.14 146 156.62 201.61 58.21 263.33 2 Soil Surface Area 3 Soil surface area is commonly determined in the laboratory using the Brunauer-Emmett- 4 Teller (BET) nitrogen absorption technique (Hodson et al., 1997). However, data from such 5 analyses are not available for all soils in the United States. Therefore, it would be necessary to 6 estimate surface area from other soil data. Within the PROFILE model, surface area is calculated 7 with soil texture and particle size distribution data (Equation 4-5) (Alveteg et al., 2004), and 8 Sverdrup and colleagues (1992), used this equation in their study of critical acid loads in 9 Maryland. This same approach could be used for mapping soil surface areas in the United States. 10 Soil texture is part of the U.S. Department of Agriculture- Natural Resources Conservation 11 Service (USDA-NRCS) Soil Survey Geographic (SSURGO) database (Table 4.0). Therefore, it March 2010 58 Draft - Do Not Quote or Cite ------- Appendix B 1 would be possible to produce a continuous coverage map of soil surface areas in the United 2 States. (4-5) 4 where 5 Aw = total exposed surface area (m2/m3) 6 x = weight fraction of clay, silt and sand when xciay + xs;it + xsand = 1 ; 7 p = soil density in kg/m3 8 Soil Mineralogy 9 Soil mineralogy is one of the most important and influential variables within PROFILE. 10 However, it is also a very time intensive and expensive measurement. Therefore, soil mineralogy 11 data in the United States is sparse, and a continuous coverage layer of soil mineralogy does not 12 exist. In most regional applications of PROFILE in Europe and other regions, the mineralogy 13 input data are based on a combination of data from soil geochemical and mineralogy analyses 14 and mineralogical composition based on output from a model such as the Analysis to Mineral 15 (A2M) model (Posch and Kurz, 2007). The A2M model estimates all possible mineral 16 compositions from total chemical analyses (Ca2+, Mg2+, K+, Na+, Ti, Al, phosphorus (P), silicon 17 (Si), iron (Fe)) of the soil and a pre-specified set of minerals that are likely to be present in the 18 soil. The highest probability mineral composition is an output of the arithmetic mean of all 19 extreme mineral modes. The resulting mineralogies are then mapped to "geological provinces" 20 (Sverdrup et al., 1990) that have the same parent material bedrock but may differ in soil 21 mineralogy in a consistent pattern (Sverdrup et al., 1990). Alternatively, the mineralogies can be 22 mapped to "mineralogy polygons" that are delineated based on probable similarities in mineral 23 compositions of the soils. Typically, the spatial borders of mineralogy polygons are determined 24 by underlying parent material geology and/or soil type groupings that are likely to have the same 25 mineralogies (H. Sverdrup personal communication, 2009b). In areas where the soils have 26 formed from transported materials, such as glacial till, it is sometimes necessary to consider the 27 surficial geology and model the origin and transport of materials to determine the parent material 28 geology (McKenzie and Ryan, 1999). March 2010 59 Draft - Do Not Quote or Cite ------- Appendix B 1 Due to the diverse geological history of the United States, it may be necessary to include 2 a variety of variables and databases in the characterization and mapping of mineralogy. Parent 3 materials underlying soils in the conterminous United States vary extremely. These materials 4 include not only bedrock beneath young soils, but also a variety of young and old regolith 5 materials that include both residuum formed in place and all varieties of transported sediments. 6 In addition, the soils are old and highly weathered in a large portion of the United States, and 7 therefore, no longer resembles the mineral composition of the parent material. For example, the 8 mineralogy of soils atop ancient residuum of the Appalachian region will vary significantly from 9 the mineralogy of younger residuum of the Western mountain ranges. Also, the mineralogy of 10 soil developed on the older loessal plain in the Mississippi basin will vary from the younger 11 glacial deposits along the northern regions of the United States. Therefore, determination of soil 12 mineralogy in the United States would require an approach that is able to recognize the varied 13 geological histories, different parent material origins, and soil mineralogies that differ from the 14 original parent material sources. Such an approach would involve the following steps be 15 conducted simultaneously: 16 1. Delineation of mineralogy polygons based on soil classification at a level supported 17 by available data 18 2. Determination of mineralogy and geochemical data availability for each mineralogy 19 polygon 20 3. Comparison of mineralogy polygons with underlying bedrock and surficial geology 21 4. Testing modeled mineralogy against actual mineralogy measurements 22 Delineation of mineralogy polygons based on soil classification at a level supported by 23 available data -_Mapping and creation of a national GIS coverage of mineralogy in the 24 conterminous United States would require the delineation of "mineralogy polygons". 25 "Mineralogy polygons" are spatially-defined polygons that are delineated based on probable 26 similarities in mineral compositions of the soils. These polygons would need to be large enough 27 in scale to be adequately covered by available mineralogy and soil analysis data, yet small 28 enough to only represent single assemblages of soil minerals. Ideally, each mineralogy polygon 29 should have at least one data point or soil profile analysis that describes the total analysis (Ca2+, 30 Mg2+, K+, Na+, Ti, Al, P, Si, and Fe) and/or mineralogy of the soil layers. Where data are 31 missing, it would be necessary to interpolate data from other locations using correlations with March 2010 60 Draft - Do Not Quote or Cite ------- Appendix B 1 surrounding and adjacent known data and other supporting criteria indicative of similar 2 mineralogies and weathering patterns (e.g., geologic and physiographic regions, bedrock geology 3 data, climatic regions, and others). 4 Within the United States, one of the most suitable coverages for the delineation of the soil 5 mineralogy polygons is the SSURGO soils database (Table 4-6). The smallest unit within this 6 database is the soil mapping unit which can consist of up to five individual soil series. A soil 7 series is defined as "soils that are similar in all major profile characteristics (Brady and Weil, 8 2002), and soils within the same series have been influenced by similar climate, topographic 9 location, biota, parent material and pedological time frame. Therefore, the soil within a series, 10 regardless of location would be expected to have identical or sufficiently similar mineralogies 11 (C. Smith personal communication, 2009). The soil groupings within the higher levels of soil 12 taxonomy may also be based on characteristics such as soil mineralogy. For example, soil orders 13 are largely classified by the degree of weathering and soil development, with Entisols 14 representing the youngest, least weathered soils, and Ultisols and Spodosols being more highly 15 weathered. Therefore, it may be possible to group the soil mapping units at a higher level of 16 taxonomy, such as the great group, family or order, as the "mineralogy polygons". However, 17 since soils are classified based on multiple formative factors, the "mineralogy polygons" could 18 be a mixture of groupings based on different levels of soil taxonomy, with all groupings based on 19 factors indicative of similar mineral assemblages in the soil. 20 Detailed soil delineations have been completed for more than 80% of the conterminous 21 United States and are used in the NRCS Soil Survey Geographic (SSURGO) dataset (Figure 4- 22 3). Data are missing for many public land areas (e.g., national forest lands), and there are 23 approximately 21,000 soil series delineations within the conterminous United States. Table 4-6 Datasets with Geochemical and Mineralogy Data for U.S. Soils DATA Soil Survey Geographic Database (SSURGO) SOURCE httpV/soils.usda.gov/survey/g eography/ssurgo/ RESOLUTION 1:12,000 to 1:63,360 INCLUDED DATA SSURGO is linked to a National Soil Information System (NASIS) attribute database. The attribute database gives the proportionate extent of the component soils (i.e., u soil series) and their properties for each map unit. The SSURGO map units consist of 1 to 3 components each. There are approximately 15,000 and 20,000 soil series polygons delineated across the United States March 2010 61 Draft - Do Not Quote or Cite ------- Appendix B DATA SOURCE RESOLUTION INCLUDED DATA U.S. General Soil Map (STATSGO http://soils.usda.gOV/survey/g eography/statsgo/ 1:250,000 The tabular data contain estimated ranges (low, high, and representative values) of physical and chemical soil properties, soil interpretations depicting the range for the geographic extent of the map unit. Soil map units are linked to attributes in the tabular data, which give the proportionate extent of the component soils and their properties. Surficial Geology of the United States (1977) (also Map of Surficial Deposits and Materials in the Eastern and Central United States (East of 102° West Longitude)) http://tin.er.usgs.gov/geology/ state/ or http://water.usgs.gov/GIS/met adata/usgswrd/XML/ofr99- 77_geol75m.xml also http://pubs.usgs.gov/imap/i- 2789/ 1:7,500,000 (Eof 102° W Longitude: 1:2,000,000) Provides approximate areal extent of about 45 categories of regolith types across the conterminous United States. Compilation East of 102° West Longitude has further classified deposits generally within original polygons. Element Concentrations in Soils and Other Surficial Materials of the Conterminous United States (Shacklette Data, 1977) USGS, Denver Federal Center Offices Sampling density: 1 sample per 6,000km2.; equivalent to the collection of samples on a 75-km grid. Ultra-low-density geochemical baseline data from 1,323 samples locations characterizing soils and other surficial materials in the conterminous United States. Elements analyzed included: Ag, Al, Ba, Be, B, Ca, Ce, Cr, Co, Cu, Ga, Ge, Hg, Fe, La, Li, Pb, Mg, Mn, Mo, Na, Nd, Ni, Nb, P, K, Rb, S, Sc, Se, Sr, Th, Ti, U, V, Yb, Y, Zn, Zr, and total carbon. The National Geochemical Survey - Database and Documentation (Version 5.0, on- going) http://tin.er.usgs.gov/geoche m/doc/home.htm and http://tin.er.usgs.gov/geoche m/ Nominal grid spacing of 17 by 17 kilometers (i.e., minimum sample density of 1 sample per 289 km2 in all land areas of the country Stream-sediment-based geochemical survey for the United States; Analytical methods include a 40-element ICP package plus single-element determinations of As, Se, and Hg by atomic absorption for every sample. about 60,000 stream-sediment samples that have been analyzed. Digital data files are presented in 6 categories. In total there are 43 individual data files for the Unites States. Some of the data has also been processed into vector data to produce maps showing the elemental concentration of As, Se, Hg, Pb, Zn, Cu, Al, Na, Mg, P, Ca, Ti, Mn, and Fe at the county level. Database contains 287 attributes (77,212 records). Integrated Geologic Map Databases for the United States (1998-2007) http://gsa.confex.com/gsa/20 06AM/finalprogram/abstract_ 110914.htm and http://tin.er.usgs.gov/geology/ state/ 1:100,000 Seamless national-scale geologic spatial data-layer and database to support national and regional level projects, including mineral resource and geoenvironmental assessments. Data include general geologic unit age, dominant lithology (rocktypel must be >50% of unit) and second most dominant lithology (rocktype2). March 2010 62 Draft - Do Not Quote or Cite ------- Appendix B DATA Soil Pedon Pit Data (on-going) Physiographic Regions of the United States (Fenneman, 1946) SOURCE USDA NRCS http://water.usgs.gov/GIS/dsd I/physio. eOO.gz RESOLUTION Area covered by a pedon varies from 10 - 100 square feet; approximately 30,000 soil pits/pedons in the NRCS database 1:7,000,000 INCLUDED DATA Geochemical elements: Al, Ca, Fe, K, Mg, Mn, Na, P, Si, Sr, Ti, and Zr. X-ray diffraction for clay mineralogy by horizon; optical mineralogy analysis is performed on the dominant sand fractions of the soil from the A-horizon, B-horizon, and C- horizon, or the most dominant horizon. More than 60 fields describing the minerals are listed in the database. The dataset is not uniform in that elemental analyses were routinely done through the 1970's but then these analyses were suspended through the 1980's. Elemental analyses were resumed during the early 1990's. It is estimated that as much as one third of the 30,000 soil pedons have geochemical data. Likewise, optical mineralogy is not performed for all pedons and the NRCS staff estimate that approximately as many as one third of the 30,000 soil pits have optical analysis results. Even though the number of pedons with data are similar for geochemical and optical analysis results, the data are not necessarily associated with the same set of pedons or even soil series. Geomorphic / physiographic broad-scale subdivisions based on terrain texture, rock type, and geologic structure and history. Nevin Fenneman's (1946) three-tiered classification of the United States - by division, province, and section. March 2010 63 Draft - Do Not Quote or Cite ------- Appendix B I mage source: http://soils.usda.gov/survev/geography/ssurgo/ Spatial and Tabular Tabular Only No Data Figure 4-3. Map Showing the Distribution and Status of SSURGO Data March 2010 64 Draft - Do Not Quote or Cite ------- Appendix B 1 Determination of mineralogy and total analysis data availability for each mineralogy 2 polygon - The soils for each spatially-defined "mineralogy polygon" would require % 3 mineralogy to determine BCW with the PROFILE model. The relative abundances of 14 4 dominant minerals are required as model input, and this % mineralogy can be based on direct 5 measurements of soil mineralogy or can be determined with the A2M model. As outlined earlier, 6 the A2M model is able to estimate the most probable % mineral composition, or proportion of 7 mineral phases, of a soil based on total analysis data and the identification of the minerals that 8 are likely to be present in the soil. Therefore, it would be necessary to determine the availability 9 of such data for each of the "mineralogy polygons" in the conterminous United States. 10 Currently, there are potentially three consistent national-scale datasets that contain 11 various levels of mineralogy and total analysis data to serve as inputs for the A2M and PROFILE 12 models. These include: 13 • Chemical Analyses of Soils and other Surficial Materials of the Conterminous United 14 States (Shacklette dataset) and accompanying Geochemical Landscapes Project data, 15 • the more recent National Geochemical Survey data, and 16 • the United States Department of Agriculture (USDA) NRCSpedon soil pit dataset. 17 A summary of these datasets is outlined in Table 4-6. 18 Chemical Analyses of Soils and other Surficial Materials of the Conterminous United 19 States (Shacklette Data) and the Geochemical Landscapes Project datasets provide geochemical 20 baseline data for soils and other surficial materials in the conterminous United States. The 21 original Shacklette dataset contains geochemical data from soils and other regolith collected and 22 analyzed by Hans Shacklette and colleagues beginning in 1958 and continuing until about 1976. 23 This dataset has approximately 1,323 samples, at a sampling density of approximately 1 sample 24 per 6,000 square kilometers (Figure 4-4). The soil samples within this dataset were analyzed for 25 a large number of elements, including Ca, Fe, Mg, Na, P, K, and Ti (Gustavsson et al., 2001), 26 that are required by the A2M model. However, assessments of mineralogy were not included in 27 these original analyses. An additional drawback with the data set is its extremely low numbers of 28 samples for the entire conterminous United States. However, more recent high-resolution studies 29 (e.g., Smith et al., 2005) for select elements (e.g., Calcium) have illustrated that the regional 30 patterns established by the Shacklette data are generally maintained except where areas have 31 been affected by anthropogenic factors (Smith, 2006). March 2010 65 Draft - Do Not Quote or Cite ------- Appendix B •••• o Dr., •Blackdotsindicatesamplesfrom sample-collection •* A -. '°otf •whitedotsindicatesamplesfromsample-collection " •- Phase2, *L O •gray dotsindicatesampleswhoseplacement into • Q phase 1 or phase 2 is uncertain. (All archived sample have been reanalyzed (personal communication with David Smith, 1-4-2009) Shacklette, HansfordT., and Josephine G. Boerngen, 1984 Figure 4-4. Soil Sampling Locations Included in the USGS Shacklette Dataset March 2010 66 Draft - Do Not Quote or Cite ------- Appendix B The Geochemical Landscapes Project was begun in 1998 with most work occurring after 2002. The purpose of the data collection is to increase the density of the Shacklette data locations to produce a high resolution geochemical dataset for North American soils of 6,000 data points (D. Smith personal communication, 2009). This is an on-going collaborative effort by the USGS, USDA Natural Resource Conservation Service, other federal agencies, and academia to build a national-scale soil geochemical survey. The project has just completed a third year of continental-sampling and completed sample collection for approximately 80% of the conterminous United States (D. Smith personal communication, 2009). The USGS anticipates that sampling may be completed for the conterminous United States in 2010; or 2011 at the latest. Both total and mineralogy analyses are being performed on these samples. Mineralogy analyses include x-ray diffraction on the clay fraction and optical analyses on the fine sands and silts. In addition, the original Shacklette data have been re-analyzed for mineralogy. The National Geochemical Survey (NGS) dataset is being built by on-going efforts by the USGS to produce a new stream-sediment-based geochemical survey for the United States at a spacing of 17 by 17 kilometers (i.e., minimum sample density of 1 sample per 289 km2 in all land areas of the country) (Figure 4-5). The project has sought to capitalize on existing datasets and archived samples. For this reason the NGS is based primarily on analyses of stream sediments to build on the massive archives of data and samples from DOE's National Uranium Evaluation (NURE) program. Much of the survey has entailed reanalysis of approximately 35,000 archival samples from the NURE program. Where NURE samples do not exist, USGS has been working with cooperators to obtain new samples. In total, the project is expecting to have more than 60,000 samples. Most or all of the sampling has been completed for the conterminous United States and only few analyses are left to complete (D. Smith personal communication, 2009). The samples are being analyzed for 40 elements, including all of the elements which are necessary input for the A2M model. In addition, for a select number of samples mineralogy analyses (x-ray diffraction of clay fraction and optimal analyses of fine sands and silts) are also being conducted. March 2010 67 Draft - Do Not Quote or Cite ------- Appendix B Spacing Density (km) N/1QOkmA2) <3 3-4 4-5 5-7 7-9 9-14 14-20 >20 63-110 40-63 20-40 12-20 5-12 3-5 <3 Image source: http://tin.er.usgs.gov/geochem/doc/status.htm Figure 4-5. Sample Density of USGS National Geochemical Survey March 2010 68 Draft - Do Not Quote or Cite ------- Appendix B 1 USD A NRCS Soil Pedon Pit Data were collected by the USD A NRC S for data required 2 for delineation of soil series, map units, and associated attributes. The data are contained in the 3 NRCS USSOILS database that provides data for the SSURGO database. There are currently 4 approximately 30,000 soil pits/pedons in the NRCS database, and soil samples from these pits or 5 pedons have been analyzed for a large variety of physical and chemical properties. These 6 analyses include total chemical analysis, which includes elements required by A2M (e.g., Al, Ca, 7 Fe, K, Mg, Na, P, Si, Ti). In addition, mineralogy has been characterized through two analyses: 8 x-ray diffraction, which identifies clay mineralogy, and optical mineralogy which determines the 9 mineral composition of the fine sand and silt fractions of the soil (C. Smith personal 10 communication, 2009). However, these three analyses have not been conducted on all soils. Only 11 11,747 of the 30,000 soil pits have been analyzed for at least one of the three parameters (Figure 12 4-6), and only 4,710 soils have been analyzed for all three (Figure 4-7). March 2010 69 Draft - Do Not Quote or Cite ------- Appendix B 1 2 3 4 Figure 4-6. NRCS Soil Pedon Sample Pit Locations (30,000 total) (Image created by RTI using data provided by NRCS on 12/2009) March 2010 70 Draft - Do Not Quote or Cite ------- Appendix B ?*•*. . • r 1 2 3 Figure 4-7. NRCS Soil Pedon Pit Sample Locations with Geochemical and Mineralogy Data (Image created by RTI using data provided by NRCS on 12/2009) March 2010 71 Draft - Do Not Quote or Cite ------- Appendix B 1 In summary, three main datasets have been identified that could provide the necessary 2 total analyses and mineralogy data for each "mineralogy polygon." These datasets would be 3 combined into a single database and overlaid on top of the "mineralogy polygon" datalayer to 4 determine the degree to which each polygon is covered by mineralogy and total analysis data. At 5 the scale of a nationwide analysis, the data from each of these datasets is considered comparable 6 given the sampling and analysis protocols that have been used (D. Smith personal 7 communication, 2009). Although the combined database would be large and offer over 75,000 8 data points, it is not likely that data would be available for all "mineralogy polygons." In such 9 cases, it would be necessary to determine the mineralogy through alternate methods. Potentially, 10 interpolation between data points could be conducted using numerical probabilistic methods. In 11 addition, it may be possible to determine probable mineralogy based on underlying bedrock or 12 surficial geology (described further in next session). Methods involving professional judgment 13 could also be used to interpret patterns and assign reasonable and appropriate values to express 14 the apparent condition. If such an approach were taken, it would be necessary to work with soils 15 experts who are familiar with the SSURGO database (e.g., NRCS Regional Staff) to make such 16 judgments. 17 Comparison of mineral polygons with underlying bedrock and surficial geology - In 18 many locations in the United States, soils have developed from the underlying bedrock or 19 surficial materials. Therefore, it may be possible to validate, support or identify the mineralogies 20 of each of the "mineralogy polygons" through a comparison with the physiographic regions of 21 the United States and the underlying bedrock and surficial geology. The physiographic provinces 22 are based on geology and topography. Therefore, these provinces relate geology and geological 23 history with expected soil characteristics, and the locations of "mineralogy polygons" should 24 broadly follow the patterns within these province boundaries. Similarly, the "mineralogy 25 polygons" could be compared against the underlying geologies to determine the accuracy of the 26 soil taxonomy groupings that were used to delineate the polygons. In addition, overlays of the 27 "mineraology polygon" datalayer and bedrock or surficial geology could support the estimation 28 of probable mineralogies and percent compositions for "mineralogy polygons" that are missing 29 soil mineralogy and or total analysis data. 30 The USGS 1:100,000 scale bedrock geology GIS cover would be first used for the 31 comparison between the "mineralogy polygons" and bedrock geology (Table 6.0). Most rock March 2010 72 Draft - Do Not Quote or Cite ------- Appendix B 1 types are typically characterized by less than 4 mineral types. Correlating the "mineralogy 2 polygons" with the bedrock type can be used to obtain a gross approximation of mineral phases 3 that would be expected in the residual parent materials and the corresponding soils. This would 4 be a particularly useful protocol to apply to areas where the soils have formed in place from the 5 weathering of the bedrock. For example, this approach could be used in unglaciated regions 6 where Entisols or residuum predominant. 7 The 1:7,500,000 scale USGS surficial geology layer could be used as the source for 8 comparison between "mineralogy polygon" and surficial geology (refer to Table 6.0). The 9 surficial geology layer would identify the type of regolith on which the soil has developed. 10 Regolith is defined here as any unconsolidated materials on top of bedrock, and consists of 11 residuum which has formed in place and transported materials that have been deposited by 12 gravity, wind, water or ice. Therefore, this layer will indicate the type of parent material that 13 supported the development of the soil and will provide an indication of the potential mineral 14 composition of the soil. Specific correlation of mineral types can be more difficult for 15 transported deposits. However, an association is still possible as correlated with general up-grade 16 areas that relate the likely origin, or areas of parent material, for the transported deposit. Even 17 though more generalized approaches to determining the mineralogy are suggested by the 18 available data, modeling the geologic source of parent materials by applying techniques similar 19 to soil-landscape modeling or environmental correlation modeling could be conducted 20 (McKenzie and Ryan, 1999). 21 Test modeled mineralogy against actual mineralogy measurements - To validate and test 22 the accuracy of the % mineral values assigned to the "mineralogy polygons", comparisons 23 should be made between the "mineralogy polygon" data layer and areas with detailed mineralogy 24 soil analyses. Such sites may include the LTER sites outlined in Table 4-7 or those detailed 25 within the scientific and geological literature. In addition, there may be locations where detailed 26 mineralogy assessments have been conducted by mining companies or research groups that could 27 be used to test the "mineralogy polygon" data layer. Comparisons would be particularly 28 important for mineralogy polygons with % mineralogy determined by the A2M model. March 2010 73 Draft - Do Not Quote or Cite ------- Appendix B Table 4-7. Long-Term Ecological Research (LTER) sites that could potentially be suitable as "field test" sites to validate BCW estimates generated with the regional application of the PROFILE model (version 5.0). LTER STUDY H.J. Andrews Experimental Forest Coweeta LTER Harvard Forest Hubbard Brook Experimental Forest Kellogg Biological Station Konza Prairie LTER Niwot Ridge Santa Barbara Coastal LTER Sevilleta LTER LOCATION Cascade Mountains, Oregon Southern Appalachian mountains, North Carolina Massachusetts White Mountain National Forest, New Hampshire Southwest Michigan Northeastern Kansas Colorado California New Mexico ADDITIONAL INFORMATION http://andrewsforest.oregonstate.edu/ http://www.lternet.edu/sites/cwt/ http://www.lternet.edu/sites/hfr/ http://www.lternet.edu/sites/hbr/ http://www.lternet.edu/sites/kbs/ http://www.lternet.edu/sites/knz/ http://www.lternet.edu/sites/nwt/ http://www.lternet.edu/sites/sbc/ http://www.lternet.edu/sites/sev/ 1 2 PROFILE Input Parameters Assigned Default Values 3 Development of national datasets or default values for the % base cation and nitrogen 4 uptake (by layer), litterfall, soil water content, logKgibbsite, and DOC input parameters would 5 most likely require the use of data from the literature and research conducted in the United 6 States. The % base cation and nitrogen uptake by soil layer variables are a function of the 7 distribution of fine roots, and rooting distributions are typically entered as one of four classes 8 into PROFILE (H. Sverdrup personal communication, 2009b). These root distribution classes are 9 based on data from an extensive literature search on the rooting habitats of common tree species 10 in Europe (Sverdrup and Stjernquist, 2002). A similar literature search could be conducted for 11 the main species within the 21 forest types in the United States, and the four root distribution 12 classes could be adjusted accordingly. 13 The litterfall parameter within PROFILE characterizes the amounts of N, Ca2+, Mg2+ and 14 K+ returned to the soil with the senescence of leaves, branches and stems. It is calculated as a 15 function of the site-specific growth rates of individual tree species and the nutrient content of the 16 different litter components. Since PROFILE is a steady-state model, the growth rates are 17 averaged over the rotation of the stand. Litterfall values have been determined for European tree 18 species (Sverdrup et al, 1990; Sverdrup and Stjernquist, 2002), and the same procedure could be 19 used for estimating values for the main species in the 21 forest types in the United States. The March 2010 74 Draft - Do Not Quote or Cite ------- Appendix B 1 Tree Chemistry Database (Pardo et al., 2004) could be serve as the source of litter nutrient 2 content and the USFS FIA database could potentially supply species-specific growth rates. 3 Soil water content is highly variable. However, since PROFILE is a steady-state model, it 4 is necessary to use a single value representative of the water content throughout the year. In 5 Sweden, a default value of 0.2 m2/m3 is often used (Halveteg et al., 2004), and it would be 6 necessary to establish a similar default value or set of default values for the United States. Such 7 values could be obtained from the literature. In addition, it may be possible to estimate a set of 8 soil water content estimates based on a simple water balance model that includes the influences 9 of precipitation, run-off, soil texture and/or soil drainage classes (H. Sverdrup personal 10 communication, 2009b). Data outlined in Table 4 and soil texture and the six drainage classes 11 (Well Drained; Excessive; Moderately Well; Poorly; Somewhat Excessively; Somewhat Poorly) 12 included within the SSURGO soils database could potentially be used in this simple water 13 balance model. 14 Soil dissolved organic carbon (DOC) and the logKgibbsite coefficient would also require 15 the use of values from the literature. Currently, with the application of PROFILE within Europe, 16 DOC is entered as 20 mg/L in the organic layers but drops rapidly with depth in the mineral soil 17 horizons (Alveteg et al., 2004). These values are based on a compilation of data from European 18 field sites (H. Sverdrup personal communication, 2009b) and are a function of the organic matter 19 content of the soil (Sverdrup et al, 1990). Similar values and relationships would need to be 20 established for forest systems in the United States based on available data and studies outlined in 21 the literature. LogKgibbsite is a coefficient that describes the concentration of Al in the soil 22 solution. It depends on the soil solution pH and differs by soil layer. Two sets of values have 23 been developed for the application of PROFILE within Europe, with one set being used for clay 24 soils and the other for non-clay soils (Sverdrup and Stjernquist, 2002). These values and 25 grouping by soil clay content were based on data from the literature and the consistent trends in 26 the gibbsite coefficients within and between soils (H. Sverdrup personal communication, 2009b). 27 For the application of PROFILE within the United States, it would be necessary to review the 28 logKgibbsite values, review the literature and potentially adjust the values as necessary to be 29 representative of conditions found in the United States. March 2010 75 Draft - Do Not Quote or Cite ------- Appendix B 1 Regionally-based Built-in PROFILE Input Parameters 2 There are nine variables that are currently built into the calculations of BCW for the 3 regional application of PROFILE (version 5.0) and do not require input data from the user. The 4 values of these variables were determined by field research in Europe, and are thought to vary 5 minimally between sites or are calculated based on the input data. These variables include: forest 6 canopy, net mineralization, % precipitation entering layer, % precipitation leaving layer, CC>2 7 pressure, immobilization, nitrification, denitrification, nutrient uptake kinetic variables (C. 8 Aksehson personal communication, 2009). Prior to applying PROFILE to map BCW throughout 9 the United States, the values and equations used for each of these variables should be examined. 10 It may be necessary to modify the model equations and/or replace the current values with 11 those from the literature to ensure that the values within PROFILE are representative of 12 conditions and processes in the United States (H. Sverdrup personal communication, 2009). 13 Forest canopy, within PROFILE, accounts for the Ca2+, Mg2+, K+ and N (as NH4+) that is 14 absorbed from or leached into the precipitation that is in contact with the canopy. Potassium, 15 Ca2+, Mg2+ are typically leached from the foliage and NH4+ is absorbed. The default values 16 within PROFILE are currently based on the results of field studies in Europe and are divided by 17 forest type (deciduous versus non-deciduous). Net mineralization within the model is a function 18 of soil organic matter content. Currently, net mineralization is set to "0" within PROFILE 19 assuming that the forests are managed sustainably and net mineralization is at an equilibrium; 20 Ca2+, Mg2+, K+ and NH4+ released through mineralization of organic matter is taken up by the 21 vegetation, returned as litter and remineralized. Therefore, there is no net loss or gain of nutrients 22 through mineralization. However, the net mineralization default value of "0" can be changed if 23 forest management is not sustainable and involves short rotations and/or practices such as whole 24 tree harvesting which remove the foliage and a large pool of "mineralized" nutrients from the 25 site. 26 The % of precipitation entering and leaving the soil layers variables within PROFILE are 27 determined based on the fine root distribution in the soil profile. Carbon dioxide pressure in the 28 soil is estimated from a small dataset of measurements conducted in different regions of the 29 world. The values that are used within PROFILE are a function of soil particle size. 30 Immobilization of nitrogen within PROFILE is currently set to range between 0.5 - 1.0 kg 31 N/ha/yr. This range of values was determined by the amount of the amount of nitrogen that has March 2010 76 Draft - Do Not Quote or Cite ------- Appendix B 1 accumulated in Northern European soils since the last glaciations. Denitrification and 2 nitrification are currently determined by mathematical equations that include the influences of 3 temperature, available soil nitrogen, soil moisture and soil pH. Nutrient uptake kinetics within 4 PROFILE consists of coupled versus uncoupled uptake of nitrogen and base cations. Within 5 PROFILE, uptake is set to "coupled" as a default because the uptake of Ca, Mg and Al and K 6 and NH4+ are coupled (Sverdrup et al., 1990). Uptake kinetics within the model are also 7 described as unspecified or vanselow depending on the uptake dynamics of base cations and Al 8 absorbed to the root surface. Currently, within PROFILE, deciduous species and domestic crops 9 are defaulted to vanselow kinetics and grasses, and conifers use unspecified kinetics (H. 10 Sverdrup personal communication, 2009). Unspecified kinetics indicates that the ion exchange 11 matrix on the root surface is indifferent to the valence of the absorbing ions (Sverdrup and 12 Warfvinge, 1993). 13 Step 2. Determination of polygon layer to spatially define the BCW rates and 14 development of continuous coverage map of calculated BCW values. 15 Following the establishment of continuous coverage databases and national datasets and 16 default values for the application of PROFILE (version 5.0) within the United States, it would be 17 necessary to construct a spatially-explicit continuous datalayer for mapping BCW throughout the 18 48 states. The resolution of the datalayer should be small scale and provide the highest level of 19 detail permitted by the data. In addition, the location of individual BCW polygons should be tied 20 to a variable or set of variables which strongly influence BCW. Since soil attributes including 21 mineralogy, bulk density, volumetric water content and exposed surface area of minerals 22 (discussed further in Section 4.3.6) are the largest sources of variability in the BCW calculations, 23 it may be most appropriate to map BCW according to mineralogy, soil series or a higher level of 24 soil taxonomy. Input data and default values for the 26 PROFILE variables would then be 25 mapped to the delineated BCW polygon layer. When multiple or sections of multiple polygons of 26 the same datalayer are present in a BCW polygon, a weighted average value for the data would be 27 calculated. All the data for each BCW polygon would then be formatted according to the 28 requirements of PROFILE and the PROFILE regional model would be run to produce maps of 29 BCW. March 2010 77 Draft - Do Not Quote or Cite ------- Appendix B 1 4.3.5 Potential limitations of proposed methodology 2 Although PROFILE is arguably the most suitable model currently available for 3 estimating and mapping BCW for terrestrial critical acid load determinations in the United States, 4 the model does have some limitations that should be acknowledged and potentially remedied 5 prior to application. The model and algorithms contained therein were developed in Sweden 6 using Swedish soils as the basis for the soil chemical and physical relationships (Hodson et al., 7 1997). The soils in Sweden are comparatively young, having formed since the last glaciations, 8 approximately 10,000 years ago (Sverdrup and Warfvinge, 1988). Therefore, there is some 9 concern that PROFILE may not accurately model base cation release in older soils (C. Smith 10 personal communication, 2009). As discussed by Hodson and Langan (1999), PROFILE does 11 not take into account the decreasing reactivity of minerals with duration of dissolution, and 12 assumes that the reaction rates are constant regardless of time and duration of dissolution. In 13 addition, the model assumes a constant versus decreasing reactive surface area as total surface 14 area increases. According to the authors, these shortfallings were two of the main reasons that 15 PROFILE did not show a decreased weathering rate with soil age relative to other models. 16 However, at the same time PROFILE has been used to estimate BCW in multiple locations with 17 older, more weathered soils, such as Maryland, China, Thailand, Argentina and Greece, and has 18 performed with apparent success (Duan et al., 2002; Sverdrup et al., 1992; H. Sverdrup). 19 PROFILE currently accounts for the weathering of 14 different minerals, with the 20 potential to include 13 additional minerals, if necessary. Potentially, there may be minerals 21 within the United States that are not represented within the 27 that are currently included within 22 PROFILE. However, a total of 48 minerals have been investigated by the researchers that 23 developed the model (H. Sverdrup personal communication, 2009b). Therefore, it may be 24 possible to add additional minerals to PROFILE to ensure that it is able to address BCW in all 25 regions of the United States. 26 Additional limitations and concerns regarding the application of PROFILE to estimate 27 BCW rates have been identified in a thorough review by Hodson and colleagues (1997). Some of 28 the main issues brought up by the authors include: the need for a more consistent set of constants 29 for the weathering rate equations; inaccuracies in the mineral compositions; errors in the 30 calculation to determine surface area; and confounding influences of soil particles greater than 31 2mm in size on soil bulk density. Hodson and colleagues (1997) point out the need to reexamine March 2010 78 Draft - Do Not Quote or Cite ------- Appendix B 1 the reaction rate coefficients associated with hornblende, tourmaline, staurolite, kaolinite, garnet, 2 augite, biotite and chlorite, arguing that coefficients assigned to these minerals are not correct. 3 Similarly, the authors claim that the compositions of the minerals used within PROFILE may be 4 incorrect in some applications and may need to be modified by the user to more accurately 5 reflect the soil being modeled. Hodson and colleagues (1997) also demonstrate the potential to 6 over and underestimate BET surface area using the soil texture equation provided within 7 PROFILE. They claim that the equation underestimated the surface area of a British soil by 65%. 8 In part, the authors attributed these inaccuracies to the development of the soil texture - surface 9 area relationship from only 92 mineral soil samples from Sweden. Lastly, Hodson and colleagues 10 (1997) point out the need to recognize soil particles greater than 2mm in size in the soil bulk 11 density estimates, as such particles can impact the density by as much as 50% for stony soils. 12 The concerns raised by Hodson and colleagues (1997) appear to be valid and should be 13 considered by users of the PROFILE model. However, the authors of the review critiqued an 14 early version of PROFILE (version 3.01) and the most recent version of PROFILE may have 15 already addressed some of these limitations. For example, the abundance of particle sizes greater 16 than 2 mm is included in the current regional model of PROFILE (version 5.0). It should also be 17 noted that Hodson and colleagues (1997) did acknowledge that despite the apparent weaknesses 18 of PROFILE, BCW rates calculated with the model are comparable to those calculated using other 19 methods. 20 In addition to the potential limitations of PROFILE as a model, application of PROFILE 21 to map BCW rates throughout the United States may also present some drawbacks or restrictions. 22 There may be areas of the United States where input data required by the model is not available. 23 In such situations, it would be necessary to extrapolate data from areas with similar soil, biotic or 24 abiotic conditions. Similarly, if data for specific variables are limited in many areas, it may be 25 necessary to adopt best available default values over large areas, until more data and better 26 coverage across the states is available. 27 4.3.6 "Field Tests" of model and uncertainty analyses 28 As outlined in the preceding sections, the proposed methodology to map BCW throughout 29 the United States would involve the use of the regional application of PROFILE (version 5.0), 30 continuous coverage data, and in some cases, input and default values from the literature. 31 Therefore, at least a portion of the input data would not be site specific and would be entered as March 2010 79 Draft - Do Not Quote or Cite ------- Appendix B 1 class values or generated by sub-models or mathematical relationships. Soil water content and 2 soil mineralogy are examples of such data. It is largely unknown to what degree, if any, this 3 proposed methodology designed for mapping large areas would influence and potentially distort 4 the estimates of BCW. Therefore, to validate the weathering estimates from the proposed mapping 5 methodology, it would be worthwhile to conduct "field tests" of the model output in different 6 regions of the United States. Such "field tests" could consist of comparing the regional estimates 7 of BCW with those determined with the single site version of PROFILE and site-specific data. 8 (No actual on-the-ground field research required.) In addition, where available, the PROFILE- 9 generated BCW rate estimates could be compared with weathering rates determined by other 10 methods. Both approaches would provide an indication of the quality and accuracy of estimates 11 from the mapping methodology and regional application of PROFILE. Sites within the Long- 12 Term Ecological Research (LTER) network would be good locations for the "field tests" due to 13 the large amounts of data available at many of these sites. In addition, at some sites, such as 14 Hubbard Brook, base cation weathering has been determined using methods other than the 15 PROFILE model. A list of LTER sites within the conterminous 48 states that could potentially 16 serve as "field test" sites is presented in Table 6.0. A sub-set of these sites representing different 17 regions and conditions within the United States should be selected to validate the BCW estimates. 18 In addition to the validating the proposed methodology with "field test" site comparisons, 19 uncertainty analyses should also be conducted on the BCW estimates that are generated with the 20 methodology. There are a total of 26 parameters within the regional application of PROFILE 21 (version 5.0) that require data entry by the user or review prior to applying the model, and each 22 of these parameters could be expected to have a level of uncertainty. Therefore, cumulatively, 23 the uncertainties associated with the BCW estimates could be quite large. In addition, because 24 BCW is one of the most influential terms in the calculation of terrestrial critical acid loads, and 25 critical loads can be used as a measure of the impact of acidifying nitrogen and sulfur deposition 26 on terrestrial ecosystems, it is important to gain a good understanding of the uncertainty 27 associated with the BCW estimates. Critical acid loads could potentially be used by decision 28 makers to set policy and NOX and SOX emission standards within the United States. Furthermore, 29 uncertainty analyses can reveal which parameters are the most influential in the BCW estimates, 30 thereby guiding which parameters should receive the greatest attention in the development of the 31 datasets and national coverages for the PROFILE model. March 2010 80 Draft - Do Not Quote or Cite ------- Appendix B 1 Earlier versions of the PROFILE model (version 3.01) have already been reviewed and 2 analyzed by researchers based on the application of the model to sites in Norway, Sweden, 3 Scotland and Wales (Jonsson et al., 1995; Hodson et al., 1996; Zak et al., 1996). Monte Carlo 4 analyses testing the uncertainty associated with user defined input variables indicated that 5 varying input parameter errors individually and simultaneously (within the range of values 6 reported in the literature) resulted in a variation in model output of+/- 40% (Jonsson et al., 7 1995). The authors also determined that bulk density, volumetric water content and exposed 8 surface area of minerals were the largest source of variation in the output values. The least 9 sensitive parameters were soil stratification, precipitation and percolation. Similar analyses were 10 conducted by Hodson and colleagues (1996) who determined the influence of single input 11 parameters, one at a time. Based on their analyses, BCW estimates could vary by over 100% using 12 the ranges in parameters values measured in field studies. The authors also found that some 13 minerals, such as K-feldspar, were particularly sensitive to variation in input values, and soil 14 temperature, moisture content and exposed mineral surface area caused the largest amounts of 15 variation in the BCW estimates. These results based on an earlier version of PROFILE suggest 16 that ranges in input values can cause the BCW estimates from the model to vary by moderate to 17 large amounts. However, the level of uncertainty associated with outputs from the most current, 18 regional application of PROFILE (version 5.0) is still unknown. In addition, there has yet to be 19 an assessment of the performance of the model in the United States and a determination of how 20 ranges in data from different regions in the country would impact the variation in model output. 21 Therefore, uncertainty analyses should be conducted as a component of the proposed 22 methodology, to provide bounds to the range of output values associated with the BCW estimates 23 for terrestrial critical acid load calculations in the United States. 24 5. CONCLUSIONS AND RECOMMENDATIONS 25 The goal of this task was to inform EPA about the tools and data available to develop 26 maximum deposition loads across the United States for aquatic and terrestrial acidification. In 27 particular, this effort focused on methodologies to estimate Bcw, a parameter that plays a crucial 28 role in predicting an ecosystem's ability to neutralize acid deposition. Based on the findings of 29 this literature review, discussions with experts, evaluation of tools, and assessment of data March 2010 81 Draft - Do Not Quote or Cite ------- Appendix B 1 availability, two process-based models are recommended: MAGIC for aquatic acidification (with 2 extrapolation through regional regression modeling) and PROFILE for terrestrial acidification. 3 It is clear that addressing limitations on soil data availability in the United States will 4 require considerable effort to populate both models; however, resources invested to satisfy this 5 data need can be leveraged to the benefit of both terrestrial and aquatic modeling goals. It is also 6 clear that the MAGIC and PROFILE models' application in the United States has focused on 7 select regions; however, model developers believe these models can be applied successfully in 8 other regions, particularly regions with more sensitive ecosystems. Finally, neither MAGIC nor 9 PROFILE models are readily accessible for public use. Therefore, it will not be practical to 10 assume states and regions could operate the models. Rather, it would be more manageable for the 11 models to be run at the Agency level with states and regional offices providing the needed input 12 data. 13 It is recommended that following EPA review of this report, candidate regions of the 14 United States be identified for modeling and levels of effort be estimated to prepare the MAGIC 15 and PROFILE models for operation and to collect and/or predict their input data. As part of the 16 effort, it is recommended that RTI collaborate with recognized experts in the development and 17 application of these two models. 18 6. REFERENCES 19 Aherne, J. Personal communication. 2009. Communication between Julian Aherne (Trent 20 University, Canada) and Jennifer Phelan (RTI International, USA) by telephone. 21 December 2009. 22 Akselsson, C. Personal communication. 2009. Communication between Cecilia Akselsson (Lund 23 University, Sweden) and Jennifer Phelan (RTI International, USA) by telephone. 24 December 2009. 25 Akselsson, C., H.U. Sverdrup, and J. Holmqvist. 2006. Estimating weathering rates of Swedish 26 forest soils in different scales, using the PROFILE model and affiliated databases. 27 Sustainable Forestry in Southern Sweden: The SUFOR Research Project. Linking Basics 28 and Management, p. 119-131. March 2010 82 Draft - Do Not Quote or Cite ------- Appendix B 1 Arp, P. Personal communication. 2009. Communication between Paul Arp (University of New 2 Brunswick, Canada) and Jennifer Phelan (RTI International, USA) during the Canadian 3 Council of Ministers of the Environment Acid Deposition Critical Loads: Status of 4 Methods and Indicators Workshop, March 18 and 19, 2009. Ottawa, Ontario, Canada. 5 Bailey, S.W., S.B. Horsley, R.P. Long, and R.A. Hallett. 2004. Influence of edaphic factors on 6 sugar maple nutrition and health on the Allegheny plateau. Soil Science Society of 1 America Journal 68:243-252. 8 Brady, N.C. and R.R.Weil. 2002. The Nature and Properties of Soils: Thirteenth Edition. 9 Prentice Hall. Upper Saddle River, New Jersey. 10 Brakke, D. F., A. Henriksen, and S.A. Norton. 1990. A variable F-factor to explain changes in 11 base cation concentrations as a function of strong acid deposition. Verb. Internal. Verein. 12 Limnol. 24: 146-149. 13 Brakke, D. F., A. Henriksen, and S.A. Norton. 1989. Estimated background concentrations of 14 sulfate in dilute lakes. Water Resources Bulletin 25(2): 247-253. 15 Bricker, O. P., B.F. Jones, and C.J. Bowser. 2003. Mass-balance Approach to Interpreting 16 Weathering Reactions in Watershed Systems. Treatise on Geochemistry, Volume 5. 17 Editor: James I. Drever. Executive Editors: Heinrich D. Holland and Karl K. Turekian. 18 pp. 605. ISBN 0-08-043751-6. Elsevier, 2003., p.l 19-132 19 Bulger, A.J., B.J. Cosby, C.A. Dolloff, K.N. Eshleman, J.R. Webb, and J.N. Galloway. 1999. 20 SNP:FISH, Shenandoah National Park: Fish in Sensitive Habitats, Volumes 1 through 21 IV. Project final report. Project Completion Report to the National Park Service. 22 Cooperative Agreement CA-4000-2-1007, Supplemental Agreement #2. University of 23 Virginia, Department of Environmental Sciences, Charlottesville, VA. 24 Chapin, F.S., III., L. Moilanen, and K. Kielland. 1993. Preferential use of organic nitrogen growth by a 25 non-mychorrizal arctic sedge. Nature 361:150-153. 26 Chen C.W., S.A. Gherini, N.E. Peters, P.S. Murdoch, R.M. Newton, and R.A. Goldstein. 1984. 27 Hydrologic analyses of acidic and alkaline lakes. Water Resources Research 20: 1875- 28 1882. March 2010 83 Draft - Do Not Quote or Cite ------- Appendix B 1 Chen, L., C. T. Driscoll, S. Gbondo-Tugbawa, M. J. Mitchell, and P. S. Murdoch. 2004. The 2 application of an integrated biogeochemical model (PnET-BGC) to five forested 3 watersheds in the Adirondack and Catskill regions of New York. Hydrological Processes 4 75:2631-2650. 5 Clair, T.A., Dennis, E.F., Cosby, BJ. 2003. Probable changes in lake chemistry in Canada's 6 Atlantic Provinces under proposed North American emission reductions. Hydrology and 7 Earth System Sciences Special Issue 7, 574-583. 8 Cosby, J. Personal communication. 2009. Communication between Jack Cosby (University of 9 Virginia, USA) and Jennifer Phelan (RTI International, USA) during the Critical Load 10 Workshop, Fall 2009 National Atmospheric Deposition Program Meeting. October 5 and 11 6, 2009. Saratoga Springs, New York, USA. 12 Cosby, B.J., A. Jenkins, R.C. Ferrier, J.D. Miller, and T.A. B. Walker. 1990 Modeling stream 13 acidification in afforested catchments: Long-term reconstructions at two sites in central 14 Scotland. Journal of Hydrology 120: 143-162.Cosby, B.J., G.M. Hornberger, P.F. Ryan, 15 and D.M. Wolock. 1989a. MAGIC/DDRP final report, vol. 1, Model, calibration, results, 16 uncertainty analysis, QA/QC, internal report, Environ. Res. Lab., U.S. Environ. Prot. 17 Agency, Corvallis, OR. 18 Cosby, B.J., G.M. Hornberger, and R.F. Wright. 1989b. Regional simulation of surface water 19 acidification: Uncertainty due to specification of atmospheric deposition, in Atmospheric 20 Deposition (Proceedings of the Baltimore Symposium, May, 1989), IAHS Publ., 179, 21 153-161. 22 Cosby, B.J., G.M. Hornberger, and R.F. Wright. 1989c. Estimating time delays and extent of 23 regional de-acidification in southern Norway in response to several deposition scenarios, 24 in Regional Acidification Models: Geographic Extent and Time Development, edited by 25 J. Kamari et al., pp. 151- 166, Springer-Verlag, New York. 26 Cosby, B.J., R.F. Wright, G.M. Hornberger, and J.N. Galloway. 1985a. Modeling the effects of 27 acid deposition: Assessment of a lumped parameter model of soil water and streamwater 28 chemistry. Water Resources Research 21:51-63. March 2010 84 Draft - Do Not Quote or Cite ------- Appendix B 1 Cosby, B.J., R.F. Wright, G.M. Hornberger, and J.N. Galloway. 1985b. Modeling the effects of 2 acid deposition: Estimation of long-term water quality responses in a small forested 3 catchment. Water Resources Research 21:1591-1601. 4 DeHayes D.H., P.O. Schaberg, GJ. Hawley, and G.R. Strimbeck, 1999. Acid rain impacts on 5 calcium nutrition and forest health. Bioscience ¥9(10):789-800. 6 Dennis, T.E., and AJ. Bulger. 1995. Condition factor and whole-body sodium concentrations in 7 a freshwater fish: Evidence for acidification stress and possible ionoregulatory 8 overcompensation. Water, Air, and Soil Pollution 55:377-382. 9 Driscoll, C.T., G.B. Lawrence, AJ. Bulger, T.J. Butler, C.S. Cronan, C. Eagar, K.F. Lambert, 10 G.E. Likens, J.L. Stoddard, and K.C. Weathers. 2001. Acidic deposition in the 11 northeastern United States: sources and inputs, ecosystem effects, and management 12 strategies. BioScience 57:180-198. 13 Driscoll, C.T., R.M. Newton, C.P. Gubala, J.P. Baker, and S.W. Christensen. 1991. Adirondack 14 mountains. Pp. 133-202 in Acidic Deposition and Aquatic Ecosystems: Regional Case 15 Studies. Edited by D.F. Charles. New York: Springer-Verlag. 16 Drohan, J.R. and W.E.Sharpe, 1997. Long-term changes in forest soil acidity in Pennsylvania, 17 U.S.A. Water, Air, and Soil Pollution 95:299-311. 18 Drohan, P.J., S.L. Stout, and G.W. Petersen. 2002. Sugar maple (Acer saccharum Marsh.) 19 decline during 1979-1989 in northern Pennsylvania. Forest Ecology and Management 20 770:1-17. 21 Duan, L., J. Hao, S. Xie, Z. Zhou, and X. Ye. 2002. Determining weathering rates of soils in 22 China. Geoderma. 110: 205-225. 23 Dupont, J., T.A. Clair, C. Gagnon, D.S. Jeffries, J.S. Kahl, SJ. Nelson, and J.M. Peckenham. 24 2005. Estimation of critical loads of acidity for lakes in northeastern United States and 25 eastern Canada. Environmental Monitoring and Assessment 109: 275-291. 26 Eilers, J.M., G.E. Glass, K.E. Webster, and J.A. Rogalla. 1983. Hydrologic control of lake 27 susceptibility to acidification. Canadian Journal of Fish and Aquatic Sciences 40: 1896- 28 1904. March 2010 85 Draft - Do Not Quote or Cite ------- Appendix B 1 Fenn, M.E., Huntington, T.G. McLaughlin, S.B., Eager, C., Gomez, A., and Cook, R.B. , 2006. Status of 2 soil acidification in North America. Journal of Forest Science 52 (special issue):3-\3 3 Fincher, J., J.R. Cumming, R.G. Alscher, G. Rubin, and L. Weinstein. 1989. Long-term ozone 4 exposure affects winter hardiness of red spruce (Picea rubens Sarg.) seedlings. New 5 Phytologist 113:85-96.Fennemaris 1946 6 Gbondo-Tugbawa, S. S., C.T. Driscoll, MJ. Mitchell, J.D. Aber, and G.E. Likens. 2002. A 7 model to simulate the response of a northern hardwood forest ecosystem to changes in S 8 deposition, Ecol. Appl, 12: 8- 23. 9 Gbondo-Tugbawa, S. S., C. T. Driscoll, J. D. Aber, and G. E. Likens. 2001. Evaluation of an 10 integrated biogeochemical model (PnET-BGC) at a northern hardwood forest ecosystem. 11 Water Resources Research 37:1057-1070. 12 Hartman, M.D., Baron, J.S., Clow, D.W., Creed, IF., Driscoll, C.T., Ewing, H.A., Haines, B.D., 13 Knoepp, J., Lajtha, K., Ojima, D.S., Parton, W.J., Renfro, J., Robinson, R.B., Van 14 Miegroet, H., Weathers, K.C., and Williams, M.W., 2009, DayCent-Chem simulations of 15 ecological and biogeochemical processes of eight mountain ecosystems in the United 16 States: U.S. Geological Survey Scientific Investigations Report 2009-5150, 174 p. 17 Hartman, M.D., J.S. Baron, and D.S. Ojima. 2007. Application of a coupled ecosystem-chemical 18 equilibrium model, DayCent-Chem, to stream and soil chemistry in a Rocky Mountain 19 watershed. Ecological Mode ling 200: 493-510. 20 Helsel, D.R., and R.M. Hirsch. 1992. Statistical Methods in Water Resources. Studies in 21 Environmental Science. New York: Elsevier Science. 22 Henriksen, A., PJ. Dillon and J. Aherne. 2002. Critical loads of acidity to surface waters in 23 south-central Ontario, Canada: Regional application of the steady-state water chemistry 24 model. Can. J. Fish. Aquat. Sci. 59 : 1287-1295. 25 Henriksen, A., and M. Posch. 2001. Steady-state models for calculating critical loads of acidity 26 for surface waters. Water, Air, and Soil Pollution: Focus 7:375-398. 27 Henriksen, A., J. Kamari, M. Posch, and A. Wilander. 1992. Critical loads of acidity: Nordic 28 surface waters. Ambio 27:356-363. March 2010 86 Draft - Do Not Quote or Cite ------- Appendix B 1 Henriksen, A. 1984. Changes in base cation concentrations due to freshwater acidification. Verh. 2 Int. Verein. Limn. 22: 692-698. 3 Hodson, M.E., and SJ. Langan. 1999a. The influence of soil age on calculated mineral 4 weathering rates. Applied Geochemistry 14: 387-394. 5 Hodson, M.E. and SJ. Langan. 1999b. Considerations of uncertainty in setting critical loads of 6 acidity of soils: the role of weathering rate determination. Environmental Pollution. 106: 7 73-81. 8 Hodson, M.E., SJ. Langan, and MJ. Wilson. 1997. A critical evaluation of the use of the 9 PROFILE model in calculation mineral weathering rates. Water, Air and Soil Pollution. 10 98: 79-104. 11 Hornberger, G.M., K J. Beven, and P.P. Germann. 1990. Inferences about solute transport in 12 macroporous forest soils from time series models. Geoderma 46: 249-262. 13 Hornberger, G.M., B J. Cosby, and R.F. Wright. 1989. Historical reconstructions and future 14 forecasts of regional surface water acidification in southernmost Norway. Water 15 Resources Research 25: 2009-2018. 16 Horsley, S.G>, R.P. Long, S.W. Bailey, R.A. Hallett, and T.J. Hall . 2000. Facotrs associated with the 17 decline disease of sugar maple on the Allegheny Plateau. Canadian Journal of Forest Research 30:1365- 18 1378. 19 Jonsson, C., P. Warfvinge, and H. Sverdrup. 1995. Uncertainty in predicting weathering rate and 20 environmental stress factors with the PROFILE model. Water, Air and Soil Pollution. 81: 21 1-23. 22 Juice, S.M., T J. Fahey, T.G. Siccama, C.T. Driscoll, E.g. Denny, C. Eagar, N.L. Cleavitt, R. 23 Minocha, and A.D. Richardson. 2006. Response of sugar maple to calcium addition to 24 northern hardwood forest. Ecology 87:1267-1280. 25 Killam, K. 1994, Soil Ecology. Cambridge, UK: Cambridge University Press. 26 Kolka, R.K., D.F. Grigal, and E.A. Nater. 1996. Forest soil mineral weathering rates: use of 27 multiple approaches. Geoderma. 73: 1-21. March 2010 87 Draft - Do Not Quote or Cite ------- Appendix B 1 Kretser, W., J. Gallagher, and J. Nicolette. 1989. Adirondack Lakes Study, 1984-1987: An 2 Evaluation of Fish Communities and Water Chemistry. Ray Brook, NY: Adirondack 3 Lakes Survey Corporation. Langan, S., Hodson, M., Kennedy, F., Hornung, M., 4 Reynolds, B., Hall, J. & Donald, L 2001. The role of weathering rate determinations in 5 generating uncertainties in the calculation of critical loads of acidity and their 6 exceedance. Water, Air and Soil Pollution: Focus 1: 299-312. 7 Langan, S.J., M. Hodson, D.C Bain, M. Hornung, B. Reynolds, J. Hall, and L. Johnston. 2001. 8 The role of mineral weathering rate determinations in generating uncertainties in the 9 calculation of critical loads of acidity and their exceedance. Water, Air and Soil 10 Pollution: Focus 1, 299-312. 11 Langan, S.J., B. Reynolds, and D.C. Bain. 1996. The calculation of base cation release from 12 mineral weathering in soils derived from Paleozoic greywackes and shales in upland UK. 13 Geoderma. 69: 275-285. 14 Langan, S.J., M.E. Hodson, D.C. Bain, R.A. Skeffington, and MJ. Wilson. 1995. A preliminary 15 review of weathering rates in relation to their method of calculation for acid sensitive soil 16 parent materials. Water, Air and Soil Pollution. 85: 1075-1081. 17 Li, H. and S.G. McNulty. 2007. Uncertainty analysis on simple mass balance model to calculate 18 critical loads for soil acidity. Environmental Pollution 149:315-326. 19 Lien, L., G.G. Raddum, and A. Fjellheim. 1992. Critical Loads of Acidity to Freshwater: Fish 20 and Invertebrates. The Environmental Tolerance Levels Programme. Rep. No. 23/1992. 21 Norwegian Ministry of Environment, Oslo, Norway. 22 Lynch, J. Personal communication. 2009. Communication between Jason Lynch (U.S. 23 Environmental Protection Agency, Clean Air Market Division, USA) and Jennifer Phelan 24 (RTI International, USA) during the Critical Load Workshop, Fall 2009 National 25 Atmospheric Deposition Program Meeting. October 5 and 6, 2009. Saratoga Springs, 26 New York, USA. March 2010 88 Draft - Do Not Quote or Cite ------- Appendix B 1 Matuszek, I.E., and G.L. Beggs. 1988. Fish species richness in relation to lake area, pH, and 2 other abiotic factors in Ontario lakes. Canadian Journal of Fisheries and Aquatic 3 Sciences 45:1931-1941. 4 McKenzie, NJ. and PJ. Ryan. 1999. Spatial prediction of soil properties using environmental 5 correlation. Geoderma 89, 67-94. 6 McNulty, S.G., B.C. Cohen, H. Li, and J.A. Moore-Myers. 2007. Estimates of critical acid loads 7 and exceedances for forest soils across the conterminous United States. Environmental 8 Pollution 149:281-292. 9 Miller, H.G. 1988. Long-term effects of application of nitrogen fertilizers on forest sites. Pp. 07- 10 106 in Forest Site Evaluation and Long-term Productivity. Edited by D.W. Cole and S.P. 11 Gessel. Seattle, WA: University of Washington Press. 12 Miller, E. 2001. Estimating soil weathering rates. Appendix 3 in NEG/ECP Forest Mapping 13 Group (Conference of New England Governors and Eastern Canadian Premiers Forest 14 Mapping Group). 2001. Protocol for Assessment and Mapping of Forest Sensitivity to 15 Atmospheric S and N Deposition: Acid Rain Action Plan - Action Item 4: Forest 16 Mapping Research Project. Prepared by NEG/ECP Forest Mapping Group. Available at 17 http://www.nrs.fs.fed.us/clean_air_water/clean_water/critical_loads/local- 18 resources/docs/NEGECP_F orest_Sensitivity_Protocol_5_21_04.pdf. 19 Nanus, L., M.W. Williams, D.H. Campbell, K.A. Tonnessen, T. Blett, and D.W. Clow. 2009. 20 Assessment of lake sensitivity to acidic deposition in national parks of the Rocky 21 Mountains. Ecological Applications 19(4): 961-973. 22 NEG/ECP Forest Mapping Group (Conference of New England Governors and Eastern 23 Canadian Premiers Forest Mapping Group). 2001. Protocol for Assessment and Mapping 24 of Forest Sensitivity to Atmospheric S and N Deposition: Acid Rain Action Plan - Action 25 Item 4: Forest Mapping Research Project. Prepared by NEG/ECP Forest Mapping 26 Group. Available at 27 http:www.nrs.fs.fed.us/clean_air_water/clean_water/critical_loads/local- 28 resources/docs/NEGECP_Forest_Sensitivotu_Protocol_5 _21_04.pdf. March 2010 89 Draft - Do Not Quote or Cite ------- Appendix B 1 Olsson, M., K. Rosen, and P.A. Melderud. 1993. Regional modeling of base cation losses from 2 Swedish forest soils due to whole-tree harvesting. Applied Geochemistry. Suppl. Issue 3 No. 2: 189-194. 4 Ouimet, R. Personal communication. 2009. Communication between Rock Ouiment (Ministere 5 des Resources naturelles et de la Faune, Canada) and Jennifer Phelan (RTI International, 6 USA) by email. December 2009. 7 Ouimet, R. 2008. Using compositional change within soil profiles for modeling base cation 8 transport and chemical weathering. Geoderma. 145: 410-418. 9 Ouimet, R., P.A. Arp, S.A. Watmough, J. Aherne, and I. DeMerchant. 2006. Determination and 10 mapping critical loads of acidity and exceedances for upland forest soils in Eastern 11 Canada. Water, Air and Soil Pollution 772:57-66. 12 Ouimet R. and L. Duchesne. 2005. Base cation mineral weathering and total release rates from 13 soils in three calibrated forest watersheds on the Canadian Boreal Shield. Canadian 14 Journal of Soil Science. 85: 245-260. 15 Pardo, L.H., and N. Duarte. 2007. Assessment of Effects of Acidic Deposition on Forested 16 Ecosystems in Great Smoky Mountains National Park using Critical Loads for Sulfur and 17 Nitrogen. U.S. Department of Agriculture, Forest Service. Prepared for Tennessee Valley 18 Authority, S. Burlington, VT. 19 Pardo, L.H., M. Robin-Abbott, N. Duarte, E.K. Miller. 2004. Tree Chemistry Database (Version 20 1.0). General Technical Report NE-324. U.S. Department of Agriculture, Forest Service, 21 Northeastern Research Station, Newton Square, PA, 45 pp. 22 Posch, M., M. Forsius, and J. Kamari. 1993. Critical loads of sulphur and nitrogen for lakes I: 23 Model description and estimation of uncertainties. Water, Air, and Soil Pollut. 66, 173- 24 192. 25 Posch, M., J. Kamari, M. Forsius, A. Henriksen, and A. Wilander. 1997. Exceedance of critical 26 loads for lakes in Finland, Norway and Sweden: Reduction requirements for acidifying 27 nitrogen and sulfur deposition. Environmental Management 21: 291-304. March 2010 90 Draft - Do Not Quote or Cite ------- Appendix B 1 Rago, P. J., and J.G. Wiener. 1986. Does pH affect fish species richness when lake area is 2 considered? Transactions of the American Fisheries Society 775:438-447. 3 Rapp, L. and Bishop, K. 2009. Surface water acidification and critical loads: exploring the F- 4 factor. Hydrology and Earth System Sciences 13: 2191 -2201 5 Reynolds, B., and D.A. Norris. 2001. Freshwater critical loads in Wales. Water, Air, and Soil 6 Pollution: Focus 1:495-505. 7 Schindler, D.W. 1988. Effects of acid rain on freshwater ecosystems. Science 239:232-239. 8 Schreck, C.B. 1982. Stress and compensation in teleostean fishes: response to social and physical 9 factors, pp. 295-321 in Stress and Fish. Edited by A.D. Pickering. New York: Academic 10 Press. 11 Schreck, C.B. 1981. Stress and rearing of salmonids. Aquaculture 28:241-249. 12 Smith, C. Personal communication. 2009. Communication between Chris Smith (U.S. 13 Department of Agriculture Natural Resources Conservation Service, USA) and Jennifer 14 Phelan (RTI International, USA) by telephone. December 2009. 15 St. Clair, S.B., I.E. Carlson, and J.P. Lynch. 2005. Evidence for oxidative stress in sugar maple 16 stands growing on acidic, nutrient imbalanced forest soils. Oecologia 7 ¥5:258-369. 17 Starr, M., AJ. Lindroos, T. Tarvainen, and H. Tanskanen. 1998. Weathering rates in the 18 Hietajarvi integrated monitoring catchment. Boreal Environment Research. 3: 275-285. 19 Stoddard, J., J.S. Kahl, F.A. Deviney, D.R. DeWalle, C.T. Driscoll, A.T. Herlihy, J.H. Kellogg; 20 P.S. Murdoch, J.R. Webb, and K.E. Webster. 2003. Response of Surface Water Chemistry 21 to the Clean Air Act Amendments of 1990. EPA 620/R-03.001. U.S. Environmental 22 Protection Agency, Office of Research and Development, National Health and 23 Environmental Effects Research Laboratory, Research Triangle Park, NC. 24 Sullivan, T.J., BJ. Cosby, J.R. Webb, R.L Dennis, AJ. Bulger, and F.A. Deviney Jr. 2008. 25 Streamwater acid-base chemistry and critical loads of atmospheric sulfur deposition in 26 Shenandoah National Park, Virginia. Environmental Monitoring and Assessment 137: 27 85-99. March 2010 91 Draft - Do Not Quote or Cite ------- Appendix B 1 Sullivan, T.J., C.T. Driscoll, BJ. Cosby, IJ. Fernandez, A.T. Herlihy, J. Zhai, R. Stemberger, 2 K.U. Snyder, J.W. Sutherland, S.A. Nierzwicki-Bauer, C.W. Boylen, T.C. McDonnell, 3 and N.A. Nowicki. 2006. Assessment of the Extent to which Intensively-Studied Lakes are 4 Representative of the Adirondack Mountain Region. Final report. New York State Energy 5 Research and Development Authority (NYSERDA), Albany, NY. Available at 6 http://nysl.nysed.gOv/uhtbin/cgisirsi/Qcwd6NzFby/NYSL/138650099/8/4298474 7 (accessed November 1, 2007). 8 Sullivan, T.J., BJ. Cosby, A.T. Herlihy, C.T. Driscoll, IJ. Fernandez, T.C. McDonnell, C.W. 9 Boylen, S.A. Nierzwicki-Bauer, and K.U. Snyder. 2007a. Assessment of the Extent to 10 Which Intensively-studied Lakes are Representative of the Adirondack Region and 11 Response to Future Changes in Acidic Deposition. Water, Air, & Soil Pollution 185: 279- 12 291. 13 Sullivan, T J., J.R. Webb, K.U. Snyder, A.T. Herlihy, and B J. Cosby. 2007b. Spatial 14 Distribution of Acid-sensitive and Acid-impacted Streams in Relation to Watershed 15 Features in the Southern Appalachian Mountains. Water, Air, & Soil Pollution 182: 57- 16 71. 17 Sullivan, T.J., BJ. Cosby, K.U. Snyder, A.T. Herlihy, B. Jackson. 2007c. Model-Based 18 Assessment of the Effects of Acidic Deposition on Sensitive Watershed Resources in the 19 National Forests of North Carolina, Tennessee, and South Carolina. Report Prepared for 20 USDA Forest Service, Asheville, NC. 21 Sullivan, T. J. Cosby, B J. Herlihy, A.T. Webb, J.R. Bulger, AJ. Snyder, K.U. Brewer, P.F. 22 Gilbert, E.H. Moore, D.L. 2004. Regional model projections of future effects of sulfur 23 and nitrogen deposition on streams in the southern Appalachian Mountains. Water 24 Resources Research 40, W02101, doi:10.1029/2003WR001998. 25 Sullivan, T. J., B. J. Cosby, J. R. Webb, K. U. Snyder, A. T. Herlihy, A. J. Bulger, E. H. Gilbert, 26 and D. Moore. 2002. Assessment of effects of acid deposition on aquatic resources in the 27 southern Appalachian Mountains, report, E&S Environ. Chem., Inc., Corvallis, OR. 28 Sullivan, TJ. 2000. Aquatic Effects of Acidic Deposition. Lewis Publishers: Washington, D.C. March 2010 92 Draft - Do Not Quote or Cite ------- Appendix B 1 Sverdrup, H. Personal communication. 2009a. Communication between Harald Sverdrup (Lund 2 University, Sweden) and Jennifer Phelan (RTI International, USA) during the Canadian 3 Council of Ministers of the Environment Acid Deposition Critical Loads: Status of 4 Methods and Indicators Workshop, March 18 and 19, 2009. Ottawa, Ontario, Canada. 5 Sverdrup, H. Personal communication. 2009b. Communication between Harald Sverdrup (Lund 6 University, Sweden) and Jennifer Phelan (RTI International, USA) by telephone. 7 December 2009. 8 Sverdrup, H. and I. Stjernquist. (editors) 2002. Managing Forest Ecosystems. Developing 9 Principles and Models for Sustainable Forestry in Sweden. Kluwer Academic Publishers, 10 The Netherlands. 321 pp. 11 Sverdrup, H., W. de Vries, and A. Henriksen. 2001. Mapping Critical Loads. A guidance to the 12 criteria, calculations data collection and mapping of critical loads. Milforapport 13 (Environmental Report) 1990: 14. Nordic Council of Ministers, Copenhagen, NORD: 14 1990: 98, 124 pp. 15 Sverdrup, H., P. Warfvinge, and T. Wickman. 1998. Estimating the weathering rate at Gardsjon 16 using different methods. In: Hultberg, H. and R. Skeffington editors. 1998. Experimental 17 Reversal of Acid Rain Effects: The Gardsjon Roof Project. John Wiley & Sons Ltd. 18 p. 232-249. 19 Sverdrup, H., W. de Vries, M. Hornung, M.S. Cresser, S.J. Langan, B. Reynolds, R. Skeffington, 20 and W. Robertson. 1995. Modification of the simple mass balance equation for 21 calculating of critical loads of acidity. In: M. Hornung, M. Sutton, and R.B. Wilson 22 (Editors), Mapping and Modeling Critical Loads for Nitrogen. A Workshop Report, 23 October 1994. Published by Institute of Terrestrial Ecology, Bush Estate, Edinburgh. 24 Sverdrup, H.U. and P. Warfvinge. 1995. Estimating field weathering rates using laboratory 25 kinetics. IN: White, A. and S. Brantley editors. Weathering Kinetics of Silicate Minerals. 26 Volume 8. Reviews in Mineralogy, Mineralogical Society of America, p. 485-541. March 2010 93 Draft - Do Not Quote or Cite ------- Appendix B 1 Sverdrup, H., De Vries, W., Hornung, M., Cresser, M.S., Langan, S.J., Reynolds, B., 2 Skeffmgton, R., and Robertson, W. 1995. Modification of the simple mass balance 3 equation for calculating of critical loads of acidity. In: M. Hornung, M. Sutton, and R.B. 4 Wilson (Editors), Mapping and Modeling Critical Loads for Nitrogen. A Workshop 5 Report, October 1994. Published by Institute of Terrestrial Ecology, Bush Estate, 6 Edinburgh. 7 Sverdrup, H. and P. Warfvinge. 1993a. Calculating field weathering rates using a mechanistic 8 geochemical model PROFILE. Applied Geochemistry. 8: 273-283. 9 Sverdrup, H., and P. Warfvinge. 1993b. The effect of soil acidification on the growth of trees, 10 grass and herbs as expressed by the (Ca+ Mg+ K)/Al ratio. Reports in Ecology and 11 Environmental Engineering 2. Lund University, Department of Chemical Engineering, 12 Lund, Sweden. 13 Sverdrup, H., P. Warfvinge, M. Rabenhorst, A. Janicki, R. Morgan, and M. Bowman. 1992. 14 Critical loads and steady-state chemistry for streams in the state of Maryland. 15 Environmental Pollution. 77: 195-203. 16 Sverdrup H.U. 1990. The Kinetics of Base Cation Release Due to Chemical Weathering. Lund 17 University Press, 246pp. 18 Sverdrup, H., W. de Vries, and A. Henriksen. 1990. Mapping Critical Loads. Miljorapport 14. 19 Nordic Council of Ministers, Copenhagen, Denmark. 20 Sverdrup, H. and P. Warfvinge. 1988. Weathering of primary silicate minerals in the natural soil 21 environment in relation to a chemical weathering model. Water, Air and Soil Pollution. 22 38: 387-408. 23 UNECE (United Nations Economic Commission for Europe). 2004. Manual on Methodologies 24 and Criteria for Modeling and Mapping Critical Loads and Levels and Air Pollution 25 Effects, Risks, and Trends. Convention on Long-Range Transboundary Air Pollution, 26 Geneva Switzerland. Available at http://www.icpmapping.org (accessed August 16, 27 2006). March 2010 94 Draft - Do Not Quote or Cite ------- Appendix B 1 U.S. EPA, 2008. Integrated Science Assessment (ISA) for Oxides of Nitrogen and Sulfur- 2 Ecological Criteria (Final Report) (ISA). U. S. Environmental Protection Agency, Office 3 of Research and Development, National Center for Environmental Assessment, Research 4 Triangle Park, NC. EPA/600/R-08/082. 5 U. S. EPA, 2009. Risk and Exposure Assessment for Review of the Secondary National Ambient 6 Air Quality Standards for Oxides of Nitrogen and Oxides of Sulfur. Final. U.S. 7 Environmental Protection Agency, Office of Air Quality Planning and Standards, 8 Research Triangle Park, NC. EPA-452/R-09-008b. 9 Van Sickle, 1, J.P. Baker, H.A. Simonin, B.P. Baldigo, W.A. Kretser, and W.E. Sharpe. 1996. 10 Episodic acidification of small streams in the northeastern United States: Fish mortality 11 in field bioassays. Ecological Applications 6:408-421. 12 Velbel, M.A. and J.R. Price. 2007. Solute geochemical mass-balances and mineral weathering 13 rates in small watersheds: Methodology, recent advances, and future directions. Applied 14 Geochemistry 22: 1682-1700. 15 Warfvinge, P. and H. Sverdrup. 1992. Calculating critical loads of acid deposition with 16 PROFILE, a steady-state soil chemistry model. Water, Air and Soil Pollution. 63: 119- 17 143. 18 Watmough S.A., J. Aherne, and P. J. Billion. 2004. Critical Loads Ontario: Relating Exceedance of the 19 Critical Load with Biological Effects at Ontario Forests. Report 2. Environmental and Resource 20 Studies, Trent University, ON, Canada. 21 Watmough, S.A., J. Ahern, and PJ. Dillon. 2005. Effect of declining base cation concentrations 22 on freshwater critical load calculations. Environmental Science & Technology. 39: 3255- 23 3260. 24 Watmough, S., J. Aherne, P. Arp, I. DeMerchant, and R. Ouimet. 2006. Canadian experiences in 25 development of critical loads for sulphur and nitrogen. Pp. 33-38 in Monitoring Science 26 and Technology Symposium: Unifying Knowledge for Sustainability in the Western 27 Hemisphere Proceedings RMRS-P-42CD. Edited by C. Aguirre-Bravo, PJ. Pellicane, 28 D.P. Burns, and S. Draggan. U.S. Department of Agriculture, Forest Service, Rocky 29 Mountain Research Station, Fort Collins, CO. March 2010 95 Draft - Do Not Quote or Cite ------- Appendix B 1 Webb, J.R., Deviney, F. A., Galloway, J. N., Rinehart, C. A., Thompson, P. A., & Wilson, S. 2 (1994). The acid-base status of native brook trout streams in the mountains of Virginia; a 3 regional assessment based on the Virginia trout stream sensitivity study. Charlottesville, 4 VA: University of Virginia. 5 Webster K.L., I.F. Creed, N.S. Nicholas, and H.V. Miegroet. 2004. Exploring interactions between 6 pollutant emissions and climatic variability in growth of red spruce in the Great Smoky 7 Mountains National Park. Water, Air, and Soil Pollution 759:225-248. 8 Wedemeyer, G.A., B.A. Barton, and DJ. MeLeay. 1990. Stress and acclimation, pp. 178-198 in 9 Methods for Fish Biology. Edited by C.B. Schreck and P.B. Moyle. Bethesda, MD: 10 American Fisheries Society. 11 Whitfield, C.J., S.A. Watmough, J. Aherne, and PJ. Dillon. 2006. A comparison of weathering 12 rates for acid-sensitive catchments in Nova Scotia, Canada and their impact on critical 13 load calculations. Geoderma. 136: 899-911. 14 Yin, X. and P.A. Arp. 1993. Predicting forest soil temperatures from monthly mean air 15 temperature and precipitation records. Canadian Journal of Forest Research, 23: 2521- 16 2536. 17 Zhai, J., C. T. Driscoll, T. J. Sullivan, and B. J. Cosby. 2008. Regional application of the PnET- 18 BGC model to assess historical acidification of Adirondack lakes. Water Resources 19 Research 44, W01421, doi:10.1029/2006WR005532. 20 March 2010 96 Draft - Do Not Quote or Cite ------- Appendix B i APPENDIX 1 2 Potentially Applicable National-Scale Geochemical Data 3 Currently, there are potentially three consistent national-scale data sets that are most 4 appropriate for use in this project: the Shacklette data, the more recent National Geochemical 5 Survey data, and the NRCS pedon soil pit (i.e., LIMS database). 6 Chemical Analyses of Soils and other Surficial Materials of the Conterminous United 7 States (Shacklette Data) & the Geochemical Landscapes Project 8 These data provide an ultra-low-density geochemical baseline for soils and other surficial 9 materials in the conterminous United States. It is the most widely cited reference for 10 geochemical background data and the data are most appropriately used to provide information on 11 background concentrations of elements in soil for areas represented by small map scales. 12 The data set contains geochemical data from soils and other regolith collected and 13 analyzed by Hans Shacklette and colleagues beginning in 1958 and continuing until about 1976. 14 Originally compiled as a paper record, the data was later included as part of the original USGS 15 PLUTO database. Approximately 1,323 samples were collected through 1976. The 1,323 sample 16 locations that comprise the Shacklette data represent a sampling density of approximately 1 17 sample per 6,000 square kilometers (metadata); equivalent to the collection of samples on a 75- 18 km grid across the country. 19 The sampling protocol called for removal of loose organic debris from the surface and 20 then collection of soil from a depth of 0-20 cm (Smith et al., 2005). Where possible, sample 21 locations were selected where surficial materials had been altered very little from their natural 22 condition as evidenced by the presence of native plants. The sample material at most sites could 23 be termed "soil" because it was a mixture of disintegrated rock and organic matter. Some of the 24 sampled deposits, however, were not soils as defined above, but were other regolith types. These 25 included desert sands, sand dunes, some loess deposits, and beach and alluvial deposits that 26 contained little or no visible organic material. 27 This national-lev el geochemical data set of 1,323 samples has been collected and 28 analyzed according to standardized protocols. This is considered one of the principal strengths of 29 the data set overall. The samples were chemically analyzed by various but compatible techniques 30 in the U.S. Geological Survey laboratories in Denver, CO. Geochemical point-symbol maps were 31 plotted for 40 elemental results and published as USGS. Professional Paper 1270 (Shacklette and 32 Boerngen, 1984). The original elements analyzed included: Ag, Al, Ba, Be, B, Ca, Ce, Cr, Co, 33 Cu, Ga, Ge, Hg, Fe, La, Li, Pb, Mg, Mn, Mo, Na, Nd, Ni, Nb, P, K, Rb, S, Sc, Se, Sr, Th, Ti, U, 34 V, Yb, Y, Zn, Zr, and total carbon. A newer set of national-level interpolated maps displaying 35 the geochemical distribution for 22 elements using the Shacklette data has since been published 36 (Gustavsson, et al, 2001). Using weighted-median and Bootstrap procedures for interpolation and March 2010 97 Draft - Do Not Quote or Cite ------- Appendix B 1 smoothing, full-color maps were produced for seven major elements (Al, Ca, Fe, K, Mg, Na, and 2 Ti) and 15 trace elements (As, Ba, Cr, Cu, Hg, Li, Mn, Ni, Pb, Se, Sr, V, Y, Zn, and Zr). 3 The major drawback with the data set is its extremely low numbers of samples for the 4 entire conterminous United States. However, more recent high-resolution studies (e.g., Smith et 5 al., 2005) have illustrated that the regional patterns established by the Shacklette data are 6 generally maintained except where areas have been affected by anthropogenic factors (Smith, 7 2006). 8 Efforts are also on-going to build upon the Shacklette data by increasing the density of 9 the sample locations and producing a high resolution geochemical data set for North America. 10 Also referred to as the Geochemical Landscapes Project, this is a collaborative effort by the 11 USGS, USDA Natural Resource Conservation Service, other federal agencies, and academia to 12 build a national-scale soil geochemical survey that will eventually increase the sample density of 13 the Shacklette data set. The Geochemical Landscapes project began in October 2002 in 14 collaboration with partners in Canada (Geological Survey of Canada; Agriculture and Agri-Food 15 Canada) and Mexico (Consejo de Recursos Minerales/Servicio Geologico de Mexico; Institute 16 Nacional de Estadistica Geografia e Informatica) that has as its long-term goal a soil 17 geochemical survey of North America (Smith et al., 2005). A 3-year pilot project was completed 18 n 2004. During the pilot project soil samples were collected for major- and trace-elements from 19 265 soil samples collected from two continental-scale transects in North America (Smith et al., 20 2005). The project has just completed a third year of continental-sampling and completed sample 21 collection for approximately 60% of the conterminous United States (D. Smith personal 22 communication, 2009). The state areas that have been completed to date are: ME, NH, VT, CT, 23 RI, MA, NY, MO, AR, MS, LA, NV, UT, CO, WY, KS, NJ, MD, WV, DE, NE, FL, SC, GA, 24 AL, OK, NM, MT, ID, MN, and SD. The USGS anticipates that sampling may be completed for 25 the conterminous US in 2010; or 2011 at the latest. However, funding doesn't allow for analyses 26 to be completed for a number of samples and several hundred grams of each sample is being 27 archived for on-going and future analysis. 28 National Geochemical Survey (NGS) 29 Efforts are on-going by the USGS to produce a new stream-sediment-based geochemical 30 survey for the United States at a nominal spacing of 17 by 17 kilometers (i.e., minimum sample 31 density of 1 sample per 289 km2 in all land areas of the country). Project mapping shows that the 32 work is either complete or nearly completed. Unlike other national geochemical data collection 33 efforts, the analytical routines and standards will be consistent throughout the survey. Analytical 34 methods include a 40-element ICP package plus single-element determinations of As, Se, and Hg 35 by atomic absorption for every sample. 36 The project has sought to capitalize on existing datasets and also achieved samples. For 37 this reason the NGS is based primarily on analyses of stream sediments to build on the massive March 2010 98 Draft - Do Not Quote or Cite ------- Appendix B 1 achieves of data and samples from DOE's National Uranium Evaluation (NURE) program. 2 Much of the survey has entailed reanalysis of approximately 35,000 archival samples from the 3 NURE program. Where NURE samples do not exist, USGS has been working with cooperators 4 to obtain new samples. The project website reports a total of about 50,000 stream-sediment 5 samples that have been analyzed for 42 elements, including arsenic, selenium, and mercury. Last 6 reported during 2004, only about 10,000 more samples needed to be collected and analyzed to 7 complete the national survey. Samples are generally categorized as follows: 8 > Inherited Data: Much of the RASS and PLUTO data were inherited into the NGS; 9 > Independent Reanalyses of NURE samples: These sample were reanalyzed by USGS 10 projects other than the NGS. Prior to the NGS, numerous USGS projects reanalyzed 11 samples from the NURE archives. Other USGS projects have continued to reanalyze 12 NURE samples in parallel with the NGS. In the majority of these cases, most or all of the 13 NURE samples in an area were reanalyzed. 14 > NURE-Systematic. Systematic reanalyses of NURE samples done by the NGS. An 15 archive of stream sediment and soil samples collected by the NURE program is stored at 16 the USGS in Denver, Colo. Rules were established to select a subset of samples for 17 reanalysis that maintains the NGS coverage. 18 > NURE-Targeted. Targeted reanalyses of NURE samples done by the NGS for various 19 reasons. 20 > USGS-Re sampling. Reanalyses of USGS archived project samples done by the NGS. The 21 archive includes most of the samples for which there are analytical data in the National 22 Geochemical Database, including those collected by USGS programs. 23 > Collaborative Sampling with State Programs. Collaborative sampling programs by the 24 USGS and states. 25 Digital data files are presented in 6 categories. In total there are 43 individual data files 26 for the United States. Some of the data has also been processed into vector data to produce maps 27 showing the elemental concentration of As, Se, Hg, Pb, Zn, Cu, Al, Na, Mg, P, Ca, Ti, Mn, and 28 Fe at the county level. 29 USDA NRCS Soil Pedon Pit Data 30 The USDA NRCS measures soil geochemical characteristics along with performing 31 quantitative and bulk mineralogy tests and other physiochemical measurements for soil series 32 delineated across the United States. This data set and associated detail was discovered through 33 communication with NRCS staff (C. Smith personal communication, 2009; T. Reinsch personal 34 communication, 2009). Most of the geochemical and mineralogy data is associated with 35 individual soil pedons. The NRCS defines a pedon as the smallest unit that can be called a soil. It 36 is a three-dimensional sample that extends from the soil surface to the deepest roots or genetic 37 soil horizons. The area covered by a pedon varies from 10-100 square feet, depending on 38 changes in soil properties. Pits are dug to expose the pedons and the NRCS generally refers to March 2010 99 Draft - Do Not Quote or Cite ------- Appendix B 1 the data associated with the pedons as soil pit data. There are currently approximately 30,000 soil 2 pits/pedons in the NRCS database. 3 Groups of pedons with very similar characteristics that are closely associated in the 4 landscape are called polypedons. Polypedons that have a common set of characteristics that fall 5 within a particular range are delineated as a basic soil unit referred to as a soil series which have 6 been identified as the basic unit of the proposed data framework, as previously discussed. The 7 same soil series delineations can occur in different and distant areas (i.e., across county areas, 8 states, or regions). A variety of data are used to define a soil (e.g., geomorphic position in the 9 landscape, relationship to the water-table, supported flora, geology, number and type of horizons, 10 sediment texture, sediment color variations, etc.), and therefore geochemical and mineralogy 11 data has not been collected from every soil pedon associated with an individual series of the 12 same name since associations can be made based on a number of these other related 13 characteristics. However, geochemical and quantitative mineralogy data has been measured for a 14 significant number of pedons and soil series locations across the country. 15 Since soils of the same series name possess enough similarities to be classified as similar 16 soils it is thought that the geochemistry data can also be extrapolated to pedons of like soil series 17 (C. Smith personal communication, 2009). Assigning mineral phases to the soil series that do not 18 have either geochemical or mineralogy data associated with their pedons will require 19 professional judgment by researchers familiar with the soil pit data and soil taxonomy to make 20 geochemical data extrapolations with a degree of confidence. In these cases the characteristics of 21 surrounding soils would be used to extrapolate geochemistry or mineralogy, or another data set 22 could be used to aid in the characterization. GIS tools would be used to help automate these 23 determinations where necessary. NRCS staff would aid RTI in determining rules and developing 24 database relationship tables that could be used in automating any extrapolation of this data. The 25 NRCS would also aid RTI in evaluating the reasonableness of the results. 26 Since soil series are delineated across the conterminous United States the pit data could 27 potentially provide a complete geochemical and mineralogy data layer for determining 28 mineralogy. Since mineralogy is already associated with the geochemical data a more accurate 29 assignment of mineral modes may be possible using this data set. Laboratory analysis includes 30 the major geochemical elements: Al, Ca, Fe, K, Mg, Mn, Na, P, Si, Sr, Ti, and Zr. In addition, x- 31 ray diffraction is used to indentify clay mineralogy generally for each horizon of a pedon, and 32 optical mineralogy analysis is performed on the dominant sand fractions of the soil from the A- 33 horizon, B-horizon, and C-horizon, or the most dominant horizon. More than 60 fields describing 34 the minerals are listed in the database. The dataset is not uniform in that elemental analyses were 35 routinely done through the 1970's but then these analyses were suspended through the 1980's. 36 Elemental analyses were resumed during the early 1990's. It is estimated that as much as one 37 third of the 30,000 soil pedons have geochemical data. Likewise, optical mineralogy is not 38 performed for all pedons and the NRCS staff estimate that approximately as many as one third of 39 the 30,000 soil pits have optical analysis results. Even though the number of pedons with data are March 2010 100 Draft - Do Not Quote or Cite ------- Appendix B 1 similar for geochemical and optical analysis results the data is not necessarily associated with the 2 same set of pedons or even soil series. 3 References 4 Smith, Chris, 2009. Personal communication between Chris Smith, (affiliation) and Scott 5 Guthrie, RTI International, 12/16/09. 6 Reinsch, Thomas, 2009. Personal communication between Thomas Reinsch, (affiliation) and 7 Scott Guthrie, RTI International, 12/18/09. 8 9 March 2010 101 Draft - Do Not Quote or Cite ------- Appendix B i APPENDIX 2 2 References for Table 3-2: Applications of the MAGIC Model 3 Aherne, J, PJ. Dillon, and BJ. Cosby. 2003. Acidification and recovery of aquatic ecosystems in 4 south central Ontario, Canada: regional application of the MAGIC model. Hydro!. Earth 5 Syst. Sci 7: 561-573. 6 Bernett, J.A., J.M. Eilers, and BJ. Cosby. 1997. Overview, Libby Lake Modeling Workshop. 7 E&S Environmental Chemistry, Inc., Corvallis, OR. 8 Bulger A.J., BJ. Cosby, and J.R. Webb. 2000. Current, reconstructed past, and projected future 9 status of brook trout (Salvelinus fontinalis) streams in Virginia. Canadian Journal of Fish 10 and Aquatic Science 57:1515-1523. 11 Bulger, A.J., C.A. Dolloff, B J. Cosby, K.N. Eshleman, J.R. Webb, and J.N. Galloway. 1995. 12 Sensitivity of Blacknose Dace (Rhinichthys Atratulus) to Moderate Acidification Events 13 in Shenandoah National Park, U.S.A. Water, Air, & Soil Pollution 753(1-4): 125-134. 14 Church, M. R. and J. Van Sickle. 1999. Potential relative future effects of sulfur and nitrogen 15 deposition on lake chemistry in the Adirondack Mountains, United States. Water 16 Resource. Res. 35:2199-2211. 17 Cosby, B J. and TJ. Sullivan. 2001. Quantification of dose-response relationships and critical 18 loads of sulfur and nitrogen for six headwater catchments in Rocky Mountain, Grand 19 Teton, Sequoia, and Mount Rainer national parks. E&S Report 97-15-01. 20 Dennis, IF., T.A. Clair, and B J. Cosby. 2005. Testing the MAGIC acid rain model in highly 21 organic, low-conductivity waters using multiple calibrations. Environmental Modeling 22 and Assessment 70(4): 303-314. 23 Eilers J.M., B J. Cosby, J.A. Bernet, and T.A. Sullivan. 1998. Analysis of the Response of Shasta 24 Lake, Idaho to Increases in Atmospheric Sulfur and Nitrogen Using the MAGIC 25 MODEL. E&S Environmental Chemistry, Inc., Corvallis, OR. March 2010 102 Draft - Do Not Quote or Cite ------- Appendix B 1 Ellis, H., and M. Bowman. 1994. Critical Loads and Development of Acid Rain Control Options. 2 Journal of Environmental Engineering 120(2): 273-289 3 Sinha, R., MJ. Small, P.P. Ryan, T.J. Sullivan, and BJ. Cosby. 1998. Reduced-Form Modelling 4 of Surface Water and Soil Chemistry for the Tracking and Analysis Framework. Water, 5 Air, & Soil Pollution 705(3-4): 617-642. 6 Sullivan, T.J. and B.J. Cosby. 2004. Aquatic Critical Load Development for the Monongahela 7 National Forest, West Virginia. Report prepared for the USDA Forest Service 8 Monongahela National Forest. 9 Sullivan, T.J. and B.J. Cosby. 2002. Critical Loads of Sulfur Deposition to Protect Streams 10 within Joyce Kilmer and Shining Rock Wilderness Areas from Future Acidification. 11 Report for the USDA Forest Service. 12 Sullivan, T. J., and J.M. Eilers. 1996. Assessment of Deposition Levels of sulfur and Nitrogen 13 Required to Protect Aquatic Resources in Selected Sensitive Regions of North America. 14 E and S Environmental Chemistry, EPA/600/R-96-123, Corvallis Environmental 15 Research Lab, OR. 16 Sullivan, T.J., B.J. Cosby, K.A. Tonnessen, and D.W. Clow. 2005. Surface water acidification 17 responses and critical loads of sulfur and nitrogen deposition in Loch Vale watershed, 18 Colorado. Water Resources Research 41: WO 1021. 19 Sullivan, T.J., B.J. Cosby, A.T. Herlihy, J.R. Webb, A.J. Bulger, K.U. Snyder, P.F. Brewer, E.H. 20 Gilbert, and D.L. Moore. 2004. Regional model projections of future effects of sulfur and 21 nitrogen deposition on streams in the southern Appalachian Mountains. Water Resour. 22 Res. 40: W02101. 23 Sullivan, T.J., B.J. Cosby, J.A. Bernert, and J.M. Eilers. 1998. Model Evaluation of 24 dose/response relationships and critical loads for nitrogen and sulfur deposition to the 25 watersheds of lower saddlebag and white dome lakes. 26 March 2010 103 Draft - Do Not Quote or Cite ------- United States Office of Air Quality Planning and Standards Publication No. EPA-452/P-10-006 Environmental Protection Health and Environmental Impacts Division March, 2010 Agency Research Triangle Park, NC ------- |