^
5
3J
V
LU
o
Model-Based Analysis
And Tracking Of
Airborne Mercury Emissions
To Assist in Watershed Planning
I I I | I I I I I I I II | I I I I I I I I I | I I I I I I I I I | I I I I I I I I I | I I I-M
-------
This page deliberately left blank.
-------
Model-Based Analysis
And Tracking Of
Airborne Mercury Emissions
To Assist in Watershed Planning
August 2008
Watershed Branch (4503-T)
Office of Wetlands, Oceans, and Watersheds
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, NW
Washington, DC 20460
Document posted at: http://www.epa.qov/owow/tmdl/techsupp.html
-------
Acknowledgments
This document was prepared for the U.S. Environmental Protection Agency's Office of
Wetlands, Oceans, and Watersheds under EPA Contracts EP-W-08-018 and 68-W-03-
028.
Dwight Atkinson and Ruth Chemerys, the EPA project managers, would like to extend
special thanks to Tom Myers, YiHua Wei, and Sharon Douglas of ICF International, San
Rafael, CA, for their excellent work in preparing this document and conducting the
supporting modeling and technical analyses.
-------
Table of Contents
Executive Summary ES-1
Purpose ES-1
Modeling Protocols for TMDL Applications ES-1
Overview of the Deposition Modeling Tools ES-1
Particle and Precursor Tagging Methodology ES-2
REMSAD/PPTM Application Procedures ES-3
Overview of Model Performance ES-3
PPTM Results ES-4
Comparison of REMSAD PPTM, CMAQ PPTM, and Other Source Attribution Techniques
ES-7
Summary of Key Findings ES-8
1. Introduction 1-1
1.1. Background and Objectives 1-1
1.2. Overview of the Air Deposition Modeling Tools 1-2
1.3. Overview of the Modeling Approach 1-4
1.4. Report Contents 1-6
2. Description of the Deposition Modeling Tools 2-1
2.1. REMSAD Modeling System 2-1
2.1.1. Input File Requirements 2-2
2.12. Carbon-Bond V Chemical Mechanism 2-3
2.13. Mercury Chemistry 2-7
2.14 Mercury Re-Emission Treatment 2-8
2.15. Wet Deposition 2-10
2.1.6. Dry Deposition 2-13
2.1.7. Particle and Precursor Tagging Methodology (PPTM) for Mercury 2-16
2.1.8. Summary of Outputs and Information Provided by REMSAD 2-17
2.1.9. REMSAD/PPTM Application Procedures 2-17
2.2. CMAQ Modeling System 2-18
2.3. CTM, GRAHM and GEOS-Chem Models 2-19
3. Meteorological and Geographical Inputs 3-1
3.1. Meteorological Inputs 3-1
3.11 Description of the Meteorological Inputs 3-1
3.2. Preparation of Geographical Inputs 3-16
4. Emissions Inputs 4-1
4.1. Emission Inventory Preparation for Non-mercury Particulate and Gaseous Species .4-1
411 Emissions Data 4-1
4.1.2. Emissions Processing 4-2
4.1.3. Summary of the Criteria Pollutant Emissions 4-3
4.2. Emission Inventory Preparation for Mercury 4-8
42.1 Review and Revision of the CAMR Mercury Emissions Inventory 4-9
4.2.2. Mercury Emissions Processing for REMSAD 4-20
42.3. Quality Assurance of the Mercury Emission Inventory 4-20
42.4 Summary of the Mercury Emissions 4-21
5. Initial and Boundary Condition Inputs 5-1
5.1. Specification of Initial and Boundary Conditions for the PM Simulation 5-1
5.2. Specification of Initial and Boundary Conditions for the Mercury Simulations 5-1
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Table of Contents
6. REMSAD Base Simulation Results 6-1
7. REMSAD PPTM Results: Mercury Deposition Contribution Analysis 7-1
7.1. PPTM Application Procedures 7-1
7.2. Mercury PPTM Results 7-8
7.2.1 Contributions to Statewide Maximum Deposition 7-8
7.2.2. Example Analysis Using the Contribution Charts 7-10
7.2.3. Comparison with a Source Apportionment Study for Ohio 7-11
8. Summary of Key Findings and Recommendations for Future Study 8-1
8.1. Summary of Findings 8-1
8.2. Recommendations 8-2
References R-1
Appendix A: Model Performance Evaluation for Non-Mercury Species A-1
Appendix B: Preliminary Comparison of CMAQ and REMSAD Tagging Simulations B-1
Appendix C: Details of Emissions Revisions for Selected States C-1
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation Amount
D-1
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S E-1
List of Figures
Figure ES-1. Summary of Mercury Tagging Results for 2001 for Wisconsin, New York, Virginia, and Texas
(with Average Background) at the Location of the Maximum Simulated Annual Mercury Deposition from Sources
within the State ES-5
(a) Wisconsin ES-5
(b) New York ES-5
(c) Virginia ES-5
(d) Texas ES-5
Figure ES-2. Summary of Source-Specific Mercury Tagging Results for 2001 for Arkansas, Michigan, and Vermont at the Location
of the Maximum Simulated Annual Mercury Deposition from Sources within the State ES-6
(a) Arkansas ES-6
(b) Michigan ES-6
(c) Vermont ES-7
Figure 1-1. REMSAD Modeling Domain forthe Mercury PPTM Simulations 1-5
Figure 3-1a. Observed Annual Rainfall Totals (cm) for 2001 3-2
Figure 3-1 b. MM5-derived Annual Rainfall Totals (cm) for 2001 3-3
Figure 3-2. Monthly Average Rainfall amount (in) Based on Observed and Simulated Daily Precipitation Values 3-4
(a) Portland 3-4
(b) Baltimore 3-4
(c) Baton Rouge 3-5
(d) Madison 3-5
(e) Oakland 3-6
Figure 3-3. Monthly Average Observed and Simulated 850 mb Temperature (°C) forthe Time of the Morning Sounding 3-6
(a) Portland 3-6
(b) Baltimore 3-7
(c) Baton Rouge 3-7
(d) Madison 3-8
(e) Oakland 3-8
Figure 3-4. Monthly Average Observed and Simulated 850 mb Dew-Point Temperature (°C) forthe Time of the Morning Sounding 3-9
(a) Portland 3-9
(b) Baltimore 3-9
(c) Baton Rouge 3-10
(d) Madison 3-10
(e) Oakland 3-11
Figure 3-5. Monthly Average Bias and Error Statistics for 850 mb Wind Speed (ms"1) forthe Time of the Morning Sounding 3-11
ii August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Table of Contents
(a) Portland 3-11
(b) Baltimore 3-12
(c) Baton Rouge 3-12
(d) Madison 3-13
(e) Oakland 3-13
Figure 3-6. Monthly Average Bias and Error Statistics for 850 mb Wind Direction (degrees) for the Time of the Morning Sounding 3-14
(a) Portland 3-14
(b) Baltimore 3-14
(c) Baton Rouge 3-15
(d) Madison 3-15
(e) Oakland 3-16
Figure 4-1 a. Low-level NOX Emissions (tons) for the 36-km Grid fora Summer Weekday 4-5
Figure 4-1 b. Low-level SO2 Emissions (tons) for the 36-km Grid fora Summer Weekday 4-6
Figure 4-2a. Elevated Point-source NOX Emissions (tons) for the 36-km Grid fora Summer Weekday 4-7
Figure 4-2b. Elevated Point-Source SO2 Emissions (tons) for the 36-km Grid fora Summer Weekday 4-8
Figure 4-3a. Spatial Distribution of Low-Level Mercury Emissions (tons) for the 2001 REMSAD Emissions Inventory
forthe 36-km Grid fora Summer Weekday: Elemental Mercury 4-23
Figure 4-3b. Spatial Distribution of Low-Level Mercury Emissions (tons) forthe 2001 REMSAD Emissions Inventory
forthe 36-km Grid for a Summer Weekday: Divalent Gaseous Mercury 4-24
Figure 4-3c. Spatial Distribution of Low-Level Mercury Emissions (tons) for the 2001 REMSAD Emissions Inventory
forthe 36-km Grid for a Summer Weekday: Particulate Mercury 4-25
Figure 4-3d. Spatial Distribution of Low-Level Mercury Emissions (tons) forthe 2001 REMSAD Emissions Inventory
forthe 36-km Grid fora Summer Weekday: All Species 4-26
Figure 4-4a. Spatial Distribution of Elevated Point-Source Mercury Emissions (tons) forthe 2001 REMSAD Emissions Inventory
forthe 36-km Grid for a Summer Weekday: Elemental Mercury 4-27
Figure 4-4b. Spatial Distribution of Elevated Point-Source Mercury Emissions (tons) forthe 2001 REMSAD Emissions Inventory
forthe 36-km Grid for a Summer Weekday: Divalent Gaseous Mercury 4-28
Figure 4-4c. Spatial Distribution of Elevated Point-Source Mercury Emissions (tons) forthe 2001 REMSAD Emissions Inventory
forthe 36-km Grid fora Summer Weekday: Particulate Mercury 4-29
Figure 4-4d. Spatial Distribution of Elevated Point-Source Mercury Emissions (tons) forthe 2001 REMSAD Emissions Inventory
forthe 36-km Grid fora Summer Weekday: All Species 4-30
Figure 5-1. Comparison of CTM, GRAHM, and GEOS-CHEM Derived Boundary Concentrations (ppt)
forthe REMSAD Modeling Domain for HGO, HG2, and HGP: February and July 2001 5-3
Figure 6-1 a. Annual Simulated versus Observed Mercury Wet Deposition (g km"2) forthe REMSAD 12-km Grid at MDN Sites:
CTM Boundary Conditions 6-3
Figure 6-1 b. Annual Simulated versus Observed Mercury Wet Deposition (g km"2) forthe REMSAD 12-km Grid at MDN Sites:
GRAHM Boundary Conditions 6-3
Figure 6-1 c. Annual Simulated versus Observed Mercury Wet Deposition (g km"2) forthe REMSAD 12-km Grid at MDN Sites:
GEOS-CHEM Boundary Conditions 6-4
Figure 6-2a. Annual Average Simulated Mercury Concentration (ng m"3) forthe REMSAD 12-km Modeling Domain
(with Average Boundary Conditions): Elemental Mercury (HGO) 6-5
Figure 6-2b. Annual Average Simulated Mercury Concentration (ng m ) forthe REMSAD 12-km Modeling Domain
(with Average Boundary Conditions): Divalent Mercury (HG2) 6-6
Figure 6-2c. Annual Average Simulated Mercury Concentration (ng m"3) for the REMSAD 12-km Modeling Domain
(with Average Boundary Conditions): Particulate Mercury (HGP) 6-7
Figure 6-3a. Simulated Annual Mercury Deposition (g km"2) forthe REMSAD 12-km Modeling Domain
(with Average Boundary Conditions): Dry Deposition 6-8
Figure 6-3b. Simulated Annual Mercury Deposition (g km"2) forthe REMSAD 12-km Modeling Domain
(with Average Boundary Conditions): Wet Deposition 6-9
Figure 6-3c. Simulated Annual Mercury Deposition (g km"2) forthe REMSAD 12-km Modeling Domain
(with Average Boundary Conditions): Total (Dry + Wet) Deposition 6-10
Figure 7-1. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km2) for Connecticut 7-13
Figure 7-2. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km) for Maine 7-15
Figure 7-3. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Massachusetts 7-17
Figure 7-4. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for New Hampshire 7-19
Figure 7-5. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Rhode Island 7-21
Figure 7-6. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Vermont 7-23
Figure 7-7. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for New Jersey 7-25
Figure 7-8. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for New York 7-27
Figure 7-9. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Delaware 7-29
Figure 7-10. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Washington, D.C 7-31
Figure 7-11. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Maryland 7-33
Figure 7-12. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Pennsylvania 7-35
Figure 7-13. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Virginia 7-37
Figure 7-14. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for West Virginia 7-39
Figure 7-15. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Alabama 7-41
Figure 7-16. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Florida 7-43
Figure 7-17. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Georgia 7-45
Figure 7-18. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Kentucky 7-47
ill August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Table of Contents
Figure 7-19. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Mississippi 7-49
Figure 7-20. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for North Carolina 7-51
Figure 7-21. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for South Carolina 7-53
Figure 7-22. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Tennessee 7-55
Figure 7-23. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Illinois 7-57
Figure 7-24. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Indiana 7-59
Figure 7-25. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Michigan 7-61
Figure 7-26. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Minnesota 7-63
Figure 7-27. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Ohio 7-65
Figure 7-28. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Wisconsin 7-67
Figure 7-29. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Arkansas 7-69
Figure 7-30. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Louisiana 7-71
Figure 7-31. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for New Mexico 7-73
Figure 7-32. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Oklahoma 7-75
Figure 7-33. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Texas 7-77
Figure 7-34. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Iowa 7-79
Figure 7-35. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Kansas 7-81
Figure 7-36. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Missouri 7-83
Figure 7-37. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Nebraska 7-85
Figure 7-38. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Colorado 7-87
Figure 7-39. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Montana 7-89
Figure 7-40. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for North Dakota 7-91
Figure 7-41. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for South Dakota 7-93
Figure 7-42. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Utah 7-95
Figure 7-43. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Wyoming 7-97
Figure 7-44. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Arizona 7-99
Figure 7-45. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for California 7-101
Figure 7-46. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Nevada 7-103
Figure 7-47. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Idaho 7-105
Figure 7-48. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Oregon 7-107
Figure 7-49. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Washington 7-109
Figure 7-50. REMSAD PPTM Results for Steubenville, Ohio for Annual Total (Wet and Dry) Mercury Deposition (a)
and Annual Wet Mercury Deposition (b) 7-111
Figure 7-51. Location of the Steubenville, Ohio study site (indicated by the blue triangle) 7-113
Figure B-1. Summer Total Wet and Dry Deposition of Mercury Simulated by (a) CMAQ and (b) REMSAD 4
Figure B-2. Estimates of Contributions to Total of Wet and Dry Deposition at the Location of Greatest Impact
from Powerton Based on PPTM Simulations by CMAQ and REMSAD B-5
Figure B-3. Estimates of Contributions to Total of Wet and Dry Deposition at the Location of Greatest Impact
from Joliet 29 Based on PPTM Simulations by CMAQ and REMSAD B-6
Figure B-4. Estimates of Contributions to Total of Wet and Dry Deposition at the Location of Greatest Impact
from Joppa Steam Based on PPTM Simulations by CMAQ and REMSAD B-7
Figure B-5. Estimates of Contributions to Total of Wet and Dry Deposition at the Location of Greatest Impact
from Illinois Sources Based on PPTM Simulations by CMAQ and REMSAD B-8
Figure D-1a. Observed Annual Rainfall Totals (cm) for 2001 D-2
Figure D-1b. MM5-Derived Annual Rainfall Totals (cm) for 2001 D-3
Figure D-2. Comparison of Observed and Simulated Annual Rainfall Amount (in) D-7
Figure D-3a. Daily Average Rainfall Amount (in) Based on Observed and Simulated Daily Precipitation Values: Salt Lake City, UT.
D-7
Figure D-3b. Daily Average Rainfall Amount (in) Based on Observed and Simulated Daily Precipitation Values: Logan, UT D-8
Figure D-3c. Daily Average Rainfall Amount (in) Based on Observed and Simulated Daily Precipitation Values: Ogden, UT D-8
Figure D-3d. Daily Average Rainfall Amount (in) Based on Observed and Simulated Daily Precipitation Values: Vernal, UT D-9
Figure D-3e. Daily Average Rainfall Amount (in) Based on Observed and Simulated Daily Precipitation Values: Evanston, WY. ..D-9
Figure D-3f. Daily Average Rainfall Amount (in) Based on Observed and Simulated Daily Precipitation Values: Rock Springs, WY D-10
Figure D-4. REMSAD Grid Cells Corresponding to Northern Utah, Southwestern Wyoming, and Northwestern Colorado D-11
Figure D-5. Daily Average Rainfall Amount (in): Observed, Simulated (MM5), and Adjusted (Adj) D-12
Figure D-6. Summer Dry Deposition of Mercury: (a) Base Simulation; (b) Sensitivity Simulation with Reduced Rainfall D-13
Figure D-7. Summer Wet Deposition of Mercury: (a) Base Simulation; (b) Sensitivity Simulation with Reduced Rainfall D-14
Figure D-8. Summer Total Deposition of Mercury: (a) Base Simulation; (b) Sensitivity Simulation with Reduced Rainfall D-15
Figure D-9. Difference in Summer Wet Deposition of Mercury: Sensitivity Simulation minus Base Simulation D-16
Figure D-10c. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Utah D-20
Figure D-11. Alternate Location for Analysis of Deposition Contributions D-21
Figure E-1. U.S. Climate Regions for Temperature and Precipitation (Source: NCDC, 2008) E-3
Figure E-2. Mercury Deposition Analysis Locations, by Climate Region E-6
Figure E-3. Annual Mercury Wet Deposition (g km"2) for 2002-2005 for the Mesa Verde (CO99), Camp Ripley (MN23),
Great Smoky Mountains (TN11), Hammond (LA28) and Waccamaw State Park (NCOS) MDN Monitoring Sites E-9
Figure E-4a. Relative Importance of the Independent Parameter Categories for the CART Wet Deposition Analyses E-16
Figure E-4b. Relative Importance of the Independent Parameter Categories forthe CART Dry Deposition Analyses E-16
iv August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Table of Contents
List of Tables
Table ES-1. REMSAD Model Performance Statistics for Mercury Wet Deposition (g km) for the 12-km Resolution Grid
at MDN Sites: 2001 Annual Simulation Period ES-4
Table 2-1. REMSAD Input Files 2-2
Table 2-2. CB-V Reaction Set 2-3
Table 2-3. Mercury Chemical Mechanism in REMSAD, Version 8 2-8
Table 2-4. Mercury Re-Emission Coefficients 2-9
Table 2-5. Henry's Law Coefficients Used in REMSAD (at 298 K) 2-11
Table 2-6. Estimated Fractions of Sulfate Wet Scavenging Rate, by Species, as Used in REMSAD 2-12
Table 3-1. Land-Use Categories Recognized by REMSAD 3-16
Table 4-1. REMSAD Emissions Species forthe Criteria Pollutant Emissions Inventory 4-1
Table 4-2. Average Seasonal Daily Emissions (tpd) forthe U.S. Portion of the REMSAD Modeling Domain: 2001 Base Case 4-3
Table 4-3. Summary of Mercury Emissions Totals by Species for Each U.S. State and for the 48 U.S. States, Canada, and Mexico 4-21
Table 6-1 a. REMSAD Model Performance Statistics for Mercury Wet Deposition (g km"2) forthe 12-km Resolution Grid
at MDN Sites: CTM Boundary Conditions 6-1
Table 6-1 b. REMSAD Model Performance Statistics for Mercury Wet Deposition (g km"2) for the 12-km Resolution Grid
at MDN Sites: GRAHM Boundary Conditions 6-2
Table 6-1 c. REMSAD Model Performance Statistics for Mercury Wet Deposition (g km"2) forthe 12-km Resolution Grid
at MDN Sites: GEOS-CHEM Boundary Conditions 6-2
Table 7-1. Tags used for the REMSAD PPTM Application for Mercury forthe Annual 2001 Simulation Period 7-2
Table A-1. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Daily Maximum Ozone Concentration (|J,gnv3) at AQS Sites A-2
Table A-2. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Monthly Average SO2 Concentration (|0,gnv3) at AQS-STN Sites A-3
Table A-3. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Monthly Average PM2.s Concentration (|J,gnv3) at AQS-STN Sites A-3
Table A-4. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Monthly Average PM2.5 Concentration (|J,gnv3) at IMPROVE Sites A-4
Table A-5. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Monthly Average Sulfate Concentration (|0,gnv3) at CASTNet Sites A-5
Table A-6. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Monthly Average Sulfate Concentration (|0,gnv3) at IMPROVE Sites A-5
Table A-7. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Monthly Average Nitrate Concentration (|J,gnv3) at CASTNet Sites A-6
Table A-8. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Monthly Average Nitrate Concentration (|J,gnv3) at IMPROVE Sites A-7
Table A-9. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Monthly Average Ammonium Concentration (|J,gnT3) at CASTNet Sites A-7
Table A-10. Month-by-Month REMSAD Model Performance Statistics forthe 12-km Resolution Grids:
Monthly Sulfate Deposition (kg ha-1) at NADP Sites A-8
Table A-11. Month-by-Month REMSAD Model Performance Statistics forthe 12-km Resolution Grids:
Monthly Nitrate Deposition (kg ha-1) at NADP Sites A-8
Table A-12. Month-by-Month REMSAD Model Performance Statistics forthe 12-km Resolution Grids:
Monthly Ammonia Deposition (kg ha-1) at NADP Sites A-9
Table A-13. Annual REMSAD Model Performance Statistics for the 12-km Resolution Grids for the Various Observational Networks A-10
Table C-1. Summary of Emissions Revisions Made to IPM Sources in State of Illinois C-1
Table C-2. Summary of Emissions Revisions Made to Non-IPM Sources in State of Illinois C-3
Table C-3. Summary of Emissions Revisions Made to IPM Sources in State of Indiana C-4
Table C-4. Summary of Emissions Revisions Made to Non-IPM Sources in State of Indiana C-5
Table C-5. Summary of Emissions Revisions made to IPM Sources in State of Iowa C-6
Table C-6. Summary of Emissions Revisions made to non-IPM and MWI Sources in State of Iowa C-6
Table C-7. Summary of Emissions Revisions made to IPM Sources in State of New Jersey C-8
Table C-8. Summary of Emissions Revisions made to Non-IPM Sources in State of New Jersey C-9
Table C-9. Summary of Emissions Revisions Made to Non-IPM Sources in State of Nevada C-10
Table C-10. Summary of Origin of Data for Mercury Emissions and Speciation for Nevada Gold Mines C-11
Table C-11. Summary of Emissions Revisions Made to IPM Sources in State of Nevada C-16
Table C-12. Summary of Emissions Revisions Made to Non-IPM Sources in State of Utah C-16
Table C-13. Summary of Emissions Revisions Made to IPM Sources in State of Utah C-17
Table C-14. Summary of Emissions Revisions Made to MWI Sources in State of Utah C-17
Table C-15a. Point Sources Included in the Collective Sources Tag for California C-19
Table C-15b. Non-point Sources Included in the Collective Sources Tag for California C-21
Table C-16a. Point Sources Included in the Collective Sources Tag for District of Columbia C-22
Table C-16b. Non-point Sources Included in the Collective Sources Tag for District of Columbia C-22
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Table of Contents
Table C-17a. Point Sources Included in the Collective Sources Tag for Illinois C-23
Table C-17b. Non-point Sources Included in the Collective Sources Tag for Illinois C-25
Table C-18a. Point Sources Included in the Collective Sources Tag for Kansas C-26
Table C-18b. Non-point Sources Included in the Collective Sources Tag for Kansas C-26
Table C-19a. Point Sources Included in the Collective Sources Tag for Kentucky C-28
Table C-19b. Non-point Sources Included in the Collective Sources Tag for Kentucky C-29
Table C-20a. Point Sources Included in the Collective Sources Tag for Mississippi C-29
Table C-20b. Non-point Sources Included in the Collective Sources Tag for Mississippi C-30
Table C-21. Non-point Sources Included in the Collective Sources Tag for New Jersey C-30
Table C-22a. Point Sources Included in the Collective Sources Tag for Ohio C-31
Table C-22b. Non-point Sources Included in the Collective Sources Tag for Ohio C-32
Table C-23a. Point Sources Included in the Collective Sources Tag for Rhode Island C-33
Table C-23b. Non-point Sources Included in the Collective Sources Tag for Rhode Island C-34
Table C-24a. Point Sources Included in the Collective Sources Tag for South Dakota C-35
Table C-24b. Non-point Sources Included in the Collective Sources Tag for South Dakota C-35
Table C-25a. Point Sources Included in the Collective Sources Tag for Tennessee C-36
Table C-25b. Non-point Sources Included in the Collective Sources Tag for Tennessee C-37
Table C-26a. Point Sources Included in the Collective Sources Tag for Utah C-38
Table C-26b. Non-point Sources Included in the Collective Sources Tag for Utah C-38
Table C-27a. Point Sources Included in the Collective Sources Tag for Virginia C-39
Table C-27b. Non-point Sources Included in the Collective Sources Tag for Virginia C-41
Table C-28a. Point Sources Included in the Collective Sources Tag for Wyoming C-44
Table C-28b. Non-point Sources Included in the Collective Sources Tag for Wyoming C-45
Table D-1. Locations and Identifiers of Surface and Upper-Air Monitoring Sites Used for the Statistical Evaluation of MM5 for Utah D-2
Table D-2a. Summary of Annual Average Metrics and Statistical Measures for MM5 for the 2001 Annual Simulation Period:
Salt Lake City, UT (Surface) D-3
Table D-2b. Summary of Annual Average Metrics and Statistical Measures for MM5 for the 2001 Annual Simulation Period:
Logan, UT (Surface) D-4
Table D-2c. Summary of Annual Average Metrics and Statistical Measures for MM5 for the 2001 Annual Simulation Period:
Ogden, UT (Surface) D-4
Table D-2d. Summary of Annual Average Metrics and Statistical Measures for MM5 for the 2001 Annual Simulation Period:
Vernal, UT (Surface) D-4
Table D-2e. Summary of Annual Average Metrics and Statistical Measures for MM5 for the 2001 Annual Simulation Period:
Evanston, WY (Surface) D-5
Table D-2f. Summary of Annual Average Metrics and Statistical Measures for MM5 for the 2001 Annual Simulation Period:
Rock Springs, WY (Surface) D-5
Table D-3a. Summary of Annual Metrics and Statistical Measures for MM5 for the 2001 Annual Simulation Period: Salt Lake
City, UT (Upper-Air) D-6
Table D-3b. Summary of Annual Metrics and Statistical Measures for MM5 for the 2001 Annual Simulation Period:
Riverton, WY (Upper-Air) D-6
Table E-1. Normal Precipitation and Temperature for the Nine NCDC Climate Regions and the Contiguous U.S.,
Based on Observed Data for 1961-1990 E-3
Table E-2. Mercury Deposition Analysis Locations, by Climate Region E-5
Table E-3. MDN and Meteorological Site Pairings for Use in Estimating Mercury Deposition Based on Meteorological Parameters E-8
Table E-4a. Range in Daily Mercury Wet Deposition (g km"2) for Each CART Wet Deposition Classification Category E-11
Table E-4b. Range in Daily Mercury Dry Deposition (g km"2) for Each CART Dry Deposition Classification Category E-12
Table E-5. Surface Meteorological Parameters used in the Mercury Deposition CART Analysis E-13
Table E-6. Upper-Air Meteorological Parameters used in the Mercury Deposition CART Analysis E-13
Table E-7a. Number of CART Classification Bins and Classification Accuracy for Mercury Wet Deposition, Using Meteorological Inputs
and Simulated Deposition Values from the REMSAD Base-Case simulation for 2001 E-14
Table E-7b. Number of CART Classification Bins and Classification Accuracy for Mercury Dry Deposition, Using Meteorological Inputs
and Simulated Deposition Values from the REMSAD Base-Case Simulation for 2001 E-15
Table E-8a. Estimated Annual Mercury Wet Deposition (g km"2) for 1997-2006 for Selected MDN Sites E-18
Table E-8b. Estimated Annual Mercury Dry Deposition (g km"2) for 1997-2006 for Selected MDN Sites E-19
Table E-8c. Estimated Annual Total Mercury Deposition (g km"2) for 1997-2006 for Selected MDN Sites E-19
Table E-9. Ten-Year Average Estimated Annual Mercury Deposition (g km"2) for 1997-2006
for Selected MDN Sites: Wet, Dry, and Total E-20
Table E-10a. Summary of Observed and Estimated (Based on Adjusted REMSAD Results) Variability in Annual Wet Deposition
(g km"2) for MDN Monitoring Sites with Five or More Years of Data During the Analysis Period E-22
Table E-1 Ob. Summary of Observed and Estimated (Based on Unadjusted REMSAD Results) Variability in Annual Wet Deposition
(g km"2) for MDN Monitoring Sites with Five or More Years of Data During the Analysis Period E-22
vi August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Table of Contents
List of Acronyms
ARL = Air Resources Laboratory
CAIR = Clean Air Interstate Rule
CAMR = Clean Air Mercury Rule
CART = Classification and Regression Tree
CB-V = Carbon-Bond chemical mechanism, Version 5
CMAQ = Community Multiscale Air Quality
CTM = Chemical Transport Model
EPA = Environmental Protection Agency
EPS2.5 = Emissions Preprocessing System, version 2.5
GEOS = Goddard Earth Observing System
GIS = Geographical Information Systems
GRAHM = Global/Regional Atmospheric Heavy Metals
IPM = Integrated Planning Model
MACT = Maximum Available Control Technology
MDN = Mercury Deposition Network
NADP= National Acid Deposition Program
NAMMIS = North American Mercury Model Inter-comparison Study
NEI = National Emissions Inventory
NOAA = National Oceanographic and Air Administration
NWS = National Weather Service (NWS)
PPTM = Particle and Precursor Tagging Methodology
PSU/NCAR MM5 = Pennsylvania State University/National Center for Atmospheric Research Fifth Generation
Mesoscale Model
RADM = Regional Acid Deposition Model
REMSAD = REgional Modeling System for Aerosols and Deposition
RUC = Rapid Update Cycle
SCC = Source Category Code
TMDL = Total Maximum Daily Loads
USGS = U.S. Geological Survey
vii August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Table of Contents
This page deliberately left blank.
viii August 2008
-------
Executive Summary
Purpose
This study was undertaken in partnership with state and EPA Regional Air and Water Offices in
order to assist developing and implementing strategies to achieve state water quality standards
for mercury. This report summarizes the application of several air quality modeling systems and
data analysis tools to support an assessment of the sources of airborne mercury and their
contribution to water quality impairment and fish contamination throughout the continental U.S.
The objective of this study was to use atmospheric deposition modeling to quantify contributions
of specific sources and source categories to mercury deposition within each of the lower 48
states. It is expected that the results of this study will provide state and local air and water
quality agencies with 1) an improved understanding of the sources and mechanisms
contributing to mercury deposition; 2) supporting information for future development of Total
Maximum Daily Loads (TMDLs); and 3) assistance in developing implementation plans for
TMDLs and related activities designed to help achieve water quality standards.
Modeling Protocols for TMDL Applications
The modeling protocols followed in this nationwide study were based upon a pilot project
conducted focusing on Wisconsin (Myers et al., 2006). The Devil's Lake TMDL Pilot Project was
conducted by EPA's Offices of Water, Air and Radiation, and Region 5 together with personnel
from the Wsconsin Department of Natural Resources. The use of mercury deposition modeling
to estimate mercury deposition in a TMDL context was examined. The conclusions and
recommendations of an external peer review of the pilot study, including emissions,
meteorological inputs, grid resolution, and source attribution (tagging) were incorporated into the
nationwide study discussed in this report.
These same basic modeling protocols, including application of the tagging feature in the primary
model used in this study, were also followed by others (e.g., Myers and Wei, 2004) in
developing publicly reviewed TMDLs - most recently the Northeastern States Mercury TMDL
approved by EPA in December 2007. (http://www.nescaum.org/focus-areas/science-and-
technology/regional-air-qualitv-modeling-program).
Overview of the Deposition Modeling Tools
The primary modeling system used for this study is the REgional Modeling System for Aerosols
and Deposition (REMSAD). REMSAD is a three-dimensional grid model designed to calculate the
concentrations of both inert and chemically reactive pollutants by simulating the physical and
chemical processes in the atmosphere that affect pollutant concentrations. REMSAD is designed
to support a better understanding of the distributions, sources, and removal processes relevant to
fine particles and other airborne pollutants, including soluble acidic components and several toxic
species (mercury, cadmium, dioxin, polycyclic organic matter (POM), atrazine, and lead).
Mercury may be present in the atmosphere both in the gas and particulate phases. The mercury
species included in REMSAD are HGO (elemental mercury vapor), HG2 (divalent mercury
compounds in gas phase), and HGP (divalent mercury compounds in particulate phase). These
species represent the oxidation state of mercury, and the gas and particulate phases. The
reactions in REMSAD, which are based on Lin and Pehkonen (1999) and other recently
published studies, simulate the transfer of mercury mass from one of these states to another.
REMSAD simulates both wet and dry deposition of mercury. Wet deposition occurs as a result
of precipitation scavenging. Dry deposition is calculated for each species based on land-use
characteristics and meteorological parameters. REMSAD also includes algorithms for the re-
ES-1 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Executive Summary
emission of previously deposited mercury (originating from anthropogenic and natural sources)
into the atmosphere from land and water surfaces.
The mercury treatment in REMSAD can be expanded to include additional, tagged mercury
species. The Particle and Precursor Tagging Methodology (PPTM) feature allows the user to tag
or track emissions from selected sources or groups of sources, and quantify their contribution to
mercury deposition throughout the modeling domain and simulation period.
Results from the Community Multiscale Air Quality (CMAQ) modeling system were used to
enhance the analysis of the effects of global background on mercury deposition. The CMAQ
model is a state-of-the-science, regional air quality modeling system that supports the detailed
simulation of the emission, chemical transformation, transport, and wet and dry deposition of
elemental, divalent, and particulate forms of mercury. For this study, CMAQ was also applied with
PPTM to provide a basis for assessing the uncertainty of the REMSAD PPTM results.
The outputs from three global models were used to specify the boundary conditions for both
REMSAD and CMAQ and thus represent a plausible range of global background contributions
based on current scientific understanding. The results from these models were made available
as part of the North American Mercury Model Inter-comparison Study (NAMMIS) (Bullock et al.,
2008).
Particle and Precursor Tagging Methodology
The Particle and Precursor Tagging Methodology (PPTM) was used in this study to track
emissions from selected sources and source categories and to quantify their contribution to
simulated annual mercury deposition totals for each of the 48 states that comprise the
coterminous U.S.
PPTM for mercury tracks emitted mass from its source through the modeling system processes.
Mercury species in the emissions and initial and boundary condition files are tagged and tracked
throughout the REMSAD simulation. Tags can be applied to emissions from selected source
regions, source categories, and individual sources, both separately and in combination. PPTM
quantifies the contribution of the tagged emissions sources (and/or initial/boundary conditions)
to the simulated species concentrations and deposition, for each mercury species considered by
the model. The emissions from each selected source, source category, or grouping are tagged
in the simulation and each grouping is referred to as a "tag."
Within the model, tagging (PPTM) is accomplished by the addition of duplicate model variables for
each species and tag. The tagged species have the same properties and are subjected to the
same processes (e.g., advection, chemical transformation, deposition) as the actual (or base)
species. PPTM was developed to utilize model algorithms as much as possible to track simulated
tag species concentrations. At each time step in the simulation, the effects of linear processes,
such as advection and dry deposition, are calculated directly for all tagged species. Potentially
non-linear processes, such as gas-phase chemistry, aqueous chemistry, and particle dynamics
are calculated for the overall (or base) species and apportioned to the tagged species.
Some example uses of the mercury PPTM methodology include 1) quantifying the contribution of
mercury emissions from various source sectors to mercury deposition at selected locations
throughout the modeling domain, 2) quantifying the contribution from boundary conditions to
mercury deposition throughout the modeling domain, 3) examining the range of influence of
emissions from selected facilities, and 4) tracking the fate of mercury emissions from a specific
source category to estimate the contribution to deposition to water bodies throughout the modeling
domain.
ES-2 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Executive Summary
REMSAD/PPTM Application Procedures
The REMSAD modeling domain for this study encompasses the continental U.S. and portions of
Canada and Mexico, with 12-km horizontal grid resolution over the entire U.S. portion of the
domain. The annual simulation period is 2001. The baseline emissions data and the
meteorological databases used for the modeling were provided by EPA.
The modeling analysis included a detailed review and revision of the mercury emissions for
each state in order to better represent the 2001 time period. This review was conducted by ICF,
EPA, and state agencies; revisions were incorporated based on information provided by the
states. Model-ready emission inventories were prepared using the revised emissions.
A total of 18 REMSAD simulations were conducted. The first simulation utilized the full Carbon Bond
Version V (CB-V) chemical mechanism to simulate ozone, particulate matter (PM), and related
species. The simulated concentrations of ozone, OH radicals, and other species that react with
mercury in the atmosphere were stored and used as input to the remaining 17 mercury tagging
(REMSAD/PPTM) simulations.
Each of the remaining 17 annual REMSAD/PPTM simulations included approximately 15 to 20
tags (for a total of approximately 300 tagged sources). The tags were defined on a state-by-
state basis, based on the emissions sources within each state. The general procedure was to
assign the first three tags for each state to the top three emitters of divalent gaseous mercury.
Then the top total mercury emitter not already tagged was assigned the fourth tag. Additional
tags were assigned to the remaining larger sources and source categories, in order to capture
the high emitters as well as the range of source types in each state and potentially important
contributors to local and regional mercury deposition in areas with known or suspected mercury
water quality problems. State agencies and EPA regional offices were involved in selecting the
sources for application of PPTM.
PPTM was also used to estimate the contribution from global background to mercury deposition.
Three alternate specifications of the boundary conditions based on global model simulations
were used in the REMSAD simulations. Each of the three global models, the Chemical
Transport Model (CTM) (developed and applied by AER), the Global/Regional Atmospheric
Heavy Metals model (GRAHM) (developed and applied by Environment Canada), and the
GEOS-Chem model (developed and applied by researchers at Harvard University), utilized the
same year 2000 emissions inventory.
Overview of Model Performance
A variety of graphical analyses and statistical measures were used to evaluate REMSAD model
performance. This evaluation focused on concentrations for ozone, sulfur dioxide (SO2) and fine
particulate matter (PM2.s) and deposition for selected PM species and mercury on a monthly
and/or annual basis, depending on the pollutant. The goal was to examine the ability of the
REMSAD modeling system to replicate the observed concentration and deposition
characteristics of the 2001 annual simulation period.
For mercury, the simulated spatial distribution of mercury deposition were found to be consistent
with the emissions and annual transport and rainfall patterns. Wet deposition accounts for much
of the deposition that occurs throughout the domain and this emphasizes the importance of
rainfall in determining mercury deposition patterns.
The REMSAD wet deposition values were compared to data from the Mercury Deposition
Network (MDN), as available from the National Acid Deposition Program (NADP). There are a
total of 98 MDN monitors in the modeling domain, although annual total deposition was
available for only 53 of those monitors for 2001. Model performance for mercury was evaluated
for each set of boundary conditions. These results are summarized in Table ES-1.
ES-3 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Executive Summary
Table ES-1. REMSAD Model Performance Statistics for Mercury Wet Deposition (g km"2)
for the 12-km Resolution Grid at MDN Sites: 2001 Annual Simulation Period.
Boundary
Conditions
CTM
GRAHM
GEOS-CHEM
Mean Observed
(g km-2)
9.26
9.26
9.26
Mean Simulated
(g km-2)
14.21
12.91
15.58
Normalized
bias (%)
59.7
45.7
73.8
Normalized
gross error (%)
65.8
55.3
78.7
Correlation
(R2)
0.74
0.73
0.75
The statistical measures of model performance indicate that the REMSAD simulation results tend to
overestimate wet deposition of mercury, as compared to the MDN monitoring data, using each of the
three sets of boundary conditions. The simulated values derived using the GRAHM boundary
conditions are consistently better matched with the observed values. It should be noted that emerging
research suggests that the MDN wet deposition data may underestimate wet deposition of mercury by
approximately 16 percent (Miller et al., 2005). It was not possible to evaluate the simulated dry
deposition results because an adequate network of dry deposition monitoring data does not exist.
PPTM Results
For each state, the contributions to mercury deposition were examined for the location of greatest
deposition from sources located within that same state. Displays summarizing 1) the contributions
from U.S., Canadian, and Mexican emissions as well as re-emissions (collectively referred to as
"emissions" in the figures in this report) versus background (for all three sets of boundary
conditions calculated using both REMSAD and CMAQ), 2) wet versus dry deposition of emissions
as simulated in REMSAD and wet versus dry deposition of average background deposition
simulated by both REMSAD and CMAQ, 3) contributions from global background, various source
regions and natural re-emissions, 4) primary source contributions from in-state sources. Example
displays of the contribution by category and region for Wisconsin, New York, Virginia, and Texas
are given in Figure ES-1 and illustrate some of the variations in percent contributions from the
selected categories that are found in the full set of modeling results. Note again that these figures
summarize deposition at a particular grid cell where the sources within the given state contribute
the most to deposition in that state. These are not, therefore, statewide summaries. The pie charts
display the percent contributions to total deposition from 1) global background (average of the
three sets of boundary conditions), 2) emissions from sources within the state, 3) emissions from
sources in neighboring states, 4) emissions from all other U.S. states, 4) emissions from Canada
and Mexico, and 5) re-emission processes.
ES-4 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Executive Summary
Figure ES-1. Summary of Mercury Tagging Results for 2001 for Wisconsin, New York, Virginia, and
Texas (with Average Background) at the Location of the Maximum Simulated Annual Mercury
Deposition from Sources within the State.
(Note that these summaries apply to the single grid cell in each state where source in that state contributed the most to
deposition and should not be assumed to apply statewide.)
(a) Wisconsin
(b) NewYork
D Avg Background
(REMSAD) 42.3%
Wisconsin 51.7%
D Neighboring states
2.1%
Other U.S. 2.1%
D Canada & Mexico
0.2%
Reemission 1.6%
Background
(REMSAD) 33.1%
New_York 47.0%
D Neighboring states
4.8%
Other U.S. 7.6%
D Canada & Mexico
5.9%
Reemission 1.5%
(c) Virginia
(d) Texas
Background
(REMSAD) 25.5%
D Virginia 68.0%
D Neighboring states
2.4%
Other U.S. 2.5%
D Canada & Mexico
0.1%
Reemission 1.4%
Background
(REMSAD) 22.5%
Texas 75.8%
D Neighboring states
0.4%
Other U.S. 0.5%
O Canada & Mexico
0.1%
Reemission 0.8%
The REMSAD-derived simulated contributions at the location of maximum deposition for many
of the states are dominated by one local source. For other states, several sources or statewide
emissions from one or more categories are major contributors. Three diverse examples of the
source-specific contributions from in-state sources are given in Figure ES-2. The examples are
for Arkansas (one dominant source), Michigan (multiple contributing sources), and Vermont
(collective statewide source categories). Note that, in each case, the in-state contributions
comprise a different percentage of the total contribution.
The pie-in-pie charts in Figure ES-2 highlight the percent contributions from the in-state sources
to the grid cell with maximum impact from in-state sources. The larger pie gives the proportion
of the overall contribution from emissions sources that are located both outside of and in the
state, and the smaller pie details the contributions from the in-state sources (specifically, the
largest in-state contributors as well as all other in-state sources). If there are five or fewer tags
for a given state, all of the tagged source contributions are displayed. The names of the sources
are given in the legend. The "Collective Sources" tag for each state includes all point and area
sources in the state that are not tagged individually, as part of a source category, or part of a
region. The legend also includes the percentages represented by the various segments of the
ES-5
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Executive Summary
pie charts. Note that the percentages for the cut-out pie chart segments are calculated based on
the total represented only in the smaller, cut-out pie chart. "Other sources within" a state refers
to sources that were tagged, but for the particular location displayed contributed only a small
amount to deposition. They were therefore aggregated in order to simplify the chart.
Figure ES-2. Summary of Source-Specific Mercury Tagging Results for 2001 for Arkansas, Michigan,
and Vermont at the Location of the Maximum Simulated Annual Mercury Deposition from Sources
within the State.
(Note that these summaries apply to the single grid cell in each state where sources in that state contributed the most
to deposition and should not be assumed to apply statewide.)
(a) Arkansas
2nd pie
Sources outside AR
(12.4%)
IAR sources (87.6%)
DAR Ash Grove Cement Co
96.4%
DAR Collective Sources
3.4%
DAR White Bluff 0.1%
AR Carle Bailey Gen Stn
0.1%
DAR Independence 0.0%
Other tagged sources
within AR 0.0%
(b) Michigan
2nd pie
Sources outside Ml (13.4%)
I Ml sources (86.6%)
DMI Central Wayne Co
Sanitation 68.1%
Ml Sources in Detroit Metro
26.0%
DMI Monroe Power Plant
4.4%
Ml Collective Sources 1.0%
DMI J. H. Campbell 0.3%
I Other tagged sources
within Ml 0.2%
ES-6
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Executive Summary
(c) Vermont
2nd pie
I Sources outside VT
(87.8%)
I VT sources (12.2%)
DVT Residential Fuel Comb.
75.9%
DVT Collective Sources
24.1%
DVT Health Services 0.0%
I Other tagged sources
within VT 0.0%
The modeling results contain much more information than is presented in this report. To
facilitate future analysis, the tagging results have also been incorporated into a database tool
(an enhanced version of ARC-Hydro, developed by ESRI under a separate effort) that allows
users to calculate the simulated contribution from each tagged source or source category to any
area of interest, such as a body of water, watershed, or county.
Comparison of REMSAD PPTM, CMAQ PPTM, and Other
Source Attribution Techniques
Source apportionment studies founded on observed data are limited in number, but the
REMSAD PPTM results were compared to a study by Keeler et al. (2006). The Keeler study
used air monitoring and wet deposition data along with statistical receptor modeling to estimate
contributors to wet deposition of mercury at their Steubenville, Ohio site. Keeler's study
estimated that about 70 percent of the mercury wet deposition at the Steubenville site came
from coal combustion. Analyzing the REMSAD PPTM results for the same location indicate that
55 percent or more of the wet deposition at this site comes from coal-fired utilities. Given that
the present methodology did not tag all coal combustion sources, the REMSAD PPTM results
are quite consistent with the conclusions of the Keeler study.
To compare PPTM results for the REMSAD and CMAQ models, CMAQ was applied for a 12-km
domain around Illinois and a summer 2001 simulation period. Seven tags were included in a
simulation representing a mix of individual sources, groups of sources, and source regions
(including global background). The relative contributions from the tagged sources as derived from
the REMSAD and CMAQ simulations were compared for several locations within Illinois and were
found to be consistent, for the area and time period considered.
ES-7
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Executive Summary
Summary of Key Findings
This study has provided an improved understanding of the sources and mechanisms
contributing to mercury deposition throughout the U.S. Key sources and source categories
contributing to mercury deposition within each state were identified and their contribution to total
mercury deposition quantified for an annual 2001 simulation period. It is expected that the
modeling results will provide supporting information for the future assessment of control
measures and development and implementation of Total Maximum Daily Loads (TMDLs).
Based on available data for the 2001 simulation period, REMSAD is able to reasonably replicate
the observed concentration patterns for ozone and PM2.s, and the observed deposition patterns
for PM2.s and mercury wet deposition.
PPTM gives expected results and the simulated contributions are consistent with the emissions
data (including magnitude and speciation characteristics), source locations, source types, and
current knowledge/theories regarding the contribution from global background. The REMSAD
PPTM results are consistent with those obtained using the CMAQ model and are also
consistent with results from a recent receptor modeling study for a specific location in Ohio.
The relative proportion of global, regional, and local (general and source-specific) contributions
varies widely among the states at the location of maximum deposition by sources within the
same state. The REMSAD results for these higher deposition areas indicate that the source
contributions are frequently dominated by
One or more nearby sources (this finding may be linked to horizontal grid resolution and, in
this case, the use of relatively high-resolution (12-km) grids), or
"Collective" sources within the state (defined in this study as all point and area sources in the
state that are not tagged individually, as part of a source category, or as part of a region).
Overall the results are characterized by contributions from a greater number of diverse sources
when large sources are not present in the state.
ES-8 August 2008
-------
1. Introduction
This report summarizes the application of several air quality modeling systems and data
analysis tools to support an assessment of the sources of airborne mercury and their
contribution to water quality impairment and fish contamination throughout the continental U.S.
The primary modeling system used for this study is the REgional Modeling System for Aerosols
and Deposition (REMSAD). The REMSAD Particle and Precursor Tagging Methodology
(PPTM), sometimes referred to as simply "tagging," was used to track emissions from selected
sources and source categories and to quantify their contribution to simulated annual mercury
deposition totals for each of the 48 states that comprise the coterminous U.S. Results from the
Community Multiscale Air Quality (CMAQ) modeling system were used to enhance the analysis
of the effects of global background on mercury deposition, and CMAQ was applied with PPTM
to provide a basis for assessing the uncertainty of the REMSAD PPTM results. The outputs
from three global models were used to specify the boundary conditions for both REMSAD and
CMAQ and thus represent a plausible range of global background.
The REMSAD modeling results were also used to estimate the variability in wet and dry
deposition for several locations throughout the U.S. Classification and Regression Tree (CART)
analysis was used to link the modeled results to observed data and to estimate mercury
deposition for the selected locations for a ten-year period.
1.1. Background and Objectives
The primary route of human exposure to mercury is through the consumption of contaminated
fish. Due to high levels of mercury in fish, all 50 states, 1 territory, and 2 tribes in the U.S. have,
in recent years, issued fish consumption advisories. These advisories may suggest limits on the
consumption of certain types offish, including limits on consumption by certain groups (e.g.,
prospective and new mothers), or not eating fish from certain bodies of water because of unsafe
levels of mercury contamination. In addition, under the Clean Water Act, states must identify
waters not meeting state water quality standards, or impaired waters. States have identified
more than 8,800 individual bodies of water as mercury-impaired.
Once a body of water is listed as impaired by a state, the Clean Water Act calls for the calculation
of a Total Maximum Daily Load (TMDL). TMDLs identify the pollutant reductions or limits that are
needed in order to achieve water quality standards. TMDLs also allocate the reductions to the
different sources of pollution, including air sources. In many parts of the U.S., atmospheric
deposition of mercury is the primary source of mercury contamination in surface waters.
In developing TMDLs, states are not required to allocate reductions to individual non-point
sources, including air sources. However, to determine which sources may need to reduce
emissions in order to achieve water quality standards, states may wish to identify the specific
categories of mercury sources within their state contributing to deposition and quantify the
contributions. From this information, the EPA and states can determine whether to assign
additional pollution limits or allocate further reductions to certain sources or categories of
sources and, ultimately, to develop appropriate management strategies for meeting water
quality criteria and protecting human health. Examples of mercury air sources include
combustion sources, medical waste incinerators, and municipal waste incinerators.
Atmospheric modeling provides an analytic method for quantifying the contributions from
sources of airborne mercury to mercury deposition. In particular, the Particle and Precursor
Tagging Methodology (PPTM), or "tagging," which is available in both the REMSAD and CMAQ
1-1 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Introduction
models) allows one to track or tag mercury emissions from selected sources, and quantify their
contribution to mercury deposition throughout the modeling domain and simulation period.
For REMSAD modeling studies, this approach has been used to estimate the relative
contribution of mercury deposition from various sources and geographic areas (e.g.,
surrounding states) to Devil's Lake in the Wisconsin TMDL Pilot. The REMSAD model was peer
reviewed in 1999 (Seigneur et al., 1999), and the modeling in the TMDL Pilot (including the
tagging application) was subjected to an external peer review resulting in an updated modeling
report in March 2006 (Myers et al., 2006). In addition to the Wisconsin pilot study, REMSAD
PPTM was used to provide total mercury deposition estimates for each of the lower 48 U.S.
states, and to provide, for each state, an estimate of the mercury contribution from in-state
versus out-of-state sources.
The REMSAD PPTM has been used in a TMDL context by EPA Region 6 (Myers and Wei,
2004) and by EPA Region 3 (Myers et al., 2004). In addition, the Northeast States for
Coordinated Air Use Management (NESCAUM) utilized REMSAD in source attribution studies in
support of the Northeastern States Mercury TMDL, approved by EPA in December 2007.
(http://www.nescaum.org/focus-areas/science-and-technology/regional-air-qualitv-modeling-
program).
CMAQ was used to support the development of the Clean Air Mercury Rule (EPA, 2005).
CMAQ with PPTM is currently being used to estimate the regional, national, and global
contributions to airborne mercury deposition for the Commonwealth of Virginia and to examine
the effects of expected future-year emissions changes on the modeled deposition amounts
(Douglas et al., 2007, 2008).
Both CMAQ and REMSAD were included in the North American Mercury Model
Intercomparison Study (NAMMIS) for mercury (Bullock et al., 2008) and the performance and
response of both models was found to be reasonable and substantially similar.
The primary objective of this study was to use the REMSAD PPTM method and other air quality
modeling tools (such as CMAQ PPTM and several global models) to quantify contributions of
specific sources of mercury to mercury deposition within each of the lower 48 states. The PPTM
feature was used to identify the major categories of air sources of mercury loadings, and to
conduct an in-depth analysis of the specific sources of mercury that may be contributing to
water quality impairment.
This study also examined the impacts of year-to-year variability in meteorological inputs on
mercury deposition modeling results. This was accomplished by using Classification and
Regression Tree (CART) analysis to link the REMSAD results to observed meteorological data
in order to estimate the amount and variability of wet and dry mercury deposition for selected
locations for a ten-year period.
1.2. Overview of the Air Deposition Modeling Tools
REMSAD is the primary modeling system used for this study. REMSAD was developed and is
maintained by ICF International (ICF, 2005). Major portions of its development were funded by the
U.S. Environmental Protection Agency (EPA). REMSAD was originally intended as a screening
toola model that could be run (quickly) for a continental-scale modeling domain (specifically the
continental U.S.) and for a full-year simulation periodto provide information on the distribution
and composition of particulate matter, the deposition of pollutant (including toxic) species onto the
1-2 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Introduction
surfaces of inland and coastal bodies of water, and the expected change in air quality and
deposition that results from changes in emissions. All of these parameters were intended to be
primarily represented in terms of seasonal or annual averages or deposition totals. What began as
a simple screening tool has evolved into a more complex one-atmosphere modeling system that
simulates the chemistry, transport, and deposition of airborne pollutants (with emphasis on
particulate matter (PM), ozone, and mercury) using algorithms that reflect the state-of-the science
and current knowledge of the important physical and chemical processes.
For this study, the representation of mercury chemistry and wet and dry deposition processes
are of primary importance. The chemical transformations of mercury included in REMSAD are
based on the review of current status of atmospheric chemistry of mercury presented by Lin and
Pehkonen (1999), with modifications and revisions based on more recent literature. Species
representing the oxidation state of mercury and the phase (gas or particulate) are tracked.
These include HGO (elemental mercury vapor), HG2 (divalent mercury compounds in gas
phase), and HGP (divalent mercury compounds in particulate phase). REMSAD simulates both
wet and dry deposition of gaseous and particulate species. Wet deposition occurs as a result of
precipitation scavenging. Dry deposition is calculated for each species based on land-use
characteristics and meteorological parameters. The mercury modeling capabilities of REMSAD,
as well as the algorithms pertaining to the other pollutants, have undergone external peer
review (Seigneur et al., 1999) independently of and prior to initiation of the EPA-sponsored
Mercury Total Maximum Daily Load (TMDL) Pilot Project for Devil's Lake, Wisconsin (Myers et
al., 2006). In the Devil's Lake Pilot, EPA Air and Water Offices and EPA Region 5 worked with
personnel from Wisconsin DNR in part to test REMSAD for potential application in TMDL
development. Specifically, the model was run with tagging of individual sources in Wisconsin
with the same type of meteorological data and at the same scale as described in this report. The
findings of the Devil's Lake Pilot underwent external peer review. Recommendations from the
peer review have been incorporated into this nationwide application, most notably a suggestion
that meteorological and emissions inputs target the same year to the extent possible.
The REMSAD Particle and Precursor Tagging Methodology (PPTM) was used to track
emissions from selected sources and source categories and to quantify their contribution to
mercury deposition. With PPTM, mercury species in the emissions and initial and boundary
condition files are tagged and tracked throughout the REMSAD simulation. Tags can be applied to
emissions from selected source regions, source categories, and individual sources, both
separately and in combination. PPTM quantifies the contribution of the tagged emissions sources
(and/or initial/boundary conditions) to the simulated species concentrations and deposition, for
each mercury species considered by the model.
CMAQ was used in this study primarily to enhance and provide perspective to the REMSAD
simulation results. The CMAQ model is a state-of-the-science, regional air quality modeling system
that is designed to simulate the physical and chemical processes that govern the formation,
transport, and deposition of gaseous and particulate species in the atmosphere (Byun and Ching,
1999). The CMAQ model was designed as a "one-atmosphere" model and can be used to simulate
ozone, particulate matter, and mercury. Mercury simulation capabilities were first incorporated
into the CMAQ model by adding gaseous and aqueous chemical reactions involving mercury to
the CMAQ chemical mechanism (Bullock and Brehme, 2002). CMAQ supports the detailed
simulation of the emission, chemical transformation, transport, and wet and dry deposition of
elemental, divalent, and particulate forms of mercury (HGO, HG2 and HGP) (Bullock et al., 2008).
CMAQ also includes PPTM for mercury (Douglas et al., 2006) which provides detailed,
1-3 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Introduction
quantitative information about the contribution of selected sources, source categories, and/or
source regions to simulated mercury concentrations and (wet and dry) deposition.
Finally, the outputs from three global models were used to specify the boundary conditions for
both REMSAD and CMAQ and thus represent global background. These include the Chemical
Transport Model (CTM) (Shia et al., 1999; Seigneur et al., 2001), the Global/Regional
Atmospheric Heavy Metals model (GRAHM) (Dastoor and Larocque, 2004; Ariya et al., 2004),
and the GEOS-Chem model (Selin et al., 2007). Estimates of boundary concentrations of
elemental mercury, divalent gas mercury, and particulate mercury prepared using each of these
models and intended for use in continental scale modeling were made available as part of the
North American Mercury Model Inter-comparison Study (NAMMIS) (Bullock et al., 2008). All
three global models are included in this study in part because the NAMMIS evaluation
concluded that they were each based upon sound science, and, absent a much more
widespread mercury monitoring network including dry deposition, it is not now possible to
determine which best replicates actual global contributions.
1.3. Overview of the Modeling Approach
For this study, 18 REMSAD simulations were conducted. The first simulation utilized the full
chemical mechanism to simulate ozone, PM, and related species. The simulated concentrations
of ozone, OH radicals, and other species that react with mercury in the atmosphere were stored
and used as input to an additional 16 mercury tagging simulations. A meteorological sensitivity
simulation was also made with REMSAD.
A CMAQ simulation was made using the tagging feature implemented by ICF for a sub-domain
surrounding Illinois. The results of this simulation are contrasted with the REMSAD results for
the same area.
Additional CMAQ simulation results for the continental US were provided by EPA. These results
were for simulations using each of the three available global model based boundary conditions.
Again, these results are compared and contrasted with the REMSAD results.
The emissions data and the meteorological databases used for the modeling were provided by
EPA. The REMSAD modeling domain encompasses the continental U.S. and portions of
Canada and Mexico, with a 36-km resolution outer grid; in addition two 12-km resolution nested
grids are located approximately over the eastern three quarters and western quarter of the U.S.,
respectively. The entire U.S. is encompassed by the 12-km grids. The modeling domain is
depicted in Figure 1-1. The annual simulation period is 2001.
1 -4 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Introduction
Figure 1-1. REMSAD Modeling Domain for the Mercury PPTM Simulations.
Horizontal Resolution is 36 km for the Outer Grid and 12km for the Two Inner Grids.
Km
-2736 -2016 -1296 -576 144 864 1584 2304
uoF
100
90
80
70
60
50
40
[TUT ITIT"! nT|!!Trnrn IF fn TITT!fin n (I
iihiiiiiiylmimyiiiHniiiliiiimiihiimiKliiiiiiuilmiiHHliiiiiff
10 20 30 40 50 60 70 60 90 100 110 120 130 140
1800
1368
938
504
72
-360 g
X
-792
-1224
-1656
-2068
2001 Domain for OW 300 tag Hg modeling
The modeling analysis included a detailed review and revision of the mercury emissions for
each state. This review was conducted by ICF, EPA, and state agencies and revisions were
incorporated based on information provided by the states. Model-ready emission inventories
were prepared using the revised emissions.
The modeling analysis also included a review of the meteorological inputs and a sensitivity
simulation in order to examine the effects of elevated precipitation inputs on the REMSAD
results. This simulation focused on a high simulated precipitation area over northern Utah.
In the mercury tagging simulations, the initial conditions, boundary conditions, and
approximately 300 emissions sources were tagged using PPTM. The tags represent various
source categories and sources within each of the lower 48 states, with emphasis on the largest
sources of mercury emissions for each state. In addition, tags were allocated to emissions from
Canada and Mexico. Post-processing software was used to combine information from the
tagging simulations. Since use of the tagging technique does not affect the simulation results,
the results from the separate simulations can be combined and compared.
The REMSAD mercury tagging results are summarized in this report and were also processed
for incorporation into an interactive Geographical Information Systems (GIS) database tool (an
enhanced version of ARC-Hydro, developed by ESRI under a separate effort). This tool allows
users to extract the modeling results for any grid cell or combination of grid cells and calculate
the simulated contribution from each tagged source or source category to any area of interest in
the modeling domain, e.g., a reservoir, watershed, or tribal area.
A CMAQ simulation was also conducted as part of this study, in order to compare the PPTM
results for REMSAD with those for CMAQ. CMAQ was applied for a three-month subset of the
annual simulation period (summer 2001) using PPTM for seven tags. The CMAQ modeling
1-5
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Introduction
domain includes an outer grid with 36-km horizontal resolution (the same as used for REMSAD)
and a one-way nested (inner) grid with 12-km resolution, covering Illinois and portions of several
surrounding states. The emissions and meteorological inputs for the two models are the same,
accounting for grid resolution.
The remaining model outputs from CMAQ and the global models were obtained from EPA and
the application procedures used to generate these results are described by Bullock et al. (2008).
1.4. Report Contents
This report summarizes the methods and results of the mercury deposition modeling analysis
conducted to quantify the potential loadings of airborne mercury emissions on bodies of water in
the lower 48 states. The modeling tools are described in Section 2. The meteorological and
geographical inputs are summarized in Section 3, and the emissions inputs are summarized in
Section 4. The base case simulations for mercury are evaluated and discussed in Section 5.
The mercury PPTM results are presented in Section 6. Finally, a summary of key findings is
provided in Section 7.
The report also includes several appendixes. Appendix A presents additional information related
to model performance for non-mercury species. Appendix B compares the CMAQ and REMSAD
PPTM results. Appendix C gives a detailed summary of the revisions made to the national-scale
emission inventory as part of this study. Appendix D presents the results of the meteorological
sensitivity simulation. Appendix E presents the CART-based assessment of mercury deposition
variability.
1-6 August 2008
-------
2. Description of the Deposition Modeling Tools
Several air quality models were used to support this assessment of mercury deposition. These
tools are described in this section of the report. Most of the modeling was conducted using
REMSAD. Therefore, most of this section contains a detailed technical description of the
REMSAD model, including the Particle and Precursor Tagging Methodology (PPTM) for
mercury. CMAQ and several global models were also used in this assessment. These are more
briefly described at the end of this section; references are provided for a more detailed
discussion of each model.
2.1. REMSAD Modeling System
Version 8 of the Regional Modeling System for Aerosols and Deposition (REMSAD) was used for
this modeling analysis. REMSAD is a three-dimensional grid model designed to calculate the
concentrations of both inert and chemically reactive pollutants by simulating the physical and
chemical processes in the atmosphere that affect pollutant concentrations. REMSAD is designed
to support a better understanding of the distributions, sources, and removal processes relevant to
fine particles and other airborne pollutants, including soluble acidic components and several toxic
species (mercury, cadmium, dioxin, polycyclic organic matter (POM), atrazine, and lead).
REMSAD provides estimates of the concentrations and deposition of the simulated pollutants at
each grid location in the modeling domain. Both wet and dry deposition processes are simulated.
Post-processing can provide concentration averages and deposition totals for any subset of the
time span of the simulation for any location within the domain.
REMSAD was designed to account for the many factors that affect the concentration and
distribution of aerosols and mercury, including:
Spatial and temporal distribution of toxic and particulate emissions (both anthropogenic and
non-anthropogenic),
Composition of the emitted particulate and mercury species,
Spatial and temporal variations in the wind fields,
Dynamics of the boundary layer, including stability and the level of mixing,
Chemical reactions involving SO2, NOX, mercury and other important precursor species,
Diurnal variations of solar insulation and temperature,
Loss of primary and secondary aerosols and toxics by dry and wet deposition, and
Ambient air quality immediately upwind and above the region of study.
The basis for the REMSAD model is the atmospheric diffusion or species continuity equation.
This equation represents a mass balance in which all of the relevant emissions, transport,
diffusion, chemical reactions, and removal processes are expressed in mathematical terms. The
REMSAD system consists of a series of preprocessor programs, the core model, and several
post-processing programs.
The REMSAD model is capable of "nesting" one or more finer-scale subgrids within a coarser
overall grid. The fully interactive two-way nesting capability permits high resolution over selected
source and/or receptor regions of interest. The modeling system may be applied at scales ranging
from a single metropolitan area to a continent containing multiple urban areas.
2-1 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
This version of REMSAD (Version 8) utilizes version V (five) of the carbon-bond chemical
mechanism (CB-V) to simulate gas-phase photochemical processes in the atmosphere and also
includes a chemical mechanism to calculate the transformations of mercury.
The particulate matter species modeled by REMSAD include a primary coarse fraction (corresponding
to particulates in the 2.5 to 10 micron size range), a primary fine fraction (corresponding to particulates
less than 2.5 microns in diameter), and several secondary particulates (e.g., sulfates, nitrates, and
organics). The sum of the primary fine fraction and all of the secondary species is assumed to be
representative of PM2.s. This is calculated as part of a post-processing step.
For the simulation of mercury, REMSAD carries the species HGO (representing elemental
mercury), HG2 (representing divalent gas mercury), and HGP (representing divalent particulate
mercury). Mercury simulations can be run separately from the full PM simulations, provided that
the full PM simulation has been run and the outputs have been saved for use in the mercury
only simulations. This substantially reduces the computer requirements for REMSAD.
Of particular importance to the current work is that the mercury treatment in REMSAD can be
expanded to include additional, tagged mercury species. The PPTM feature allows the user to
tag or track emissions from selected sources, and quantify their contribution to mercury
deposition throughout the modeling domain and simulation period.
2.1.1. Input File Requirement
There are seventeen input files for REMSAD. These fall into the general categories of
emissions, initial and boundary conditions, meteorological fields, surface characteristics,
chemical parameters, and simulation control parameters. The files are listed and briefly
described in Table 2-1.
Table 2-1. REMSAD Input Files.
File Type/Name
Description
Source of Data/Information for
REMSAD PPTM Application
Emissions
EMISSIONS
PTSOURCE
Low-level (surface-layer) emissions for area, mobile, low-level
point, non-road, and biogenic sources
Elevated (upper-layer) point-source emissions
EPA OAQPS
EPA OAQPS
Initial and Boundary Conditions
AIRQUALITY Initial species concentrations for each grid cell within the modeling
domain
BOUNDARY Species concentrations along the lateral boundaries of the
modeling domain
Derived from global model simulations
of mercury, or estimated from the
literature
Derived from global model simulations
of mercury, or estimated from the
literature
CHLORINE Surface chlorine concentrations
Meteorological Fields
Wl N D u- and v- wi nd components
TEMPERATURE Temperature
PSURF Surface pressure
H20 Water vapor concentration
VDIFFUSION Vertical diffusivities or exchange coefficients
CLW Cloud-water mixing ratio
RLW Rain-water mixing ratio
RAIN Rainfall rate
Estimated from the literature
MM5 files from EPA OAQPS
MM5 files from EPA OAQPS
MM5 files from EPA OAQPS
MM5 files from EPA OAQPS
MM5 files from EPA OAQPS
MM5 files from EPA OAQPS
MM5 files from EPA OAQPS
MM5 files from EPA OAQPS
2-2
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
.... T ... n . .. Source of Data/Information for
Fie Type/Name Description mr.^-. A »
*v |_ REMSAD PPTM Application
Surface Characteristics
SURFACE Land-use characteristics USGS LULC data
TERRAIN Terrain heights MM5 files from EPA OAQPS
Chemistry Parameters
CHEMPARAM Chemical reaction rates and other CB-V parameters Standard REMSAD file
RATES Photolysis rates Standard REMSAD file
Simulation Control
SIMCONTROL Simulation control parameters and option specifications User specified
2.1.2. Carbon-Bond V Chemical Mechanism
The carbon-bond V (five) photochemical mechanism (CB-V) is an updated version of CB-IV
(Gery et al., 1989) as enhanced to include radical-radical termination reactions. The CB-V
mechanism is derived from the mechanism implemented in UAM-V (SAI, 1999) with some
specific adaptations for REMSAD.
Secondary organic aerosols (SOA) are known to result from the reactions of hydrocarbons in
the atmosphere, and version 8 of REMSAD includes a calculation of the yield of SOA from both
anthropogenic and biogenic hydrocarbon species. The REMSAD mechanism accounts for the
anthropogenic contribution of toluene (TOL) and xylene (XYL) reactions to SOA formation. In
addition, the mechanism includes reactions with biogenic monoterpenes (TERP), which are the
principal biogenic precursors of SOA.
Table 2-2 lists the CB-V gas-phase reactions. Of these, only the reaction of SO2 with the OH
radical to form sulfate directly affects particulate concentrations. However, a number of the gas-
phase species affect the production of particulates in aqueous phase. Peroxide, which is a
product of the gas-phase chemistry, is important in the aqueous production of sulfate. To a
lesser degree, ozone also affects the production of sulfate in aqueous phase. Nitric acid
produced in gas phase can later be converted to particulate via reaction with ammonia. Radical
species such as OH and HO2 can affect the evolution of toxics such as POM and mercury. The
gas-phase products (ozone, peroxide, nitric acid, and radicals) are the result of a complex
interaction of many reactions in the mechanism. SOA is also a known product of gas-phase
interactions, and the gas-phase production of SOA is included in the mechanism.
Table 2-2. CB-V Reaction Set.
1
2
3
4
5
6
7
8
9
10
11
REACTION
N02 = NO + 0 - NOXY
0 = 03
03 + NO = N02 + NOXY
03 + N02 = N03
03 + OH = H02
03 + H02 = OH
03 + N03 = N02
03 = 0
03 = 010
NO + NO = 2.0000 N02 +2.0000 NOXY
NO + N02 + H20 = 2.0000 HN02 - NOXY
RATE CONSTANT*
(ppnv1-min-1)
4.926E-01
4.641 E+06
2.808E+01
4.726E-02
1.149E+02
2.957E+00
1.499E-02
2.611E-02
1.681E-03
1.499E-04
2.997E-08
2-3 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
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
48
49
50
51
52
53
54
55
56
57
58
59
REACTION
NO + 0 = N02 + NOXY
NO + OH = HN02
NO + H02 = N02 + OH + NOXY
NO+X02 = N02 + NOXY
NO+X02N = NTR
NO + N03 = 2.0000 N02 + NOXY
N02 + 0 = NO - NOXY
N02 + 0 = N03
N02 + OH = HN03 - NOXY
N02 + H02 = PNA-NOXY
N02 + N03 = N02 + NO - NOXY
N02 + N03 = N205
HN02 + HN02 = N02 + NO + NOXY
HN02 + OH = N02 + NOXY
HN02 = NO + OH
HN03 + OH = N03 + NOXY
PNA + OH = N02 + NOXY
PNA = N02 + H02 + NOXY
PNA = 0.6100 N02 + 0.6100 H02 + 0.3900 OH + 0.3900 N03 + NOXY
N205 + H20 = 2.0000 HN03 -2.0000 NOXY
N205 = N02 + N03
H202 + OH = H02
H202 = 2.0000 OH
01D = 0
01 D + H20 = 2.0000 OH
CO + OH = H02
H2 + OH = H02 + H20
H02 + OH =
H02 + H02 = H202
H02 + H02 + H20 = H202
H02+X02 =
H02+X02N =
X02+X02 =
X02+X02N =
X02N + X02N =
N03 + OH = N02 + H02
N03 + H02 = HN03-NOXY
N03 + N03 = 2.0000 N02
N03 = 0.8900 N02 + 0.8900 0 + 0.1 100 NO -0.1 000 NOXY
FORM + 0 = OH + H02 + CO
FORM + OH = H02 + CO
FORM + N03 = HN03 + H02 + CO - NOXY
FORM = 2.0000 H02+ CO
FORM = CO
ACET + 0 = C203 + OH
ACET + OH = C203
ACET + N03 = C203 + HN03 - NOXY
ACET = FORM + 2.0000 H02 + CO + X02
RATE CONSTANT*
(ppnv1-miiT1)
2.458E+03
1.104E+04
1.196E+04
1.139E+04
1.139E+04
3.840E+04
1.433E+04
2.325E+03
1.346E+04
2.053E+03
9.691 E-01
1.741E+03
1 .499E-05
6.644E+03
9.729E-02
2.191E+02
6.793E+03
5.173E+00
2.775E-04
2.997E-06
2.262E+00
2.511E+03
4.020E-04
4.362E+10
3.247E+05
3.544E+02
9.891 E+00
1.625E+05
5.259E+03
3.662E+03
1.199E+04
1.199E+04
1.998E+02
3.996E+02
1.998E+02
3.247E+04
5.195E+03
3.397E-01
1.670E+01
2.358E+02
1.477E+04
8.592E-01
1.652E-03
2.297E-03
6.644E+02
2.068E+04
3.497E+00
2.855E-04
2-4
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
REACTION
ALDX + 0 = CX03 + OH - PAR
ALDX + OH = CX03-PAR
ALDX + N03 = CX03 + HN03 - PAR - NOXY
ALDX = ACET + 2.0000 H02 + CO + X02 - PAR
C203 + NO = N02 + FORM + H02 + X02 + NOXY
C203 + N02= PAN - NOXY
C203 + H02 = 0.2600 03
C203 + C203 = 2.0000 FORM + 2.0000 X02 + 2.0000 H02
CX03 + NO = N02 + ACET + H02 + X02 + NOXY
CX03 + N02 = PANX-NOXY
CX03 + H02 = 0.3300 03
CX03 + C203 = ACET + FORM + 2.0000 X02 + 2.0000 H02
PAN = N02+C203 + NOXY
PANX = N02 + CX03 + NOXY
PANX + OH = N02 + ACET + NOXY
CH4 + OH = FORM+X02 + H02
PAR + OH = 0.8700 X02 + 0.1300 X02N + 0.1 100 H02 -0.1 100 PAR + 0.0600 ACET +
0.7600 ROR + 0.0500 ALDX
ROR = 0.9600 X02 + 0.6000 ACET + 0.9400 H02 -2.1 000 PAR + 0.0400 X02N + 0.0200
ROR + 0.5000 ALDX
ROR = H02
ROR + N02 = NTR- NOXY
ETH + 0 = 0.9500 FORM + 1 .5500 H02 + 0.9500 CO + 0.6000 X02 + 0.3500 OH
ETH + OH = X02 + 1 .5600 FORM + 0.2200 ALDX + H02
ETH + 03 = 1 .0200 FORM + 0.3300 CO + 0.0800 H02 + 0.0200 H202
OLE + 0 = 0.1900 ACET + 0.2900 H02 + 0.1900 X02 + 0.2000 CO + 0.2000 FORM +
0.0070 X02N+ 0.6100 PAR + 0.3000 ALDX + 0.1000 OH
OLE + OH = 0.7100 FORM + 0.3600 ACET + 0.5900 ALDX -0.7100 PAR + 0.71 00 X02 +
0.9500 H02
OLE + 03 = 0.2000 ACET + 0.8600 FORM + 0.4500 X02 - PAR + 0.3100 OH + 0.4000
CO + 0.4200 H02 + 0.3200 ALDX + 0.0800 H202
OLE + N03 = 0.91 00 X02 + FORM + 0.0900 X02N - PAR + 0.3500 ACET + 0.5600 ALDX
+ N02
IDLE + 0 = 1.1400 ACET + 0.7600 ALDX + 0.1000 H02 +0.1000 X02 +0.1000 CO +
0.1000 PAR
IDLE + OH = 1 .2000 ACET + 0.8000 ALDX + H02 + X02
IDLE + 03 = 0.6000 ACET + 0.4000 ALDX + 0.2500 FORM + 0.2500 CO + 0.5000 0 +
0.5000 OH + 0.5000 H02
IDLE + N03 = 1 .0900 ACET + 0.7300 ALDX + H02 + N02
TOL + OH = 0.4400 H02 + 0.0800 X02 + 0.3600 CRES + 0.5600 T02 + an SV1 + an
SV2
T02 + NO = 0.9000 N02 + 0.9000 H02 + 0.9000 OPEN + 0.1000 NTR + 0.9000 NOXY
T02 = CRES + H02
CRES + OH = 0.4000 CRO + 0.6000 X02 + 0.6000 H02 + 0.3000 OPEN
CRES + N03 = CRO + HN03 - NOXY
CRO + N02 = NTR -NOXY
CRO + H02 = CRES
OPEN + OH = X02 + 2.0000 CO + 2.0000 H02 + C203 + FORM
OPEN + 03 = 0.0300 ALDX + 0.6200 C203 + 0.7000 FORM + 0.0300 X02 + 0.6900 CO
+ 0.0800 OH + 0.7600 H02 + 0.2000 MGLY
RATE CONSTANT*
(ppnv1-miiT1)
1.010E+03
2.957E+04
8.393E+00
1.006E-03
2.658E+04
1.283E+04
2.008E+04
2.215E+04
2.957E+04
1.369E+04
2.088E+04
2.448E+04
2.289E-02
2.443E-02
1.699E+03
1.689E+01
1.196E+03
1.318E+05
9.592E+04
2.215E+04
1.077E+03
1.249E+04
2.398E-03
5.907E+03
3.948E+04
1.499E-02
1.409E+01
3.397E+04
9.422E+04
3.097E-01
5.725E+02
8.752E+03
1.199E+04
2.518E+02
6.055E+04
3.247E+04
1.998E+04
8.093E+03
4.426E+04
1.499E-02
2-5
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
REACTION
OPEN = C203 + H02+CO
XYL + OH = 0.7000 H02 + 0.5000 X02 + 0.2000 ORES + 0.8000 MGLY + 1 .1 000 PAR +
0.3000 T02+ 321 SV1+ 322 SV2
MGLY + OH=X02 + C203
MGLY = C203 + H02 + CO
ISOP + 0 = 0.2500 H02 + 0.2500 X02 + 0.7500 ISPD + 0.2500 CX03 + 0.2500 PAR +
0.5000 FORM
ISOP + OH = 0.9100 ISPD + 0.9900 X02 + 0.9100 H02 + 0.6290 FORM + 0.0880 X02N
ISOP + 03 = 0.6000 FORM + 0.6500 ISPD + 0.1500 ALDX + 0.2000 CX03 + 0.3500 PAR
+ 0.2700 OH + 0.2000 X02 + 0.0700 H02 + 0.0700 CO
ISOP + N03 = 0.2000 ISPD + X02 + 0.8000 H02 + 0.2000 N02 + 0.8000 ALDX + 2.4000
PAR + 0.8000 NTR -0.8000 NOXY
ISOP + N02 = 0.2000 ISPD + X02 +0.8000 H02 + 0.2000 NO + 0.8000 ALDX + 2.4000
PAR + 0.8000 NTR -NOXY
ISPD + OH = 0.3300 CO + 0.2500 ACET + 0.1 700 FORM + 1 .5650 PAR + 0.1 680 MGLY
+ 0.5000 H02 +0.7130 X02 + 0.2100 C203 + 0.2900 CX03
ISPD + 03 = 0.0200 ACET + 0.1500 FORM + 0.2300 CO + 0.8500 MGLY + 0.3600 PAR +
0.1100 C203 + 0.0640 X02 + 0.2700 OH +0.1500 H02
ISPD + N03 = 0.6400 CO + 0.2800 FORM + 0.3600 ALDX + 1 .2820 PAR + 0.9250 H02 +
0.0800 CX03 + 0.0750 X02 + 0.8500 NTR + 0.1500 HN03 - NOXY
ISPD = 0.3300 CO + 0.0700 ACET + 0.9000 FORM + 0.8320 PAR + 1 .0330 H02 + 0.7000
X02 + 0.2670 C203 + 0.7000 CX03
TERP + 0 = 0.1500 ALDX + 0.5100 PAR + 993.0000 SV3 + 994.0000 SV4
TERP + OH = 0.7500 H02 + 1 .2500 X02 + 0.2500 X02N + 0.2800 FORM + 0.4700 ALDX
+ 995.0000 SV3 + 996.0000 SV4
RATE CONSTANT*
(ppnv1-miiT1)
1.493E-02
3.697E+04
2.508E+04
1.084E-02
5.315E+04
1.469E+05
1.898E-02
9.951 E+02
2.198E-04
4.962E+04
1.049E-02
1.477E+00
9.195E-05
5.263E+04
9.915E+04
TERP + 03 = 0.5700 OH + 0.0700 H02 + 0.7600 X02 + 0.1800 X02N + 0.2400 FORM +
115 0.0010 CO + 7.0000 PAR + 0.2100 ALDX + 0.3900 CX03 + 997.0000 SV3 + 998.0000 1.099E-01
SV4
116
TTT
TERP + N03 = 0.4700 N02 + 0.2800 H02 +1.0300 X02 + 0.2500 X02N + 0.4700 ALDX
+ 0.5300 NTR -0.5300 NOXY + 999.0000 SV3 + 990.0000 SV4
9.811E+03
S02 + OH = SULF + H02
1.321E+03
ETOH + OH = 0.9500 ACET + 0.1000 X02 + H02 + 0.1000 FORM
4.726E+03
* For single reactant processes, rate is in min-1. Reaction rates are for noon (zenith angle 17 degrees)
at approximately 100W longitude, 40N latitude, 298 K, 1 atm.
As an option, REMSAD also includes a reduced-form version of the CB_V , termed "micro-CB"
(|o,CB). This form of the mechanism was used in REMSAD prior to version 8. The mechanism is
based on a drastic reduction in the speciation of the organic compounds; the inorganic and
radical parts of the mechanism are identical to CB-V. Further details on micro-CB can be found
in the REMSAD user's manual (ICF, 2005).
REMSAD requires information on solar radiation in order to calculate photolysis rates for the
photochemical reactions that drive the formation of OH radical and the steady-state
concentrations of NO, NO2, and ozone.
Photolysis rates are calculated as a preprocessing step using a parameterized light model
developed by Schippnick and Green (1982). Lookup tables for photolysis rates for NO2 and
ozone for various zenith angles and altitudes are generated and used by the model (through
linear interpolation) to assign a photolysis rate to each grid cell and time step.
2-6
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
The ratios of ozone, formaldehyde, and acetaldehyde to NO2 photolysis rate have been defined for
two altitudes (1380 m and 10000 m) that approximately represent the average altitude of the
boundary layer and the upper troposphere. These ratios, and ozone photolysis rates, were used to
generate OH lookup tables for the lower and upper atmospheres. Photolysis rates for several other
photochemical reactions are derived from these values using scaling factors (see Gery et al., 1989).
Cloud cover is not treated by the photolysis rate preprocessor, but at the end of each time step
all photolysis rates are corrected for cloud cover using the algorithm developed by Chang
(1987) for the Regional Acid Deposition Model (RADM). The cloud cover scaling factors applied
to the photolysis rates vary from 1 for clear skies to 0.01 for completely overcast conditions
(under some cloud vertical distribution this scaling factor can be higher than 1). This procedure
assumes that the cloud cover effects for all photolysis rates are the same as for NO2.
2.1.3. Mercury Chemistry
Mercury (Hg) is volatile in elemental form but involatile in many oxidized inorganic forms and
therefore may be present both in the gas and particulate phases. Gaseous mercury species other
than elemental Hg may be present in the atmosphere (e.g., organo-mercury compounds). Estimates
of mercury emissions include a significant fraction of gaseous, oxidized mercury (EPA, 1997).
The chemical transformations of mercury included in REMSAD are based on the review of
current status of atmospheric chemistry of mercury presented by Lin and Pehkonen (1999) with
a number of updates based on more recent literature. The mercury species included in
REMSAD are HGO (elemental mercury vapor), HG2 (divalent mercury compounds in gas
phase), and HGP (divalent mercury compounds in particulate phase). These species represent
the oxidation state of mercury, and the gas and particulate phases. The reactions in REMSAD
cause transfer of mercury mass from one of these states to another.
In cloud water, HGP is assumed to dissolve with the solubility of HgO (mercury (II) oxide). Some
HG2 is assumed to be adsorbed to soot particles (e.g., see Seigneur et al., 1998). In REMSAD,
the treatment is parameterized using a simple formula. The species PEC (primary elemental
carbon) is used as an indicator of the amount of soot present. Fifty-five percent (based on the
upper limit suggested by Seigneur et al., 1998) of the dissolved divalent mercury (Hg2+) in
aqueous phase is assumed to be adsorbed to soot particles when PEC is 55 |o,g/(mole of air) or
greater. When PEC is zero, no adsorption takes place. Between these two extremes, the fraction
of adsorbed Hg2+ is linearly interpolated. This parameterization is based in part on results of
comparing the REMSAD aqueous mercury chemistry with published information on the aqueous
mercury chemistry in other models such as CMAQ (Myers, 2004; Ryaboshapko, 2002).
REMSAD does not have an internal estimate of chlorine concentrations, which is important in
many of the aqueous phase reactions. Therefore, an input file is required to specify chlorine.
The chlorine pathway is considered to be active only at night and chlorine at upper levels is
typically set to zero. Chlorine concentrations are supplied for the surface level with differing
values over the ocean and over land. A typical value used for chlorine over the ocean is 125 ppt
(Tokos et al., 1998). Chlorine over land areas is much lower. A value of 5 ppt over land was
chosen. Chlorine concentrations are reduced linearly from the surface to zero at a height of
2000 m over the ocean or at a height of 1000 m over land. The ocean value for chlorine was
used approximately 18 km inland in addition to over the ocean.
In order to treat reduction of HG2 by sulfur compounds, the average amount of dissolved SO2 is
estimated during the calculation of the aqueous formation of sulfate (via reaction of SO2 with
H2O2, O3, and O2). Equilibrium concentrations of HgSO3 and Hg(SO3)22" are calculated and then
2-7 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
the production rate of HGO from HgSO3 is calculated. The pH of cloud water is needed in order to
calculate the Henry's law coefficients of some species. In these cases, pH is assumed to be 4.5.
Some of the individual species-specific reactions such as photoreduction (for halo-compounds
of divalent Hg) and reactions of dimethylmercury by O3, OH, and other radicals have been
neglected, since their effects are expected to be small.
The routine that calculates chemical transformations of mercury is provided with total
concentrations of HGO, HG2, and HGP. The routine calculates the fraction in gas and aqueous
phases of each of these categories. Gas and aqueous chemical transformations are calculated
independently. The routine then recombines the gas and aqueous fractions to return the new
total concentrations of HGO, HG2, and HGP.
Table 2-3 lists the reactions that are included in the REMSAD mechanism for mercury.
Table 2-3. Mercury Chemical Mechanism in REMSAD, Version 8.
Reaction
Rate (unit)
For HGO
Gas phase
HGO + 03Hx1/2HGP + 1/2HG2
HGO + H202 -> HG2
HGO + OH^1/2HGP + 1/2HG2
3.0e-20 (cm3molecule-1s-1)
8.5e-19(cm3molecule-1s-1)
7.7e-14 (cm3molecule-1s-1) (lower bound of range in Pal &
Ariya, 2004; also within range in Sommaretal., 2001)
Aqueous phase
HGO + 03 -> HG2
HGO + OHnxHG2
HGO + Claq -» HG2
HgSOS -> HGO
4.7e+7 (M-1s-1)
2.0e+9 (M-is-1)
(See eq. 8 in Lin and Pehkonen, 1999)
T e(3i97i T- 12595)/T s-i (Van Loon et al, 2000)
For HG2
Aqueous phase
HG2 + H02 -> HGO
HG2 + 80s2- o HgSOs
HgSOs + 80s2- <-> Hg(SOs)22-
Hg2 + OH- <-> Hg(OH)+
Hg2 + 20H-^Hg(OH)2
Hg2 + OH- + Cl- o HgOHCI
Hg2 + Cl- <-> HgCh
Hg2 + 2CI-^HgCI2
Hg2 + 3CI-^HgCls-
Hg2 + 4CI-^HgCI42-
1.7e+4(M-1s-1)
5.e+12 (M-1)
2.5e+11 (M-i)
4.27e+10 (M-1)
1.74e+22(M-1)
1.78e+1 8 (M-2)
2.0e+7 (M-i)
1.e+14(M-2)
1.e+15(M-3)
3.986+1 5 (M-*)
Source: Lin and Pehkonen, 1999, except as noted.
2.1.4. Mercury Re-Emission Treatment
Re-emission of mercury from land or water surface is believed to occur but has not been
accurately quantified. Sofiev and Galperin (2000) note that mercury can be reduced (or
methilated) and re-emitted back into the air after oxidation and deposition. Syrakov (1998) finds
2-8
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
that airborne mercury (both anthropogenic and natural) deposited on land and water surfaces is
re-emitted back to the atmosphere through natural processes, such as microbial activity. He
also reports that natural emissions and the re-emission of mercury are mainly in the form of
HgO. Very small amounts are in the form of organic mercury compounds, since these are very
quickly reduced to metal vapor in the atmosphere. Other modelers (Shia, et al., 1999) note that
the emissions from land and ocean surfaces consist of 1) cycling of mercury with its natural
budget estimated to be 2000 Mg yr"1 and 2) recycling of previously deposited mercury of
anthropogenic origin which is estimated to also be on the order of 2000 Mg yr"1. All of their
natural and re-emitted mercury emissions are in the form of HgO.
Syrakov describes a methodology for incorporating re-emission into a transport model, and this
methodology is used in REMSAD. This method estimates the rate at which mercury becomes
fixed (and therefore unavailable for re-emission) and the rate at which mercury is re-emitted. A
re-emission mass, which is a measure of the amount of mercury that could be re-emitted, is
tracked. Syrakov suggests the following parameterization and constants:
dQav/dt = D + W- areemisQav - afixQav,
dQfiX/dt = afixQav,
RE =
Here D is the dry deposition flux, W is the wet deposition flux, Qav is the re-emission mass, Qfix
is the fixed (unavailable) mass, RE is the re-emission flux, and afix and areemis are fixation and re-
emission coefficients (see Table 2-4).
Table 2-4. Mercury Re-Emission Coefficients.
Coefficient
Sea 0.005 0.000002
Land 0.0002T 0.000002T
is and afix in hr1. T is temperature in C.
areemis and afix over land are zero when T < 0.)
It is clear from the magnitude of these coefficients that the rate of re-emission of newly
deposited material will be much faster than its rate of fixation. (The time to fix half of deposited
mercury mass is on the order of years while the time to re-emit half of the deposited mercury is
on the order of weeks.) Therefore, although conceptually attractive, initializing the Qav mass with
the deposition results of an existing simulation would result in an apparent over estimation of
Qav. (Simulation results show annual deposition of between 10 and 100 g/km2 while Syrakov
estimates Qav at only 0.2 g/km2 over water and between 1 .7 and 3.9 g/km2 over land.) It was
decided therefore to initialize Qav to 0.2 g/km2 over water and 2.0 g/km2 over land. Syarkov's
treatment is followed, except that Qfix is not tracked since it does not affect the evolution of Qav.
Because of the uncertainties inherent in virtually all of the parameters required to implement this
treatment, the base calculation was not made dependent on the re-emission calculation.
However, because of the availability of the mercury tagging species, the re-emitted mercury can
be tracked as a separate species. Calculation of Qav is dependent on deposition of all emissions
2-9 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
and boundary concentrations. Re-emission takes place into one specific tag as elemental
mercury.
2.1.5. Wet Deposition
Wet deposition is the scavenging of gasses and particulates from the atmosphere by
precipitation, and their subsequent deposition (via rainwater) to the surface. Wet deposition is
one of the mechanisms for the removal of pollutants from the atmosphere that is represented in
REMSAD. Separate treatments are used for gasses and particulates.
Gaseous Wet Deposition
The gaseous wet scavenging algorithm in REMSAD is based on Henry's law and specifically
Hales and Sutter (1973). According to Henry's Law, the dissolved concentration of a gas in
water is proportional to the partial pressure of the gas over the water. Mathematically, this can
be expressed as [X\(aq)] = HAPA, where [X\(aq)] is the aqueous concentration of the gas in
mole/liter, HA is the Henry's Law constant for the gas, and PA is the partial pressure of the gas in
atmospheres. Some gases, such as SO2, react with the hydrogen ions present in water and are
effectively more soluble than predicted by the above law. The increased solubility is accounted
for by using the effective Henry's law coefficient, which is dependent on the hydrogen ion
concentration. Expressions for the effective Henry's law coefficients are included in REMSAD
for the standard simulated species. The approach developed by Hales and Sutter involves
calculating the amount of gas dissolved in water and the rate at which the liquid water rains out
of the system. It originally focused on sulfur dioxide. As used in REMSAD, it has been
generalized for any gaseous species (assuming low concentration).
The REMSAD wet deposition algorithm considers six gaseous (NO, NO2, SO2, NH3, VOC,
HNO3) plus seven toxic species. Temperature dependencies for Henry's Law constants are
incorporated for all species. Solubility (KH), ionization (K1D, K2D), and vapor pressure constants
for the toxic species were obtained from recent literature.
The following scavenging rate was derived by Hales and Sutter (for derivation, see below).
RWET = RANM I H(LWC + SOL),
where SOL is solubility, LWC is liquid water content, H is the layer depth, and RANM is rainfall in
m/hr. This scavenging rate is used to adjust (reduce) the species concentration in each model layer.
Scavenging is applied successively to each layer and the total flux (wet deposition to the surface) is
the sum of the mass removed from all layers that extend from near cloud top to the ground.
According to Hales and Sutter, the ratio of the gaseous concentration of a given species to the
liquid-phase concentration (C/C,) in the atmosphere can be expressed as:
SOL=Cg/Cf=l/(KH+^-
Where
C/ = liquid-phase concentration (mol/l water)
Cg = gaseous-phase concentration (atm)
[hT] = concentration of hydrogen ions (mol/l water)
KH = species-dependent Henry's Law equilibrium constant (mol/l-atm)
)KII)K
2-10 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
and KID and K2D are species-dependent first and second ionization constants. For the purpose
of estimating the hydrogen ion concentration, REMSAD assumes a cloud water pH of 4.5. For
the purpose of the current modeling, the species of interest are elemental mercury and divalent
gas mercury. The Henry's law coefficients (at 298 K) used for these species in REMSAD are
shown in Table 2-5.
Table 2-5. Henry's Law Coefficients Used in REMSAD (at 298 K).
Species KH (mol/l-atm)
HGO 0.112
HG2 1.4X106
Conservation of mass requires the following relationship between gas-phase concentration,
liquid-phase concentration, and total (liquid and gas) concentration (C0):
Cg + (Ci x LWC) = C0
where LWC is the liquid water content (of the atmosphere). Substituting from above and
rearranging terms the liquid-phase concentration is expressed as:
Ci = C0/(LWC + SOL)
where LWC is taken from the rain liquid water input data file. The scavenging of pollutant mass
by precipitation can be generally expressed by the equation:
Wi = pw RANM Xi/#
Where
W, = mass of species / material scavenged per unit volume of air per unit time
w = the water density
i = mass of species i scavenged per unit mass of water
H = the layer depth.
The scavenging rate for gaseous removal, RWET, can be expressed by dividing the above
equation for \N\ by the species concentration, C0:
RWET=Wi/C0=(pv*RANM*Xi)/HC0
By substituting into the above expression and using the alternative definition for C/= pw j..
RWET = RANM I H(LWC + SOL)
Particulate Wet Deposition
Wet deposition of aerosols in REMSAD utilizes many of the relationships established by Scott
(1978), which relate rainfall rate and cloud type to fraction of ambient sulfate within rainwater
reaching the ground. The equations have been expanded from sulfate only to treat any aerosol
species. Non-sulfate aerosols are assumed to scavenge at a constant fraction of the sulfate
rate. This fraction can be specified by the user in the CHEMPARAM file and is dependent upon
each species' hygroscopic nature and its affinity to exist with other hygroscopic species.
Settings for this fraction in current CHEMPARAM files are given in Table 2-6.
2-11 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
Table 2-6. Estimated Fractions of Sulfate Wet Scavenging Rate, by Species, as Used in REMSAD.
PN03
GS04
AS04
NH4N
NH4S
SOA
POA
PEC
PMFINE
PMCOARS
1.0
1.0
1.0
1.0
1.0
0.5
0.2
0.2
0.2
0.2
For aerosols smaller than 1|o,m in diameter, it is assumed that the capture of aerosols by phoretic
attachment or Brownian motion is negligible and that the principal scavenging mechanism is the
nucleation of cloud droplets around aerosols followed by particle growth through coalescence and
accretion of cloud droplets to sizes large enough to fall through the cloud as precipitation.
Aerosols larger than 1|o,m are removed strictly by impaction with falling raindrops.
For the portion of clouds in which Bergeron (mixed cloud of ice crystals and liquid water)
processes for rain initiation occurs, it is expected that aerosols do not participate in the
nucleation of cloud ice crystals, and therefore are not present in ice crystals as they coalesce
into larger precipitating crystals. In warm clouds, including both stratiform and convective
clouds, nucleation and coalescence are assumed to be the dominant process for cloud droplet
growth with aerosols acting as nuclei. However, the cloud layer depth over which these
processes occur is treated separately for stratiform and convective clouds.
Relationships between rainfall rate, median drop size and fallspeed, and precipitable water
content have been developed by Scott (1978) and Kessler (1969) as:
V= 130D05
D = 8.95x 10-4R°'21
M= 0.071ft088
where V is fallspeed or velocity (m s"1), D is median drop diameter (m), M is precipitable water
content (g m"3), and R is the rainfall rate converted to mm h"1.
These relationships are used in REMSAD to determine the precipitable water content, drop
diameter, and velocity of the hydrometeor at cloud base and at top of the riming zone. The
depth over which active hydrometeor growth is occurring is estimated based on the particular
layer structure of the cloud and rainfall rate within the cell. A typical residence time t in the
riming zone of 384 s for stratiform clouds and 769 s for convective clouds has been used to
initially estimate the depth of the riming zone. Final residence time is determined once the top of
the riming zone has been determined. The residence time is then used to calculate the vertically
averaged cloudwater mass, m whose relationship is expressed as:
m = (1/C^)(3.12 + 0.88 In R)
2-12 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
where d = 5.2 x 10"3 m3g"V1 (for raindrops) and t is time. The washout rate wrat for the layer
just above the riming zone is given by:
where -t= 435R"071 + 1200 based on continental warm phase-clouds and
M0 = 3.15 x 10-3grrr3.
If the cloud type is stratiform and the layer just above the riming zone is freezing, then the
washout rate is considered negligible and the contribution from this layer is ignored.
To determine the washout rate for each cloud layer within the riming zone, the precipitable
water content Mk is determined for each layer from the expression:
where
Mfc+i = precipitable water within layer k + 1 ,
Mk = precipitable water within layer k,
tk = time hydrometeor is within layer k.
The wrat for each layer k is calculated in an analogous manner to the top layer,
wrai= Mk/mtk
For the layer just below the cloud base down to layer one the wrat removes particles strictly
through impaction with falling raindrops. The expression for their removal is given by:
wrat = 2.14 10-2C7fi°-88
with an assumed inertial impaction efficiency of 0.3.
The effective washout rate fraction for each species is then adjusted logarithmically for the
hygroscopic affinity (fc) and aerosol size distribution of each species and the aerosol available
for incorporation into the cloud water. This washout or scavenging rate is used to adjust
(reduce) the aerosol concentration in each model layer. Washout is applied successively to
each layer and the total flux (wet deposition to the surface) is the sum of the mass removed
from all layers that extend from near cloud top to the ground.
2. 1. 6. Dry Deposition
The dry deposition algorithm in REMSAD is based on the scheme in the Regional Acid
Deposition Model (RADM) as described by Wesely (1989). A more complete description of this
algorithm is provided by Scire (1991). In this methodology, the flux of pollutant material to the
surface (the lower boundary of the modeling domain), F0, is expressed as a product of the
concentration in the lowest model layer (C/) and the deposition velocity (Vd):
Fo = -C, Vd
Thus dry deposition of a given species is directly proportional to the concentration of that
species within the lowest model layer.
2-13 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
The deposition velocity is estimated as an inverse sum of a series of resistances (such that the greater
the resistance, the lower the deposition velocity). For gaseous species this is expressed as follows:
Vd =
Ra+Rb+RS
where Ra is aerodynamic resistance, Rb is boundary-layer resistance, and Rs is surface resistance.
These represent the effects of turbulent diffusion (within the lowest layer), molecular diffusion (that
occurs very near the surface), and finally uptake at the surface (once the surface is reached).
The aerodynamic resistance (fia) is dependent on the surface characteristics and atmospheric
stability conditions. It is calculated from two surface-layer similarity parameters: the friction
velocity and the Monin-Obukhov length (see Gray et al., 1991).
The boundary or quasi-laminar layer resistance (Rb) represents the process of molecular
diffusion of the transport of pollutants through the laminar layer around solid objects and is
highly dependent on the Schmidt number (the ratio of air kinematic viscosity to the molecular
diffusivity of the pollutant in air; see Gray et al., 1991). Note that molecular diffusion is inversely
proportional to the molecular weight.
The surface resistance (fis) is actually a set of parallel resistances associated with (1) leaf
stomata, (2) leaf cuticles, (3) lower canopy resistances (e.g., bark, stems, etc.), and (4) surface
soil, litter, and water (see Wesely, 1989). Surface resistance (resistance to uptake) is both
species and surface dependent.
The deposition velocity of particulate species also depends on particle size distribution and
density. Particles have a sedimentation velocity (Vsed) or fall-out rate that can be a significant
component of the deposition velocity for large particles. Very small particles have a negligible
sedimentation velocity and behave in a manner similar to gases. In REMSAD particle deposition
velocity is calculated as:
T/- T7- I
V d ~ V sed ^
Ra+Rb+RaRbVsed
where Vsed (m/s) is given by the equation
VSed = g dp2 (p - pair) C/18jU,
where p is the particle density (gnT3), pa/ris the air density, g is the acceleration due to gravity (9.8
ms"2), dp is particle diameter (m), and \i is the viscosity of air. C is the slip correction factor given by
C = 1 + 2(lldp)[A1 + A2exp(-A3dpll)l
where / is the mean free path, and A-,, A2, and A3 are 1.257, 0.4, and 0.55 (Friedlander, 1977).
Calculation of Micrometeorological Parameters
Two meteorological scaling parameters (with a basis in similarity theory) are needed for the
calculation of the aerodynamic resistance term used in the dry deposition algorithm. These are the
friction velocity and the Monin-Obukhov length and are calculated within REMSAD from the gridded
wind, temperature, and pressure input fields. These scaling parameters for velocity and length are
essentially invariant with the atmospheric surface layer and enable the calculation of various
turbulence-related effects. The approach to calculation of these parameters is based on similarity
theory. Temperature and pressure for the surface and the lowest model layer are used to calculate
2-14 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
a potential temperature gradient, which is then combined with wind speed for the lowest model layer
to determine stability within the layer. Friction velocity and Monin-Obukov length are then calculated
following the formulation of Louis (1979) for each land-use category. These parameters vary
according to land-use category due to differences in roughness length, which is also considered in
the calculation. Wind speed at a height of 10 m above ground level (agl) is also estimated.
Recognizing that the roughness of a water surface depends on surface stress (i.e., wind speed),
water roughness length is specifically calculated from friction velocity and subsequently used in
the calculation of the resistance terms for grid cells containing water surfaces.
Calculation of Resistance Terms
The surface (10 m) wind speed, friction velocity, and Monin-Obukhov length for each land-use
category within a grid cell are used to calculate aerodynamic and boundary resistances for each
land-use type in that cell. The resistances are combined with the land-use-dependent surface
resistance to obtain a land-use-dependent deposition velocity. The velocities are then weighted
by fractional area covered by each land-use type within the grid cell to obtain a single deposition
velocity for each grid cell for each species. Other key effects that are incorporated into the
calculation of the resistance terms include moisture stress, differences due to water surfaces,
and surface moisture.
EFFECTS OF MOISTURE STRESS ON STOMATAL RESISTANCE
Stomatal resistance, which controls daytime gaseous dry deposition to vegetated surfaces via
the surface resistance term, increases markedly during periods of moisture stress (Scire, 1991).
The deposition algorithm in REMSAD identifies three vegetation states for each grid cell: active
unirrigated vegetation in unstressed conditions or irrigated vegetation (State A), active
unirrigated vegetation in stressed conditions (State B); and inactive vegetation (State C). Of
these states, however, State A is used almost exclusively since data indicating one of the other
two states are usually not available. The resistance is approximated for each state as follows:
For State A, stomatal resistance is parameterized in terms of a reference resistance (which is
season and land-use dependent), solar radiation, and surface air temperature. Solar flux is
calculated as a function of solar zenith angle, and adjusted directly by the percentage of
cloud cover for each cell. A surface air temperature correction factor to stomatal resistance is
also included. Default values for minimum, maximum, and optimum temperatures for
stomatal closing of 0, 40, and 20°C, respectively, are used.
For State B, which by definition corresponds to minimum stomatal opening, stomatal
resistance is arbitrarily set to a multiple of the resistance for State A. The multiplication factor
is equal to 10.
For State C, stomatal resistance is set to a large value (1.0 X 10s) that effectively prevents
deposition.
For applications in which a lack of data does not allow either accurate determination of moisture
stress conditions or the breakdown of irrigated versus unirrigated vegetation (most cases), only
state A is considered.
2-15 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
DEPOSITION TO WATER SURFACES
To accommodate that the deposition to water surfaces can be rapid for many soluble gases a
formulation for surface resistance over water based on the work of Slinn et al. (1978) is used by
REMSAD. In liquid-phase, resistance is given by
R. - -^- (1)
a* ki
where H is the Henry's law coefficient, «*is an effective enhancement of solubility of each gas
in water, and /c, is the liquid-phase transfer velocity, which includes the effects of surface stress.
Slinn et al. (1978) expressed /c, in terms of surface friction velocity u* over water as:
ki = 4.8 x io~4 u*
EFFECTS OF SURFACE MOISTURE
The REMSAD dry deposition algorithm includes modifications to the surface resistances for dew- and rain-
wetted surfaces per Wesely (1989). The extent of dew is estimated internally by the REMSAD based on
relative humidity and wind speed. As suggested by Scire (1991), a formula given by Wesley and Lesht
(1988) is used to determine that dew is present when the quantity: (100 - RH) (u + 0.6) is less than 19,
where RH is the relative humidity (%) and u is the wind speed (m/s). As recommended by Wesley, dew
wetted surfaces have enhanced deposition for SO2 and other soluble species but increased resistance for
ozone. For rain wetted surfaces, resistance to uptake is increased for all species.
2.1.7. Particle and Precursor Tagging Methodology (PPTM) for
Mercury
Using the PPTM approach, mercury species in the emissions and initial and boundary condition
files are tagged and tracked throughout the REMSAD simulation. Tags can be applied to
emissions from selected source regions, source categories, and individual sources, both
separately and in combination. PPTM quantifies the contribution of the tagged emissions
sources (and/or initial/boundary conditions) to the simulated species concentrations and
deposition, for each mercury species considered by the model.
PPTM for mercury tracks emitted mass from its source through the modeling system processes.
Wthin the model, tagging (PPTM) is accomplished by the addition of duplicate model variables for
each species and tag. The tagged species have the same properties and are subjected to the same
processes (e.g., advection, chemical transformation, deposition) as the actual (or base) species.
Typically, each tag includes all of the species necessary to keep track of the mercury emissions from a
particular source or source grouping, but, the different species that comprise mercury emissions (e.g.,
elemental, divalent, and particulate) can also be tagged separately. Because the tagged species are
separate from the base species, PPTM does not alter or affect the base simulation results.
The emissions from each selected source, source category, or grouping are tagged in the simulation
and each grouping is referred to as a "tag." The tagged species are differentiated from the regular
species used in the REMSAD model by a suffix added to the species name. Each individual species
from a given source or source grouping is tagged and the combination of all of the individual species
represents the tag. As an example, in order to track the mercury emissions from incinerators, the
species HGO_t1, HG2J1, and HGPJ1, referring to elemental (HG), divalent (HG2), and particulate
(HGP) emissions from incinerators, will be created. The "t" refers to tagging and the number one is
the tag number. Collectively, these species (are referred to as the incinerator tag.
2-16 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
PPTM was developed to utilize model algorithms as much as possible to track simulated tag
species concentrations. At each time step in the simulation, the effects of linear processes, such
as advection and dry deposition, are calculated directly for all tagged species. Potentially non-
linear processes, such as gas-phase chemistry, aqueous chemistry, and particle dynamics are
calculated for the overall (or base) species and apportioned to the tagged species. The results for
the tagged species are not normalized to ensure that the sum of the tagged species equals the
total. Thus, the difference between the sum of all tags and the overall concentration gives an
estimate of the numerical uncertainty in the calculated contribution.
Some example uses of the mercury PPTM methodology include 1) quantifying the contribution of
mercury emissions from various source sectors to mercury deposition at selected locations throughout
the modeling domain, 2) quantifying the contribution from boundary conditions to mercury deposition
throughout the modeling domain, 3) examining the range of influence of emissions from selected
facilities, and 4) tracking the fate of mercury emissions from a particular source category estimate the
contribution to deposition to water bodies throughout the modeling domain.
2.1.8. Summary of Outputs and Information Provided by REMSAD
Key REMSAD output files contain information on the simulated concentrations and deposition
totals. Specifically, the average file contains time-averaged (typically hourly averaged)
concentrations for each species for each grid cell for the entire modeling region. The deposition
file contains wet and dry deposition (g/km2) for a selected output interval for each species in
each grid cell for the entire modeling domain.
The simulation results are typically displayed using spatial distribution plots and a variety of
other graphical analysis products. The base case simulation results are compared with
observed data using scatter plots and a variety of statistical measures.
The tagged species are included as additional species in the model output files and the results
can be post-processed and displayed in the same way as the standard species. Spatial plots of
the tagged species can be used to show the extent and magnitude of the contributions from the
tagged sources within the modeling domain. The contribution from each tag at individual
locations throughout the domain can also be extracted from the gridded model output. Finally,
the tags can be summed and compared with the base simulation results to quantify the
numerical accuracy of the results.
2.1.9. REMSAD/PPTM Application Procedures
Application of REMSAD with PPTM for a single base year includes the following steps:
REMSAD Application Procedures
Select a modeling domain (considering extent and horizontal and vertical grid resolution)
Select a simulation period (typically an annual period, and preferably with typical (not
extreme) meteorological conditions)
Prepare emissions inventory, meteorological, and initial and boundary condition, and
geographical input files for the selected domain and simulation period
Apply REMSAD for criteria pollutants (ozone, PM, etc.) and evaluate model performance
Apply REMSAD for mercury and evaluate model performance.
2-17 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
PPTM Application Procedures
Identify sources for tagging, conceptually and within the emissions inventory
Prepare the tagged emissions inventory and initial/boundary condition files.
Apply REMSAD for mercury only with PPTM for up to approximately 20 tags (repeat, as
needed, to accommodate all identified tags)
Post-process and analyze the PPTM contributions.
Other applications of REMSAD may include future-year emissions projections and model runs.
Tagging can be applied for the future-year scenarios as well.
2.2. CMAQ Modeling System
The CMAQ model is a state-of-the-science, regional air quality modeling system that is designed to
simulate the physical and chemical processes that govern the formation, transport, and deposition of
gaseous and particulate species in the atmosphere (Byun and Ching, 1999). The CMAQ model was
designed as a "one-atmosphere" model and can be used to simulate ozone, particulate matter, and
mercury. For mercury, CMAQ supports the detailed simulation of the emission, chemical
transformation, transport, and wet and dry deposition of elemental, divalent, and particulate forms of
mercury. Version 4.6 of CMAQ was used for this study.
According to Bullock et al. (2008), the CMAQ model reflects the current state-of-the-science in
simulating the atmospheric processes that influence the dispersion, advection, chemical
transformation, and deposition of mercury. The CMAQ model includes three mercury (Hg) species;
elemental mercury (Hg° or HG in CMAQ), reactive gaseous mercury (RGM or HGIIGAS in CMAQ),
and particulate-bound mercury (PHg or APHGJ and APHGI in CMAQ). Reactive gaseous mercury
is known to be comprised almost entirely of divalent mercury (Hg2+), since Hg compounds at other
valence states tend to be chemically unstable in the atmosphere. Particulate-bound mercury is also
primarily comprised of divalent mercury, but may also include elemental mercury.
Mercury simulation capabilities were first incorporated into the CMAQ model by adding gaseous
and aqueous chemical reactions involving mercury to the CMAQ chemical mechanism (Bullock
and Brehme, 2002). Since that time, the chemical mechanism has been further updated to
include additional reactions and updated information on reaction rates. The most recent
changes to CMAQ for mercury include an updated dry deposition algorithm and the
incorporation of natural mercury emissions. The CMAQ modeling system, including the mercury
modeling component, has been peer reviewed (e.g., Amar et al., 2005).
In addition to the state-of-the science chemical mechanism for mercury, other key features of
the CMAQ model in simulating mercury deposition include state-of-the-science advection,
dispersion and deposition algorithms, the latest version of the Carbon Bond chemical
mechanism (CB05), and the CMAQ Particle and Precursor Tagging Methodology (PPTM).
PPTM for mercury (Douglas et al., 2006) provides detailed, quantitative information about the
contribution of selected sources, source categories, and/or source regions to simulated mercury
concentrations and (wet and dry) deposition. Mercury emissions from selected sources, source
categories, or source regions are (numerically) tagged and then tracked throughout a
simulation, and the contribution from each tag to the resulting simulated concentration or
deposition for any given location can be quantified. By tracking the emissions from selected
sources or source locations, the methodology also provides information on the fate of the
emissions from these sources.
2-18 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
The CMAQ model has been used by EPA to support the development of the Clean Air Mercury
Rule (CAMR) (EPA, 2005a). This study included the evaluation of global modeling results to
prescribe boundary conditions for CMAQ, evaluation of simulated mercury deposition vs. MDN
data, and assessment of the contribution of mercury emissions from coal-fired power plants on
mercury deposition in the U.S.
CMAQ was also included in the North American Mercury Model Intercomparison Study
(NAMMIS) for mercury (Bullock et al., 2008) and the performance and response of CMAQ was
found to be reasonable and also consistent with that for REMSAD).
Additional information on the CMAQ modeling system can be found in Byun and Ching (1999).
2.3. CTM, GRAHM and GEOS-Chem Models
Three global simulation models provided boundary conditions for the continental scale
modeling. The results from these models were made available as part of the North American
Mercury Model Inter-comparison Study (NAMMIS) (Bullock et al., 2008). The three models
include the Chemical Transport Model (CTM), the Global/Regional Atmospheric Heavy Metals
model (GRAHM), and the GEOS-Chem model. Results from these three global modeling tools
were used to prepare three estimates of boundary concentrations of elemental mercury, divalent
gas mercury, and particulate mercury for the REMSAD and CMAQ simulations conducted for
this study and also for the CMAQ simulations conducted by EPA and used to support this study.
According to Bullock et al. 2008), all three of the global models are based on reasonable
scientific definitions and assumptions.
The CTM model (Shia et al., 1999; Seigneur et al., 2001) simulates the emission and transport
of mercury, mercury transformation processes, and wet and dry deposition, particularly of HG2
and HGP. To generate initial and boundary conditions for regional-scale modeling, the CTM is
run for several years using the same set of annual meteorological conditions until a steady state
is achieved. As used in this study, CTM was applied with a horizontal resolution of 8 degrees
latitude by 10 degrees longitude, and nine vertical layers extending to the stratosphere. For this
study, monthly average CTM concentrations were calculated and used in preparing the initial
and boundary conditions for the REMSAD (and CMAQ) model simulations.
The GEOS-Chem model (Selin et al., 2007) also simulates the emission and atmospheric
transport of mercury on the global scale. It includes HGO, HG2, and primary HGP. The
chemistry includes HGO oxidation to HG2 by OH and ozone as well as aqueous-phase
photochemical reduction of HG2 to HGO (Bullock et al, 2008). Wet and dry deposition are also
simulated. GEOS-Chem version 7.01 (http://www-as.harvard.edu/chemistry/trop/geos/) was
used for this study with a horizontal resolution of 2 degrees latitude by 2.5 degrees longitude.
For this study, 3-hour outputs from GEOS-Chem were used to calculate monthly average
concentrations and prepare the initial and boundary conditions for the REMSAD (and CMAQ)
model simulations.
The Global/Regional Atmospheric Heavy Metals (GRAHM) model (Dastoor and Larocque, 2004;
Ariya et al., 2004) is an extended version of the Canadian operational weather forecasting
model, the Global Environmental Multiscale (GEM) model. The GRAHM model simulates HGO,
HG2, and HGP. The model simulates the oxidation of HGO by ozone to HG2 and HGP, several
aqueous-phase chemical transformations, and wet and dry deposition, primarily of HG2 and
HGP. As used in this study, GRAHM was applied with a horizontal resolution of 5 degrees
latitude by 5 degrees longitude, and 28 vertical layers. For this study, 6-hourly outputs from
2-19 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
GEOS-Chem were used to calculate monthly average concentrations and thus prepare the
initial and boundary conditions for the REMSAD (and CMAQ) model simulations.
2-20 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Description of the Deposition Modeling Tools
This page deliberately left blank.
2-21 August 2008
-------
3. Meteorological and Geographical Inputs
In this section, the meteorological inputs and the data and methods used to prepare the
geographical input files for the application of REMSAD are discussed. Similar datasets and
methods were used to prepare the meteorological and geographical inputs for the supporting
application of CMAQ.
3.1. Meteorological Inputs
For this application of REMSAD, the existing meteorological inputs used are those that were
developed by EPA for use in their evaluation of emissions rules. The meteorological data files
were prepared by EPA using the outputs from the Fifth Generation Pennsylvania State
University/National Center for Atmospheric Research (PSU/NCAR) Mesocale Model (MM5).
The gridded meteorological fields provided by EPA cover the full REMSAD modeling domain
with approximately 36-km horizontal resolution. The simulation period includes the calendar
year 2001 (plus the last several days of December 2000, as a model "spin-up" period). These
2001 meteorological input fields were used in the EPA's evaluation of the Clean Air Interstate
Rule (CAIR) and the Clean Air Mercury Rule (CAMR) (EPA, 2005a and b). The MM5-derived
meteorological fields were mapped directly to the REMSAD grids shown in Figure 1-1, such that
the 36-km MM5 results were interpolated to the 12-km REMSAD grids and used directly to
specify the meteorological inputs throughout the remainder of the modeling domain.
Since these input fields were evaluated by EPA and utilized in prior applications, the files were
not subjected to extensive evaluation or quality assurance. However, some comparisons with
observed data and with a set of alternative meteorological input fields derived from Rapid
Update Cycle (RUC) model output for 2001 were made as part of the meteorological dataset
selection process (Douglas et al., 2005). This evaluation showed that the relative performance
of the two models varies by month, by geographic region, and among the key meteorological
parameters. For consistency with other EPA studies (including the CAMR modeling) and to
possibly take advantage of higher-resolution MM5-based meteorological fields for future
applications, the MM5-derived fields were selected for use in this study.
3.1.1. Description of the Meteorological Inputs
Simulated MM5-derived surface pressure patterns for the 15th day of each month were visually
compared with the surface weather analyses prepared by the National Weather Service (NWS).
Of interest are the locations of high and low-pressure systems, the simulated and observed
surface pressure values near the locations of the primary highs and lows, and the overall
patterns across the continental U.S.
The MM5-derived surface pressure fields represent the overall pressure patterns and the
general distribution of high and low pressure centers. The MM5 results show good agreement
with the observed patterns and pressures for the middle days of January through April and
December, and less skill during the summer and autumn months. Overall, the MM5-derived
fields show less spatial variability in surface pressure than is suggested by the data (i.e., the
observed range in pressure is not captured).
The MM5-derived heights corresponding to constant pressure surfaces for 850 and 700 mb
were compared with analyses from the National Oceanographic and Air Administration (NOAA)
Air Resources Laboratory (ARL). A comparison time of 0700 EST was used, since the upper-air
data are available at that time. The analyses were derived using upper-air radiosonde
observations. The high and lows in the height fields reflect high and low pressure systems aloft.
MM5 depicts the range of patterns (including the distinctive ridge/trough and zonal patterns) that
characterize the middle days of each month. The high-pressure ridge over the central U.S. in
mid-July and mid-September is depicted by MM5.
3-1 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Meteorological and Geographical Inputs
Observed and MM5-derived annual rainfall totals are presented and compared in Figure 3-1.
The observed precipitation plot was obtained from the National Atmospheric Deposition
Program (NADP) web site (http://nadp.sws.uius.edu) (NADP, 2005).The annual rainfall patterns
from the MM5 model show the highest precipitation areas over British Columbia, Canada and
over the Atlantic Ocean (off the coast of Florida and northward). The MM5 results show a
significant amount of precipitation over the southern Appalachian Mountains, along the eastern
seaboard, and in southern Arizona. Compared to observation-derived precipitation amounts, the
model appears to overestimate total precipitation over the interior western states (the amounts
are better in line with the observations for the coastal western states). MM5 produces far too
much precipitation in southern Arizona but gives reasonable annual precipitation totals for the
central states and throughout the Northeast. MM5 underestimates the area of highest observed
annual precipitation over Louisiana, Mississippi, southern Arkansas, southwestern Tennessee,
and eastern Texas, and significantly overestimates precipitation over Alabama, Georgia, the
Carolinas, and portions of peninsular Florida.
Figure 3-1 a. Observed Annual Rainfall Totals (cm) for 2001.
Total precipitation, 2001
Sites not pictured:
AK01 35 cm
AK03 24 cm
HI99 374 cm
VI01 105 cm
National Atmospheric Deposition Program/National Trends Network
http://nadp.sws.uiuc.edu
3-2
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Meteorological and Geographical Inputs
Figure 3-1 b. MM5-derived Annual Rainfall Totals (cm) for 2001.
LEVEL 1 RAIN (cm)
Time: 0 Apr 11, 2010-0 Apr 10, 2020
100 -
+ MAXIMUM = 544.4 cm (145,5)
- MINIMUM = 0.4 cm (30,32)
15B4. 2304.
2736. -2016. -1296. -576.
30
40
60
80
100
130
140
1512.
792.
73.
-648.
-1368.
-2088.
-j
C >
0 ?0 Q
O -O -O -O o O O -O O
Total Precipitation for 2001 MM5 REMSAD input file (cm)
3-3
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Meteorological and Geographical Inputs
Figure 3-2 shows the variation in monthly mean rainfall amounts, observed and simulated by
MM5 for five geographically diverse areas (Portland, ME; Baltimore, MD/Washington, D.C.;
Baton Rouge/Slidell, LA; Madison/Green Bay, Wl; and Oakland, CA). The month-to-month
variations and rainfall amounts are generally well represented.
Figure 3-2. Monthly Average Rainfall amount (in)
Based on Observed and Simulated Daily Precipitation Values.
(a) Portland
Portland
-Obs
-MM5
10 11 12
(b) Baltimore
Baltimore
-Obs
-MM5
10 11 12
3-4
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Meteorological and Geographical Inputs
(c) Baton Rouge
ra
Baton Rouge
Obs
-MM5
10 11 12
Month
(d) Madison
-= 0.3
Madison
-Obs
-MM5
10 11 12
3-5
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Meteorological and Geographical Inputs
(e) Oakland
-=- 0.3
Oakland
Obs
MM5
11 12
Figure 3-3 provides mean observed and simulated 850 mb upper-air temperatures for the time
of the morning observation for these same sites. For all five sites, the input fields agree very
well with the observed upper-air temperature data for the 850 mb level. The MM5-based
temperatures tend to be slightly higher than observed.
Figure 3-3. Monthly Average Observed and Simulated 850 mb Temperature (°C)
for the Time of the Morning Sounding.
(a) Portland
o 10
oo
Portland
-Obs
MM5
Month
3-6
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Meteorological and Geographical Inputs
(b) Baltimore
o
.a
E
o
in
00
Baltimore
Obs
-MM5
Month
(c) Baton Rouge
20
O 15
o
in
00
Baton Rouge
Obs
MM5
6 7
Month
10 11 12
3-7
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Meteorological and Geographical Inputs
(d) Madison
20
15
O 10
o
in
00
-5
-10
-15
Madison
Obs
MM5
_5 R
R Q
-44-
Month
(e) Oakland
Oakland
Obs
-MM5
10 11 12
3-8
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Meteorological and Geographical Inputs
Figure 3-4 displays the mean observed and simulated 850 mb upper-air dew-point
temperatures, also for the morning observation time. For four of the five example sites
(Portland, Baltimore/Washington, Baton Rouge/Slidell, and Madison/Green Bay), the input fields
agree well with the observed upper-air dew-point temperature data for the 850 mb level. For
Oakland, MM5 does not capture the lower dew-point temperatures and thus the drier air over
this part of the country during the summer months.
Figure 3-4. Monthly Average Observed and Simulated 850 mb Dew-Point Temperature (°C)
for the Time of the Morning Sounding.
(a) Portland
Portland
-Obs
MM5
in -
00
Month
(b) Baltimore
Baltimore
Obs
MM5
O
Month
3-9
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Meteorological and Geographical Inputs
(c) Baton Rouge
Baton Rouge
Obs
MM5
Month
(d) Madison
Madison
Obs
MM5
Month
3-10
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Meteorological and Geographical Inputs
(e) Oakland
Oakland
-Obs
MM5
Month
Figure 3-5 compares the bias and error values for 850 mb wind speed for the time of the
morning sounding. The MM5-derived wind speeds tend to be higher than observed, with a bias
on the order of 2 to 4 ms"1.
Figure 3-5. Monthly Average Bias and Error Statistics for 850 mb Wind Speed (ms"1)
for the Time of the Morning Sounding.
(a) Portland
Portland
-MM5 (Bias) - - - MM5 (Error)
Month
3-11
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Meteorological and Geographical Inputs
(b) Baltimore
A
1
*-2 o
E. 2
(0
> n
> U
E
O 9
10 *-
CO
Baltimore
MM5 (Bias) - -- - MM5 (Error)
^^1 "^
>^J * *~ ^
y*-^/ ^^^.^-^^^^^B^,
^/23455789 T^.
*-^^
:1/^
w
Month
(c) Baton Rouge
Baton Rouge
MM5 (Bias) - -- - MM5 (Error)
Month
3-12
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Meteorological and Geographical Inputs
(d) Madison
Madison
MM5 (Bias) - - - MM5 (Error)
00
Month
(e) Oakland
-Sf 4
1 2
£
§ 0
.a
£ -2
o
10
co -4
Oakland
-MM5 (Bias) - -- - MM5 (Error)
Month
3-13
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Meteorological and Geographical Inputs
Figure 3-6 compares the bias and error values for 850 mb wind direction, again for the time of
the morning sounding. A bias on the order of 20 to 40 degrees appears to be present in the
MM5 fields, although this varies by site and by month.
Figure 3-6. Monthly Average Bias and Error Statistics for 850 mb Wind Direction (degrees)
for the Time of the Morning Sounding.
(a) Portland
0)
0)
o
a
>
.a
F
o
10
fin -i
40 _
on
on
40
fin
«--*-'
1 2
Portland
MM5 (Bias) - »- - MM5 (Error)
*.---+-+-. + --+'^*^' +
34567891
Month
-^ ^
D 11 12
(b) Baltimore
fin -,
O)
0)
"° 9n
a
> n
> u
Eon
o
10 .
fin
Baltimore
MM5 (Bias) - - - MM5 (Error)
^^" * **^ ""' » ^
..^--^
4 * "
1 2 3 4 5 6 7 8 9 10 11 12
Month
3-14
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Meteorological and Geographical Inputs
(c) Baton Rouge
fin
O)
0)
a
Eon
o
8 40
fin
Baton Rouge
MM5 (Bias) - - - MM5 (Error)
^1
- -*. ^'^ * *""^-^-..
"^ "«'
1 2 3 4 5 6 7 8 9 10 11 12
Month
(d) Madison
fin -i
o
0)
Q
> n
> U
£}
Eon
o
fin
Madison
MM5 (Bias) - -- - MM5 (Error)
^ -*-«--*-.
123456789 10
Month
11 12
3-15
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Meteorological and Geographical Inputs
(e) Oakland
3
0)
o
a
&
E
o
in
fin -i
Aft
on
on
Aft
fin
« »-.,
1 2
Oakland
MM5 (Bias) - - - MM5 (Error)
*---*^ ,>»"""""^-^
' **- -A"*
34567891
Month
-'*
D 11 12
3.2. Preparation of Geographical Inputs
Geographic input fields used by the REMSAD model are terrain height and land-use category.
The terrain heights were derived from terrain data provided by EPA (along with the MM5-
derived meteorological files). Land-use information, which defines the fractional coverage of
several different land-cover types for each grid cell, was derived from USGS Land Use and
Land Cover (LULC) data. The land use data used for this project are available at approximately
200 m horizontal resolution. A description of the data is available at http://edcwww.cr.usgs.gov/
products/landcover/lulc.html. Land-use inputs were prepared for the REMSAD 36- and 12-km
grids. The 200 m data were averaged over the area of each 36-km or 12-km grid cell in order to
derive the land use fractions within each grid cell.
The land use categories used by REMSAD and the associated surface roughness lengths for
each category are presented in Table 3-1.
Table 3-1. Land-Use Categories Recognized by REMSAD.
Category
Number
1
2
3
4
5
6
7
8
9
10
11
Land-Use
Category
Urban
Agricultural
Range
Deciduous forest
Coniferous forest including wetland
Mixed forest
Water
Barren land
Nonforest wetlands
Mixed agricultural and range
Rocky (low shrubs)
Surface
Roughness
(meters)
3.00
0.25
0.05
1.00
1.00
1.00
0.0001
0.002
0.15
0.10
0.10
3-16
August 2008
-------
4. Emissions Inputs
In this section, the preparation of the REMSAD-ready emission inventories are summarized for
both the criteria pollutants and mercury. Similar datasets and methods were used to prepare the
emissions inputs for the application of CMAQ (as applied for this study). The detailed
corrections to the emission inventories were specific to this study and were not included in
CMAQ simulations conducted by EPA.
4.1. Emission Inventory Preparation for Non-mercury
Participate and Gaseous Species
The REMSAD base emissions inventory for the criteria pollutants includes anthropogenic and
biogenic emissions for the species listed in Table 4-1.
Table 4-1. REMSAD Emissions Species for the Criteria Pollutant Emissions Inventory.
REMSAD (Version 6) Emissions Species Species ID
Nitrogen oxide NO
Nitrogen dioxide N02
Primary organic aerosols POA
Primary elemental carbon PEC
Gaseous sulfate GS04
Participate nitrate PN03
Volatile organic carbon VOC
Sulfur dioxide S02
Particulate matter with a diameter PMFINE
less than 2.5 microns
Particulate matter with a diameter greater than PMCOARS
2.5 but less than 10 microns
Ammonia NH3
Carbon monoxide CO
Carbonyl CARB
Monoterpenes TERP
Isoprene ISOP
4.1.1. Emissions Data
The REMSAD base emissions inventory for the criteria pollutants was prepared using the EPA
2001 CAIR database including the emissions data and associated PM speciation profile and
cross reference files (except the splits for VOC and CARB), temporal profile and cross-
reference files, and surrogate data and cross-reference files. The various CAIR emissions
inventory files were converted to the formats required for processing with EPS2.5. A brief
description of each emission component is provided below.
4-1 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
U.S.
AREA SOURCE DATA
2001 area source data includes the following sectors: fugitive dust, agricultural (NH3
emissions from livestock and fertilizer application), fires (2001-specific wildfires, prescribed
burning, agricultural burning, and open burning), and other area (airport and onroad mobile
refueling emissions, and other area source emissions).
County-specific transportable fraction adjustments were applied to the emissions included in
the fugitive dust sector.
NONROAD MOBILE SOURCE DATA
2001 nonroad mobile source data includes monthly emissions calculated by the National
Mobile Inventory Model (NMIM), and 2001 annual emissions for airports, railroads,
commercial marine vessels (ARM).
ONROAD MOBILE SOURCE DATA
2001 onroad mobile source data includes monthly emissions calculated by the National
Mobile Inventory Model (NMIM) for all states except for State of California; and annual
mobile source emissions data for State of California.
California-specific temporal profiles were applied to the annual emissions to obtain the
monthly emissions.
POINT SOURCE DATA
2001 point source data includes point source and fugitive dust sectors.
County-specific transportable fraction adjustments were applied to the emissions included in
the fugitive dust sector.
Canada
Area, nonroad, and mobile source data are for 1995.
Point source data are for 1995 and only available for Eastern Canada.
Mexico
Area, nonroad, onroad mobile, and point source data were as used in the CAIR modeling
studies (see EPA. 2004. CAIR Emissions Inventory Overview).
Offshore
Offshore emissions data for a small portion of the Gulf of Mexico provided by the Texas
Commission on Environmental Quality (TCEQ). (Emissions for the bulk of sources in the Gulf
are not included.)
Offshore point source emissions were prepared for each season.
4.1.2. Emissions Processing
The emissions were processed using version 2.5 of the Emissions Preprocessing System
(EPS2.5). EPS2.5 consists of series of computer modules that incorporate spatial, temporal,
4-2 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
and chemical resolution into an emission inventory used for modeling. Anthropogenic point-,
area-, and mobile-source emissions data are processed separately through the EPS2.5 system.
The key processing steps include:
Chemical speciationPoint, area, and mobile source emissions are chemically speciated
from VOC into the CB-V species.
Temporal allocationEmissions are temporally allocated based on temporal profiles. These
include monthly, weekly, and/or diurnal profiles for on-road motor-vehicle emissions, and in
some cases, operating schedule (months/year, days/week, hours/day, and start hour)
information for point sources.
Spatial allocationPoint-source emissions are directly assigned to grid cells based on the
source location coordinates included in the input emissions data for each source. Area- and
mobile-source emissions are allocated to grid cells using gridded spatial allocation surrogates.
For this study, area source emissions were prepared for each month and spatially allocated in
the modeling domain using the EPA provided surrogates. The nonroad mobile source emissions
were prepared for each month except for the ARM data (which were prepared for each season)
and spatially allocated in the modeling domain using the EPA provided surrogates. The onroad
mobile source emissions were prepared for each month and spatially allocated in the modeling
domain using the EPA provided surrogates. Point source and offshore emissions were prepared
for each season.
Following processing of each inventory component, the area-, mobile-, and low-level point
source emissions were merged with the biogenic emissions to form the low-level emissions
input file for REMSAD. Emissions associated with elevated points sources were incorporated
into the separate point-source emissions file. In this file, each stack or facility is treated as a
separate emissions source.
4.1.3. Summary of the Criteria Pollutant Emissions
The 2001 base-year criteria pollutant emissions for the U.S. portion of the modeling domain are
presented by major source category and by season in Table 4-2. This table highlights the key
source categories for each component and the variations in the emissions throughout the year.
Table 4-2. Average Seasonal Daily Emissions (tpd) for the U.S. Portion
of the REMSAD Modeling Domain: 2001 Base Case.
Species
NOx
VOC
S02
PM2.5
PMC
NH3
CO
NOx
VOC
S02
Point
21,304
4,448
38,157
3,606
978
249
12,135
21,398
4,466
37,647
Area Mobile
Winter
5,508
23,984
4,451
9,453
22,807
5,333
34,061
Spring
4,453
21,257
3,466
32,224
19,465
1,699
996
169
707
242,859
33,121
19,418
1,897
Biogenic
2,626
28,280
6,842
6,643
120,914
Total
61,662
76,178
44,308
14,055
23,955
6,289
295,897
65,615
166,054
43,011
4-3 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
Species
PM2.5
PMC
NH3
CO
NOx
VOC
S02
PM2.5
PMC
NH3
CO
NOx
VOC
S02
PM2.5
PMC
NH3
CO
Point
3,519
986
246
12,028
22,254
4,481
38,722
3,588
1,020
248
12,111
21,462
4,514
37,452
3,575
999
250
12,210
Area Mobile Biogenic Total
9,053
23,101
11,212
34,638
Summer
3,618
19,161
2,817
8,642
23,529
11,031
34,703
Fall
4,123
20,516
3,398
8,106
22,983
8,720
25,248
1,235
185
756
225,649
34,800
24,160
2,221
1,677
217
832
233,602
32,764
19,054
1,909
1,233
185
767
218,147
13,808
24,272
12,214
20,066 292,381
9,070 69,742
325,886 373,688
43,760
13,907
24,766
12,111
46,824 327,240
5,551 63,900
117,805 161,890
42,758
12,914
24,167
9,737
23,244 278,848
The spatial distribution of emissions throughout the domain is illustrated in Figures 4-1 and 4-2.
These examples focus on NOX and SO2 emissions for a typical summer weekday, but the spatial
distributions are characteristic of most anthropogenic species. Figures 4-1 a and b show the
distribution of low-level sources of NOX and SO2, respectively. Figures 4-2a and b illustrate the
spatial distribution of NOX and SO2 emissions from elevated (point) sources. In the plots, the
different colors designate different ranges of total daily emissions. Note that the NOX and SO2
plots use different intervals for the color representation of the emissions. Figures 4-1 and 4-2
display the emissions for the 36-km (coarse) grid. Emissions files were also produced for the
12-km resolution nested grids (not shown).
4-4 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
Figure 4-1 a. Low-level NOX Emissions (tons) for the 36-km Grid for a Summer Weekday.
Max value: 300303.6 (kg/day) at (130. 66)
-2736 -2016 -1296
1 1 Qjiilim lnjnvwil|ri'Hini !'l/b>
Km
-570
144
804 1584 2304
1000
1368
IliL-lfHiT lilln .1 ii l l:li mil ill
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
-2088
|LowLevel NOx Emissions: Summer Weekday
g o O g 0 N0x
°° >°OcT **£ *0oo s°0o *<*£ '0% Total: 463B94BO (Kg/day)
4-5
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
Figure 4-1 b. Low-level SO2 Emissions (tons) for the 36-km Grid for a Summer Weekday.
Max value: 750419.4 (kg/day) at (110. 62)
Km
-2736 -2016 -1296 -576
144
864 1584 2304
E 1000
11 11 wifiTrJTi 11111: liiiiiiilimiilin mil in
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
-1656
-2088
o
5c°
I LowLevel SOS Emissions: Summer Weekday
S02
Total: 10404030 (Kg/day)
4-6
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
Figure 4-2a. Elevated Point-source NOX Emissions (tons) for the 36-km Grid for a Summer Weekday.
Max value: 304336,3 (kg/day) at { 67. 24)
Km
-2736 -2016 -1296 -576 144 864 1584 2304
i i Q}-I11111 i\wiI*AII11' 'Tiiirrrn \\\\\\ IIITTTTIII \\
1 00
eoi-
40
10
in\[\ iluiLjiilii
llllllHlllllllllfi
0 10 30 30 40 50 60 70 80 90 100 110 120 130 140
1000
1368
936
504
72
-360 |
i
-792
-1224
-1656
-2088
ivated Point NOx Emissions: Summer Weekday
NOx
Total: 13578840 (Kg/day)
4-7
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
Figure 4-2b. Elevated Point-Source SO2 Emissions (tons) for the 36-km Grid fora Summer Weekday.
Max value: 769670.9 (kg/day) at (111. 62)
Km
-2736 -2016 -1296 -576
110
144
864
1584
2304
MIII:illlli*rtlTnTillllll!i ! HI 1:1 i 'ill ill
'0 10 30 30 40 50 60 70 80 90 100 110 120 130 140
-2088
n -
°° 7
°0
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
and mercury emissions data should target the same year to the extent possible. This
recommendation was made in order to facilitate model validation by comparisons to measured
data. The meteorological data used in this project was the same as that used by EPA as part of
the Clean Air Mercury Rule (CAMR) modeling and represented the year 2001. However, with
the exception of medical waste incinerator emissions, other mercury emissions from the CAMR
were from the year 1999; medical waste incinerator emissions in CAMR were from 2002.
Therefore, summaries of the mercury emissions from the CAMR inventory were prepared for
each state and asked the EPA Regional offices to lead a review of those data, with participation
from their member states where necessary, and determine the representativeness of the
emissions for the year 2001. The following discussion details the changes, additions, and
deletions made to the CAMR inventory as a result of this effort to produce a more accurate
picture of 2001 emissions. A summary of state specific, speciated mercury emissions is
presented in Section 7.
4.2.1. Review and Revision of the CAMR Mercury Emissions
Inventory
Detailed summaries of the top emitters in the CAMR mercury inventory were prepared for each
state, and this information was provided to the appropriate EPA regional offices and state
agencies for review. In some cases, the state agencies asked to be allowed to conduct a more
detailed review of emissions beyond the list of top emitters. In these cases, the states were
provided the complete inventory for the state, and changes to the inventory were incorporated
as directed by the state. For cases where these changes were extensive, detailed lists of the
changes are included in Appendix C. The revisions to the mercury emissions are summarized in
the remainder of this section, by EPA region and by state. For those regions/states not listed
here, there were no revisions.
EPA Region 1
STATE OF CONNECTICUT
Revisions were made to the top five emitters of divalent gaseous mercury and the top five
emitters of total mercury in the state based on the Connecticut Department of Environmental
Protection (CTDEP) 2001 mercury emission estimates. The revised emissions were provided by
NESCAUM (2006a) and CTDEP (2006).
In addition, per the request of CTDEP, the speciation profile used for Naugatuck Treatment Co.
and Mattabassett Regional Sewage Authority was changed from the default (50 percent
elemental, 30 percent gaseous divalent, and 20 percent particulate mercury) to that for sewage
sludge incineration (22 percent elemental, 58 percent gaseous divalent, and 20 percent
particulate mercury). In the CAMR database, the MACT code assigned to the facilities is 00000,
and the SCC code is for miscellaneous industrial process.
The revisions resulted in total mercury emission increases of 0.04 tons/year for non-IPM point
sources, compared to the CAMR base emissions for Connecticut.
STATE OF MAINE
The emissions for Holtra Chemical Manufacturing Co. were removed from the inventory (total
mercury emissions in the CAMR database for the facility is 0.065 tons/year). The chlor-alkali
plant closed prior to 2001 (NESCAUM, 2006a).
STATE OF NEW HAMPSHIRE
Revisions were made to the top five emitters of divalent gaseous mercury and the top five total
mercury emitters in the state, based on the New Hampshire Department of Environmental
Conservation (NHDEC) 2001 mercury emission estimates (NESCAUM, 2006b). The mercury
4-9 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
speciation profile was also revised for coal-fired units for two point sources (Merrimack and
Schiller) using NHDEC's stack test data.
The revisions resulted in 0.04 tons/year total mercury emissions increases for IPM sources, and
0.06 tons/year total mercury emissions decreases for non-1 PM sources.
The location for Design Contempo Inc provided in the CAMR database puts the facility outside the
state. The coordinates were changed to the centroid of Lisbon, NH where the facility is located.
STATE OF VERMONT
The Vermont Air Pollution Control Division (VT APCD) provided 2001 mercury emissions
estimates for the top ten fuel-combustion facilities and the top five process/manufacturing
facilities in the state with total mercury emissions of about 7 Ibs per year (NESCAUM, 2006c).
Four of the 15 facilities from the VT APCD list could be matched to sources in the CAMR
database, and the emissions were revised (total emissions from the four facilities are 1.7
Ibs/year). The emissions for remainder of the sources were not added to the 2001 inventory,
since the emissions may already be included in the CAMR database as non-point sources.
EPA Region 2
STATE OF NEW JERSEY
The revisions were made to the point- and area-source emissions based on the New Jersey
Department of Environmental Protection (NJDEP) 2001 mercury emission estimates (NJDEP, 2006a).
Revisions are summarized below. For additional detail, see Appendix C.
The revisions for the point sources included the following:
Total mercury emissions were revised for several facilities. The revisions resulted in a net
increase total mercury emissions of 0.03 tons/year (resulting from an increase of 0.14
tons/year for IPM sources and a decrease of 0.11 tons/year for non-IPM point sources).
In the CAMR database, the MACT code assigned to U. S. Pipe & Foundry Co. and CO Steel
Raritan is 0107 for Industrial/Commercial/lnstitutional Boilers & Process Heaters. Thus the
mercury emissions of the facilities were speciated as 50, 30, and 20 percent elemental,
gaseous divalent, and particulate mercury, respectively. The speciation for CO Steel Raritan
was changed to the profile for iron and steel foundries, which is 80, 10, and 10 percent for
elemental, divalent gaseous, and particulate mercury, respectively. In addition, the speciation
for U. S. Pipe & Foundry Co. was changed based on stack test data for the facility (to 62.3
percent elemental, 37.5 percent divalent gaseous, and 0.2 percent particulate mercury).
An iron manufacturing plant, Griffin Pipe Products, is not included in the CAMR database.
The facility was added to the inventory based on the information provided by NJDEP
(NJDEP, 2006b and c) including stack parameters and total mercury emissions of 0.025
tons/year. The emissions were speciated using the iron foundries profile.
The locations for Stepan Chemical Company, Geon Company, and Owens Corning provided
in the CAMR database put the facilities outside the state. The locations for the facilities were
changed using the coordinates provided in the EPA 2001 criteria pollutant inventory.
The revisions for the non-point (area) sources included the following:
In the CAMR database, the emissions totals for human and animal cremation for the state
are identical. The emissions for animal cremation (0.052 tons/year) were removed from the
inventory.
4-10 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
Additional elemental mercury emissions for fluorescent lamp breakage (200 Ibs/year) and
miscellaneous volatilization (300 Ibs/year) were included in the inventory. The additional
emissions were distributed among the counties based on the original distribution of
fluorescent lamp breakage emissions.
EPA Region 3
STATE OF VIRGINIA
For Virginia, a steel mill with one fairly large electric arc furnace plus mill equipment (Chaparral)
in the state is not represented in the CAMR database. The facility was added to the inventory
based on emissions and stack parameter information provided by the Virginia Department of
Environmental Quality (VA DEQ). The 2001 mercury emissions for this facility are 0.14
tons/year (VA DEQ, 2006). The emissions were speciated using the steel manufacturing
speciation profile.
Also, Cogentrix of Richmond is assigned the FIPS code for Mecklenburg County, North Carolina
in the CAMR database. After double checking with the EPA Region 3, it was confirmed that the
facility is located at Richmond, VA, and the FIPS code was changed accordingly. The emissions
from the source (0.003 tons/year) were moved from the State of North Carolina to the State of
Virginia. This is an IPM point source.
STATE OF WEST VIRGINIA
The locations for CNG-Yellow Creeks and Weirton Steel Corporation provided in the CAMR
database put the facilities outside the state. The locations for the facilities were changed using
the coordinates provided in the EPA 2001 criteria pollutant inventory.
EPA Region 4
STATE OF GEORGIA
The emissions for a hazardous waste incineration facility in the state (Searle) were reduced
from 0.202 tons/year (CAMR database) to 0.002 tons/year, based on 2004 stack test results
(EPA Region 4, 2006). According to specialists at EPA Region 4, the 0.002 tons/year is a better
estimate of the 2001 emissions from the incinerator than the 0.202 tons/year, because the
facility burns a consistent waste stream generated on-site that has not changed significantly
since before 2001. Also, no process changes or physical changes were made to the incinerator
from 2000-2004 when the stack test was done.
STATE OF NORTH CAROLINA
The IPM source (Cogentrix of Richmond), incorrectly placed in Mecklenburg County, North
Carolina in the CAMR database, and was relocated to Virginia. After double checking with the
EPA Region 3, it was confirmed that the facility is located at Richmond, VA, and FIPS code was
changed accordingly. The emissions from the source (0.003 tons/year) were reassigned from
North Carolina to Virginia.
4-11 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
EPA Region 5
STATE OF ILLINOIS
Revisions were made to the IPM and non-IPM point-source emissions based on the Illinois EPA
2001 mercury emission estimates (IL EPA, 2006a, b and c; EPA OW, 2006). Details of the
revisions are included in Appendix C. A summary of the point-source revisions follows:
Emissions for all coal-fired utilities in the state were adjusted to represent 2001 operating
conditions. This accounted for a switch from bituminous to sub-bituminous coal during the
period 1999-2001 for some of the larger mercury sources, as well as an overall increase in
amount of coal combusted at many Illinois utilities. The switch to western coal plus the
increase in coal combustion resulted in increased mercury emissions and a change in the
speciation profiles for some of the plants.
Emissions for the top six mercury emitters in the cement and lime manufacturing category in
the state were revised, based on the state's 2002 NEI inventory submittal. The 2002 data are
the closest data to 2001 and indicate significantly lower emissions than the CAMR database.
The revisions resulted in a net decrease in total mercury emissions of 1.24 tons/year (from
an increase in mercury emissions of 0.17 tons/year for IPM sources, and a decrease of 1.41
tons/year for non-IPM sources).
Due to changes in coal type (bituminous vs. sub-bituminous) usage, speciation profiles were
changed for the Baldwin, Hennepin and Wood River plants.
The locations for Baxter Healthcare Corp and Radco Industries provided in the CAMR
database put the facilities outside the state. The locations for the facilities were changed
using the coordinates provided in the EPA 2001 criteria pollutant inventory.
STATE OF INDIANA
Revisions were made to the point-source emissions based on the Indiana Department of
Environmental Management (IDEM) 2001 mercury emission estimates (IDEM, 2006a). Details
of the emissions changes are included in Appendix C. Changes are summarized below.
The revisions for the point sources included the following:
Revised emissions and stack parameters for the top 20 mercury emitters in the state. The
revisions resulted in 0.16 tons/year total mercury emissions deceases for IPM sources, and
0.19 tons/year total mercury emissions increases for non-IPM sources.
Added the emissions (0.077 tons/year) for Indiana Harbor Coke Company, which is not
included in the CAMR database.
Removed two waste incinerators (Ball Memorial and Clarian Health Partners) with total
mercury emissions of 0.66 tons/year from the inventory. According to IDEM, these facilities
ceased operations and reported no emissions in 2001 (IDEM, 2006b).
STATE OF MICHIGAN
A Portland cement manufacturing plant (LaFarge Midwest Inc.) that operates five dry process
cement kilns in the state is not included in the CAMR database. The facility was added to the
inventory based on emissions and stack parameter information provided by the Michigan
Department of Environmental Quality (MDEQ). The mercury emissions for 2001 for this plant
4-12 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
total 0.29 tons/year (MDEQ, 2006). The emissions were speciated using the cement
manufacturing (dry process) speciation profile.
STATE OF MINNESOTA
Revisions were made to the point-source emissions based on the Minnesota Pollution Control
Agency (MPCA) 2001 mercury emission estimates (EPA OW, 2006b; MPCA, 2006a).
The revisions for the point sources included the following:
Revised Speciation for Coal-fired Power Plants
Sherburn County Generating Plant
Revised speciation based on the test data from the plant (Mostardi Platt, 2000).
Units 1 and 2: 90.7 percent elemental, 7.1 percent divalent gaseous, 2.0 percent particulate
mercury.
Unit 3: 96.3 percent elemental, 2.0 percent divalent gaseous, 21.7 percent particulate mercury.
Clay Boswell
Test data available for Units 3 and 4 (Roy F. Weston, 2000).
Unit 3: 98.98 percent elemental, 1.0 percent divalent gaseous, 0.02 percent particulate mercury.
Unit 4: 91.5 percent elemental, 5.9 percent divalent gaseous, 2.6 percent particulate mercury.
Allen S. King Generating Plant
Revised speciation based on the test data from the plant (MPCA, 2006c).
93.9 percent elemental, 6.1 percent divalent gaseous, 0 percent particulate.
Revised Emissions and Speciation for Sludge Incinerator
The emissions for the sludge incinerator (MCES Metropolitan WWTP - St. Paul) were
revised using the MPCA 2001 emissions estimates. Based on the documented decline in the
mercury concentration in the incinerated sludge, the emissions were reduced from 350
Ib/yearto 102 Ib/year.
The speciation profile for the incinerator was also revised in consideration of wet scrubbers in
operation during the 2001 time period. Based on the Method 29 stack testing data from the
facility, the speciation was changed from the EPA profile for sewage sludge incineration (22
percent elemental, 58 percent divalent gaseous, and 20 percent particulate) to 95.3 percent
elemental, 3.5 percent divalent gaseous, and 1.2 percent particulate mercury.
Revised Emissions and Speciation for Municipal Waste Combustors
The emissions for the municipal waste combustors were revised using the MPCA 2001
emissions estimates. The emissions for the combustors were reduced from 555 Ib/year to
95 Ib/year based on compliance stack tests, and emissions for the Perham Renewable RF
facility were omitted due to closure.
The EPA profile for municipal waste combustors (22 percent elemental, 58 percent divalent
gaseous, and 20 percent particulate mercury) was replaced with facility-specific speciation
4-13 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
profiles for Polk Co. Solid Waste Resource Recovery (0.6 percent elemental, 97.2 percent
divalent gaseous, and 2.2 percent particulate mercury) and Olmstead WTE Facility (16.9
percent elemental, 81.2 percent divalent gaseous, and 2.0 percent particular mercury).
MPCA provided speciation test data based upon Method 29 testing for the two facilities.
Revised Emissions and Speciation Profiles for Taconite Plants
Emissions for the taconite plants were revised using the MPCA 2001 emissions estimates.
The emissions for the taconite plants were reduced from 792 Ib/year to 480 Ib/year based on
a mass balance study conducted for the MPCA and accounting for one closed facility: LTV
Steel Pellet Co.
Speciation for the taconite plants was revised based on test data (using the Ontario Hydro
method) provided by MPCA (MPCA, 2006b) as follows:
- Nibbing Taconite Company: 93.31 percent elemental, 6.6 percent divalent gaseous, and
0.09 percent particulate mercury (based on test data for the company's Line 2, a straight
grate furnace fired with natural gas, tested from 9/29/98 to 10/2/98)
- United Taconite: 98.71 percent elemental, 1.08 percent divalent gaseous, and 0.21
percent particulate mercury (based on test data for the company's Line 1, a grate-kiln
furnace fired with natural gas, tested on 5/4/2005).
- National Steel: 50 percent elemental, 30 percent divalent gaseous, and 20 percent
particulate mercury. This is the only taconite facility that did not have a wet scrubber,
hence the default profile was kept (MPCA, 2006a).
- Other taconite plants: average of Nibbing Taconite Company's and United Taconite's
profile (96.01 percent elemental, 3.84 percent divalent gaseous, and 0.15 percent
particulate mercury).
STATE OF WISCONSIN
During the review of emissions for a previous mercury modeling analysis for Wisconsin (Myers
et al., 2006), the Wsconsin Department of Natural Resources (WDNR) recommended that the
emissions from the Superior Special Services site remediation be reduced from 940 Ib/year to
33 Ib/yr (WDNR, 2005). That recommendation is followed in this study.
EPA Region 7
STATE OF IOWA
Revisions were made to the IPM and non-IPM point sources and Medical Waste Incinerators
(MWI) based on the Iowa Department of Natural Resources 2001 mercury emission estimates
(Iowa DNR, 2006 and Iowa DNR, 2007). Details of the emissions changes are included in
Appendix C. Changes are summarized below.
In the CAMR database, one set of stack parameters is used for all stacks/units for a given
source. Iowa DNR provided more detailed information for stack parameters for the IPM sources,
i.e., different sets of stack parameters for the stacks/units in a facility. Iowa DNR also provided
revised emissions for the state's top 40 divalent gaseous emitters. All of the state provided
information was incorporated in the point source revisions.
The revisions resulted in total mercury emissions increases of 0.01 tons/year for IPM sources
and 0.02 tons/year for non-IPM and MWI sources.
4-14 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
STATE OF KANSAS
Revisions were made to the IPM and non-1 PM point sources based on the Kansas Department
of Heath and Environment (KDHE) data (KDHE, 2006). The revisions were for the stack
locations and parameters for the top five divalent gaseous emitters in the state. No changes
were made to the emissions totals.
STATE OF NEBRASKA
Revisions were made to the point-source emissions based on the information provided by EPA
Region 7 and the Lincoln/Lancaster County Health Department (LLCHD, 2006; EPA Region 7,
2006a and b).
The revisions made to the point source emissions files include the following:
Emissions for the Sheldon facility were changed from 0.034 tons/year to 0.0445 tons/year
Emissions for Goodyear Tire & Rubber Co. (0.058 tons/year in the CAMR database) were
omitted because there are no mercury emissions for 2001, nor are there any mercury
emissions in any of the subsequent years from the facility.
Deeter Foundry Inc. was added to the inventory with mercury emissions of 0.002 tons/year,
the EPA speciation profile for iron foundries was applied for this facility.
STATE OF MISSOURI
Revisions were made to the point-source emissions based on the information provided by EPA
Region 7 and the Missouri Department of Natural Resources (MO DNR).
Revisions to the State Top Mercury Emitters
Emissions and stack parameters for the top five divalent gaseous and top five total mercury
emitters in the state were revised (MO DNR, 2006a; EPA Region 7, 2006c). The revisions
resulted in total mercury emissions increases of 0.06 tons/year for IPM sources and a very
small amount (9.0x10"5 tons/year) for non-IPM sources
Revisions to the Lead Smelters
Considerable efforts were made to estimate the emissions from the lead smelters operated
by the Doe Run Company. EPA Region 7 and MO DNR (MO DNR, 2006b; EPA Region 7,
2006d and e) estimated that total 2001 mercury emissions from three sites of the Doe Run
Company (Glover, Herculaneum, and Buick) are 0.32 tons/year, while the EPA CAMR
database only provides emissions for one site with 0.00026 tons/year.
For Glover and Herculaneum sites, the stack parameters were provided by EPA Region 7, and
for Buick site, the stack parameters were obtained from the EPA criteria pollutant inventory.
Assuming the same mercury control efficiency and speciation as the Minnesota taconite
facilities that have been tested, the average profile for the Minnesota taconite facilities (96.01
percent elemental, 3.84 percent divalent gaseous, and 0.15 percent particulate) were used
for the speciation of the primary smelters at Glover and Herculaneum sites.
The EPA profile for secondary lead smelting used for Doe Run Company in the CAMR
database (80 percent elemental, 10 percent divalent gaseous, and 10 percent particulate)
was used for the secondary smelter at the Buick site.
4-15 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
EPA Region 8
STATE OF COLORADO
The emissions for a steel mill in the state (CF&I Steel L P DBA Rocky Mountain Steel Mills)
were increased from 0.0436 tons/year (CAMR database) to 0.2825 tons/year, based on the
information provided by EPA Region 8 (EPA Region 8, 2007).
STATE OF UTAH
Revisions were made to the point sources based on the Utah Department of Environment
Quality 2005 mercury emission estimates (Utah DEQ, 2007b). Changes are summarized below.
Further detail on these changes is included in Appendix C.
The emissions for the IPM, non-IPM and Medical Waste Incinerators included in the CAMR
database were updated to 2005 levels along with the revisions to stack parameters which
provided by Utah DEQ
There were point sources in the CAMR database that were not included in the Utah DEQ
2005 inventory. Of the sources in the CAMR inventory but not in the Utah DEQ inventory,
special attention was focused on the top 10 emitters (i.e., those emitting 3 Ib/year or more).
The total emissions of these top 10 emitters are about 98% of all emissions in CAMR but not
in the Utah DEQ inventory. After careful examination of other databases (i.e., TRI) and Utah
DEQ's confirmation, 6 of these top 10 emitters were dropped from the inventory because
independent verification could not be found that the sources exist or emitted mercury. Details
regarding the sources that were excluded are shown in Appendix C.
The mercury speciation for the coal fired boilers in Kenecott Utah Copper Corporation (50
percent elemental, 30 percent divalent gaseous, and 20 percent particulate mercury) was
revised based on the information provided by Utah DEQ (Utah DEQ, 2007a). The information
included the coal type, boiler type, and controls for each of the 4 boilers at the facility. Based
on data used in the CAMR rule making for similar facilities, a speciation profile of 37.26
percent elemental, 57.84 percent divalent gaseous, and 4.9 percent particulate mercury was used
for the Kenecott boiler.
The revisions resulted in total mercury emissions increases of 0.31 tons/year for IPM sources
and decreases of 0.45 tons/year for non-IPM and MWI sources.
EPA Region 9
STATE OF ARIZONA
Four copper mine facilities of Phelps Dodge are not represented in the CAMR database.
Following the EPA Region 9's direction (EPA Reg. 9, 2006d), mercury emissions of 0.12 tons/year
for the facilities were added to the inventory based on the 2001 TRI database. The emissions
were speciated using the EPA speciation profile for metal mining (80 percent elemental, 10
percent divalent gaseous, and 10 percent particulate). The stack parameters for one of the copper
mine facilities (Phelps Dodge Morenci Inc.) are available in the EPA 2001 criteria pollutant
inventory, and the average stack parameters for this facility were used for other three facilities.
STATE OF CALIFORNIA
Revisions were made to the non-IPM point and non-point source emissions following direction
provided by EPA Region 9 (EPA Reg. 9, 2006e). Based on information received from Barrick,
4-16 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
Inc, the Homestake goldmine in California has been closed since 2002. At EPA's direction, it
was removed from the inventory (EPA, 2007).
Cement Plants
Mercury emissions are available for ten cement plants in California in the 2001 TRI
database, and these emissions were incorporated into the inventory. The revisions resulted
in an increase in mercury emissions of 1.37 tons/year.
The emissions were speciated using the EPA speciation profile for Portland cement manufacturing
(75 percent elemental, 13 percent divalent gaseous, and 12 percent particulate mercury).
The stack parameters for Calaveras Cement Company and California Portland Cement Co.
are missing in the CAMR database. The missing stack parameters for both facilities were
replaced using the available information for Calaveras Cement Company provided in the
EPA 2001 criteria pollutant inventory.
The stack parameters for Long Beach City-SERRF Project are missing in the CAMR
database. The missing stack parameters were replaced with those for similarly named Long
Beach (SERRF) that is listed in the database, after EPA Region 9 (EPA Reg. 9, 2006f)
confirmed that they are the same facility.
Gold Mining
The mercury speciation for Home Stake Mining Company (99.86 percent elemental, 0.09
percent divalent gaseous, and 0.06 percent particulate mercury) were revised using the
speciation profile (78 percent elemental, 19.3 percent divalent gaseous, and 2.7 percent
particulate mercury) for Cortez Gold Mines #2 (aka Pipeline Mill) in Nevada. This information
was provided by the EPA Region 9 (EPA Reg. 9, 2006f).
The Homestake Gold Mine was removed from the inventory because the facility was closed
in 2002 (Barrick, 2007; EPA, 2007).
Residential Home Construction
In the CAMR database, the mercury emissions from residential home construction in State of
California are about 1.3 tons/year. However, there is no mercury emissions from the source
category included in the 2002 NEI (EPA, 2007). The emissions were removed from the
inventory.
STATE OF NEVADA
Revisions were made to the IPM and non-IPM point source emissions following direction
provided by EPA Region 9, as summarized below and detailed in Appendix C.
4-17 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
Gold Mines
Emissions and Speciation Revisions
The mercury emissions and associated speciation data for all gold mines in the State of
Nevada were revised based on information from several data sources. The final information
used was tabulated by EPA Region 9 (EPA Region 9, 2007). Further references are given in
Appendix C.
Stack Parameters Revisions
Location and stack parameters for all gold mines were revised based on the information
provided in the MACT questionnaire (NDEP, 2006).
Cement Plant
Emissions for Nevada Cement Company were added from the TRI 2001 database.
The emissions were speciated using EPA speciation profile for Portland cement
manufacturing.
Stack parameters for this plant were unavailable in the EPA inventories, so the average
stack parameters for California cement plants were used.
IPM
The mercury emissions, associated speciation and stack parameters for a coal fired utility
(Mohave) were revised based on the information provided by EPA Region 9 (NDEP. 2007).
The revisions to gold mines resulted in a decrease in total mercury emissions of 8.42 tons/year.
Adding the cement plant emissions increased total mercury emissions by 0.0095 tons/year. The
revisions to the IPM source resulted in a decrease in total mercury emissions of 0.074
tons/year.
EPA Region 10
STATE OF OREGON
The revisions were made to the non-IPM point source emissions based on information provided
by the Oregon Department of Environmental Quality (OR DEQ).
Cement Plant
The emissions, speciation, and stack parameters were revised for Ash Grove Cement
Company based results of Ontario hydro testing supplied by OR DEQ (OR DEQ, 2007a, b,
and c). The revisions resulted in mercury emissions increases of 1.14 tons/year, from 0.1154
tons/year in CAMR to 1.255 tons/year. Speciation was revised from 75/13/12 to 29/63/8
(%HGO/%HG2/%HGP).
Steel Mill
Emissions for Oregon Steel Mills, Inc and Cascade Steel Rolling Mills, Inc were revised
based on information from OR DEQ (2007a). The revisions resulted in mercury emissions
decreases of 1.71 tons/year.
The mercury speciation for Cascade Steel Rolling Mills, Inc (50 percent elemental, 30
percent divalent gaseous, and 20 percent particulate mercury) was revised using the EPA
4-18 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
speciation profile for iron and steel foundries (80 percent elemental, 10 percent divalent
gaseous, and 10 percent particulate).
Municipal Waste Incinerator
The emissions for Covanta Marion, Inc. were decreased by 0.09 tons/year (OR DEQ, 2006c).
The locations for Alsea Veneer Inc and Pacific Softwood Co. provided in the CAMR database
put the facilities outside the state. The locations for the facilities were changed using the
coordinates provided in the EPA Facility Registry System web site.
STATE OF WASHINGTON
Revisions were made to the non-IPM point source emissions based on information provided by
the Northwest Clean Air Agency and Puget Sound Clean Air Agency.
Refineries
Emissions for Tesoro (formerly Shell Oil Co.) were changed from 43 Ib/year to 9 Ib/year.
Emissions for the Puget Sound facility (Equilon Enterprises) were changed from 9.2 Ib/year
to zero.
Cement Plant
Two cement plants located in King County, WA are not included in the CAMR database.
Mercury emissions of 44 Ib/year and 68 Ib/year were added to the inventory for Ash Grove
Cement Company and Lafarge Corporation, respectively. The emissions were speciated
using the EPA speciation profile for Portland cement manufacturing (75 percent elemental,
13 percent divalent gaseous, and 12 percent particulate). The stack parameters for the
facilities from the EPA 2001 criteria pollutant inventory were used.
Georgia Pacific West Inc
Emissions were lowered by 83 percent for this facility, because the facility only operated in
February and March of 2001 (WA DE 2006).
Tacoma
Emissions were lowered by 83 percent for this facility, because the facility only operated two
months in 2001 (WA DE 2007).
Canada
One location correction was made to the Canadian point sources. The coordinates provided for
Ontario Hydro - Lambton TGS (with total mercury emissions of 0.19 tons/year) in the CAMR
database put the facility outside the province and modeling domain. Based on the information
provided in Canadian web sites, the facility is located on the St. Clair River, approximately 21
kilometers south of Sarnia, Ontario. The correct coordinates for the facility were obtained and
incorporated into the inventory.
One incinerator (Burnaby Refuse Incinerator) was added to the Canadian point source
emissions based on the information provided by Environment Canada (Canada, 2007a) and
4-19 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
Metro Vancouver (Canada 2007b). The revisions resulted in mercury emissions increases of
0.06 tons/year.
Mexico
The Mexico point source emissions, which are not available in the CAMR database, were added
to the 2001 inventory used for REMSAD modeling. The point source emissions data were
provided by Commission for Environmental Cooperation (CEC, 2001). These data include
speciated mercury emissions for 268 point sources in Mexico along with source locations and
stack height. Default stack diameter, exit velocity and temperature information was added to the
data for processing the emissions for Mexico.
Total mercury emissions from the point source data for Mexico are 29.42 tons/year with 20.89
tons/year of elemental, 5.76 tons/year of divalent gaseous, and 2.77 tons/year of particulate
mercury. About 69 percent of these emissions are from point sources located within the
modeling domain (13.07 tons/year of elemental, 4.68 tons/year of divalent gaseous and 2.43
tons/year of particulate mercury).
4.2.2. Mercury Emissions Processing for REMSAD
For this application, the mercury emissions data were processed using the Emissions
Preprocessing System (EPS2.5).
The revised 2001 CAMR mercury emissions data for area and point sources were speciated
into particulate, divalent, and elemental emissions. For sources with speciated emissions data,
the speciation was retained. For sources reporting total mercury emissions (unspeciated),
speciation was accomplished using the speciation profiles and cross-reference file provided by
EPA. The emissions for each of the three species were gridded and temporally allocated by
month, day, and hour using the EPS2.5 algorithms. The emissions were prepared for weekday,
Saturday, and Sunday for each season. For the low-level (area) sources, separate emissions
files were prepared for the 36-km grid and for each of the 12-km grids.
4.2.3. Quality Assurance of the Mercury Emission In ventory
The goal of the quality assurance procedures was to ensure that the emission estimates from
the EPA CAMR inventory and all additions and corrections provided by the states and EPA
regional offices are properly represented in the REMSAD input files. As noted above, the
modified CAMR emissions inventory is referred to as the updated emission inventory.
The QA procedures included:
Cross checks of emissions totals in the updated inventory files compared to the REMSAD
input files.
- These types of checks were used to ensure that the processing did not result in emissions
being left out of the inventory and that there were no errors in converting the units of
emissions.
Displays of emissions density of area sources
- These displays were used to check for inconsistencies in the emissions among the states.
Plots of point source emissions by emissions category or by individual state
- These displays were used to confirm that the elevated point sources of mercury were
located within the correct state.
4-20 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
- For the tagged emissions, these same plots were used to verify that the emissions tagged
for a particular state were properly located in the domain.
Some additional checks were made in order to identify possible unrealistic values or parameters
in the updated emissions inventory files.
The 2001 base-year criteria pollutant emissions for the U.S. portion of the modeling domain are
presented by major source category and by season in Table 4-2. This table highlights the key
source categories for each component and the variations in the emissions throughout the year.
4.2.4. Summary of the Mercury Emissions
The mercury emissions are briefly summarized in this section of the report. Table 4-3 lists
annual mercury emissions for each state, the 48 U.S. states, Canada, and Mexico. The
emissions totals are provided for each species and for the sum of the three mercury species.
Table 4-3. Summary of Mercury Emissions Totals by Species for Each U.S. State
and for the 48 U.S. States, Canada, and Mexico.
State
Alabama
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
District of Columbia
Florida
Georgia
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
HGO
(tpy)
2.146
0.887
0.805
3.815
0.564
0.200
0.593
0.004
1.510
1.466
0.562
3.336
1.895
0.806
0.963
1.818
1.910
0.247
0.753
0.320
1.484
1.139
0.518
1.519
0.475
0.362
1.765
0.075
1.043
HG2
(tpy)
1.351
0.106
0.350
1.329
0.155
0.202
0.108
0.001
1.242
0.969
0.155
1.462
1.511
0.313
0.178
1.299
0.335
0.105
0.931
0.454
1.124
0.286
0.326
0.594
0.070
0.095
1.243
0.124
0.408
HGP
(tpy)
0.228
0.049
0.164
0.951
0.057
0.084
0.015
0.001
0.462
0.137
0.118
0.318
0.343
0.032
0.081
0.446
0.115
0.038
0.236
0.152
0.261
0.104
0.122
0.118
0.022
0.005
0.075
0.033
0.192
Total
(tpy)
3.726
1.043
1.320
6.095
0.776
0.485
0.716
0.006
3.214
2.572
0.835
5.116
3.749
1.150
1.222
3.564
2.359
0.389
1.920
0.926
2.869
1.530
0.966
2.230
0.567
0.462
3.082
0.232
1.643
4-21 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
State
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylvania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
West Virginia
Wisconsin
Wyoming
U. S. Total
Canada
Mexico
Grand Total
HGO
(tpy)
1.083
1.252
1.019
0.930
2.901
0.875
0.734
3.003
0.085
0.709
0.050
1.251
5.610
0.488
0.017
0.722
0.408
1.458
1.579
0.923
58.074
4.540
13.07
75.679
HG2
(tpy)
0.056
0.919
1.273
0.159
2.066
0.298
0.870
3.766
0.062
0.658
0.021
0.882
2.115
0.219
0.007
0.790
0.149
1.564
0.520
0.096
33.319
2.909
4.676
40.903
HGP
(tpy)
0.019
0.361
0.239
0.032
0.313
0.064
0.208
0.773
0.035
0.246
0.002
0.148
0.596
0.065
0.004
0.209
0.042
0.155
0.094
0.024
8.588
0.869
2.430
11.886
Total
(tpy)
1.157
2.532
2.531
1.122
5.280
1.237
1.812
7.541
0.182
1.613
0.074
2.281
8.321
0.772
0.028
1.721
0.599
3.177
2.193
1.043
99.981
8.318
20.17
128.469
The spatial distribution of mercury emissions throughout the domain is illustrated in Figures 4-3
and 4-4. Figure 4-3 illustrates the distribution of low-level mercury emissions corresponding to
the 2001 REMSAD emissions inventory for each of the three species and for all three species
combined. Figure 4-4 similarly displays the elevated point source emissions. Note that, although
these displays present the emissions at 36 km resolution, the emissions were processed for
each sub-domain to 12-km resolution.
4-22
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
Figure 4-3a. Spatial Distribution of Low-Level Mercury Emissions (tons) for the 2001 REMSAD
Emissions Inventory for the 36-km Grid for a Summer Weekday: Elemental Mercury.
Max value: 12036.7 (g/day) at ( 61, 6)
-2736 -2016
110
864 1584 2304
E 1800
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
-2088
LowLevel HgO Emissions: July Weekday
HGO_1
Total: 67130 (g/day)
4-23
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
Figure 4-3b. Spatial Distribution of Low-Level Mercury Emissions (tons) for the 2001 REMSAD
Emissions Inventory for the 36-km Grid for a Summer Weekday: Divalent Gaseous Mercury.
Max value: 7221.7 (g/day) at ( 61, 6)
-2736 -2016 -1296
2304
E 1800
= 1368
11111111111111111111 u+tfiTim 11111111111111111111111111111111111111 ff
-1656
10 20 30 40 50 60 70 80 90 100 110 120 130 140
-2088
^
10 75 So
LowLevel Hg2 Emissions: July Weekday
HG2_1
Total: 20651 (g/day)
4-24
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
Figure 4-3c. Spatial Distribution of Low-Level Mercury Emissions (tons) for the 2001 REMSAD
Emissions Inventory for the 36-km Grid for a Summer Weekday: Particulate Mercury.
Max value: 4814.3 (g/day) at ( 61, 6)
Km
-2736 -2016 -1296 -576 144 864 1584 2304
E 1800
= 1368
uttfmffti 1111111111111111111111111111111111111 ff
Tl 111111111111111111111111111111111111111111111111
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
-2088
LowLevel Hgp Emissions: July Weekday
HGP_1
Total: 10497 (g/day)
4-25
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
Figure 4-3d. Spatial Distribution of Low-Level Mercury Emissions (tons) for the 2001 REMSAD
Emissions Inventory for the 36-km Grid for a Summer Weekday: All Species.
Max value: 24072.7 (g/day) at ( 61, 6}
Km
-2736 -2016 -1296 -576 144 864 1584 2304
-. 1800
1368
11111111111111111111 uttfmfil 11111111111111111111111111111111111111 ff
THIN 11 In ii in ii miiiliiimiii! in ii il
i -1666
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
-2088
I LowLevel Total Hg Ems: July Weekday
i THG
3oS° Total: 98277 (g/day)
4-26
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
Figure 4-4a. Spatial Distribution of Elevated Point-Source Mercury Emissions (tons) for the 2001
REMSAD Emissions Inventory for the 36-km Grid for a Summer Weekday: Elemental Mercury.
Max value: 3055.8 (g/day) at ( 65, 4)
Km
-2736 -2016 -1296 -576 144 864 1584 2304
^ 1800
= 1368
"0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
_ pnoo
"O ?5
70 - 75
Jevated Point HgO Emissions: Summer Weekday
HGO_1
° Total: 119747 (g/day)
4-27
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
Figure 4-4b. Spatial Distribution of Elevated Point-Source Mercury Emissions (tons) for the 2001
REMSAD Emissions Inventory for the 36-km Grid for a Summer Weekday:
Divalent Gaseous Mercury.
Max value: 3233.2 (g/day) at (118, 65)
Km
-2736 -2016 -1296 -576 144 864 1584 2304
-E 1800
-E 1368
u+tfmffti 1111111111111111111111111111111111111 ff
Tl 111111111111111111111111111111111111111111111111
'0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
I -1224
-1656
PORH
^
e ^ e
Jevated Point Hg2 Emissions: Summer Weekday
75 3. ., HG2_1
7s"-?o° Total: 81070 (g/day)
4-28
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
Figure 4-4c. Spatial Distribution of Elevated Point-Source Mercury Emissions (tons) for the 2001
REMSAD Emissions Inventory for the 36-km Grid fora Summer Weekday: Particulate Mercury.
Max value: 387.3 (g/day) at ( 24, 50)
Km
-2736 -2016 -1296 -576
144
864
1584 2304
E 1800
-E 1368
11111 nn null i nn iiu+KmfTtiiliiin MI i ill MI MI i ill MI i n i ill miff
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
-1666
-2088
}- 6
~~ 6
TO
|evated Point Hgp Emissions: Summer Weekday
HGP_1
3° Total: 17926 (g/day)
4-29
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Emissions Inputs
Figure 4-4d. Spatial Distribution of Elevated Point-Source Mercury Emissions (tons) for the 2001
REMSAD Emissions Inventory for the 36-km Grid for a Summer Weekday: All Species.
Max value: 4742.1 (g/day) at (118, 65)
Km
-2736 -2016 -1296 -676 144 864 1584 2304
E 1800
-E 1368
II I IIIIII11II11111111 LJKrtTnli'r 11111111111111 11111111111111111111
"0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
_ OflflO
0 S 4 6
3 f 8
10
evated Point Total HG Ems: Summer Weekday
i THG
3°?° Total: 218744 (g/day)
4-30
August 2008
-------
5. Initial and Boundary Condition Inputs
In this section, the preparation of the initial and boundary condition input files for the application
of REMSAD is discussed. The same datasets and similar methods were used to prepare the
initial and boundary conditions inputs for the CMAQ simulations conducted by EPA (that are
referenced later in this report).
The initial conditions assign the species concentrations for each grid cell the initial simulation
time. The boundary conditions define the concentrations along the lateral boundaries of the
modeling domain for each hour of the simulation period. A default value of 1 x 10"20 ppm is
assigned to concentrations of all species at the top of the modeling domain.
5.1. Specification of Initial and Boundary Conditions for the
PM Simulation
The initial and boundary conditions for the criteria pollutant simulation were based on those
used by EPA for the CAIR modeling (EPA, 2005a). The initial and boundary conditions fields are
derived from global modeling using the GEOS-CHEM model (Yantosca, 2004). The boundary
concentration files vary in both space and time.
5.2. Specification of Initial and Boundary Conditions for the
Mercury Simulations
It is expected that global background concentrations of mercury are high enough to influence the
magnitude of mercury deposition within the U.S. The magnitude of global background
concentrations is not, however, well known. In particular, the concentrations of the oxidized forms
of mercury are very uncertain. Background concentrations of about 1.6 nanograms per cubic
meter (ng m"3) of elemental mercury have been used in past modeling exercises (Pai et al., 1999;
Myers et al., 2003). These exercises have indicated that background mercury may make up more
than 50 percent of the total airborne mercury in some areas, and support for this estimate can be
found in experimental studies (e.g., Blanchard et al., 2002). Estimates of the fraction of oxidized
mercury in total gaseous mercury range from a small fraction of a percent to as much as five
percent. Given the potentially large influence of background mercury and the meager set of
observations, other methods for setting boundary and initial concentrations have been sought.
During the North American Mercury Model Inter-comparison Study (NAMMIS) (Bullock et al.,
2008), the results of several global simulation models were made available for the purpose of
preparing boundary conditions for continental scale modeling. The results from three models,
the Chemical Transport Model (CTM) (developed and applied by AER), the Global/Regional
Atmospheric Heavy Metals model (GRAHM) (developed and applied by Environment Canada),
and the GEOS-Chem model (developed and applied by researchers at Harvard University),
were used to prepare three estimates of boundary concentrations of elemental mercury, divalent
gas mercury, and particulate mercury. All three global models utilized a year 2000 global
emissions inventory. Resolution varied among the global models: 2 by 2.5 degrees for GEOS-
CHEM, 5 by 5 degrees for GRAHM, and 8 by 10 degrees for CTM. Formulations of the models
differ in the specific chemical reactions included for mercury species and in other details, as
reported by Bullock et al. (2008). Further discussion of the global model estimates of boundary
concentrations is included in Section 6.
For the current modeling, three baseline simulations were prepared, one using each of the
global model derived boundary concentrations. The files were prepared as monthly average
5-1 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Initial and Boundary Condition Inputs
files and the simulated horizontal and vertical variation in the concentrations was retained in the
files. Each of the three versions of boundary concentrations was simulated as a separate tag to
allow analysis of differences in results due to variations in the boundary concentrations. This will
be discussed further in the section on the results of the tagging simulations.
Three sets of initial concentrations were also prepared (one for each global model) based on an
average of the simulated boundary concentrations for January. The initial conditions do not vary
horizontally, but the vertical variation from the boundary files was retained. Since the simulation
is initialized ten days prior to the beginning of the analysis period (January 1, 2001), the initial
concentrations are not expected to have a large influence on the results.
A summary and comparison of the boundary concentrations derived from the three global
models is presented in Figure 5-1. The concentrations depicted in the plots represent the
average around the perimeter of the outermost REMSAD domain (see Figure 1-1), for the
species HGO (elemental mercury), HG2 (divalent gas mercury), and HGP (particulate mercury).
The boundary conditions are compared for February and July in order to examine the temporal
variation of concentrations. The units are parts per trillion (ppt). At standard temperature and
pressure conditions (STP), 0.2 ppt is approximately 1.8 ng m"3 of mercury.
Although these charts do not show all of the variation present in the boundary concentrations
(e.g., the horizontal variation is not depicted), the variation with height and differences among
the models are apparent. The order of magnitude is similar among the global models for HGO
and HG2, although the GRAHM model has lower concentrations of these species than the other
two models. HGP concentrations are low for the CTM and GEOS-CHEM models, but the
GRAHM model has a considerable amount of divalent mercury as particulate. The vertical
profiles of concentrations are rather different among the models for HG2 and HGP.
The differences in boundary concentrations derived from the three models can lead to some
differences in the simulation results for mercury, as will be discussed later in the section on
tagging results.
5-2 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Initial and Boundary Condition Inputs
Figure 5-1. Comparison of CTM, GRAHM, and GEOS-CHEM Derived Boundary Concentrations (ppt)
for the REMSAD Modeling Domain for HGO, HG2, and HGP: February and July 2001.
12000
0.
V * \
\ "A
\ \
\ A
N \
\ V.
\ \
V\
\ V
00 0.05 0.10 0.15 0.20 0
Feb. HgO cone, (ppt)
CTM
GRAHM
25
12000
0.
i *
/ ''
r
^7 -""
s /
^ 1 --"
/ t
1 7
X/
tz
00 0.01 0.02 0.03 0.04 0.
Feb. Hg2 cone, (ppt)
GRAHM
GEOS-CHEM
05
12000 -
0-
0.0
^^
^^
~^L
s
/
I
\
\
/
00 0.002 0.004 0.006 0.008 0.010 0.012 O.C
Feb. HgP cone, (ppt)
CTM
GRAHM
GEOS-CHEM
14
12000
0.
, \
x- V
\ \\
V\
\V
\\
\\
V>
00 0.05 0.10 0.15 0.20 0.
July HgO cone, (ppt)
CTM
GRAHM
GEOS-CHEM
25
18000
16000
14000
12000
« 8000
6000
4000
2000
0
0
^ "
f 7
I
j* .'
s' .-"'
^ ,-'
//
K
/-'
CTM
GRAHM
00 0.01 0.02 0.03 0.04 0.05
July Hg2 cone, (ppt)
12000 -
0.0
^~~
^*- "*-
^^
f
\
\
$
s
00 0.002 0.004 0.006 0.008 0.010 0.012 O.C
July HgP cone, (ppt)
GRAHM
GEOS-CHEM
14
5-3
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Initial and Boundary Condition Inputs
This page deliberately left blank.
5-4 August 2008
-------
6. REMSAD Base Simulation Results
In this section, the ability of the REMSAD modeling system to replicate the observed deposition
characteristics of mercury is examined, on a seasonal and annual basis. Model performance for
ozone, sulfur dioxide (SO2) and fine particulate matter (PM2.s) is presented in Appendix A.
Model performance for mercury is evaluated for total wet deposition of mercury against the
monitors in the Mercury Deposition Network (MDN) available from the National Acid Deposition
Program (NADP). There are a total of 98 MDN monitors in the modeling domain. It should be
noted that some emerging research suggests that the MDN measurement techniques may
underestimate wet deposition by about 16 percent (Miller et al., 2005). Nevertheless, the evaluations
below use the MDN observations without any adjustment.
Simulated values in this comparison include the contribution from re-emitted mercury, using the
method discussed in Section 2 of this report. Three alternate sets of boundary conditions were
used, CTM, GRAHM, and GEOS-CHEM, as discussed in Section 5. Model performance for
mercury is evaluated for each set of boundary conditions.
The following metrics and statistical measures were used to quantify model performance:
Mean observed deposition = 1/A/ZO,
Mean simulated deposition = 1//V ZS,
Mean residual = 1//V Z (S,- O,)
Normalized bias (expressed as percent) = 100 -1//V 2 (S,- O,)/ O,
Normalized gross error (expressed as percent) = 100 -1//V Z |S/- O/|/ O,
Where S is the simulated concentration, O is the observed concentration, and N is the number
simulation-observation pairs used in the calculation.
In preparing the statistics and scatter plots, simulated and observed wet deposition were
compared for 1) each site and 2) the average over all sites. All results are for 12-km resolution.
Table 6-1 summarizes seasonal and annual model performance, considering all MDN sites within the
modeling domain. Here, winter is defined as January, February and December, spring is March, April,
and May, and so forth. Results for the three sets of boundary conditions are presented separately.
Table 6-1 a. REMSAD Model Performance Statistics for Mercury Wet Deposition (g km"2)
for the 12-km Resolution Grid at MDN Sites: CTM Boundary Conditions.
Period
Winter
Spring
Summer
Autumn
Mean Observed
(g km-*)
1.30
2.31
3.66
1.89
Mean
Simulated
(g km-*)
1.65
3.52
6.32
3.37
Mean residual
(g km-*)
0.35
1.21
2.66
1.48
Normalized
bias (%)
29.6
70.3
91.2
121.9
Normalized
gross error (%)
74.6
92.8
96.7
129.2
Annual 9.26 14.84 5.58 59.7 65.8
6-1 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD Base Simulation Results
Table 6-1 b. REMSAD Model Performance Statistics for Mercury Wet Deposition (g km"2)
for the 12-km Resolution Grid at MDN Sites: GRAHM Boundary Conditions.
Mean Observed
Period
Winter
Spring
Summer
Autumn
(g km-2)
1.30
2.31
3.66
1.89
Mean
Simulated
(g km-')
1.55
3.05
5.74
3.25
Mean residual
(g km-2)
0.25
0.74
2.08
1.35
Normalized
bias (%)
25.0
46.3
72.7
115.6
Normalized
gross error (%)
69.8
78.5
81.1
123.6
Annual 9.26 13.55 4.29 45.7 55.3
Table 6-1 c. REMSAD Model Performance Statistics for Mercury Wet Deposition (g km"2)
for the 12-km Resolution Grid at MDN Sites: GEOS-CHEM Boundary Conditions.
Period
Winter
Spring
Summer
Autumn
Mean Observed
(g km-2)
1.30
2.31
3.66
1.89
Mean
Simulated
(gkm-2)
2.01
4.09
6.51
3.60
Mean residual
(g km-2)
0.72
1.79
2.85
1.70
Normalized
bias (%)
58.7
97.4
98.2
139.6
Normalized
gross error (%)
94.7
116.7
103.3
145.8
Annual 9.26 16.20 6.94 73.8 78.7
The statistical measures of model performance indicate that the REMSAD simulations tend to
overestimate the observed mercury deposition values using each of the three sets of boundary
conditions, on an annual and seasonal basis. Performance tends to degrade throughout the year,
perhaps due to a build up of excess mercury over time. The simulated values derived using the
GRAHM boundary conditions are consistently better matched with the observed values. In
interpreting the evaluations against the MDN monitoring data, it should be kept in mind that some
emerging research suggests that the MDN measurement techniques may underestimate wet
deposition by about 16 percent (Miller et al., 2005).
Scatter plots showing the REMSAD simulated total annual wet deposition of mercury versus the
observed values at the MDN sites are presented in Figure 6-1. For all three sets of boundary
conditions, there is some positive bias in the simulation results, and there is some scatter about
the 1:1 line, primarily for the lower range of values. The R2 correlation values are similar and on
the order of 0.75 for all three sets of boundary conditions. As indicated by the plots and the
statistics, the GEOS-CHEM boundary conditions result in the greatest amount of overestimation
of mercury deposition. The GRAHM boundary conditions give the best overall model
performance.
6-2 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD Base Simulation Results
Figure 6-1 a. Annual Simulated versus Observed Mercury Wet Deposition (g km )
for the REMSAD 12-km Grid at MDN Sites: CTM Boundary Conditions.
y=1.6422x- 0.3663
R2 = 0.74
10 15 20 25
Observed (gm/Km2)
30 35
Figure 6-1 b. Annual Simulated versus Observed Mercury Wet Deposition (g km")
for the REMSAD 12-km Grid at MDN Sites: GRAHM Boundary Conditions.
35
.it
30 -
25 -
15 -
10 -
5 -
0
y=1.4979x-0.3213
^ = 0.7282
5 10 15 20 25 30 35
Observed fgm km2)
6-3
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD Base Simulation Results
Figure 6-1 c. Annual Simulated versus Observed Mercury Wet Deposition (g km"2)
for the REMSAD 12-km Grid at MDN Sites: GEOS-CHEM Boundary Conditions.
35
30 -
ST 25 -
S
i> 20 -
^ 15 -
3
E
w 10 -
5 -
0
y =1.81 85X- 0.6329
^ = 0.7511
10 15 20 25 30 35
Observed ignvkm2)
Spatial distribution plots of the REMSAD simulated mercury concentrations are provided for the
annual simulation period and for the region covered by the 12-km grids in Figure 6-2. The
patterns are similar for all three sets of boundary conditions and the plot gives the results with
the average of the three sets of boundary conditions. The plots show annual average
concentrations for elemental, divalent, and particulate mercury, respectively. Divalent mercury
concentrations are the highest of the three species, especially over portions of the mid-Atlantic
states and California. This spatial distribution is consistent with the emissions and annual
transport patterns.
6-4
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD Base Simulation Results
Figure 6-2a. Annual Average Simulated Mercury Concentration (ng m ) for the REMSAD 12-km
Modeling Domain (with Average Boundary Conditions): Elemental Mercury (HGO).
l.KVKI. 1 HC;() (ng/rn:i)
Time: 0 Jan l[ 3001-0 Dec 31, 3001
MAXIMUM - 0.7 nR./m3 (80,63)
MTNTMtJM - H.S i!g/ii]
-330-1. -isa-i -861
240 M I I I i M i i | i i i i i i i i i | i i
ISO
120 -
Km
-Ml
576.
1K96.
| I I I | I I IT- 1308.
I I I I I I I I I I I I I I I I I I 1 I I I i I I I I I rS-ii I I i I I I I [ I I I 1 1 ii I I I I I i t I I I i i I I I I r i
O 'J 'O :
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD Base Simulation Results
Figure 6-2b. Annual Average Simulated Mercury Concentration (ng m ) for the REMSAD 12-km
Modeling Domain (with Average Boundary Conditions): Divalent Mercury (HG2).
LEVEL 1 IIG2 (ng/m3)
Time: 0 Jan 1, HOOT 0 Dec HI, HOO1
T- MAXIMUM - 1.UHO ng/m3 (fi'l.lfid)
MINIMUM = 0.010 iiK/mS (1,243)
-1564.
Km
-144.
576.
ISO
120
i 1 i i i i
60
100
a oo
aeo
see.
132.
-S52.
-Ifi72.
ooooooooo
%
^> ^? & G Q O O
je concentration of HC2 Annual 2001
DivaleuL gay Tag 53 grp 16
S001.iSk.hfi
6-6
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD Base Simulation Results
Figure 6-2c. Annual Average Simulated Mercury Concentration (ng m ) for the REMSAD 12-km
Modeling Domain (with Average Boundary Conditions): Particulate Mercury (HGP).
LEVEL 1 IIGP (iifi/ni3)
Time: 0 Jan 1. 2OO1-U Deo in, HOO1
+ MAXIMUM - O.'t'SS [ijj/mS (31.1.31)
- MINIMUM - 0.001 iiK/mS (84,343)
§
?
3
1 as.
"/., Q., "O, O-, tt, Vn 'i> SU -ffl,,
^ ^fe "O *0 S0 °0 °0 °0 °0
Avc^rfigt; concenLralion of TICP Anniifil 2001
Particulate Tag 53 grplQ
3001.12k.bg
Spatial distribution plots of the REMSAD simulated total mercury deposition are provided for the
annual simulation period and for the region covered by the 12-km grids in Figure 6-3. Again, the
display incorporates the average boundary conditions. The plots show annual average
concentrations for dry, wet, and total deposition, respectively. Clearly wet deposition accounts
for much of the deposition that occurs throughout the domain. This spatial distribution is
consistent with the emissions and annual transport and rainfall patterns. These displays
emphasize the importance of rainfall in determining mercury deposition patterns.
6-7
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD Base Simulation Results
Figure 6-3a. Simulated Annual Mercury Deposition (g km ) for the REMSAD 12-km Modeling Domain
(with Average Boundary Conditions): Dry Deposition.
CO
^
>
LEVEL 1 THG (g/km2)
Time: 0 Jan 1, 2001-0 Dec 31, 3001
+ MAXIMUM - 355.3 g/kmE (314.31)
- MINIMUM - 1.4 g/kma (31,232)
1308.
l l l i i rt-tV i I I I I I I I I I I I 1 I 1 I I I I I
120 1BO 240 300
360
132.
- -852.
-1572.
\\X
Total dry deposition of THG_53 -- Annual 2001
Total Hg Tag 53 grp!6
200 1.1 3k.hg
6-8
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD Base Simulation Results
Figure 6-3b. Simulated Annual Mercury Deposition (g km ) for the REMSAD 12-km Modeling Domain
(with Average Boundary Conditions): Wet Deposition.
I.FVKl, I THG (g/km2)
Time: 0 Jan 1. 2001-0 Dec 31, 2001
-2304. -1584
240H MM Mll|
100 -
120 -I
+ MAXIMUM - 109.3 g/km3 (317.151)
- MINIMUM - 0.0 g/krn2 (215,172)
1296
[308.
- 500.
- 133.
360
- -852.
I ! I " -1572.
I
Total wet deposition of THG_53 - - Annual 2001
Total Hg Tag 53 grp!6
2001.12k.bg
6-9
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD Base Simulation Results
Figure 6-3c. Simulated Annual Mercury Deposition (g km ) for the REMSAD 12-km Modeling Domain
(with Average Boundary Conditions): Total (Dry + Wet) Deposition.
IFVEl, I THG (g/km2)
Time: 0 Jan 1, K001-0 Dec 31, 2001
MAXIMUM - 450.6 g/km2 (314.31)
MINIMUM - 0.0 g/kmS (215,172)
1296
£016.
1308.
- SUB.
360
852.
- 1572.
&
c
o
S
E
9
Total wet + dry deposition of THG_53 - - Annual 2001
Total Hg Tag 53 grp!6
2001.12k.hg
CMAQ model performance for wet mercury deposition for this same simulation period was
assessed by Bullock et al. (2008) as part of the NAMMIS study. Overall, Bullock found CMAQ
performance to be comparable to that for REMSAD, with some statistical measures better for
CMAQ, when both models were applied with 36-km horizontal resolution. A key finding of the
NAMMIS study related to model performance is that performance for both the CMAQ and REMSAD
models is influenced by the specification of boundary conditions, but the differences in model
performance among the three sets of boundary conditions is different for the two models. CMAQ
shows better agreement with observed wet deposition data with the CTM-derived boundary
conditions and REMSAD performance is better with the GRAHM-derived boundary conditions.
Another key finding of the NAMMIS study is that model performance is limited by the ability of the
meteorological inputs to accurately represent the location and amount of precipitation.
6-10
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD Base Simulation Results
Researchers involved in the NAMMIS study also concluded that each global model was based
upon sound science, and, absent a much more extensive monitoring network that includes dry
deposition measurements, it is not now possible to conclude which global model performs the
best. It is for these reason that it was chosen to present results from all three global models,
driving both REMSAD and CMAQ (see Section 7), so that the reader can more thoroughly
understand the likely range of background contributions.
6-11 August 2008
-------
7. REMSAD PPTM Results: Mercury Deposition
Contribution Analysis
The results of the REMSAD mercury tagging simulations are presented in this section,
beginning with an overview of the application procedures. The tagging results are then
summarized for selected locations, and examine the contributions in relation to a source
receptor study.
For ease of reading, all figures follow the text of this section.
7.1. PPTM Application Procedures
The entire modeling analysis included 18 annual REMSAD simulations, and each simulation included
approximately 15 to 20 tags (for a total of approximately 300 tagged sources). The tags were defined
on a state by state basis. As noted in Section 4, summaries of the mercury emissions inventory for
each state were provided to the EPA regional offices and to each state to facilitate their review of
emissions. These same summaries were used to identify candidate sources for mercury tagging. The
summaries listed and ranked the top five divalent gas mercury emitters and the top five total mercury
emitters (excluding those already in the divalent gas ranking). The rankings were made on a facility
basis, not on an individual stack basis, to avoid using multiple tags for a single facility.
Nominally, five tags were to be defined for each state. Four of these tags were to be assigned to
individual sources or source categories, and the fifth tag was reserved to collectively tag all
remaining sources. Since approximately 300 tags were to be simulated, this would leave about
50 additional tags that could be used for a more detailed breakdown of the source contributions
for selected states and/or source categories. These 50 tags were assigned based on
recommendations from EPA Regional and state personnel.
The general procedure was to assign the first three tags to the top three emitters of divalent
gaseous mercury. Then the top total mercury emitter not already tagged was assigned the
fourth tag. On occasion, there was deviation from this approach due to wide disparities in
magnitude of emissions or other extraordinary circumstances. In a few cases, states with very
low mercury emissions were assigned fewer than five tags. States with multiple large sources
and thus a greater need for detail in the emissions breakdown were assigned more than five
tags. Because of the magnitude of the emissions and in order to allow an analysis of the
importance of speciation, three tags were assigned to individual species emitted from a
goldmine in Nevada.
The states and EPA regional offices were informed of this general procedure when asked for their
input. Some states and regional offices made specific requests for tags and recommendations for
changes in the assignment. Where possible, these requests were accommodated.
The tags used for the application of PPTM are listed and summarized in Table 7-1.
7-1 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Table 7-1. Tags used for the REMSAD PPTM Application for Mercury
for the Annual 2001 Simulation Period.
Tags are Ordered by Region Number, State, and Emissions Totals for Elemental (HGO), Divalent Gaseous
(HG2), and Particulate (HGP) Mercury Emissions.
Mercury Emissions (tpy)
Region/State
Region 1
Connecticut
Maine
Massachusetts
Rhode Island
Vermont
New Hampshire
Region 2
New Jersey
New York
Source Name/ Description
Bridgeport RES CO
Mattabassett Reg. Sewage
Mid-CT Project (CRRA)
Naugatuck Treatment Co.
Southeastern CTt RRF
Collective Sources (remaining sources in state)
Dragon Products Co.
Greater Portland Region RRF
Mid Maine Waste Action Corp.
Penobscot Energy Recovery
Collective Sources (remaining sources in state)
Brayton Point
Pittsfield RRF
SE Mass RRF
Springfield RRF
Collective Sources (remaining sources in state)
Narragansett Bay Comm.
Rhode Island Hospital
WoonsocketWWTF/NETCo
Zambarano Memorial Hospital
Collective Sources (remaining sources in state)
Health Services
Residential Fuel Comb.
Collective Sources (remaining sources in state)
Merrimack
Schiller
SES Claremont RRF
Wheelabrator Concord
Collective Sources (remaining sources in state)
Camden RRF
Co Steel Sayreville
Essex Co. RRF
Hudson
NY/NJ Harbor Counties
Collective Sources (remaining sources in state)
American Ref-Fuel Co Niagara
Niagara Falls
Niagara Mohawk Pwr Corp
Wheelabrator Westchester
Counties along Lake Ontario
Counties along NY/NJ Harbor
Collective Sources (remaining sources in state)
Source Type
Incineration
Incineration
Incineration
Incineration
Incineration
Miscellaneous industrial processes
Incineration
Incineration
Incineration
Coal fired utility
Incineration
Incineration
Incineration
Incineration
Incineration
Incineration
Incineration
Health Services
Residential Fuel Comb.
Coal fired utility
Coal fired utility
Incineration
Incineration
Incineration
Miscellaneous industrial processes
Incineration
Coal fired utility
Utility - Other fuel
Incineration
Miscellaneous industrial processes
Incineration
HGO
0.018
0.008
0.010
0.006
0.010
0.146
0.013
0.003
0.015
0.001
0.215
0.031
0.045
0.021
0.027
0.196
0.005
0.020
0.013
0.019
0.029
0.004
0.009
0.004
0.003
0.001
0.015
0.002
0.054
0.011
0.179
0.047
0.011
0.334
0.462
0.130
0.035
0.137
0.024
0.144
0.242
0.540
HG2
0.049
0.021
0.027
0.017
0.027
0.061
0.002
0.008
0.039
0.002
0.053
0.083
0.119
0.056
0.070
0.126
0.012
0.012
0.008
0.011
0.019
0.0
0.005
0.001
0.052
0.003
0.039
0.005
0.026
0.029
0.022
0.123
0.028
0.077
0.129
0.078
0.093
0.082
0.064
0.077
0.060
0.464
HGP
0.017
0.007
0.009
0.006
0.009
0.035
0.002
0.003
0.013
0.001
0.019
0.007
0.041
0.019
0.024
0.060
0.004
0.008
0.005
0.007
0.010
0.0
0.004
0.001
2.0E-04
2.7E-04
0.013
0.002
0.017
0.010
0.022
0.042
0.002
0.048
0.067
0.052
0.032
0.055
0.022
0.031
0.039
0.130
7-2
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Mercury Emissions (tpy)
Region/State
Region 3
Delaware
District of Columbia
Maryland
Pennsylvania
Virginia
West Virginia
Region 4
Alabama
Florida
Source Name/ Description
Edge Moor
Indian River
Motiva Enterprises
Occidental Chemical Corp.
Collective Sources (remaining sources in state)
Benning
Collective Sources (remaining sources in District)
Baltimore Res Co
Brandon Shores
Chalk Point
Lehigh Portland Cement
Morgantown
Phoenix Servicesjnc.
Collective utilities
Collective Sources (remaining sources in state)
Bruce Mansfield
General Electric Co.
Harrisburg WTE
Homer City
Keystone
Montour
Shawville
Collective utilities
Collective Sources (remaining sources in state)
Chesapeake Energy Center
Chesterfield Power Station
Jewel Coke Company LLP
NASA Refuse-fired Steam Gen.
Norfolk Navy Yard
Collective Sources (remaining sources in state)
Fort Martin
John E Amos
Mitchell (VW)
Mt. Storm Power Station
Philip Sporn
PPG Industries, Inc
Collective utilities
Collective Sources (remaining sources in state)
Gaston
Gorgas
Miller
Occidental Chem Muscle Shoal
Mobile Bay area
Collective Sources (remaining sources in state)
Crist
Crystal River
F. J. Gannon
St. Josephs Hospital
Pensacola Bay area
South-FL urban area
Source Type
Coal fired utility
Coal fired utility
Petroleum refineries & related
industries
Inorganic chem processes (chlor-
alkali)
Oil fired utility
Incineration
Coal fired utility
Coal fired utility
Miscellaneous industrial processes
Coal fired utility
Incineration
Coal fired utility
Mineral Products
Incineration
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Ferrous Metals Processing
Incineration
Incineration
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Inorganic chem processes (chlor-
alkali)
Coal fired utility
Coal fired utility
Coal fired utility
Inorganic chem processes (chlor-
alkali)
Coal fired utility
Coal fired utility
Coal fired utility
Incineration
HGO
0.011
0.019
0.041
0.510
0.013
0.001
0.003
0.027
0.100
0.055
0.030
0.050
0.003
0.108
0.381
0.315
0.210
0.070
0.238
0.238
0.157
0.119
0.719
0.936
0.023
0.047
0.135
0.025
0.021
0.470
0.058
0.124
0.058
0.168
0.071
0.537
0.352
0.090
0.164
0.136
0.604
0.380
0.287
0.574
0.028
0.071
0.054
0.492
0.019
0.141
HG2
0.023
0.049
0.006
0.027
0.003
3.3E-04
0.001
0.071
0.154
0.137
0.005
0.132
0.047
0.136
0.250
0.173
0.126
0.185
0.631
0.631
0.415
0.316
0.834
0.456
0.062
0.125
0.017
0.066
0.056
0.464
0.153
0.329
0.153
0.294
0.188
0.028
0.384
0.035
0.254
0.291
0.189
0.020
0.211
0.386
0.075
0.188
0.089
0.296
0.014
0.130
HGP
0.003
0.005
0.005
0.0
0.002
2.2E-04
0.001
0.024
0.013
0.014
0.005
0.012
0.012
0.013
0.143
0.016
0.084
0.064
0.057
0.057
0.037
0.028
0.128
0.302
0.006
0.011
0.017
0.023
0.019
0.133
0.014
0.029
0.014
0.026
0.017
0.000
0.033
0.022
0.022
0.026
0.002
0.000
0.048
0.131
0.007
0.017
0.007
0.197
0.007
0.056
7-3
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Mercury Emissions (tpy)
Region/State
Georgia
Kentucky
Mississippi
North Carolina
South Carolina
Tennessee
Region 5
Illinois
Indiana
Michigan
Source Name/ Description
Collective Sources (remaining sources in state)
Bowen
Olin Corp
Scherer
Wansley
Collective Sources (remaining sources in state)
Big Sandy
Ghent
H. L. Spurlock
Paradise Fossil Plant
Collective Sources (remaining sources in state)
Chambers of MS Inc-Clearview
Jack Watson
Pascagoula ERF
Victor J. Daniel
Collective Sources (remaining sources in state)
Belews Creek
BMWNC
Marshall
Roxboro
Waccama Lake area
Collective Sources (remaining sources in state)
Foster Wheeler Charleston RRF
Safety Disposal Systems
Wateree
Winyah Generating Station
Collective Sources (remaining sources in state)
Gallatin Fossil Plant
Johnsonville Fossil Plant
Kingston Fossil Plant
Olin Corp.
Collective Sources (remaining sources in state)
Joliet29
Joppa Steam
Powerton
Waukegan
Util in Chicago
Util outside Chicago
Non-util in Chicago
Collective Sources (remaining sources in state)
Cliffy Creek
Gibson Generating Station
Lehigh Portland Cement Kilns
Rockport
Tanners Creek
Utilities in Gary, IN
Utilities outside Gary, IN
Collective sources in Gary, IN
Collective Sources (remaining sources in state)
Central Wayne Co Sanitation
J. H. Campbell
Monroe Power Plant
StClair Power Plant
Sources in Detroit Metro
Lafarge Midwest Inc
Source Type
Coal fired utility
Miscellaneous industrial processes
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Landfills
Coal fired utility
Incineration
Coal fired utility
Coal fired utility
Incineration
Coal fired utility
Coal fired utility
Incineration
Incineration
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Miscellaneous industrial processes
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Cement kilns
Coal fired utility
Coal fired utility
Incineration
Coal fired utility
Coal fired utility
Coal fired utility
Mineral Products
HGO
0.705
0.088
0.587
0.375
0.045
0.370
0.072
0.104
0.062
0.125
1.456
0.071
0.047
0.019
0.062
0.319
0.067
0.107
0.058
0.116
0.223
0.448
0.021
0.008
0.032
0.049
0.599
0.048
0.056
0.069
0.615
0.463
0.222
0.210
0.440
0.310
0.359
0.530
0.621
0.644
0.120
0.108
0.059
0.121
0.031
0.105
0.522
0.089
0.740
0.037
0.095
0.159
0.070
0.440
0.218
HG2
0.451
0.233
0.031
0.219
0.120
0.366
0.190
0.121
0.096
0.149
0.743
0.009
0.084
0.050
0.037
0.146
0.177
0.064
0.155
0.259
0.124
0.494
0.055
0.124
0.083
0.037
0.358
0.128
0.148
0.182
0.032
0.392
0.099
0.094
0.197
0.090
0.149
0.417
0.178
0.239
0.113
0.162
0.010
0.321
0.082
0.049
0.469
0.029
0.276
0.097
0.149
0.229
0.050
0.279
0.038
HGP
0.170
0.021
0.000
0.007
0.011
0.098
0.017
0.010
0.008
0.013
0.398
0.009
0.007
0.017
0.002
0.087
0.016
0.043
0.014
0.023
0.030
0.114
0.019
0.033
0.008
0.003
0.183
0.011
0.013
0.016
0.000
0.107
5.0E-04
4.7E-04
0.001
4.6E-04
0.001
0.022
0.132
0.161
0.007
0.014
0.009
0.029
0.007
0.001
0.039
0.030
0.208
0.034
0.012
0.017
0.002
0.058
0.035
7-4
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Mercury Emissions (tpy)
Region/State
Minnesota
Ohio
Wisconsin
Region 6
Arkansas
Louisiana
New Mexico
Oklahoma
Texas
Source Name/ Description
Collective Sources (remaining sources in state)
Clay Boswell
Olmstead WTE Facility
Pope-Douglas Waste
Sherburne Co. Generating Plant
Taconite Facilities
Collective "wetland" sources near Lake Superior
Collective Minn./St. Paul area
Collective Sources (remaining sources in state)
ASHTA Chemicals Inc.
Cardinal
Conesville
Eastlake
Kyger Creek
W. H. Sammis
Collective utilities
Collective Sources (remaining sources in state)
Columbia
Pleasant Prairie
South Oak Creek
Vulcan Materials Chem Div
Collective Sources (remaining sources in state)
Ash Grove Cement Co
Carle Bailey Gen Stn
Independence
White Bluff
Collective Sources (remaining sources in state)
Big Cajun 2
Pioneer Americas Inc.
PPG Industries-lnc.
R. S. Nelson
Collective Sources (remaining sources in state)
Escalante
Four Corners
Los Alamos Natl Lab
San Juan
Collective Sources (remaining sources in state)
AES Shady Point-lnc.
Holnam-lnc.
Muskogee
Sooner
Collective Sources (remaining sources in state)
ALCOA AI& Chem
Big Brown
Chemical Waste Mgmt
Monticello
Collective Sources (remaining sources in state)
Source Type
Coal fired utility
Solid waste incineration
Solid waste incineration
Coal fired utility
Taconite
Miscellaneous industrial processes
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Inorganic chem processes (chlor-
alkali)
Incineration
Miscellaneous industrial processes
Coal fired utility
Coal fired utility
Coal fired utility
Miscellaneous industrial processes
Inorganic chem processes (chlor-
alkali)
Coal fired utility
Coal fired utility
Coal fired utility
Miscellaneous industrial processes
Coal fired utility
Coal fired utility
Mineral products
Coal fired utility
Coal fired utility
Non-ferrous metals processing
Coal fired utility
Incineration
Coal fired utility
HGO
0.465
0.154
0.004
0.009
0.269
0.226
0.028
0.333
0.117
0.786
0.096
0.175
0.076
0.066
0.073
1.103
0.526
0.125
0.282
0.077
0.514
0.580
0.145
0.100
0.129
0.172
0.259
0.187
0.572
0.580
0.073
0.499
0.042
0.502
0.012
0.491
0.035
0.198
0.110
0.138
0.099
0.330
0.558
0.153
0.340
0.511
4.047
HG2
0.282
0.012
0.020
0.024
0.016
0.024
0.016
0.080
0.096
0.041
0.201
0.253
0.197
0.175
0.187
0.798
0.214
0.036
0.126
0.056
0.027
0.275
0.050
0.060
0.058
0.077
0.106
0.084
0.030
0.031
0.033
0.157
0.001
0.019
0.007
0.023
0.005
0.007
0.053
0.061
0.044
0.132
0.077
0.280
0.117
0.533
1.108
HGP
0.104
0.003
0.000
0.008
0.006
0.010
0.007
0.035
0.036
0.000
0.018
0.022
0.018
0.016
0.016
0.068
0.155
1.8E-04
0.001
0.002
0.0
0.091
0.055
0.040
2.9E-04
0.001
0.068
0.001
0.0
0.001
1.7E-04
0.113
1.9E-04
0.004
0.005
0.006
0.004
0.001
0.037
3.1E-04
2.2E-04
0.027
0.074
0.001
0.129
0.003
0.389
7-5
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Mercury Emissions (tpy)
Region/State
Region 7
Iowa
Kansas
Missouri
Nebraska
Region 8
Colorado
Montana
North Dakota
South Dakota
Utah
Wyoming
Source Name/ Description
Council Bluffs
Dubuque
George Neal North
George Neal South
Collective utilities
Collective Sources (remaining sources in state)
Ash Grove Cement Co.
Jeffrey Energy Center (Westar)
La Cygne (KCP&L)
Lawrence (Westar)
Collective Sources (remaining sources in state)
Doe Run Buick
latan
Labadie
Rush Island
Sioux
Thomas Hill
Counties around Kansas City
Collective Sources (remaining sources in state)
Gerald Gentlemen Station
Nebraska City
North Omaha
Sheldon
Collective Sources (remaining sources in state)
CEMEX-lnc.-Lyons Cement
Comanche
Craig
Pawnee
CF & 1 Steel L P DBA Rocky Mtn Steel Mills
Collective Sources (remaining sources in state)
Colstrip
Colstrip Energy
Livingston/Park County MWC
Stone Container Corp.
Collective Sources (remaining sources in state)
Antelope Valley Station
Coal Creek
Coyote
Milton R. Young
Collective Sources (remaining sources in state)
Big Stone
Health Services
Collective Sources (remaining sources in state)
Hunter
Intermountain Power
Ash Grove
Clean Harbors (formerly Aptus)
Davis/Wasatch
Huntington
Nucor Steel
Collective Sources (remaining sources in state)
Dave Johnston
Jim Bridger
Laramie River Station
Source Type
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Incineration
Coal fired utility
Coal fired utility
Coal fired utility
Lead Smelter
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Mineral products
Coal fired utility
Coal fired utility
Coal fired utility
Ferrous Metals Processing
Coal fired utility
Coal fired utility
Incineration
Wood, Pulp& Paper, Publishing Prod.
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Coal fired utility
Health services
Coal fired utility
Coal fired utility
Mineral Products
Incineration
Incineration
Coal fired utility
Internal Combustion: Other fuel
Coal fired utility
Coal fired utility
Coal fired utility
HGO
0.114
0.002
0.137
0.097
0.362
0.094
0.085
0.411
0.163
0.063
0.241
0.134
0.066
0.252
0.120
0.076
0.123
0.062
0.686
0.134
0.073
0.062
0.040
0.054
0.117
0.007
0.075
0.008
0.226
0.166
0.407
0.003
0.009
0.008
0.049
0.163
0.219
0.095
0.189
0.264
0.037
0.006
0.007
0.172
0.069
0.044
0.016
0.008
0.050
0.059
0.070
0.123
0.297
0.239
HG2
0.045
0.005
0.052
0.043
0.122
0.045
0.029
0.013
0.043
0.009
0.085
0.017
0.029
0.112
0.054
0.072
0.055
0.029
0.225
0.019
0.032
0.028
0.004
0.012
0.020
0.035
0.004
0.040
0.028
0.032
0.022
0.008
0.024
0.005
0.011
0.024
0.034
0.022
0.034
0.045
0.018
0.0
0.003
0.017
0.037
0.015
0.005
0.021
0.065
0.007
0.051
0.030
0.009
0.008
HGP
2.9E-04
4.8E-04
0.001
2.4E-04
0.002
0.028
0.032
0.002
0.003
0.001
0.043
0.017
1.5E-04
0.001
2.7E-04
0.004
2.8E-04
0.004
0.092
9.6E-05
1.8E-04
1.4E-04
4.0E-04
0.004
0.019
0.001
0.001
0.001
0.028
0.012
0.006
4.5E-05
0.008
0.003
0.005
0.001
0.002
0.013
0.001
0.015
2.4E-04
0.0
0.002
0.002
0.007
0.017
0.006
0.007
0.006
0.007
0.013
0.001
0.001
0.001
7-6
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Mercury Emissions (tpy)
Region/State
Region 9
Arizona
California
Nevada
Region 10
Idaho
Oregon
Washington
Source Name/ Description
Naughton
Collective Sources (remaining sources in state)
Apache Station
Cholla
copper mines
Coronado
Navajo
Northstar Steel Arizona
Springerville
Collective Sources (remaining sources in state)
Calaveras Cement Co
Hanson Permanente Cement
Long Beach SERRF
PG&E-GeysersUnits13&16
Riverside Cement Co.
RMC Pacific Materials
Sierra Army Depot
Collective cement plants
Collective Sources (remaining sources in state)
Newmont Lone Tree
Newmont Mining Corporation - Gold Quarry
Operations
Newmont Mining Corporation - Twin Creeks Mine
Barrick Goldstrike Mines, Inc
Florida Canyon
Glammis Marigold
Queenstake Resources USA Inc - Jerritt Canyon
Mine
Cortez Gold Mines Mill #2
Bald Mountain Mine
Denton Rawhide
Other NV Gold Mines Collective Emissions
Mohave
Collective Sources (remaining sources in state)
Amalgamated Sugar
INEEL INTEC
P4 Production LLC
Potlatch Pulp and Paperboard
Potlatch Wood Products Div.
Collective Sources (remaining sources in state)
Ash Grove Cement Co.
Cascade Steel
Portland General Electric Co.
Oregon Steel Mills-lnc.
Weyerhaeuser Company
Boardman
Collective Sources (remaining sources in state)
Ash Grove and Lafarge
Centralia
Georgia Pacific West Inc
Long View Fibre Co.
Source Type
Coal fired utility
Coal fired utility
Coal fired utility
Copper Mines
Coal fired utility
Coal fired utility
Ferrous Metals Processing
Coal fired utility
Mineral Products
Mineral Products
Utility - Other fuel & Incineration
Internal Combustion
Mineral Products
Mineral Products
Incineration
Mineral Products
Gold mining
Gold mining
Gold mining
Gold mining
Gold mining
Gold mining
Gold mining
Gold mining
Gold mining
Gold mining
Gold mining
Coal fired utility
Industrial Boilers - Coal
Miscellaneous Industrial Processes
Inorganic Chemical Mfg
Wood, Pulp & Paper, & Publishing
Products
Industrial Boilers - Other Fuel
Mineral Products
Ferrous Metals Processing
Utility: Internal Combustion
Ferrous Metals Processing
Industrial Boilers - Internal
Combustion
Coal fired utility
Mineral Products
Coal fired utility
Inorganic Chemical Mfg
Wood, Pulp & Paper, & Publishing
Products
HGO
0.071
0.194
0.057
0.118
0.098
0.118
0.119
0.150
0.152
0.075
0.956
0.187
0.211
0.398
0.256
0.123
0.304
0.224
1.156
0.269
0.106
0.194
0.282
0.078
0.397
0.121
0.056
0.066
0.062
0.037
0.032
0.064
2.9E-04
0.004
0.367
0.125
0.010
0.056
0.358
0.043
0.030
0.012
2.5E-04
0.055
0.235
0.042
0.183
0.047
0.022
HG2
0.018
0.031
0.003
0.009
0.012
0.006
0.031
0.019
0.007
0.019
0.166
0.032
0.206
0.239
0.044
0.021
0.105
0.039
0.477
0.038
0.045
0.018
0.023
0.135
0.054
0.720
0.026
0.034
0.107
0.021
0.008
0.015
1.7E-04
0.002
0.046
0.075
0.006
0.025
0.792
0.005
0.018
0.001
1.5E-
04
0.028
0.025
0.007
0.082
0.006
0.013
HGP
2.3E-04
0.020
0.001
0.001
0.012
0.002
0.001
0.019
0.002
0.012
0.153
0.030
0.104
0.159
0.041
0.020
0.115
0.036
0.292
0.005
0.008
0.006
0.007
0.008
0.005
0.018
0.002
0.002
0.007
0.001
0.000
0.005
1.1E-04
0.002
0.046
0.050
0.004
0.017
0.106
0.005
0.012
0.001
1.0E-04
0.001
0.083
0.007
0.000
0.003
0.009
7-7
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Mercury Emissions (tpy)
Region/State
Other Region
Ontario, Canada
Canada
All Other Canada
Mexico
Source Name/ Description Source Type
Spokane Reg. Disposal Incinerator
Tacoma Incinerator
Wood, Pulp & Paper, & Publishing
Collective pulp and paper Products
Collective Whatcom County
Collective Sources (remaining sources in state)
Collective Sources
Burnaby Refuse Incinerator
Collective Sources
Collective Sources
HGO
0.003
0.001
0.032
0.007
0.072
1.191
0.012
3.337
13.065
HG2
0.007
0.002
0.019
0.002
0.011
1.021
0.036
1.852
4.676
HGP
0.002
0.001
0.013
0.001
0.006
0.314
0.012
0.543
2.430
Initial & Boundary Condition Tags
CTM Global model providing boundary conditions -
GEOS-CHEM Global model providing boundary conditions -
GRAHM Global model providing boundary conditions -
__ . u. . /u ... Global model providing boundary condition
GRAHM (HGO) confribution from HGO
GRAHM (HG2)
GRAHM (HGP)
Global model providing boundary condition
contribution from HG2
Global model providing boundary condition
contribution from HGP
-
-
-
-
-
-
7.2. Mercury PPTM Results
In this section, the REMSAD mercury tagging results are examined for selected locations
throughout the modeling domain. This section also examines, in some detail, the simulated
contributions in relation to a source receptor study in Ohio.
The modeling results contain much more information than is presented here. To facilitate future
analysis, the tagging results have been incorporated into a GIS database tool (an enhanced version
of ARC-Hydro, developed by ESRI under a separate effort) that allows users to extract the results
for any grid cell or combination of grid cells and calculate the simulated contribution from each
tagged source or source category to any area of interest in the modeling domain, such as a county,
watershed, or body of water.
7.2.1. Contributions to Statewide Maximum Deposition
For each state, the contributions to mercury deposition were examined for the location of greatest
deposition from sources located within that same state. This is not necessarily the location of overall
maximum deposition for the state, but this approach allows us to focus on the intra-state
contributions. Note that contributions are summarized only for a single location in each state.
The summaries should not be taken as necessarily representative of contributions on a
statewide basis.
Figures 7-1 through 7-49 summarize, for each of the 48 states included in the modeling domain
and the District of Columbia, the simulated contributions at the location of maximum deposition
from in-state sources. The plots are ordered by EPA region and then alphabetically by state for
each region, as follows:
Figures 7-1 through 7-6 are for Region 1 (Connecticut, Maine, Massachusetts, New
Hampshire, Rhode Island, and Vermont).
7-8
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figures 7-7 and 7-8 are for Region 2 (New Jersey and New York).
Figures 7-9 through 7-14 are for Region 3 (Delaware, District of Columbia, Maryland,
Pennsylvania, Virginia, and West Virginia).
Figures 7-15 through 7-22 are for Region 4 (Alabama, Florida, Georgia, Kentucky,
Mississippi, North Carolina, South Carolina, and Tennessee).
Figures 7-23 through 7-28 are for Region 5 (Illinois, Indiana, Michigan, Minnesota, Ohio, and
Wisconsin).
Figures 7-29 through 7-33 are for Region 6 (Arkansas, Louisiana, New Mexico, Oklahoma,
and Texas).
Figures 7-34 through 7-37 are for Region 7 (Iowa, Kansas, Missouri, and Nebraska).
Figures 7-38 through 7-43 are for Region 8 (Colorado, Montana, North Dakota, South
Dakota, Utah, and Wyoming).
Figures 7-44 through 7-46 are for Region 9 (Arizona, California, and Nevada).
Figures 7-47 through 7-49 are for Region 10 (Idaho, Oregon, and Washington).
Part (a) of each figure is a map that shows the annual total mercury deposition for the state in
question. The location of the maximum simulated contribution from in-state sources is indicated
by the blue triangle on this plot.
Part (b) gives the modeling results for the grid cell location of the triangle in Part (a). The annual
total mercury deposition as simulated by REMSAD for this grid cell is given in the caption for
Part (b). The information in Part (b) includes the REMSAD modeling results as well as CMAQ-
derived information for global background contributions. Note the contributions to mercury
deposition are displayed only for this one grid cell location within the state. Since the REMSAD
modeling domain is defined by 12-km horizontal resolution, each grid cell covers a 12 by 12 km
area. Results should not be extrapolated to indicate source contributions on a statewide basis.
Part (b) of each figure consists of four plots, as follows:
1. The bar chart in the upper left-hand corner of the display compares the contribution to total
deposition from emissions versus background conditions. The first bar represents the
contribution to total deposition from all emissions sources, i.e., sources in the U.S., Canada,
Mexico, and re-emissions. The next four bars display the contribution of global background
concentrations to total deposition as estimated using REMSAD and the three sets of
initial/boundary conditions (CTM, GRAHM and GEOS-CHEM (G-C)), which are referred to
here as background. The average of the three initial/boundary contributions for REMSAD is
also presented. The next four bars display the contribution of global background
concentrations to total deposition as estimated using CMAQ and the three sets of
initial/boundary conditions (CTM, GRAHM, and G-C). The average of the three
initial/boundary contributions for CMAQ is also presented. The average values are used in
the remaining summary charts.
2. The bar chart in the upper right-hand corner of the display illustrates and compares wet and
dry deposition amounts comprising 1) total deposition from emissions sources, 2) average
background deposition from the REMSAD simulation, 3) average background deposition for
the corresponding CMAQ simulations, and 4) total overall deposition from REMSAD.
3. The pie chart in the lower left-hand corner of the display illustrates the percent contributions
to total deposition at the selected grid cell from 1) the initial and boundary conditions or
7-9 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
background (average of the three sets of conditions for REMSAD), 2) emissions from
sources within the state, 3) emissions from sources in neighboring states, 4) emissions from
all other U.S. states, 4) emissions from Canada and Mexico, and 5) re-emission processes.
Re-emissions on the plots refers to mass that has been re-emitted and subsequently
redeposited, as described in Section 2.1.4.
4. The double (or pie-in-pie) pie chart in the lower right hand corner of the display summarizes
the contributions from emissions sources only, without including the background. This plot
highlights the percent contributions from the in-state sources. The larger pie gives the
proportion of the overall contribution from emissions sources that are located outside of vs.
inside of the state, and the smaller pie details the contributions from the in-state sources
(specifically, the five largest in-state contributors as well as all other in-state sources). If
there are five or fewer tags for a given state, all of the tagged source contributions are
displayed.
The names of the sources are given in the legend. The "Collective Sources" tag for each
state includes all point and area sources in the state that are not tagged individually, as part
of a source category, or as part of a region. When a chart refers to "Other tagged sources
within" a state, it refers to sources that were tagged but contributed only a small amount to
deposition at the location chosen. These sources were therefore aggregated for the
purposes of the chart. The legend also includes the percentages represented by the various
segments of the pie charts. Note that the percentages for the cut-out pie chart segments are
calculated based on the total represented only in the smaller, cut-out pie chart. Note also
that very small contributions sometimes appear as zero percent.
In interpreting the results presented for the tagging simulations, the reader is reminded that all
model simulation results include some uncertainty, and that uncertainty is often difficult to quantify.
Therefore, although contribution values may be repored to tenths of a percent, this is done to
differentiate values that range widely in magnitude, not because of actual precision to that level.
The contribution results should be viewed in a relative sense more than an absolute sense.
7.2.2. Example Analysis Using the Contribution Charts
To aid the use and interpretation of the contribution charts, the following discussion of the
results is provided for Connecticut (Figures 7-1a and b).
The map in Figure 7-1 a shows the location of maximum simulated mercury deposition within
Connecticut from sources located within Connecticut. The location, in the northern part of the
state, is marked with the blue triangle. All of the charts in Figure 7-1 b summarize deposition
amounts at this location and are not meant to represent an analysis for the entire state. Note
that this is not the location of maximum deposition within the illustrated area. That maximum,
indicated by a small"+" sign, is located in the southeastern part of the domain, in the NY/NJ
harbor area.
In Figure 7-1 b, mercury deposition at the marked location is broken down in various ways. The
total deposition is 46.9 g km"2, consisting of (from the upper left chart) 32.9 g km"2 from
emissions (from the U.S., Canada, Mexico, and re-emission) and 14.0 g km"2 from the average
background from REMSAD. For comparison purposes, the average background as simulated by
CMAQis 18.3gkm"2.
The chart in the upper right-hand corner of the display indicates that the contribution to total
deposition from dry deposition is approximately double (31.3 g km"2) that from wet deposition
(15.6 g km"2) at the selected location. The relative amount of wet versus dry deposition is
7-10 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
different for the emissions versus the background. The emissions contribution is characterized
by a much higher dry deposition amount (6.0 versus 26.9 g km"2 for wet and dry deposition,
respectively). The REMSAD average background contribution is characterized by higher wet
deposition (9.6 versus 4.4 g km"2 for wet and dry deposition, respectively). The CMAQ-derived
average background estimates indicate more similar amounts of wet and dry deposition (8.4
versus 9.9 g km"2, respectively).
In the pie charts, the contributions from emissions sources are broken out in detail.
The chart in the lower left-hand corner indicates that 55.8 percent of the total mercury
deposition is from sources that are located in Connecticut. Average background (boundary
conditions) contributes 29.8 percent. The remaining deposition is broken down as follows: 7.6
percent from emissions sources in neighboring states (states that share a state boundary with
Connecticut), 5.2 percent from emissions sources in other U.S. states, 0.3 percent from
emissions sources in Canada and Mexico, and 1.4 percent from re-emission processes.
The lower right pie-in-pie chart displays percent contribution to the emissions-only portion of the
total contribution, i.e. not including background, from emissions sources outside Connecticut
and sources located within the state. The first pie chart indicates that 20.5 percent of the
emissions-only contribution to mercury deposition at the selected location is from outside
sources and 79.5 percent is from in-state sources. The second pie chart indicates that, of the
contributions from in-state sources, 92.5 percent is from the Mid-CT Project (CRRA). From this
it can be computed that emissions from the Mid-Ct Project (CRRA) account for 73.5 percent
(92.5 percent of the 79.5 percent contribution from Connecticut sources) of the deposition from
emissions sources and 51.6 percent (92.5 percent of the 55.8 percent contribution from CT
sources) of the overall deposition. "CT Collective Sources" (those sources in Connecticut that
were not tagged as individual sources, as part of a category of sources, or as part of a region)
contribute the next largest amount at 4.6 percent of the in-state emissions contribution, which is
3.7 percent (4.6 percent of 79.5 percent) of the total emissions contribution and 2.6 percent (4.6
percent of 55.8 percent) of the overall deposition. "Other tagged sources within CT" contribute
0.3 percent of the in-state emissions contribution, or less than 0.2 percent (0.3 percent of 55.8
percent) of the overall deposition. Note that this terminology "Other sources within CT" refers to
sources that were tagged individually or as part of a source category but that contributed only a
small amount of the deposition at this location, and so they are aggregated for the purposes of
this chart.
An appropriate summary for this location would be that about half the deposition is from the Mid-
CT Project (CRRA), about one-third from background and re-emissions, with the remainder from
other sources located in Connecticut, neighboring states, other states, Canada, and Mexico.
7.2.3. Comparison with a Source Apportionment Study for Ohio
It is of great interest to evaluate the contribution analysis against studies of apportionment
based on observed data, but such studies are very limited in number. A study by Keeler et al.
(2006) used receptor modeling to estimate contributions to wet deposition of mercury at
Steubenville, Ohio. Although the study was for the years 2003 and 2004 (while this modeling is
for 2001), it is still interesting to compare the conclusions of that study with the results of the
contribution analysis derived from this modeling.
The Keeler study used air monitoring and wet deposition data along with statistical receptor
modeling to estimate contributors to wet deposition of mercury at their Steubenville, Ohio site. In
order to compare to the results of the Keeler study, simulation results were extracted from the
7-11 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
REMSAD 2001 tagging simulations at the location of the Steubenville site (40.379 N, 80.620
W). The simulation results are summarized in Figure 7-50a in the same format as has been
presented for the state-by-state results. Since Keeler's study dealt only with wet deposition,
simulation results are presented in Figure 7-50b for wet deposition only. The wet deposition
charts include details of the contributions from sources within Ohio (lower left chart) and sources
within neighboring states (lower right). The location of the site is shown on the spatial map in
Figure 7-51.
Keeler's study estimated that about 70 percent of the mercury wet deposition at the Steubenville site
came from coal combustion. Another 6 percent came from iron/steel production. Since the Keeler
study used the presence of trace elements as indicators of the originating processes, these
estimates would apply to all coal combustion and all iron and steel production, not to specific
sources. Although the methodology did not tag all coal combustion sources, estimates can be
developed from the REMSAD simulation results for the purpose of comparison with the Keeler
study.
Utilizing the average background contributions, the REMSAD PPTM results indicate that 49.8
percent of the wet deposition is from Ohio sources at Steubenville. Of that portion, at least 97.6
percent is from coal fired utilities. Another 13.6 percent of the overall deposition comes from
neighboring states, and, of this portion, at least 66.6 percent is from coal fired utilities. This
gives an estimate that 57.7 percent or more of the wet deposition at this site comes from coal-
fired utilities. Since all coal-fired utilities in the states neighboring Ohio or in the rest of the U.S.
were not tagged, it is likely that there is some additional contribution to deposition from coal
combustion. Iron and steel production in this area also were not tagged, so a comparison for
this source category cannot be made.
Although the Keeler study did not apportion dry deposition among industrial sectors, it was
found from the REMSAD PPTM analysis that the total wet and dry deposition contribution from
utility sources is on the order of at least 50 to 60 percent.
The REMSAD-based estimates are consistent with the results from the Steubenville study.
Additional tags are needed in order to account for a greater fraction of the coal combustion and
to ascertain whether a comparable estimate to Keeler's could be made for the contribution from
iron and steel production.
7-12 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-1. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Connecticut.
a.
LEVEL 1 THG (g/kin3)
Time: 0 Jan '.. 2001-0 Deo 31, 2001
MAXIMUM - S-1 2 g/ktnB (6.-I)
MINIMUM - 12.3 g/ktn2 (1,24)
J
I
t-
Annual total wct+dry deposition of TIIC
within Connecticut
7-13
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-1 b. Connecticut. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (46.9 g/km2).
Connecticut
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
35 -1
30 -
25-
20 -
15 -
10 -
5-
32.9
f
14.2
20.4 19.6 183
124 15-3 14.0 [~| 14.8 _ j-|
Emiss CTM GRHM G-C
REMSAD
Avg
CTM GRHM G-C Avg
CMAQ
31.3
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Connecticut Sources without Background (5,6)
JAvg Background
(REMSAD) 29.8%
I Connecticut 55.8%
D Neighboring states 7.6%
DOtherU.S. 5.2%
D Canada & Mexico 0.3%
DReemission 1.4%
2nd pie
I Sources outside CT
(20.5%)
I CT sources (79.5%)
DCT Mid-CT Project (CRRA)
92.5%
DCT Collective Sources 4.6%
DCT Mattabassett Reg.
Sewage 1.4%
CT Bridgeport RES CO
0.7%
DCT Naugatuck Treatment
Co. 0.5%
Other tagged sources within
CT 0.3%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-14
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-2. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Maine.
l.KVKI. 1 TUG (s/kmZ) * MAXIMUM - 27.6 g/krr.Z (3.8)
Time: 0 Jan 1. 3001 0 Dec 31. 30SWIMUM - 4.1 g/kro2 (5.43)
o
I
I ' ' ' ITT 1
J-lssa.
Annual total wet-dry deposition of TKG 3001
within Maine
7-15
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-2b. Maine. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (26.5 g/km2).
Maine
Emissions vs. Background Contributions (g/km2) (1,2,3,4)
Wet vs. Dry Contributions (g/km2) (1,2)
20 18-4
15-
10 -
5 -
^^^
£1 7.0 ±1
r
i
n
i i
Emiss CTM GRHM G-C
REMSAD
8.1
Avg
15.1
i 1
13.1 12.5
9.3
CTM GRHM G-C Avg
CMAQ
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
19.6
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Maine Sources without Background (5,6)
DAvg Background
(REMSAD) 30.5%
Maine 62.9%
D Neighboring states 0.5%
Other U.S. 4.2%
D Canada & Mexico 0.6%
Reemission 1.3%
2nd pie
Sources outside ME (9.4%)
I ME sources (90.6%)
D ME Mid Maine Waste
Action Corp. 96.7%
ME Collective Sources
3.0%
DME Greater Portland
Region RRF 0.3%
ME Dragon Products Co.
0.0%
DME Penobscot Energy
Recovery 0.0%
Other tagged sources within
ME 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-16
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-3. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Massachusetts.
a.
LEVEL 1 TUG (fi/km2)
Time: 0 Jan 1. 2001 0 Dec 31, 2001
+ MAXIMUM - 46.9 g/km2 (13,H)
MIKIMTJM = 7.7 g/km2 (1,23
- S36.
Annual total wet+dry deposition of T1IG 2001
within Massachusetts
7-17
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-3b. Massachusetts. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (39.1 g/km2).
Massachusetts
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
35 -1
30 -
25-
20 -
15-
10 -
5-
29.7
1
19-2 17.9
9.5 83 10.2 9.4 1 _^
1 1 ^^m ^H ^H I
LJ n LJ LJ LJ LJ
Emiss CTM GRHM G-C Avg CTM GRHM G-C
REMSAD CMAQ
16.7
1
Avg
27.8
Emissions Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Massachusetts Sources without Background (5,6)
Avg Background
(REMSAD) 24.0%
Massachusetts 67.9%
DNeighboring states 2.6%
DOtherU.S. 4.3%
D Canada & Mexico 0.3%
Reemission 0.9%
2nd pie
I Sources outside MA
(10.6%)
I MA sources (89.4%)
DMAPittsfieldRRF98.6%
MA Collective Sources
0.7%
DMA Springfield RRF 0.6%
MABrayton Point 0.1%
D MA SE Mass RRF 0.1%
I Other tagged sources within
MA 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-18
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-4. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for New Hampshire.
a.
LEVEL 1 THG (g,/km3) < MAXIMUM - Jl.O «/km2 (lfl.1)
Time: 0 Jan 1, 3001-0 Deo 31. 3OS1NIMUM - i.a g/t
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-4b. NewJHampshire. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (27.6 g/km2).
New_Hampshire
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
25 -,
20 -
15-
10 -
5-
20.2
n
7.8
16.9
I 10.9
6.3 8'° 7.4
I 1 I I I I
I I I I I I
15.2 143
|
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
21.7
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from New_Hampshire Sources without Background (5,6)
Avg Background
(REMSAD) 26.7%
New_Hampshire 64.9%
D Neighboring states 1.7%
DOtherU.S. 5.4%
D Canada & Mexico 0.4%
Reemission 0.9%
2nd pie
I Sources outside NH
(11.5%)
I NH sources (88.5%)
DNHSESCIaremontRRF
98.7%
DNH Collective Sources
0.9%
DNHMerrimackO.3%
NH Wheelabrator Concord
0.0%
DNH Schiller 0.0%
I Other tagged sources within
NH 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-20
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-5. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Rhode Island
a.
LEVEL 1 THG (g/kmS)
Time.-: D Jail ]. S001 C Ccc 31, 2001
+ MAXIMUM - 16.9 g/kmZ (1,10)
MINIMUM - 13.2 g/kmS (1.18)
Annual total wet * dry deposition of THG
within Rhode Island
2001
7-21
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-5b. Rhodejsland. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (21.3 g/km2).
Rhodejsland
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
25 1
20 -
15-
10 -
5-
19-4 18.4 172
9.9 117 10.0
12-4 11.4
13.8
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
11.2
,10.1
Emissions Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Rhodejsland Sources without Background (5,6)
DAvg Background
(REMSAD) 53.5%
Rhodejsland 17.3%
D Neighboring states 13.4%
Other U.S. 12.8%
D Canada & Mexico 0.5%
Reemission 2.4%
2nd pie
I Sources outside Rl (62.7%)
IRI sources (37.3%)
DRI Collective Sources
53.7%
DRI Zambarano Memorial
Hospital 26.2%
DRI Rhode Island Hospital
8.5%
Rl Narragansett Bay
Comm. 6.5%
DRI Woonsocket WWTF/NET
Co 5.2%
Other tagged sources within
Rl 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-22
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-6. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Vermont.
a.
I.i':v«:i. 1 TUG (g/km2) + MAXIMUM - 40.0 g/km2 (2,38)
Time: 0 Ian 1, 2001 0 Dec 31. 300LNIMUM - 7.7 g/kmS (7.16)
Km
778. 1890. 201B.
Annual total wct+dry deposition, of THG
wiLhiri Vermont.
7-23
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-6b. Vermont. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (14.4 g/km2).
Vermont
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
20 -i
15 -
10 -
5-
16.4
11.3 11-8 -|08
3.6
| |
^_^
9.2
I 1
i i
ID. I 143
11.4
__
^^
i 1
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
8.3
Emissions Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Vermont Sources without Background (5,6)
Avg Background
(REMSAD) 75.1%
Vermont 3.0%
D Neighboring states 4.8%
Other U.S. 11.7%
D Canada & Mexico 2.3%
Reemission 3.0%
2nd pie
Sources outside VT
(87.8%)
I VT sources (12.2%)
DVT Residential Fuel Comb.
75.9%
IVT Collective Sources
24.1%
DVT Health Services 0.0%
I Other tagged sources within
VT 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-24
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-7. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for New Jersey.
LEVEL 1 THG (g/km3) + MAXIMUM - IIO.S « ktn2 (S.ie)
Time: 0 Jan 1, 2001-0 Deo 31, 30HINJMUU - 14.5 g/kmZ (12.2)
Annual lotal wet+dry deposition of THG 2001
within Mew Jersey
7-25
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-7b. New_Jersey. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (39.1 g/km2).
New_Jersey
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
25 1
20 -
15-
10 -
5-
23.0
r
16.3 177
16.1
21.5
16.3
21 .C
)
19.6
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
Emissions Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from New_Jersey Sources without Background (5,6)
DAvg Background
(REMSAD) 41.2%
New_Jersey 36.0%
D Neighboring states 13.8%
Other U.S. 6.4%
D Canada & Mexico 0.3%
Reemission 2.3%
2nd pie
I Sources outside NJ (38.7%)
INJ sources (61.3%)
DNJ Collective Sources
40.0%
NJ NY/NJ Harbor Counties
33.1%
DNJ Essex Co. RRF 20.6%
I NJ Hudson 3.6%
DNJ Co Steel Sayreville
2.3%
Other tagged sources within
NJ 0.4%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-26
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-8. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for New York.
LEVEL 1 THG (K/kmS>)
Tirror 0 Jan 1, 2001-0 Don fl1, 20O1
+ MAXIMUM 12.1.8 g/kmE (13.1)
- MINIMUM = 7.7 E/kmZ (43.37)
1344. 1404.
I I I I I
Km
1564. 1704. 1S24.
I | I I I II I I I | | I I I I I |
Annual total wet+dry deposition of TIIG 2001
within New York
7-27
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-8b. New_York. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (40.5 g/km2).
New_York
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
30 -1
25 -
20 -
15 -
10-
5-
27.1
n
172 18.0
13.8 .._ 14.8 134 __ 15.0
i 1
16.8
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
,19.2
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from New_York Sources without Background (5,6)
Avg Background
(REMSAD) 33.1%
New_York47.0%
D Neighboring states 4.8%
DOtherU.S. 7.6%
D Canada & Mexico 5.9%
DReemission 1.5%
2nd pie
I Sources outside NY
(29.7%)
I NY sources (70.3%)
DNY American Ref-Fuel Co
Niagara 60.6%
DNY Collective Sources
23.0%
DNY Counties along Lake
Ontario 7.4%
NY Niagara Falls 5.3%
DNY Niagara Mohawk Pwr
Corp 3.6%
Other tagged sources within
NY 0.1%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-28
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-9. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Delaware.
:.KVKL 1 THC- (g/km3) + MAXIMUM - 110.2 g/kmS
Time; 0 Jan 1, S001-0 Dec 31, 3001 - MINIMUM - 12.7 (/kntS (5,3)
Annual total wet) dry deposition of THG SOOl
within Delaware
7-29
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-9b. Delaware. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (52.8 g/km2).
Delaware
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
40 -1
35 -
30 -
25-
20-
15 -
10-
5 -
36.2
n
Emis
1R7 18-4 1RR 18'6 18'9
16-7 14.6 16'6 _ 15.0 | 1
r n 1 n n
s CTM GRHM G-C Avg CTM GRHM G-C
REMSAD CMAQ
17.5
Avg
30.8
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Delaware Sources without Background (5,6)
DAvg Background
(REMSAD) 31.4%
D Delaware 42.9%
D Neighboring states 19.9%
Other U.S. 3.9%
D Canada & Mexico 0.2%
Reemission 1.7%
2nd pie
I Sources outside DE
(37.5%)
IDE sources (62.5%)
DDE Occidental Chemical
Corp. 85.8%
DE Motiva Enterprises
7.4%
DDE Edge Moor 5.0%
DE Collective Sources
1.6%
DDE Indian River 0.3%
I Other tagged sources within
DE 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-30
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-10. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Washington, D.C.
a.
I.KVKI. 1 THG (g/kmS)
Time: 0 Jon 1. 3001 0 Dec 31. 2001
+ MAXIMUM - 99.3 g/km2 [10,IB)
MINIMUM - IS.7 K/ltma (:3.8)
Annual lot.al wnH-dry dc:poKi(.ior. of THG 2001
within District of Columbia
7-31
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-1 Ob. Washington DC. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (31.1 g/km2).
Washington DC
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
20 1
15 -
10-
5-
17.5 18-6 18.4 1RB
f
13.6
11.9
15.1
13.6
n
13.6
n
~
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
18.8
2.3
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from DC Sources without Background (5,6)
Avg Background
(REMSAD) 43.7%
DC 1.7%
D Neighboring states 9.4%
DOther U.S. 42.2%
D Canada & Mexico 0.2%
Reemission 2.7%
1st pie
2nd pie
I Sources outside DC
(97.0%)
I DC sources (3.0%)
DDC Collective Sources
99.0%
DDCBenning 1.0%
D Other tagged sources within
DC 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-32
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-11. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Maryland.
1
LEVEL 1 THG (g/km2)
Time: 0 Jan 1, 2001-0 Dec 31, 3001
1416.
+ MAXIMUM = 124.9 g/km2 (7.26)
- MINIMUM = 11.2 g/km2 (8,11)
1B96.
160.
-72.
e" s' '-
.. ...
" ' a J'' '
Annual total wet) dry deposition of TH
within Maryland
7-33
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-11 b. Maryland. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (99.3 g/km2).
Maryland
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
100 -1
80 -
60 -
40
20 H
0
86.6
Emiss
12.8 11.1 14.0 12.7
CTM GRHM G-C Avg
REMSAD
18'°
12.8
177 16-2
CTM GRHM G-C Avg
CMAQ
60.9
Emissions Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Maryland Sources without Background (5,6)
Avg Background
(REMSAD) 12.7%
D Maryland 80.9%
D Neighboring states 4.0%
DOtherU.S. 1.5%
DCanada & Mexico 0.1%
Reemission 0.7%
2nd pie
I Sources outside MD (7.3%)
IMD sources (92.7%)
DMD Phoenix Servicesjnc.
44.2%
DMD Collective Sources
33.2%
DMD Brandon Shores 10.2%
MD Baltimore Res Co 7.7%
DMD Collective utilities 3.7%
Other tagged sources within
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-34
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-12. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Pennsylvania.
a.
T.EVTCI, 1 TUG (g/km2)
Time: 0 Jan 1, 2001 0 Dec 31, 2001
+ MAXIMUM 124.8 g/km2 (16,17)
MINIMUM - ll.S g/fciriH (19.2)
Km
1&24, 1844.
m i i i i i i i i
: I . I ; 1 L_LL J 1 ! ! J 1 ,?_!_ 1_J I M. 1 /! I I I I A! i ! I
396.
- 156.
4-0 .8-0 '.S.o'S.Q^O.o^.O^.oaa.o^.o i
*'° «- J? a«*
Annual total wet I dry deposition of THG 2001
within Pcnnsv'lvHnifi
7-35
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-12b. Pennsylvania. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (124.9 g/km2).
Pennsylvania
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
120 -1
100 -
80 -
60 -
40 -
20 -
0
113.8
Emiss
11.2 9.7 12.4 11.1
i~i , , ,
CTM GRHM G-C Avg
REMSAD
18.1 138 18.1 16.7
n n n
CTM GRHM G-C Avg
CMAQ
120 n 102.7
109.3
Emissions Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Pennsylvania Sources without Background (5,6)
Avg Background
(REMSAD) 8.9%
Pennsylvania 86.7%
DNeighboring states 2.8%
DOtherU.S. 1.1%
DCanada & Mexico 0.1%
Reemission 0.4%
2nd pie
Sources outside PA (4.8%)
I PA sources (95.2%)
D PA Keystone 53.9%
PA Homer City 42.4%
DPA Collective utilities 1.7%
PA General Electric Co.
0.8%
DPA Collective Sources 0.7%
1 Other tagged sources within
PA 0.5%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-36
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-13. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Virginia.
LEVEL 1 TUG (g/ltm2)
Time: 0 Jan 1, 2001-0 Dec 31, 20O1
4- MAXIMUM - 9B.U H/kmZ (S2.41)
- MINIMUM - IJ.K K/km2 (Uk!,34)
133
108.
Annual I.ol.nl wc:l,*dry dcipoail.ion of TMO 2001
within Virginia
7-37
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-13b. Virginia. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (51.0 g/km2).
Virginia
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
40 -1
35 -
30 -
25-
20-
15 -
10-
5 -
E
38.0
mis
13.2 113 14-5 13.0 116 12.6 -|-|4
ni I I 1 I 1 I I I 1
I I I I I I I I I I
s CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
31.8
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Virginia Sources without Background (5,6)
Avg Background
(REMSAD) 25.5%
Virginia 68.0%
DNeighboring states 2.4%
DOtherU.S. 2.5%
DCanada & Mexico 0.1%
DReemission 1.4%
2nd pie
I Sources outside VA (8.7%)
IVA sources (91.3%)
D VA Collective Sources
70.0%
DVA Chesapeake Energy
Center 17.0%
DVA Norfolk Navy Yard
11.2%
VA NASA Refuse-fired
Steam Gen. 1.4%
DVA Chesterfield Power
Station 0.3%
Other tagged sources within
VA 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-38
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-14. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for West Virginia.
+ MAXIMUM - 12-1.8 s/kmZ (26.41)
- MINIMUM - 11.2 g/kmH (27,26)
EL 1 THG (g/km3)
Time: 0 Jan 1. 200! G DC.-C 3\. 3001
i i i i i i i i-ttr-i i i i i i i i I i i i i i i i i i I i i I _
Annual LoUl wet+dry deposil.ion of THG
within West Virginia
2001
7-39
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-14b. West_Virginia. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (75.6 g/km2).
West_Virginia
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
70 -i
60 -
50-
40 -
30-
20 -
10-
57.2
I
|
Emis
lal 163 20.8 18.4 22.1 193 23* 21.7
^_ i 1 | 1 i 1 1 j
1 1 n 1 1 1 1 1 1 1 1 1 1
s CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
50 n
40-
46.1
9.5
Emissions Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from West_Virginia Sources without Background (5,6)
DAvg Background
(REMSAD) 24.3%
West_Virginia 60.6%
D Neighboring states 11.4%
Other U.S. 2.2%
DCanada & Mexico 0.1%
Reemission 1.3%
2nd pie
I Sources outside WV
(20.0%)
IWV sources (80.0%)
DWV PPG INDUSTRIES-
INC. 48.2%
WV Mitchell (WV) 27.9%
DWV Collective utilities
19.0%
WV John E Amos 1.6%
DWV Collective Sources
1.5%
Other tagged sources within
WV 1.9%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-40
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-15. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Alabama.
LEVEL 1 THG (g/km2) < MAXIMUM - 76.5 g/km2 (15.40)
Time: 0 Jan 1. 2001-0 Deo 31. 30B1NIMUM - 15.5 (!/km2 (39.22)
T 3.0 *.a i*o J«.o1?.o^.oe«.o*.oa».o * i
° a.0 ie.0 >e:a?e_a?-i:a2B:03?_a3e_a-'6
Annual total wet-dry deposition of TKG 2001
within Alabama
7-41
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-15b. Alabama. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (76.4 g/km2).
Alabama
Emissions vs. Background Contributions (g/km2) (1,2,3,4)
Wet vs. Dry Contributions (g/km2) (1,2)
70 1
60 -
50-
40 -
30-
20 -
10-
583
. |
Emis
17.6 15.g 20'9 18.1 20.2 19.5 ^ 21.2
n n n n 1 1 n
s CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
69.7
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Alabama Sources without Background (5,6)
Avg Background
(REMSAD) 23.7%
Alabama 73.3%
D Neighboring states 0.9%
DOtherU.S. 1.0%
D Canada & Mexico 0.0%
Reemission 1.0%
2nd pie
I Sources outside AL (3.8%)
IAL sources (96.1%)
DAL Gorgas 63.5%
AL Miller 34.0%
DAL Collective Sources 1.5%
AL Gaston 0.7%
DAL Mobile Bay area 0.3%
I Other tagged sources within
AL 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-42
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-16. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Florida.
a.
CD
3
I
I
t.KVEI, 1 THC (g/krn2)
Time: 0 Jnll l.'aOUl-U Dei: 31. H001
- MINIMUM - 15.3 g/kmE (33,56)
Km
8*0. 960. '.000. 1400. 1380. 1440. 1580. 1600.
-1098.
10 -
145Z.
- i..a e.Q ia.0ie.0?o.0s^0?g0.^.0%.0 >
° " "o^-o^-n^o^-o^.o^.o^.o-
Annual total wet I dry deposition of THG 2001
within Florida
7-43
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-16b. Florida. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (450.6 g/km2).
Florida
Emissions vs. Background Contributions (g/km2) (1,2,3,4)
Wet vs. Dry Contributions (g/km2) (1,2)
500 -,
400-
300 -
200-
100 -
0-
422.2
Emiss
28.6 25.6 31.0 28.4
CTM GRHM G-C Avg
REMSAD
21.9 22.7 26.5 23.7
i i i
CTM GRHM G-C Avg
CMAQ
400 n
300-
250-
200-
150-
100-
50-
0-
71.7
_
Emis
Wet DDry
350.5
_
sion:
23.6 4 8
95.3
13.710.0 |
355.3
5 Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Florida Sources without Background (5,6)
DAvg Background
(REMSAD) 6.3%
Florida 93.3%
D Neighboring states 0.0%
Other U.S. 0.1%
DCanada & Mexico 0.0%
Reemission 0.3%
2nd pie
Sources outside FL (0.4%)
IFL sources (99.6%)
DFL St. Josephs Hospital
90.6%
FL F. J. Gannon 5.2%
DFL Collective Sources 4.1%
FL Crystal River 0.0%
DFL South-FL urban area
0.0%
Other tagged sources within
FL 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-44
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-17. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Georgia.
LEVEL 1 TUG (g/km3)
Time: 0 Jan 1. 200] 0 Ccc 31. 2001
t MAXIMUM - 89.0 (Am2 (44.31)
- MINIMUM - 1B.O g/kmZ (20.&S)
1.0 3.0 iS.o'f.oaO.O&<'0Sa-Oaa-03a'0 > dry deposition of THG KOD1
within Georgia
7-45
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-17b. Georgia. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (48.5 g/km2).
Georgia
Emissions vs. Background Contributions (g/km2) (1,2,3,4)
Wet vs. Dry Contributions (g/km2) (1,2)
35 -1
30 -
25-
20 -
15-
10 -
5-
31.3
n
Emis
17.0 152 19'5 17.2 18-8 17.2 ^ 19-1
I 1
| |
s CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
42.7
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Georgia Sources without Background (5,6)
Avg Background
(REMSAD) 35.5%
Georgia 58.6%
DNeighboring states 2.8%
DOtherU.S. 1.3%
DCanada & Mexico 0.1%
DReemission 1.7%
2nd pie
I Sources outside GA (9.2%)
IGA sources (90.8%)
DGABowen95.6%
GA Collective Sources
2.7%
DGAWansley 1.2%
GA Scherer 0.5%
DGAOLIN CORP. 0.0%
I Other tagged sources within
GA 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-46
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-18. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Kentucky.
I
LEVEL 1 THC (g/km3)
Time: 0 Jan I, 2001-0 Deo 31, 2001
+ MAXIMUM - SBB.l 8,/kmS (15.13)
- MINIMUM 13.0 g/kmV (50,3)
si a.
1HO.
I I I I I I I I 1 i LLJ I I I I I I I I I I I I J_
-300.
-430.
Annual total wet I dry deposition of THG 2001
within Kentucky
7-47
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-18b. Kentucky. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (288.1 g/km2).
Kentucky
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
300 -1
250
200
150
100
50
E
67.
mis
D
21.0 18.5 23.8 21.1 21.3 18.3 23.1 20.9
I 1 i 1 1 1 1 1 1 1 i 1 1 1 ^
s CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
300 -1
250
200
150
100
50
0
Wet DDry
233.9 239.4
33.1
Emis
sion:
48.7
15.655 10.710.1 i
5 Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Kentucky Sources without Background (5,6)
Avg Background
(REMSAD) 7.3%
Kentucky 91.0%
D Neighboring states 0.7%
DOtherU.S. 0.6%
D Canada & Mexico 0.0%
Reemission 0.4%
2nd pie
I Sources outside KY (1.8%)
IKY sources (98.2%)
DKY Collective Sources
99.9%
KY Paradise Fossil Plant
0.1%
DKY Ghent 0.0%
KY H. L. Spurlock 0.0%
DKY Big Sandy 0.0%
I Other tagged sources within
KY 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-48
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-19. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Mississippi.
I-EVF.l. 1 THC (g/km2) * MAXIMUM - 76.3 t/kaig (11.6)
Time: 0 Jan 1. SOO! 0 Dec 3J. SOHltJIMUU - 0.0 g/ltmH (3.5)
Annual total wet+dry deposition of TUG S001
within Mississippi
7-49
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-19b. Mississippi. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (47.8 g/km2).
Mississippi
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
30 -1
25 -
20 -
15-
10-
5-
'
>5.5
22.2
25.0
.1
22.3
22.3 22.4
26.2
23.6
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
24.5
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Mississippi Sources without Background (5,6)
Avg Background
(REMSAD) 46.6%
Mississippi 47.0%
D Neighboring states 1.8%
DOtherU.S. 2.0%
DCanada & Mexico 0.1%
Reemission 2.5%
2nd pie
I Sources outside MS
(12.0%)
I MS sources (88.0%)
DMS Collective Sources
99.6%
MS Jack Watson 0.2%
D MS Chambers of MS I nc-
ClearviewO.1%
MS Pascagoula ERF 0.1%
QMS Victor J.Daniel 0.1%
I Other tagged sources within
MS 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-50
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-20. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for North Carolina.
a.
I
LEVEL 1 THG (g/km2)
Time: 0 Jan 1, 2001-0 Dec 31, 2001
+ MAXIMUM = 68.5 g/km2 (34,15)
- MINIMUM = 13,0 g/km2 (10,15)
1692. 1B12. 1932.
10B2. 1212. 1332. 1452. 1572.
4.0 s.O ^.0Js. 0S°. 0S4.0 S3.03S.
*'° '" « o ^.o ^.o so. o s4ro sa^o
o
Annual total wet+dry deposition of THG 2001
within North Carolina
70
204.
-324.
-444.
-564.
7-51
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-20b. North_Carolina. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (68.5 g/km2).
North_Carolina
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
50 -1
40 -
30-
20 -
10-
44.0
n
Emis
24.3 216 ^ 245 22.7 ^ ^ 23.3
s CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
36.7
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from North_Carolina Sources without Background (5,6)
Avg Background
(REMSAD) 35.8%
North_Carolina 57.3%
DNeighboring states 2.8%
DOtherU.S. 2.2%
DCanada & Mexico 0.1%
Reemission 1.9%
2nd pie
I Sources outside NC
(10.8%)
INC sources (89.2%)
NC BMW NC 91.1%
NC Collective Sources
6.6%
DNC Marshall 1.3%
NC Belews Creek 0.6%
DNCRoxboroO.4%
I Other tagged sources within
NCO.1%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-52
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-21. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for South Carolina.
a.
LEVEL 1 THG (g/km3)
riiun: 0 Jnii 1, 2001-(] DHI; HI, KDO1
I MAXIMUM - HH.U g./krnH (37,14)
- MINIMUM - 1E.O g/kmE (3,35)
Annual total wet+dry deposition of THG 2001
within South Carolina
7-53
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-21 b. South_Carolina. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (89.0 g/km2).
South_Carolina
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
80 -1
70 -
60 -
50-
40 -
30 -
20 -
10 -
67 1
r
Emis
22.0 192 244 21.9 19.8 187 22.6 20.4
n. i 1 1
1 1 n 1 1 n
s CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
58.4
Emissions Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from South_Carolina Sources without Background (5,6)
Avg Background
(REMSAD) 24.6%
South_Carolina 71.3%
D Neighboring states 1.5%
DOtherU.S. 1.2%
D Canada & Mexico 0.0%
DReemission 1.4%
2nd pie
I Sources outside SC (5.5%)
ISC sources (94.5%)
DSC Safety Disposal
Systems 97.7%
SC Collective Sources
1.7%
DSC Foster Wheeler
Charleston RRF 0.3%
SC Wateree 0.2%
DSC Winyah Generating
Station 0.1%
Other tagged sources within
SC 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-54
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-22. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Tennessee.
a.
an
a
o
CJ
LEVEL 1 THG (g/km3)
Time: 0 Jan 1, 2001-0 Dec 31, 3001
552. 672.
+ MAXIMUM = 288.1 g/kmS (20,37)
- MINIMUM = 13,0 g/km2 (55,17)
10 -
1272. 1392.
228.
348.
468.
70
-588.
Bj
8.0 ^.o n0soros^o^a 3sro 3ero
En
3
Annual total wet+dry deposition of THG 2001
within Tennessee
7-55
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-22b. Tennessee. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (65.1 g/km2).
Tennessee
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
40 n
35 -
30 -
25 -
20-
15 -
10-
5 -
36.8
28.1
25.0
I I
31.8
28.3 27.8 _ .
24.3 23.1 25'1
__
I I
I
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
44.2
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Tennessee Sources without Background (5,6)
Avg Background
(REMSAD) 43.5%
Tennessee 47.1%
D Neighboring states 4.5%
DOtherU.S. 2.6%
DCanada & Mexico 0.1%
Reemission 2.3%
2nd pie
I Sources outside TN
(16.8%)
I TN sources (83.2%)
DTN Collective Sources
67.2%
DTN Gallatin Fossil Plant
31.5%
DTN Johnsonville Fossil
Plant 1.0%
ITN Kingston Fossil Plant
0.2%
JTNOlin Corp. 0.0%
I Other tagged sources within
TN 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-56
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-23. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Illinois.
a.
LEVEL 1 THG (g/km2) * MiXlMUM - BBB.l g/km3 (33.8)
Time: 0 Jan 1. 2001 0 Dec 31. 30BINIMUM - 14.1 gArna (7.6S)
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-23b. Illinois. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (53.5 g/km2).
Illinois
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
40 -,
35-
30-
25 -
20 -
15-
10-
5 -
E
35.6
i 1
mis
17-5 16.0 ^f 17.9 19.1 16.9 ^ 18.8
s CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
36.4
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Illinois Sources without Background (5,6)
DAvg Background
(REMSAD) 33.5%
Illinois 57.8%
D Neighboring states 5.9%
Other U.S. 1.3%
DCanada & Mexico 0.1%
OReemission 1.5%
2nd pie
Sources outside IL (13.2%)
IIL sources (86.8%)
OIL Collective Sources
98.9%
IL Util outside Chicago
0.6%
DILPowertonO.2%
IL Non-util in Chicago 0.1%
DIL Util in Chicago 0.1%
I Other tagged sources within
ILO.1%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-58
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-24. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Indiana.
a.
LEVEL 1 THG (g/km3) i MAXIMUM - 94.9 g/km2 (6,1)
Time: 0 ,'un 1. 3001 0 Dec 31. 3081NIMUM - I8.O g/kmZ (En,48)
* u
Annual lolal welidry deposition of THG 2001
within Indiana
7-59
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-24b. Indiana. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (67.5 g/km2).
Indiana
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
50 -1
40 -
30-
20 -
10-
44.8
n
Emis
22.3 198 JJjJ 227 21.5 20, JJ1 21.9
n| 1
| |
s CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
56.8
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Indiana Sources without Background (5,6)
Avg Background
(REMSAD) 33.6%
Indiana 54.0%
D Neighboring states 7.1%
DOtherU.S. 3.3%
DCanada & Mexico 0.1%
Reemission 1.8%
2nd pie
I Sources outside IN (18.7%)
I IN sources (81.4%)
DINRockport93.1%
IN Utilities outside Gary, IN
4.8%
DIN Collective Sources 1.0%
UN Gibson Generating
Station 0.9%
] IN Cliffy Creek 0.1%
I Other tagged sources within
IN 0.2%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-60
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-25. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Michigan.
l.KVKI. 1 THG (g/kmB)
Time: 0 Jnn 1. K001-0 Dec 31. 2001
444. 564
t MAXIMUM * Bl.l g/km2 (64.22)
= Jj.O
Annual Lotal well-dry deposition of THG 3001
within Michigan
7-61
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-25b. Michigan. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (56.7 g/km2).
Michigan
Emissions vs. Background Contributions (g/km2) (1,2,3,4)
50 -1
40 -
30-
20 -
10-
i
40.6
16.2 143 17.9 16, p? 18.0 ^ |j
n n n u n
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
Wet vs. Dry Contributions (g/km2) (1,2)
Emissions Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Michigan Sources without Background (5,6)
Avg Background
(REMSAD) 28.5%
Michigan 61.9%
D Neighboring states 3.2%
DOtherU.S. 3.1%
D Canada & Mexico 2.0%
Reemission 1.3%
2nd pie
I Sources outside Ml (13.4%)
I Ml sources (86.6%)
DMI Central Wayne Co
Sanitation 68.1%
Ml Sources in Detroit Metro
26.0%
DMI Monroe Power Plant
4.4%
Ml Collective Sources 1.0%
DMI J. H. Campbell 0.3%
I Other tagged sources within
Ml 0.2%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-62
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-26. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Minnesota.
LEVEL 1 THG (g/km2)
Time; 0 Jan 1. 3001-0 Dec 31. 8001
+ MAXIMUM - ZSS g, km2 (55.17)
- MINIMUM - 0-0 g/kinS (23.14)
- 684. E
Annual total wet-dry dcpositior. of THG -- HOOl
within Minnesota
7-63
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-26b. Minnesota. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (27.3 g/km2).
Minnesota
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
20 -i
15 -
10-
5-
16.3
13.9 134 ITS
11.5 11.9 11 n i 111 1Z8
9.7
__.
^^
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
17.7
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Minnesota Sources without Background (5,6)
Avg Background
(REMSAD) 40.4%
D Minnesota 56.9%
D Neighboring states 0.4%
DOtherU.S. 0.6%
D Canada & Mexico 0.2%
DReemission 1.5%
2nd pie
Sources outside MN (4.5%)
IMN sources (95.5%)
DMN Taconite Facilities
97.9%
DMN Clay Boswell 1.0%
DMN Minn./St. Paul area
0.4%
MN Collective Sources
0.3%
DMN "wetland" sources Lake
Superior 0.3%
Other tagged sources within
MN 0.2%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-64
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-27. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Ohio.
a.
ISVEl 1 THG (@/km2)
Time: 0 Jan 1. 2001-0 Eec 31. 2001
MAXIMUM - 98.2
MINIMUM - 12.7 g
(-i:,27)
(42.T)
p
1DH. 5
Annual total wet ( dry deposition of THG 3001
wilhin Ohio
7-65
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-27b. Ohio. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (55.4 g/km2).
Ohio
Emissions vs. Background Contributions (g/km2) (1,2,3,4)
50 -,
40-
30 -
20-
10 -
E
40.0
mis
20.4 20.7 -in 1
15.6 135 17.1 15.4 ^ 16.4 ^ la' '
ri n n n n
1 1 1 1 1 1 1 1 1 1 1 1 1 1
s CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
Wet vs. Dry Contributions (g/km2) (1,2)
30 n
25-
20-
15-
10-
5-
0-
17.2
22.8
Wet DDry
11.4
Dfe
28.6268
9.7 9.4
m
Emissions Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD
)
Contributions to Total Deposition (2,4)
Contributions from Ohio Sources without Background (5,6)
DAvg Background
(REMSAD) 27.8%
Ohio 60.5%
D Neighboring states 7.1%
Other U.S. 2.4%
D Canada & Mexico 0.9%
OReemission 1.4%
2nd pie
Sources outside OH
(16.2%)
I OH sources (83.8%)
DOH Collective Sources
72.4%
OHEastlake19.8%
DOH Collective utilities 5.2%
OHConesville1.0%
DOH Cardinal 0.6%
I Other tagged sources within
OH 1.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-66
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-28. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Wisconsin.
LEVEL 1 TIIC (g/krnS)
Time: 0 Jan 1, S001-0 Dec 31, 2001
» MAXIMUM - «.2 g/km8 (46.1)
- MINIMUM - 0.0 g/kmS (2.81)
I
5
Annual Lol.nl welt dry deposition of THG
within Wisconsin
2001
7-67
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-28b. Wisconsin. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (29.6 g/km2).
Wisconsin
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
20 1
15 -
10-
5-
17.1
n
13.1
10.7
13.8
12.5 14°
11.0
13.9
13.0
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
18.5
Emissions Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Wisconsin Sources without Background (5,6)
Avg Background
(REMSAD) 42.3%
Wisconsin 51.7%
DNeighboring states 2.1%
DOtherU.S. 2.1%
D Canada & Mexico 0.2%
Reemission 1.6%
2nd pie
I Sources outside Wl (10.3%)
IWI sources (89.7%)
DWI Vulcan Chem Div.
89.9%
Wl Collective Sources 9.3%
DWI Columbia 0.4%
Wl Pleasant Prairie 0.3%
DWI South Oak Creek 0.1%
I Other tagged sources within
Wl 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-68
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-29. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Arkansas.
I
LEVEL 1 THC (g/kmZ)
Time: 0 Jan 1, 2001-0 Dec 31, 20O1
En
56. 276. 386
MAXIMUM = 84.0 6/kma (4,8)
MIMMliM I7H B k':-:; CH.II)
Annual total wet+dry deposition of THC -- 3001
within Arkansas
7-69
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-29b. Arkansas. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (50.8 g/km2).
Arkansas
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
35 -1
30 -
25-
20 -
15-
10 -
5-
31.1
f
19.2
17.5
1
22'4 19.7 \20A 20.0 21-7
1
1
1
1
1
1
1
1
1
1
1
Emiss CTM GRHM G-C Avg j CTM GRHM G-C Avg
REMSAD ! CMAQ
1
27.9
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Arkansas Sources without Background (5,6)
Avg Background
(REMSAD) 38.8%
DArkansas 53.6%
D Neighboring states 5.2%
DOtherU.S. 0.6%
D Canada & Mexico 0.2%
DReemission 1.5%
2nd pie
I Sources outside AR
(12.4%)
IAR sources (87.6%)
DAR Ash Grove Cement Co
96.4%
DAR Collective Sources
3.4%
DAR White Bluff 0.1%
AR Carle Bailey Gen Stn
0.1%
DAR Independence 0.0%
I Other tagged sources within
AR 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-70
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-30. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Louisiana.
LEVEL 1 TUG (g/ktn2)
Time: 0 Jan I, aoui-0 Dec 31,
MAXIMUM = 1H.3 g/kmZ (8,13)
MINIMUM = 0.0 g/lcm3 (23,18)
HIS.
336.
Annual Lnlal weLtdry deposiLion of THG
wilhi:i Louisiana
7-71
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-30b. Louisiana. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (76.9 g/km2).
Louisiana
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
50 n
40 -
30-
20 -
10-
47.5
f
29'° 25.4
33.8
29.4 97o 31-3 278
__ 24.8 ^m
n
| |
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
45.3
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Louisiana Sources without Background (5,6)
Avg Background
(REMSAD) 38.2%
Louisiana 58.4%
D Neighboring states 0.5%
DOtherU.S. 1.0%
DCanada & Mexico 0.1%
DReemission 1.7%
2nd pie
I Sources outside LA (5.5%)
I LA sources (94.5%)
DLA Pioneer Americas Inc.
56.0%
LA Collective Sources
43.5%
DLABigCajun20.4%
LA PPG Industries-lnc.
0.0%
DLAR.S. Nelson 0.0%
I Other tagged sources within
LA 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-72
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-31. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for New Mexico.
a.
LEVEL 1 THG (g/kmSJ
Time: 0 Jan 1, 3001-0 £>ec 31. 3001
4 UAXIUUU - 23.7 s/fcma (a 1.0)
- MtNIMUM - 3.8 |/lcm2 (30,60)
Annual total wet-l-dry dopciaition of TEiCj HOC11
within New Mexico
7-73
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-31 b. New_Mexico. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (15.8 g/km2).
New_Mexico
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
30 -1
25 -
20 -
15-
10-
5-
28.3
10.6 -p.
7.0 7.9 7.7 . . 8.8
n n n 1 1 n
22.0
20.6
n
23.7
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
10.5
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from New_Mexico Sources without Background (5,6)
JAvg Background
(REMSAD) 55.8%
I New Mexico 42.4%
D Neighboring states 0.2%
DOtherU.S. 0.0%
D Canada & Mexico 0.3%
Reemission 1.3%
2nd pie
I Sources outside NM (4.1%)
INM sources (95.9%)
DNM Los Alamos Natl Lab
99.2%
NM San Juan 0.3%
DNM Collective Sources
0.3%
NM Four Corners 0.2%
DNM Escalante 0.0%
I Other tagged sources within
NM 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-74
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-32. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Oklahoma.
a.
cc
c
G
LEVEL 1 THG (g/km2)
Time: 0 Jan 1, 2001-0 Dec 31, 2001
-588. -468.
+ MAXIMUM = B4.0 g/km2 (66,2)
- MINIMUM = 10.2 g/kma (1,23)
132.
253.
276.
396.
516.
70
636.
-756.
|
Annual total wet+dry deposition of THG 2001
within Oklahoma
7-75
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-32b. Oklahoma. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (51.8 g/km2).
Oklahoma
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
35 -1
30 -
25-
20 -
15-
10 -
5-
31.5
n
19.3 18.3
23.3
246 91 7
20.3 20.6 20.0 I 1 21'7
^_ ^^m I I I I
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
30.6
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Oklahoma Sources without Background (5,6)
Avg Background
(REMSAD) 39.2%
D Oklahoma 55.6%
DNeighboring states 2.8%
DOtherU.S. 0.5%
D Canada & Mexico 0.2%
DReemission 1.7%
2nd pie
I Sources outside OK (8.6%)
I OK sources (91.4%)
DOKHolnam-lnc. 99.5%
OK Collective Sources
0.3%
D OK Sooner 0.1%
IOKMuskogeeO.1%
DOKAES Shady Point-lnc.
0.0%
Other tagged sources within
OK 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-76
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-33. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Texas.
a.
LEVEL 1 THG (g/kmS)
Time: 0 Jnn 1, S001-0 Dec 31, 2001
Km
972. -732 -.192. -252.
- MAXIMUM 111.3 g/kmS (10T..15)
- MINIMUM =
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-33b. Texas. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (111.3 g/km2).
Texas
Emissions vs. Background Contributions (g/km2) (1,2,3,4)
Wet vs. Dry Contributions (g/km2) (1,2)
100n 86.3
80-
60 -
40-
20 -
24.8 221 28'1 25.0 24.8 28.0 32-1 28.3
n n n n n n
n
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
68.2
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Texas Sources without Background (5,6)
DAvg Background
(REMSAD) 22.5%
Texas 75.8%
D Neighboring states 0.4%
Other U.S. 0.5%
DCanada & Mexico 0.1%
Reemission 0.8%
2nd pie
Sources outside TX (2.2%)
ITX sources (97.8%)
DTX Chemical Waste Mgmt
94.1%
TX Collective Sources 5.8%
DTXMonticelloO.1%
TX Big Brown 0.0%
DTX ALCOA Al&Chem
0.0%
Other tagged sources within
TX 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-78
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-34. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Iowa.
a.
LEVEL 1 THG (g/km2)
Time: 0 JHII 1, SJOO1 0 I3fic Ml, 2O01
-t- MAXIMUM - Sia.h g/ltm2 (43,3)
MINIMUM - 1D.4 g/ktnS (1,38)
578.
- 360
240.
Annual total wet+dry deposJLion ol TUG
within Iowa
7-79
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-34b. Iowa. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (25.1 g/km2).
Iowa
Emissions vs. Background Contributions (g/km2) (1,2,3,4)
25 -,
20-
15 -
10-
5 -
7.4
15.8
^
19.8
17.7 17.8 162
19.6
17.9
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
Wet vs. Dry Contributions (g/km2) (1,2)
16.8
Emissions Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Iowa Sources without Background (5,6)
DAvg Background
(REMSAD) 70.4%
Iowa 22.8%
D Neighboring states 1.4%
Other U.S. 2.0%
DCanada & Mexico 0.2%
Reemission 3.2%
2nd pie
Sources outside IA (22.8%)
IIA sources (77.2%)
DIA George Neal North
57.7%
DIA George Neal South
39.6%
DIA Council Bluffs 1.7%
IA Collective Sources 0.7%
DIA Collective utilities 0.4%
I Other tagged sources within
IA 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-80
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-35. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Kansas.
a.
LEVEL 1 THG (g/km3)
Time: 0 Jan 1, 2001-0 Dec 31, 2001
+ MAXIMUM = 44.B g/km2 (60,7)
- MINIMUM = 9.3 g/km2 (2,20)
216.
Ed B4-
SO
876.
-396.
Annual total wet+dry deposition of THG -- 2001
within Kansas
7-81
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-35b. Kansas. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (34.5 g/km2).
Kansas
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
30 -, 27.1
25 -
20 -
15-
10-
5-
12.5
21.6
19.4
n
22.0
22.8
23.1
n
24.3
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
19.2
5.3
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Kansas Sources without Background (5,6)
Avg Background
(REMSAD) 63.8%
Kansas 27.3%
DNeighboring states 2.1%
Other U.S. 3.5%
D Canada & Mexico 0.2%
Reemission 3.1%
2nd pie
I Sources outside KS
(24.6%)
IKS sources (75.4%)
DKS Collective Sources
96.5%
KS Ash Grove Cement Co.
3.0%
DKSLaCygne(KCP&L)
0.3%
KS Jeffrey Energy Center
(Westar)0.1%
DKS Lawrence (Westar)
0.1%
Other tagged sources within
KS 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-82
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-36. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Missouri.
LEVEL 1 TUG (g/kmB)
Time: 0 Jan 1, 3001-0 Dec 31. 3001
£
2
n
1
I
MAXIMUM - 86.4 g/kraa (60.18)
- MINIMUM - 17.fi g/km2 (37,48)
76B.
Annual total wet I dry deposition of TKG
within Missouri
2001
7-83
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-36b. Missouri. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (29.3 g/km2).
Missouri
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
25 1
20 -
15 -
10 -
5-
18.5 19-5
16'4 146 16'5
12.8
r
i
^^
^^
17.2
21.1
19.3
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
23.0
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Missouri Sources without Background (5,6)
Avg Background
(REMSAD) 56.3%
Missouri 34.3%
D Neighboring states 3.8%
DOtherU.S. 2.6%
D Canada & Mexico 0.2%
Reemission 2.9%
2nd pie
I Sources outside MO
(21.5%)
I MO sources (78.5%)
DMOLabadie90.9%
MO Collective Sources
5.2%
DMODoeRunBuick1.5%
MO Sioux 1.2%
D MO Rush Island 0.9%
I Other tagged sources within
MO 0.4%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-84
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-37. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Nebraska.
a.
CD
O
a
CM
Tt
ra
LEVEL 1 THG (g/km2)
Time: 0 Jan 1, 2001-0 Dec 31, 2001
-648. -528.
40P
MAXIMUM = 32.3 g/km2 (65.13)
MINIMUM = 7.3 g/km2 (11,40)
192.
- 300.
130.
70
-60.
B.O
Annual total wet+dry deposition of THG -- 2001
within Nebraska
7-85
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-37b. Nebraska. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (25.6 g/km2).
Nebraska
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
25 -,
20-
15 -
10-
5 -
20.5
n
5.1 1
18.2
22.8
20.5
20.0
18.5
22.2
20.2
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
18.4
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Nebraska Sources without Background (5,6)
DAvg Background
(REMSAD) 80.1%
Nebraska 9.2%
D Neighboring states 4.2%
Other U.S. 2.5%
DCanada & Mexico 0.2%
Reemission 3.8%
2nd pie
Sources outside NE
(53.8%)
I NE sources (46.2%)
DNE North Omaha 86.6%
NE Collective Sources
6.6%
DNE Nebraska City 6.3%
I NE Sheldon 0.4%
DNE Gerald Gentlemen
Station 0.1%
Other tagged sources within
NE 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-86
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-38. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Colorado.
LEVEL 1 THG (g/klil3)
Time: 0 Jan 1, S001 0 Dec 31, 2001
t MAXIMUM - 84.1 g/kmZ (11,37)
MINIMUM - 3.B g/kma (23,13)
398.
SO -
364.
0 I i i i li I i I i i i i I i i I i i i i i i i i fi I i i i i i i i i i I T~n i _
Annual total wet I dry deposition of THG - 2001
wilhiri Colorado
7-87
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-38b. Colorado. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (14.8 g/km2).
Colorado
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
20 1
15 -
10-
5-
17.1
14.8
12 1
H.U 78 ; 1
R 1 ^^ 1 1
II II
1
14.7
i .
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
10.6
Emissions Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Colorado Sources without Background (5,6)
JAvg Background
(REMSAD) 58.9%
I Colorado 38.6%
D Neighboring states 0.3%
DOtherU.S. 0.7%
DCanada & Mexico 0.1%
Reemission 1.3%
2nd pie
I Sources outside CO (6.0%)
I CO sources (94.0%)
DCO Pueblo Rocky Mountain
Steel 77.3%
CO Comanche 21.6%
DCO Collective Sources
0.8%
CO CEMEX-lnc.-Lyons
Cement 0.2%
DCO Pawnee 0.1%
I Other tagged sources within
CO 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-88
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-39. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Montana.
KVKl, 1 THC (fi/km2)
Time: 0 Jan 1, S001-0 Dec 31, 8001
i MAXIMUM = 19.8 R/km2 (33,7)
- MINIMUM = 3.1 g/km2 (11, S9)
Km
1476. 1356. 1238. 1116. 936. 876. 756. 636. 516.
J M I I I I I I I-M M
30
pniiiiiiMliiiiiiiiiliiiiiiiiiriiitiiiiiliiiiiiiiiliiiiiiiiiliiiiiMiiliin
076.
T5G.
0 10
30 40 50 60 70
°°
ArinuaJ LoLal weL+dry deposiLion of TilG BOOl
within Montana
7-89
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-39b. Montana. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (19.5 g/km2).
Montana
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
25 1
20 -
15-
10 -
5-
10.1 97
|
20.2 21-2 19.4
10.7 qa
7.9 | 1
II
I I I I I I
i
1fi 7
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
13.7
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Montana Sources without Background (5,6)
Avg Background
(REMSAD) 48.4%
Montana 49.7%
D Neighboring states 0.3%
DOtherU.S. 1.0%
DCanada & Mexico 0.1%
Reemission 0.6%
2nd pie
I Sources outside MT (3.8%)
IMT sources (96.2%)
DMT Livingston/Park County
MWC 99.9%
DMT Collective Sources
0.1%
DMT Stone Container Corp.
0.0%
MT Colstrip 0.0%
DMTColstrip Energy 0.0%
I Other tagged sources within
MT 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-90
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-40. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for North Dakota.
a.
o
8
LEVEL 1 THG (g/km2)
Time: 0 Jan 1, 2001-0 Dec 31, 3001
-600.
40-
+ MAXIMUM = 19.5 g/km3 (58,21)
- MINIMUM = 6.4 g/km3 (7,30)
Annual total wet+dry deposition of THG 2001
within North Dakota
1080.
- 960.
- 840,
- 730,
600,
7-91
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-40b. North_Dakota. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (14.4 g/km2).
North_Dakota
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
16-,
14 -
12 -
10-
6 -
4 -
2 -
11.2 11-6 1Q6
3.8
__
| |
y.'l
^^
13.8
13.2
10.7 H
^^^
12.6
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
8.6
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from North_Dakota Sources without Background (5,6)
Avg Background
(REMSAD) 73.7%
North_Dakota21.5%
D Neighboring states 0.2%
Other U.S. 0.9%
D Canada & Mexico 0.6%
Reemission 3.1%
2nd pie
I Sources outside ND
(18.4%)
I ND sources (81.6%)
D ND Milton R. Young 76.9%
ND Collective Sources
10.0%
DND Coyote 6.2%
ND Antelope Valley Station
3.5%
DND Coal Creek 3.3%
I Other tagged sources within
ND 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-92
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-41. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for South Dakota.
a.
£
o
LEVEL 1 TUG (g/km2)
Timo: [) Jan 1. 2O01-O l)r-o Ii1 ,
+ MAXIMUM - Kh.7 g/ltma (KH,:-IH)
- MINIMUM - 7.3 g/kmS (10,17)
sas. me. anfi svs. me. ss. S4-.
_i M M i i in | i i i i i i i i i | i T i i i i i i i | i i i i i n i T i T r i i i i i i i | i i i i i i i i i f r t
10
40
57G.
456. g
K
336.
v /o «.o ls-o 'e.o*°.a^-o**>.o a«o a».
*'° " 9. ^. 'e so &t~ 3e~ 3s'
Annual total wcL+dry d«posil.inn of TUG 2001
within South Dakota
7-93
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-41 b. South_Dakota. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (12.5 g/km2).
South_Dakota
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
16-,
14 -
10-
6 -
4 -
2 -
11.4
i p 9.8
r
1.3 1
1 1
- 11.2
^_^
14.0
12.1
14.3
13.5
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
7.3
5.2
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from South_Dakota Sources without Background (5,6)
Avg Background
(REMSAD) 89.8%
South Dakota 4.5%
D Neighboring states 1.1%
Other U.S. 0.9%
D Canada & Mexico 0.3%
Reemission 3.4%
2nd pie
Sources outside SD
(56.3%)
ISO sources (43.7%)
DSD Collective Sources
99.9%
ISO Big Stone 0.1%
DSD Health Services 0.0%
I Other tagged sources within
SD 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-94
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-42. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Utah.
a.
LEVEL 1 TUG (g/kmS) + MAXIMUM - 255.4 g/kmS (8.58)
Time: 0 Jan 1, 2001-0 Dec 31. 2001 - MINIMUM = 5.0 g/kmS (4T.6E)
Annual lolal wet. i dry deposition of 'I'Ma 20!) 1
within Utah
7-95
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-42b. Utah. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (28.5 g/km2).
Utah
Emissions vs. Background Contributions (g/km2) (1,2,3,4)
,n . . 28.3
30 -
25 -
20-
15 -
10 -
5-
24.9
,^., 112 -
21.4
n
24.9
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
Wet vs. Dry Contributions (g/km2) (1,2)
18.1
Emissions Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Utah Sources without Background (5,6)
DAvg Background
(REMSAD) 45.6%
Utah 52.0%
D Neighboring states 0.7%
Other U.S. 0.7%
D Canada & Mexico 0.2%
Reemission 0.9%
2nd pie
Sources outside UT (4.5%)
IUT sources (95.5%)
OUT Collective Sources
72.6%
UT Davis/Wasatch 25.2%
OUT Intermountain Power
0.7%
UT Ash Grove 0.6%
DUTHuntingtonO.6%
I Other tagged sources within
UT 0.4%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-96
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-43. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Wyoming.
I.KVKI, 1 TUG (g/km2)
Time: 0 Jan !. 3001 0 Dec 31. 3001
+ MAXFMUM - 62.2 £ km2 (5.29)
MINIMUM - I 9 g/kmK (ll.UB)
I I I I I I I I I I I I I I I I I I I I I I I i I I I | . I I I I I I I I | i I I I I I I I I I M M
Annual LoLal wcl+dry dcposilion of TUG 3001
wilhin Wyo ruins
7-97
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-43b. Wyoming. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (17.6 g/km2).
Wyoming
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
30 -,
25 -
20 -
1b -
10-
5-
262
24.2 24.1
15'° 13.0
2.6
I 1
^^~
16'8 15.0
^^~
^~
21.9
n
~
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
12.4
Emissions Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Wyoming Sources without Background (5,6)
Avg Background
(REMSAD) 85.0%
D Wyoming 10.2%
D Neighboring states 0.8%
Other U.S. 0.9%
D Canada & Mexico 0.2%
Reemission 2.9%
2nd pie
I Sources outside WY
(32.0%)
IWY sources (68.0%)
DWY Collective Sources
96.9%
WY Dave Johnston 2.0%
DWY Laramie River Station
0.6%
WY Jim Bridger 0.3%
DWY Naughton 0.2%
I Other tagged sources within
WY 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-98
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-44. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Arizona.
LEVEL 1 TUG (g/kin2)
Time: 0 Jan 1. 2001-0 Ceo 31. 2001
Km
!T18. 1530. 1476. 1356.
MAXIMUM - 282.8 t/kmZ (10.10)
MINIMUM - 6.7 i/kmS (52.S7)
I I I I I I I I I I I I I
I i i I i i t t t i I i i i i i i i
Annual total wet dry deposition of THG 2O01
within Arizona
7-99
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-44b. Arizona. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (37.4 g/km2).
Arizona
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
35 -1
30 -
25-
20 -
15-
10 -
5-
32.5
20.0 20.3
1 1 16'8 15.2 H
i 1
17.4
27.3
22.2
27.3
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
26.2
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Arizona Sources without Background (5,6)
JAvg Background
(REMSAD) 46.5%
lArizona 50.9%
D Neighboring states 0.9%
DOtherU.S. 0.1%
D Canada & Mexico 0.5%
DReemission 1.1%
Sources outside AZ (4.7%)
AZ sources (95.3%)
2nd pie
DAZ Northstar Steel Arizona
99.8%
AZ Collective Sources 0.1 %
DAZ copper mines 0.1%
AZ Navajo 0.0%
DAZ Cholla 0.0%
I Other tagged sources within
AZ 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-100
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-45. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for California.
a.
I
t
LEVEL 1 TUG (g/kmS) + MAXIMUM - 255.4 t/kn-.Z (66.BB)
Time: 0 Jan 1, 2001-0 Deo 31. aOfllNIMUM = 3.9 S'km2 (J9.1)
Annual total wt^t-dry deposition of TUG 200)
within California
7-101
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-45b. California. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (64.9 g/km2).
California
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
60 -1
50 -
40 -
30-
20-
10 -
56.1
I
|
Emis
20.7 23'8 197
14.5 I I , .
8.6 73 105 8.8 . .
I 1 I 1 I I I 1
s CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
54.4
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from California Sources without Background (5,6)
Avg Background
(REMSAD) 13.6%
California 86.2%
D Neighboring states 0.0%
DOtherU.S. 0.0%
D Canada & Mexico 0.0%
Reemission 0.2%
2nd pie
I Sources outside CA (0.3%)
ICA sources (99.7%)
DCA Collective Sources
69.8%
CA Long Beach SERRF
29.7%
DCA Calaveras Cement Co
0.2%
CA Riverside Cement Co.
0.1%
DCA Collective cement
plants 0.0%
Other tagged sources within
CAO.1%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-102
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-46. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km"2) for Nevada.
a.
I.EVKI. 1 THG (g/km2) + MAXIMUM - 250.1 g, kn-.2 (-10.67)
Time: 0 Jan 1. 2001 0 Dec 31. 30H1NUUJM - 5.7 g/kro2 (2.3)
-8016, -1896. -1778. -1656. -IS38 -1416.
Annual total wet + dry deposition of THG 3001
\vithin N'evada
7-103
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-46b. Nevada. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (255.4 g/km2).
Nevada
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
300 -1
250 -
200 -
150 -
100 -
50 -
0
246.7
Emiss
9.1
7.1 10.0 8.7
CTM GRHM G-C Avg
REMSAD
20.1 15.3 21.4 18.9
, , ,
CTM GRHM G-C Avg
CMAQ
250 -1
200
150
100
50
0
Wet DDry
231.7 236.4
15.0
^m
Emis
sion:
4047 7.111.8 19-0
_^ ^m
s Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Nevada Sources without Background (5,6)
Avg Background
(REMSAD) 3.4%
Nevada 96.4%
D Neighboring states 0.1%
DOtherU.S. 0.0%
D Canada & Mexico 0.0%
Reemission 0.0%
2nd pie
I Sources outside NV (0.2%)
INV sources (99.8%)
DNV Jerritt Canyon Mine
99.8%
NV Barrick Goldstrike
Mines 0.0%
DNV Newmont Gold Quarry
Operations 0.0%
NV Florida Canyon 0.0%
DNV Glammis Marigold 0.0%
I Other tagged sources within
NVO.1%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-104
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-47. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Idaho.
a.
LEVEL 1 THG (g/kmS) <- MAXIMUM - 255.4 g/kn>2 (13.8)
Time: 0 Jan 1, 300]-0 Dec 31. 3OS1MMUM - 3.8 g/km2 (29.8:
i
Annual total wcl-i-dry deposition of THG 2001
within Idaho
7-105
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-47b. Idaho. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (62.2 g/km2).
Idaho
Emissions vs. Background Contributions (g/km2) (1,2,3,4)
Wet vs. Dry Contributions (g/km2) (1,2)
60 -,
50 -
40-
30 -
20 -
10-
E
54.3
mis
20.7 17n 22-9 20.2
i 1 I I i i
8.0 6.6 91 7.9 I I 1 I I I
I I I I I I I I I I I I I I I I
s CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
60 -1
50-
40-
30-
20-
10-
0-
3.5
__^m
Emis
50.8
sion:
Wet DDry
9.111-1
3.8 4.1 . '-6
54.9
5 Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Idaho Sources without Background (5,6)
DAvg Background
(REMSAD) 12.7%
Idaho 86.1%
D Neighboring states 0.9%
Other U.S. 0.1%
DCanada & Mexico 0.0%
Reemission 0.2%
2nd pie
Sources outside ID (1.3%)
I ID sources (98.7%)
D ID P4 Production LLC
99.9%
ID Collective Sources 0.0%
DID Potlatch Pulp and
Paperboard 0.0%
IDINEELINTECO.0%
DID Potlatch Wood Products
Div. 0.0%
Other tagged sources within
ID 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-106
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-48. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Oregon.
LEVEL 1 THG (g/kma)
Time: 0 Jan 1, 20O1-0 Dec 31,
MAXIMUM = B4.3 g/km2 (S8.32)
MINIMUM = 4.5 g/km2 (3B.S4)
Annual Lnlzil weL+dry dnposil.ion of THC] 2OD]
within Oregon
7-107
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-48b. Oregon. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (84.3 g/km2).
Oregon
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
80 -1
70 -
60 -
50-
40-
30 -
20-
10 -
E
75.1
mis
18.6 171 157
10.1 73 10.2 9.2 | 1 11-4 __ ,__,
rn i i rn rn I I n I I
s CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
57.3
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Oregon Sources without Background (5,6)
Avg Background
(REMSAD) 10.9%
Oregon 88.7%
D Neighboring states 0.3%
DOtherU.S. 0.0%
D Canada & Mexico 0.0%
Reemission 0.1%
1st pie
2nd pie
I Sources outside OR (0.4%)
I OR sources (99.6%)
DOR Ash Grove 99.9%
OR Collective Sources
0.1%
DOR Boardman 0.0%
OR Portland General
Electric Co. 0.0%
DOR Cascade Steel 0.0%
I Other tagged sources within
OR 0.0%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-108
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-49. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Washington.
LEVEL 1 THG (g/kmS)
Time: 0 Jan 1, £001-0 Dec 31. S001
+ MAXIMUM - 34.B g/kinE (51.10)
- MINIMUM 3.7 g/kmK (16,36)
-211Z.
-1992.
1360.
1020. £
Annual LoLal \s'cl+dry deposition ol THG 2001
within Washington
7-109
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-49b. Washington. Deposition Analysis for the Single Grid Cell (the Blue Triangle in the Accompanying Spatial Plot) Where In-State
Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (19.9 g/km2).
Washington
Emissions vs. Background Contributions (g/km2) (1,2,3,4) Wet vs. Dry Contributions (g/km2) (1,2)
20 1
15 -
10 -
5-
17.3 17.3
ii 9.1 1°-3 88 n
^_ 70 1
| |
12.1
1b.6
Emiss CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
16.7
Emissions Avg Bckgnd Avg Bckgnd Total
(REMSAD) (CMAQ) (REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Washington Sources without Background (5,6)
Avg Background
(REMSAD) 44.2%
Washington 53.9%
D Neighboring states 1.2%
DOtherU.S. 0.4%
D Canada & Mexico 0.2%
Reemission 0.2%
2nd pie
I Sources outside WA (3.4%)
IWA sources (96.6%)
DWACentralia97.1%
WA Long View Fibre Co.
1.3%
DWA Collective Sources
0.8%
WA Collective pulp and
paper 0.5%
DWA Ash Grove and Lafarge
0.2%
Other tagged sources within
WAO.1%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-110
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-50. REMSAD PPTM Results for Steubenville, Ohio for Annual Total (Wet and Dry) Mercury Deposition (a)
and Annual Wet Mercury Deposition (b).
Figure 7-50a. Deposition Analysis for the Single Grid Cell Including Steubenville, Ohio.
Simulated Annual Total Mercury Deposition for 2001 (48.1 g/km2).
Ohio
Emissions vs. Background Contributions (g/km2) (1,2,3,4)
Wet vs. Dry Contributions (g/km2) (1,2)
32.1
21.4 ... 22-7 20.8
Emiss
CTM GRHM G-C Avg
REMSAD
CTM GRHM G-C Avg
CMAQ
Emissions
Avg Bckgnd
(REMSAD)
Avg Bckgnd
(CMAQ)
35.9
Total
(REMSAD)
Contributions to Total Deposition (2,4)
Contributions from Ohio Sources without Background (5,6)
DAvg Background
(REMSAD) 33.2%
D Ohio 44.4%
D Neighboring states 17.9%
Other U.S. 2.4%
D Canada & Mexico 0.3%
Reemission 1.8%
1st pie
2nd pie
Sources outside OH
(33.5%)
I OH sources (66.5%)
D OH Cardinal 74.9%
OH Collective utilities
10.0%
DOHW. H.Sammis8.5%
OHConesville2.9%
DOH Collective Sources
1.8%
Other tagged sources
within OH 1.8%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the cell containing the location used in a specific
monitoring study. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-111
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-50b. Wet Deposition Analysis for the Single Grid Cell Including Steubenville, Ohio.
Simulated Annual Wet Mercury Deposition for 2001 (35.9 g/km2).
Ohio
Emissions vs. Background Contributions (g/km2) (1,2,3,4)
Contributions to Wet Deposition (2,4)
30 -,
25
20
15 -
10 -
5
E
24.3
mis
11-3 10.4 Iff 11-6 Q* 11'9 12'4 11-3
Dl 1
| |
s CTM GRHM G-C Avg CTM GRHM G-C Avg
REMSAD CMAQ
DAvg Background
(REMSAD) 32.2%
Ohio 49.8%
D Neighboring states 13.6%
DOtherU.S. 2.2%
D Canada & Mexico 0.1%
Reemission 2.1%
Contributions from Ohio Sources without Background (5,6)
Breakdown of contributions from sources in neighboring states(5,6)
1st pie
2nd pie
Sources outside OH
(26.6%)
I OH sources (73.4%)
D OH Cardinal 78.9%
OH Collective utilities 9.4%
DOHW. H. Sammis7.8%
OHConesville 1.5%
DOH Collective Sources
1.2%
Other tagged sources
within OH 1.2%
QWV Collective utilities
36.9%
PA Bruce Mansfield 14.8%
DWV Collective Sources
12.7%
D WV John E Amos 5.3%
DWV Philip Sporn 5.0%
PA Collective utilities 3.5%
D KY Big Sandy 3.2%
Other sources in
neighboring states 18.6%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) "Avg" uses the average of REMSAD or CMAQ simulation results for the three global model background estimates.
5) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
6) "Collective Sources" tag for each state includes all point and area sources in the state that are not individually tagged.
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the cell containing the location used in a specific
monitoring study. Results should not be extrapolated to indicate source contributions on a statewide basis.
7-112
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
Figure 7-51.Location of the Steubenville, Ohio study site (indicated by the blue triangle).
LEVEL 1 THG (g/km2)
Time: 0 Jan 1, 3001-0 Deo 31, 2001
+ MAXIMUM = 99.3 6/km2 (41,37)
- MINIMUM = 12.7 R/kmE (43,7)
Annual total wet+dry deposition of THG -- 2001
within Ohicj
7-113
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
REMSAD PPTM Results: Mercury Deposition Contribution Analysis
This page deliberately left blank.
7-114 August 2008
-------
8. Summary of Key Findings and
Recommendations for Future Study
This study has provided an improved understanding of the sources and mechanisms
contributing to mercury deposition throughout the U.S. Key sources and source categories
contributing to mercury deposition within each state have been identified, and their contribution
to total mercury deposition for an annual 2001 simulation period quantified. It is expected that
the modeling results will provide supporting information for the future assessment of control
measures and calculation of Total Maximum Daily Loads (TMDLs).
8.1. Summary of Findings
Based on available data for the 2001 simulation period, REMSAD is able to reasonably replicate
the observed concentration patterns for ozone and PM2.s, and the observed deposition patterns
for PM2.5 and mercury (wet deposition only). Wet deposition of mercury is typically overestimated
by 30 percent or more, on both an annual and seasonal basis, when compared to MDN
monitoring data. Some emerging research suggests, however, that the MDN measurement
techniques may underestimate wet deposition by about 16 percent (Miller et al., 2005).
Wet deposition accounts for much of the deposition that occurs throughout the domain. The
simulated spatial distribution of mercury deposition is consistent with the emissions and
annual transport and rainfall patterns.
Since only wet deposition monitoring data are available, the evaluation of model performance for
mercury is driven by the reliability of the meteorological input fields, and in particular precipitation.
The ability to evaluate model performance for air concentrations and dry deposition of mercury
and, thus, the overall emissions influence on mercury deposition is limited by a lack of data.
PPTM gives expected results and the simulated contributions are consistent with the emissions
data (including magnitude and speciation characteristics), source locations, source types, and
current knowledge/theories regarding the contribution from global background. The REMSAD
PPTM results are consistent with those obtained using the CMAQ model and are also
consistent with results from a recent receptor modeling study for a specific location in Ohio.
The mercury emission inventory used as the starting point for this study was that used in the
CAMR modeling and, for the most part, represented emissions levels up to several years
prior to 2001. In order to more accurately represent 2001, EPA Regions and states
recommended a number of adjustments to the emissions inventory affecting the magnitude,
stack parameters, and speciation of emissions for many sources.
Boundary conditions are an important consideration in national- and regional-scale mercury
deposition modeling. In order to address the inherent uncertainty in global contributions (i.e.,
boundary concentrations), results from three global models (CTM, GRAHM, and GEOS-
CHEM) were used to establish boundary conditions for the REMSAD simulations. At
locations where estimated deposition was dominated by local sources, contributions from
boundary conditions were relatively consistent regardless of the global model used.
CMAQ estimates of dry deposition attributable to the boundary conditions are consistently
greater than REMSAD estimates of dry deposition attributable to the boundary conditions.
For several Rocky Mountain and Southwest states, CMAQ estimates of deposition
attributable to the boundary conditions are considerably higher for both wet and dry
deposition than the REMSAD estimates. Since observed dry deposition data are not
available, it is not possible to determine which estimates better represent actual conditions.
8-1 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Summary of Key Findings and Recommendations for Future Study
Re-emission of mercury is a small but consistent contributor to mercury deposition.
Simulated wet deposition of mercury is sensitive to rainfall amount, and overestimation of
rainfall in certain areas can lead to overestimation of mercury deposition.
National-scale, annual mercury deposition and source attribution modeling at 12-km
resolution is practicable, especially if, like REMSAD, the model can be applied once for non-
mercury species and then as needed for mercury.
Based on the analysis of available meteorological and wet mercury deposition data for a ten-
year period, the 2001 annual simulation period is an average year for mercury deposition.
For the 2001 simulation period, the relative amount of wet and dry mercury deposition varies
by state, with generally a higher wet-to-dry deposition ratio for the eastern U.S. where
precipitation is greater and is distributed more evenly throughout the year.
The relative contribution of global background generally decreases from west to east.
However, in the vicinity of large emitters, particularly of divalent mercury, local sources can
dominate deposition, regardless of geographic region.
Deposition at the location of maximum deposition by sources within the same state is
frequently dominated by one or more nearby sources. This finding may be linked to
horizontal grid resolution and, in this case, the use of relatively high-resolution (12-km) grids.
There are numerous instances where deposition at the location of maximum deposition by
sources within the same state is dominated by "collective" sources within the state (defined
here as all point and area sources in the state that are not tagged individually, as part of a
source category, or as part of a region). This finding suggests that quality assurance of
emissions information for the smaller sources, as well as the larger sources, is important,
especially where multiple small sources are concentrated in a limited area.
8.2. Recommendations for Future Study
This study has generated a large amount of information and additional analysis (and mining)
of the results is needed in order to fully utilize the results in addressing state- and water-body
specific mercury deposition issues.
Additional analysis of the results for impaired water bodies is an important next step.
Extraction, synthesis, and application of the results to support specific TMDL calculations or
the identification of effective mercury deposition issues are other important areas of analysis.
Continued improvement of the mercury emissions inventory with emphasis on speciation,
motor-vehicle emissions, and emissions and stack parameter information for smaller sources
will benefit future modeling efforts.
Future modeling efforts for this simulation period should consider the use of improved or
alternative meteorological inputs, with particular emphasis on improved simulated rainfall
amounts for the western U.S.
Similar modeling for additional base years (for example, 2002 and 2005) with different
meteorological conditions and emissions would allow the assessment of year-to-year
variations in overall deposition and the contributions to deposition and would also provide a
check on the 2001 results.
8-2 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Summary of Key Findings and Recommendations for Future Study
Application of both REMSAD and CMAQ with PPTM for additional tags, including source
categories and multi-state source regions, to obtain a more complete and more detailed
understanding of the source contributions to mercury deposition, is also recommended.
The databases and tools used for this are also well suited for the analysis of the effects of
future changes in mercury emissions on mercury deposition.
8-3 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Summary of Key Findings and Recommendations for Future Study
This page deliberately left blank.
8-4 August 2008
-------
References
Ariya, P., A. Dastoor, M. Amyot, W. Schroeder, L. Barrie, K. Anlauf, F. Raofie, A. Ryzhkov, D.
Davignon, J. Lalonde, and A. Steffen. 2004. Arctic: A sink for mercury. Tellus, 56B: 397-403.
Barrick, Inc. 2007. Letter from M. Barbanell to EPA Region 9. Re: Comments on ICF Report:
Mercury-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in
Watershed Planning. 6 July 2007.
Blanchard, P., F.A. Froude, J.B. Martin, H. Dryfhout-Clark, J.T. Woods. 2002. Four years of
continuous total gaseous mercury (TGM) measurements at sites in Ontario, Canada. Atmos.
Environ., 36: 3735-3743.
Bullock, O.R., Brehme, K.A. 2002. Atmospheric mercury simulation using the CMAQ model:
formulation description and analysis of wet deposition results. Atmos. Environ., 36:
2135-2146.
Bullock, O. R., D. Atkinson, T. Braverman, K. Civerolo, A. Dastoor, D. Davignon, J. Ku. K.
Lohman, T. Myers, R. Park, C. Seigneur, N. Selin, G. Sistla, and K. Vijayaraghavan.
2008. The North American mercury model intercomparison study (NAMMIS). (Parts 1
and 2 submitted to the J. Geophys. Res.).
Byun, D. W., and J. K. S. Ching. 1999. "Science Algorithms of the EPA Models-3 Community
Multiscale Air Quality (CMAQ) Modeling System." U.S. EPA Office of Research and
Development, Washington, D.C. (EPA/600/R-99/030).
Canada. 2007a. E-mail correspondence from Environment Canada D. Davignon dated
September 18 and 19, 2007.
Canada. 2007b. E-mail correspondence from Metro Vancouver C. Allan dated September 27, 2007.
CEC. 2001. Preliminary Atmospheric Emissions Inventory of Mercury in Mexico, Report No.
3.2.1.04, Commission for Environmental Cooperation.
CT DEP. 2006. E-mail correspondence from E. Pierce dated April 18, 2006.
Dastoor, A., and Y. Larocque. 2002. Global circulation of atmospheric mercury: A modeling
study, Atmos. Environ., 38: 147-161.
Douglas, S., B. Hudischewskyj, S. Beckmann, and T. Myers. 2005. "Comparison of MM5- and
RUC-Based Meteorological Input Fields for REMSAD Mercury Modeling (Revised),"
memorandum dated 14 December, ICF International, San Rafael, California.
Douglas, S., T. Myers, and Y. Wei. 2006. "Implementation of Sulfur and Nitrogen Tagging in the
Community Multi-scale Air Quality Model." Prepared for the U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards. ICF International, San Rafael,
California (06-076).
EPA. 1997. "Mercury Study Report to Congress; Volume III: Fate and Transport of Mercury in
the Environment." EPA OAQPS/ORD, EPA-452/R-97-005,
(http://www.epa.gov/ttn/atw/112nmerc/volume3.pdf).
EPA. 2001. EPA Toxics Release Inventory for 2001 available at:
http://www.epa.gov/tri/tridata/tri01/data/index.htm.
R-1 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
References
EPA, 2004. Information and documentation on the IPM model available at
www.epa.gov/airmarkets/epa-ipm.
EPA. 2005a. "Technical Support Document for the Final Clean Air Interstate Rule: Air Quality
Modeling," EPA OAQPS (http://www.epa.gov/cair/pdfs/finaltech02.pdf).
EPA. 2005b. "Technical Support Document for the Final Clean Air Mercury Rule: Air Quality
Modeling," EPA OAQPS (http://www.epa.gov/ttn/atw/utility/agm oar-2002-0056-6130.pdf).
EPA. 2005c. "Emissions Inventory and Emissions Processing for the Clean Air Mercury Rule
(CAMR)," EPA OAQPS.
EPA. 2006. "Guidance on the Use of Models and Other Analyses for Demonstrating Attainment
of Air Quality Goals for Ozone, PM2.5, and Regional Haze," Draft 3.2. EPA Office of Air
Quality Planning and Standards, Research Triangle Park, North Carolina (September 2006).
EPA. 2007. E-mail correspondence from A. Pope dated August 22, 2007.
EPA OW. 2006a. E-mail correspondence from D. Atkinson email dated April 12, 2006.
EPA OW. 2006b. "Summary of 2001 Hg Emissions Inventory Modifications for Minnesota"
memorandum prepared by R. Dwight Atkinson dated April 8, 2005.
EPA OW. 2006c. Personal communication by phone from D. Atkinson on September 19, 2006.
EPA Region 4. 2006. E-mail correspondence from R. Gillam dated March 10, 2006.
EPA Region 7. 2006a. E-mail correspondence from M. S Jay dated March 28, 2006.
EPA Region 7. 2006b. E-mail correspondence from M. S Jay dated March 29, 2006.
EPA Region 7. 2006c. E-mail correspondence from M. S. Jay with attachment "Mercury
Comparison.xls" dated March 29, 2006.
EPA Region 7. 2006d. E-mail correspondence from M. S. Jay dated May 19, 2006.
EPA Region 7. 2006e. E-mail correspondence from M. S. Jay dated June 5, 2006.
EPA Region 8. 2007. E-mail correspondence from M. Langenfeld dated August 7, 2007.
EPA Region 9. 2006a. E-mail correspondence from S. Frey with attachment "DOC. PDF" dated
August 15, 2006.
EPA Region 9. 2006b. E-mail correspondence from S. Frey dated September 11, 2006.
EPA Region 9. 2006c. Ontario Hydro test results provided by e-mail by S. Frey on August 25, 2006.
EPA Region 9. 2006d. Phelps Dodge in AZ air dep inventory provided by e-mail by P. Tsai on
August 7, 2006.
EPA Region 9. 2006e. NEI/TRI data for R9 for OW mercury modeling study, e-mail by L.
McClain-Vanderpool on May 4, 2006.
R-2 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
References
EPA Region 9. 2006f. Hg Inventory latest: discussion of Long Beach SERRF and CA gold mine,
e-mail by S. Frey on August 21, 2006.
EPA Region 9. 2007. E-mail correspondence from S. Chilingaryan with attachment
"speciationformodelingoct31.xls" dated October 31, 2007.
Friedlander, S. K. 1977. Smoke, Dust, and Haze: Fundamentals of Aerosol Behavior. John
Wiley & Sons, New York.
Gery, M. W., G. Z. Whitten, J. P Killus, and M. C. Dodge. 1989. A Photochemical kinetics
mechanism for urban and regional computer modeling. J. Geophys. Res., 94:12,925-12,956.
Gray, H. A., M. P. Ligocki, G. E. Moore, C. A. Emery, R. C. Kessler, J. P. Cohen, C. C. Chang,
S. I. Balestrini, S. G. Douglas, R. R. Schulhof, J. P. Killus, and C. S. Burton. 1991.
"Deterministic Modeling in the Navajo Generating Station Visibility Study." Systems
Applications International, San Rafael, California (SYSAPP-91/045).
Hales, J. M., and S. L. Sutter. 1973. Solubility of sulfur dioxide in water at low concentrations.
Atmos. Environ., 7: 997-1001.
ICF. 2005. "User's Guide to the Regional Modeling System for Aerosols and Deposition
(REMSAD), Version 7." ICF International, San Rafael, CALIFORNIA. (Available at
www.remsad.com).
IDEM. 2006. E-mail correspondence from K. Watson with attachment "EPACopy of 2001 Hg top
20_w_stks_2_22_06.xls" dated February 24, 2006.
IDEM. 2006b. E-mail correspondence from J. Bates dated March 8, 2006.
IL EPA. 2006a. E-mail correspondence from A. Russo with attachment "REMSAD_review2.xls"
dated April 11, 2006.
IL EPA. 2006b. E-mail correspondence from A. Russo with attachment "REMSAD_review2.xls"
April 13,2006.
IL EPA. 2006c. E-mail correspondence from R. Kaleel dated April 4, 2006.
Iowa DNR. 2006. E-mail correspondence from M. Stein with
attachment2001_IA_top_40_hg2_emitters.xls" dated April 18, 2006.
Iowa DNR. 2007. E-mail correspondence from M. Johnson dated July 10, 2007.
KDHE. 2006. E-mail correspondence from L. Deahl with attachment "pic23754.jpg" dated
February 22, 2006.
Kessler, E. 1969. "On the Distribution and Continuity of Water Substance in Atmospheric
Circulation." Meteor. Monographs, No. 32, 84 pp.
Lin, C. A., and S.O. Pehkonen. 1999. The chemistry of atmospheric mercury: A review, Atmos.
Environ, 33: 2067-2079.
LLCHD. 2006. E-mail correspondence from G. R. Bergstrom dated March 22, 2006.
R-3 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
References
Louis, J. F. 1979. Parametric model of vertical eddy fluxes in the atmosphere. Boundary Layer
Meteor., 17:187-202.
Miller, E., A. VanArsdale, G. Keeler, A. Chalmers, L. Poissant, N. Kamman, and R. Brulotte.
2005. Estimation and mapping of wet and dry mercury deposition across northeastern North
America. Ecotoxicology, 14:53-70.
Mostardi Platt. 2000. "Speciated Mercury Testing Performed for Northern States Power
Company at the Sherburne County Generating Plant Unit 3."
http://www.epa.gOv//ttn/atw/combust/utiltox/mercury/sherburnr.pdf
MDEQ. 2006. E-mail correspondence from J. Taylor-Morgan dated March 14, 2006.
MO DNR. 2006a. E-mail correspondence from J. Wade dated March 21, 2006.
MO DNR. 2006b. E-mail correspondence from M. Freebairn dated March 30, 2006.
MPCA. 2006a. Comments from E. Swain on e-mail correspondence from D. Atkinson titled
"Suggested MN sources to tag" dated January 13, 2005.
MPCA . 2006b. E-mail correspondence from H. Jiang dated April 17, 2006.
MPCA. 2006c. E-mail correspondence from E. Swain to D. Atkinson titled "Speciation profiles
for Riverside, High Bridge, or King" dated April 13, 2005.
Myers, T. C., Y. H. Wei, G. Glass, J. Mangahas, and B. Hudischewskyj. 2002. "Application of
the REMSAD Modeling System to Estimate the Deposition of Nitrogen and Mercury." ICF
International, San Rafael, California (02-101).
Myers, T., Y. Wei, B. Hudischewskyj, and S. Douglas. 2003. "Application of the REMSAD
Modeling System to Estimate the Deposition of Mercury in Support of the Wisconsin TMDL
Pilot." ICF International, San Rafael, California (03-050).
Myers, T. 2004. "Recent revisions to REMSAD mercury chemistry routine." Memorandum to
Dwight Atkinson dated 20 January 2004.
Myers, T. and Y. Wei. 2004. "REMSAD Air Deposition Modeling in Support of TMDL
Development for Southern Louisiana." ICF International, San Rafael, California (04-038).
Myers, T., Y. Wei, B. Hudischewskyj, and S. Douglas. 2004. "Mercury Air Deposition TMDL
Development for Maryland." ICF International, San Rafael, California (04-068).
Myers, T., Y. Wei, B. Hudischewskyj, and S. Douglas. 2006. "Application of the REMSAD
Modeling System to Estimate the Deposition of Mercury in Support of the Wisconsin TMDL
Pilot (Revised)." Systems Applications International/ICF International, San Rafael, California
(06-017)].
Myers, T. and S. Douglas. 2006. "Comparison of REMSAD Model Performance for Non-Mercury
Species for 12- and 36-km Grid Resolution."Memorandum dated 13 November, ICF
International, San Rafael, California.
R-4 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
References
National Atmospheric Deposition Program (NRSP-3). 2005. NADP Program Office, Illinois State
Water Survey, 2204 Griffith Dr., Champaign, Illinois.
NDEP. 2006. "Precious Metal Mining - Mercury Air Emissions Questionnaire," Nevada Division
of Environmental Protection. March 2006.
NDEP. 2007. E-mail correspondence from F. Forsgren with attachment "NV Hg Sources.xls"
dated July 18, 2007.
NESCAUM. 2006a. E-mail correspondence from J. Graham with attachment "CT 2001 Hg Inv
Update.xls" dated April 11, 2006.
NESCAUM. 2006b. E-mail correspondence from J. Graham with attachment "2001 NH
Emissions Revision.xls" dated March 27, 2006.
NESCAUM. 2006c. E-mail corresponded from J. Graham with attachment "Copy of VT HG
emiss - JRM.ICF.xls" dated April 14, 2006.
NJDEP. 2006a. E-mail correspondence from M. Aucott with attachment
"DEP_revised_2001_NJ_Hg_sources.xls" dated March 21, 2006.
NJDEP. 2006b. E-mail correspondence from S. Agrawal dated March 22, 2006.
NJDEP. 2006c. E-mail correspondence from S. Agrawal dated April 6, 2006.
OR DEQ. 2006a. E-mail correspondence from G. Ebersole dated August 8, 2006.
OR DEQ. 2006b. E-mail correspondence from G. Ebersole dated August 9, 2006.
OR DEQ. 2006c. E-mail correspondence from G. Ebersole dated August 1, 2006.
OR DEQ. 2007a. E-mail correspondence from G. Ebersole dated April 5, 2007.
OR DEQ. 2007b. E-mail correspondence from M. Bailey dated August 28, 2007.
OR DEQ. 2007c. Personal communication by phone between EPA OW D. Atkinson and OR
DEQ M. Bailey on June 4, 2007.
Pai, P., P. Karamchandani, and C. Seigneur. 1999. Sensitivity of simulated atmospheric mercury
concentrations and deposition to model input parameters." J. Geophys. Res., 104:13,855 -
13,868.
Roy F. Weston. 2000. Emissions Test Report. Minnesota Power Company. Boswell Energy
Center. http://www.epa.goV//ttn/atw/combust/utiltox/mercurv/clayboswellr.pdf.
Ryaboshapko, A., R. Bullock, R. Ebinghaus, I. llyin, K. Lohman, J. Munthe, G. Petersen, C. Seigneur,
and I. Wangberg. 2002. Comparison of mercury chemistry models, Atmos. Env., 36:3881-3898.
SAI. 1999. "User's Guide to the Variable-Grid Urban Airshed Model (UAM-V)." Systems
Applications International, San Rafael, California.
R-5 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
References
Scire, J. S. 1991. "A Review of the UAM-V Dry Deposition Algorithm and Recommendations for
Dry Deposition Modeling in the LMOS Study Region." Sigma Research Corporation,
Westford, Massachusetts (Document A195-100).
Scott, B. C. 1978. Parameterization of sulfate removal by precipitation. J. Appl. Meteor, 17:
1375-1389.
Seigneur, C., H. Abeck, G. Chia, M. Reinhard, N. Bloom, E. Prestbo, and P. Saxena. 1998.
Mercury adsorption to elemental carbon (soot) particles and atmospheric particulate matter.
Atmos. Env., 32: 2649-2657.
Seigneur, C., G. Hidy, I. Tombach, J. Vimont, and P. Amar. 1999. "Scientific Peer-Review of the
Regulatory Modeling System for Aerosols and Deposition (REMSAD)." The KEVRIC
Company, Inc. Durham, North Carolina.
Seigneur, C., P. Karamchandani, K. Lohman, K. Vijayaraghavan, and R. Shia. 2001. Multiscale
modeling of the atmospheric fate and transport of mercury. J. Geophys. Res., 106: 27,795 -
27,809.
Selin, N., D. Jacob, R. Park, R. Yantosca, S. Strode, L. Jaegle, and D. Jaffe. 2002. Chemical
cycling and deposition of atmospheric mercury: Global constraints from observations. J.
Geophys. Res., 112.
Shia, R., C. Seigneur, P. Pai, M. Ko, and N. D. Sze. 1999. Global simulation of atmospheric
mercury concentrations and deposition fluxes. J. Geophys. Res., 99 :23,747 - 23,760.
Slinn, W.G.N., L. Hasse, B. B. Hicks, A. W. Hogan, D. Lai, P. S. Liss, K. O. Munnich, G. A.
Sehmel, and O. Vittori. 1978. Some aspects of the transfer of atmospheric trace constituents
past the air-sea interface. Atmos. Environ., 14:1013-1016.
Sofiev, M., and M. Galperin. 2000. "Long-Range Transport Modeling of the Particle-Carried and
Persistent Toxic Pollutants." Proceedings of the Millennium NATO/CCMSITM, May 15-19,
2000, Boulder, CO, AMS, pp. 148-155.
Syrakov, D. 1998. "Long-Term Calculation of Bulgarian Impact in Mercury Pollution over SE
Europe." Bulgarian contribution to 1998 Annual Report for 1998, NIMH, EMEP/MSC-E,
Sofia-Moscow.
Tokos, J. J. S., B. Hall, J. A. Calhoun, and E. M. Prestbo. 1998. Homogeneous gas-phase
reaction of HgO with H2O2, O3, CH3I, and (CH3)2S: Implications for atmospheric Hg
cycling. Atmos. Environ., 32:823-827.
Utah DEQ. 2007a. E-mail correspondence from B. LeBaron with attachment "Brock
Kennecott2.xls" dated July 27, 2007.
Utah DEQ. 2007b. E-mail correspondence from D. McMurtrie with attachment "Mercury
Inventory revised for EPA modeling083107.xls" dated August 31, 2007.
VA DEQ. 2006. E-mail correspondence from J. Schneider with attachment "51264 2000 thru
2005 Hg.pdf" dated May 31, 2006.
R-6 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
References
Van Loon, L. L, E. A. Mader, and S. L. Scott. 2001. Sulfite stabilization and reduction of the
aqueous mercuric ion: Kinetic determination of sequential formation constants." J. Phys.
Chem.A, 105:3190-3195.
WA. DE. 2006. E-mail correspondence from E. Guilfoil, Subj: cement plants, dated August 8,
2006.
WA. DE. 2007. E-mail correspondence from E. Guilfoil, Subj: changes to WA area Hg
emissions, dated September 6, 2007.
WDNR. 2006. E-mail correspondence from O. Cabrera-Rivera (forwarded by D. Atkinson) dated
February 2, 2005.
Wesley, M. L., and B. M. Lesht. 1988. Comparison of the RADM Dry Deposition Module with
Site-Specific Routines for Inferring Dry Deposition. U.S. EPA, Research Triangle Park, North
Carolina (EPA/600/54-88/027).
Yantosca, B. 2004. "GEOS-CHEMv7-01-02 User's Guide," Atmospheric Chemistry Modeling
Group, Harvard University.
R-7 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
References
This page deliberately left blank.
R-8 August 2008
-------
Appendix A: Model Performance Evaluation for
Non-Mercury Species
In this section, the ability of the REMSAD modeling system to replicate the observed
concentration and deposition characteristics of the simulation period is examined. The
assessment of model performance considers concentrations for ozone, sulfur dioxide (SO2) and
fine particulate matter (PM2.s), and deposition for selected PM species on a monthly and/or
annual basis, depending on the pollutant.
Methodology
A variety of statistical measures were used to quantify model performance for ozone, SO2 and
PM2.s, as follows:
Normalized bias (expressed as percent) = 100 -1//VE (Si- O,)/ O,
Normalized gross error (expressed as percent) = 100 -1//V Z |S/- O/|/ O,
Fractional bias (expressed as percent) = 200 -1//VE (Si- O,)/ (S, + O,)
Fractional error (expressed as percent) = 200 -1//V Z |S/- O/|/ (S, + O,)
Mean residual = 1//VZ (S,- O,)
Mean unsigned error = 1//V Z |S/- O/|
Coefficient of determination (R2) =
(Z S, O, - ZS, ZO/N)2 /[ (ZO,2 - (ZO,)2/N) (ZS,2 - (ZS,)2/N) ]
Where S is the simulated concentration, O is the observed concentration, and N is the number
simulation-observation pairs used in the calculation.
Model performance for ozone was evaluated against observations available from the EPA Air Quality
System (AQS) network. The number of sites ranges from 500 to 1100, depending on the time of year.
The sites are primarily located in urban areas. The daily maximum simulated ozone concentration for
each monitor for each day was compared to the corresponding maximum observed concentration.
Monthly values for each measure were calculated using the daily comparisons.
Model performance for SO2 was evaluated against observations available from the AQS network,
which includes more than 4,000 sites (primarily located in urban areas). The monthly average
values were compared.
Simulated concentrations of PM2.5were compared with observed values from the Speciated
Trends Network (AQS-STN), which includes more than 200 sites, and from Interagency
Monitoring of Protected Visual Environments (IMPROVE) network sites, which sample
approximately 100 Class I national parks and wilderness areas throughout the U.S. Simulated
concentrations of sulfate and nitrate were evaluated against observations from the IMPROVE
and Clean Air Status and Trends Network (CASTNet) monitoring sites. CASTNet includes more
than 70 sites that are a mix of mostly rural and suburban sites. Simulated concentrations of
ammonium were evaluated against observations from the CASTNet sites. Monthly and annual
average simulated concentrations for each monitor were compared to the corresponding
observed concentrations.
A-1 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix A: Model Performance Evaluation for Non-Mercury Species
Simulated wet deposition amounts of sulfate, nitrate, and ammonia were compared with
observed values from the National Acid Deposition Program (NADP), which includes more than
200, typically rural, sites. Monthly and annual deposition totals for each monitor were compared
to the corresponding observed totals.
Model performance measures for the REMSAD 12-km modeling domain are presented in the
tables below. The 12-km results were used for all subsequent analyses. A comparison of the 12-
and 36-km results is provided by Myers and Douglas (2006). This comparison provides some
insight into the benefits of using the 12 -km grid.
Summary of Monthly Model Performance for Gaseous and Particulate Species
Concentrations
Statistical measures of model performance for ozone are presented for each month in Table A-1.
Table A-1. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Daily Maximum Ozone Concentration (ugm3) at AQS Sites.
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Normalized
bias
(%)
25.2
19.1
9.5
11.1
7.1
13.0
14.4
16.9
13.0
2.2
21.4
25.2
Normalized
gross error
(%)
45.2
35.4
27.2
21.6
21.3
27.5
30.3
31.7
28.3
20.9
37.3
42.7
Fractional
bias
(%)
6.1
7.2
1.5
6.8
3.1
6.9
6.9
9.0
6.2
-2.3
7.9
9.1
Fractional
error
(%)
29.9
26.3
22.1
18.7
19.3
23.7
25.6
26.5
24.0
19.2
26.6
29.7
Mean
residual
(ugnr3)
3.96
4.88
0.99
6.95
2.51
5.76
8.13
10.13
4.93
-3.53
5.51
5.89
Mean
unsigned error
(ugm-3)
18.87
18.67
19.26
19.48
22.01
26.82
30.11
31.74
23.92
19.98
19.92
18.73
Coefficient of
Determination,
R2
0.130
0.135
0.169
0.185
0.316
0.377
0.259
0.266
0.307
0.248
0.247
0.106
The normalized bias ranges from 2 to 25 percent. The bias is positive for each month, indicating
some overestimation of ozone throughout the domain. The values of bias are lowest for the
transitional months (in terms of meteorology) of March through May and October and highest for
the winter months, when ozone concentrations are typically low. The values for the typical ozone
season for most areas (April through October) range from 2 to 17 percent, which indicates
reasonable model performance for ozone (EPA, 2006). Similarly the normalized gross error is
largest during the winter months, and indicative of reasonable model performance for the ozone
season. The fractional bias and error values confirm that the higher normalized bias and error
values during the winter months are driven, in part, by low concentrations. The mean residual
values tend to be largest during the summer months, when ozone concentrations are highest. The
correlation statistics indicate that the simulation-observation pairs are generally not well
correlated, but the greatest correlations occur during the ozone season months.
Statistical measures of model performance for SO2 are presented for each month in Table A-2.
A-2 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix A: Model Performance Evaluation for Non-Mercury Species
Table
A-2. Month-by-Month REMSAD Model
Performance Statistics for the 12-km Resolution Grids:
Monthly Average SO2 Concentration (ugm 3) at AQS-STN Sites.
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Normalized
bias
(%)
-12.5
-3.6
-0.3
-6.6
7.0
22.0
28.4
31.7
38.6
22.6
20.7
13.9
Normalized
gross error
(%)
46.9
48.4
50.2
51.3
59.6
66.9
69.2
69.6
73.8
63.5
58.2
57.5
Fractional
bias
(%)
-31.7
-23.4
-20.9
-28.5
-19.2
-8.3
-3.1
0.8
4.9
-3.8
-3.3
-10.6
Fractional
error
(%)
54.9
52.9
54.0
58.1
60.1
59.4
58.6
58.1
58.7
57.8
51.7
54.5
Mean
residual
(ngm-3)
-3.47
-1.56
-0.99
-1.82
-0.51
1.00
1.28
2.04
2.38
0.86
0.85
0.69
Mean
unsigned error
(ngm-3)
8.77
7.29
6.52
6.55
6.49
6.65
6.88
7.27
6.75
7.15
7.04
6.83
Coefficient of
Determination,
R2
0.323
0.171
0.142
0.199
0.171
0.176
0.128
0.133
0.141
0.173
0.268
0.268
The normalized bias ranges from -12 to nearly 40 percent. The bias is negative for January
through April, and positive for each month thereafter. The normalized gross error ranges from
approximately 45 to 75 percent. Both measures indicate that the SO2 concentrations, which are
often urban-scale in nature, are not well represented by the model. Differences between the
normalized and fractional bias values indicate that the normalized values are influenced by
overestimation of (very) low concentrations. The fractional bias is more frequently negative. The
correlation statistics indicate that the simulation-observation pairs are generally not well
correlated, but the greatest correlations occur during the winter months.
Statistical measures of model performance for total PM2.s concentration for the AQS-STN sites are
presented for each month in Table A-3.
Table A-3. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Monthly Average PM2.5 Concentration (ugm3) at AQS-STN Sites.
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Normalized
bias
5.7
6.4
15.0
4.0
-11.5
-17.6
-13.7
-6.6
12.5
14.0
12.0
36.1
Normalized
gross error
37.0
36.5
35.4
35.8
30.7
31.0
33.6
33.2
33.6
36.8
41.5
60.1
Fractional
bias
-7.1
-5.3
5.3
-6.0
-19.6
-26.2
-22.9
-16.0
3.6
4.7
-0.5
15.3
Fractional
error
36.7
36.2
32.1
35.3
34.6
37
38.0
34.2
29.7
34.1
40.8
50.4
Mean
residual
(ngm-3)
-0.16
0.48
1.51
0.41
-1.24
-2.20
-2.06
-1.46
0.98
1.39
1.09
3.11
Mean
unsigned error
5.76
4.54
3.89
3.94
3.49
3.85
4.25
4.53
3.35
3.88
5.61
6.33
Coefficient of
Determination,
R2
0.130
0.119
0.319
0.259
0.417
0.598
0.394
0.402
0.261
0.320
0.123
0.070
A-3
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix A: Model Performance Evaluation for Non-Mercury Species
The normalized bias ranges from 4 to approximately 18 percent for most months, but is higher for
December (36 percent). The bias is negative for the summer months (May through August), and
positive for the remaining months, indicating some underestimate of the typically higher summertime
PM concentrations. The normalized gross error ranges from approximately 30 to 40 percent, except
again for December. The correlation statistics indicate that the simulation-observation pairs are fairly
well correlated, and the correlation values are highest for the summer months.
Statistical measures of model performance for total PM2.s concentration for the IMPROVE sites
are presented for each month in Table A-4.
Table A-4. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Monthly Average PM2.5 Concentration (jigm3) at IMPROVE Sites.
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Normalized
bias
(%)
46.6
10.5
8.6
-23.0
-37.0
-36.3
-26.5
-3.6
-10.2
-4.7
3.1
56.1
Normalized
gross error
(%)
56.7
39.3
32.9
42.8
42.2
43.7
38.7
42.2
34.7
31.6
43.1
75.5
Fractional
bias
(%)
24.9
0.9
0.6
-37.7
-54.0
-54.1
-39.8
-19.8
-20.5
-12.4
-9.7
26.4
Fractional
error
(%)
36.5
36.5
31.6
52.9
58.5
60.0
49.5
38.3
39.9
33.3
43.9
51.1
Mean
residual
(ngm-3)
1.51
0.31
0.57
-1.36
-2.34
-2.75
-2.39
-0.80
-0.66
-0.04
0.51
1.82
Mean
unsigned error
(|igm-3)
1.97
1.59
1.81
2.61
2.88
3.07
3.09
2.96
2.16
2.01
2.39
2.48
Coefficient of
Determination,
R2
0.669
0.642
0.719
0.429
0.588
0.671
0.508
0.411
0.430
0.570
0.635
0.629
The normalized bias ranges from -40 to approximately 10 percent for most months, but is higher
for January (47 percent) and December (56 percent). A lower fractional bias for these months
suggests that the high values are driven by overestimation of low concentrations. The normalized
bias is negative for the warmer months (April through October), and positive for the remaining
months, and the values indicate some underestimate of the typically higher summertime PM
concentrations. The normalized gross error ranges from approximately 30 to 40 percent, except
again for January and December. The correlation statistics indicate that the simulation-
observation pairs are fairly well correlated, and this does not vary much by month.
Statistical measures of model performance for sulfate concentration for the CASTNet sites are
presented for each month in Table A-5.
A-4 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix A: Model Performance Evaluation for Non-Mercury Species
Table A-5. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Monthly Average Sulfate Concentration (ugm3) at CASTNet Sites.
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Normalized
bias
(%)
-32.4
-28.9
-22.4
-26.0
-26.7
-41.1
-47.9
-40.6
-30.6
-13.2
-18.3
17.6
Normalized
gross error
(%)
36.2
33.8
24.7
29.1
33.6
42.5
49.3
41.5
31.3
22.2
25.2
32.5
Fractional
bias
(%)
-42.0
-37.8
-28.3
-34.5
-36.1
-56.2
-67.9
-55.3
-41.1
-19.8
-23.6
9.8
Fractional
error
(%)
45.4
42.1
30.4
37.3
42.0
57.3
69.1
56.1
41.8
28.1
29.1
28.5
Mean
residual
(ngm-3)
-0.89
-0.79
-0.56
-0.90
-0.84
-2.00
-2.42
-1.96
-0.82
-0.19
-0.46
0.13
Mean
unsigned error
(ngm-3)
0.92
0.82
0.58
0.95
0.99
2.00
2.44
1.97
0.84
0.38
0.50
0.35
Coefficient of
Determination,
R2
0.880
0.792
0.850
0.498
0.824
0.661
0.783
0.824
0.880
0.642
0.880
0.819
The normalized bias ranges from approximately -50 to 20 percent. The largest (negative) bias
values are associated with the summer months, when sulfate concentrations tend to be highest.
This indicates that regional-scale sulfate concentrations are underestimated during these
months. The normalized gross error also ranges from approximately 20 to 50 percent;
consistency with the absolute value of the bias indicates that the concentrations are generally
over- or underestimated and not mixed. Despite the differences between the simulated and
observed values, the values are well correlated.
Statistical measures of model performance for sulfate concentration for the IMPROVE sites are
presented for each month in Table A-6.
Table A-6. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Monthly Average Sulfate Concentration (ugm3) at IMPROVE Sites.
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Normalized
bias
(%)
-27.8
-23.0
-18.7
-17.7
-33.4
-38.8
-45.7
-39.3
-28.9
-25.2
-23.7
44.3
Normalized
gross error
(%)
32.0
25.9
24.8
32.7
39.7
49.5
48.4
44.2
43.3
34.8
31.5
58.8
Fractional
bias
(%)
-35.6
-29.0
-25.0
-25.4
-46.0
-59.2
-66.9
-55.6
-44.9
-35.8
-32.3
20.0
Fractional
error
(%)
39.3
31.7
30.1
38.2
51.4
67.0
69.3
59.7
54.9
44.0
39.0
38.0
Mean
residual
(ngm-3)
-0.31
-0.34
-0.32
-0.31
-0.61
-0.90
-1.18
-1.01
-0.46
-0.29
-0.29
0.06
Mean
unsigned error
(ngm-3)
0.35
0.36
0.37
0.52
0.71
0.99
1.23
1.07
0.65
0.46
0.39
0.22
Coefficient of
Determination,
R2
0.845
0.778
0.859
0.719
0.767
0.848
0.736
0.837
0.689
0.797
0.841
0.861
The normalized bias ranges from approximately -45 to 45 percent. The bias values are all
negative, except for December, indicating that the sulfate concentrations at the more remote
approximately 30 to 45 percent, but is higher for December. Differences between the normalized
IMPROVE sites are generally underestimated. The normalized gross error ranges from gross
A-5 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix A: Model Performance Evaluation for Non-Mercury Species
error and the absolute value of the bias indicate a mix of over- and underestimation. Despite the
differences between the simulated and observed values, the values are well correlated.
Statistical measures of model performance for particulate nitrate concentration for the CASTNet
sites are presented for each month in Table A-7.
Table A-7. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Monthly Average Nitrate Concentration ((igm3) at CASTNet Sites.
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Normalized
bias
(%)
75.1
64.1
114.3
49.6
-10.9
158.1
42.7
87.9
99.8
93.6
125.3
107.7
Normalized
gross error
(%)
98.7
96.1
146.2
95.6
70.9
219.2
124.0
163.4
169.8
140.3
149.0
128.5
Fractional
bias
(%)
28.2
14.2
27.7
1.4
-51.8
6.2
-36.9
-22.5
-13.6
14.6
39.9
39.3
Fractional
error
(%)
59.1
62.4
77.2
73.5
86.4
117.8
111.8
117.6
114.5
96.9
77.6
73.1
Mean
residual
(ngm-3)
0.34
0.21
0.58
0.40
-0.15
0.27
-0.02
0.04
0.17
0.58
1.00
0.79
Mean
unsigned error
(ligm-3)
1.00
0.72
1.02
0.87
0.39
0.56
0.35
0.41
0.52
0.86
1.20
0.94
Coefficient of
Determination,
R2
0.648
0.585
0.584
0.214
0.343
0.055
0.001
0.000
0.027
0.333
0.630
0.671
Since the values of nitrate can be very low for many sites (especially in the eastern U.S.) the
calculation of the relative measures of model performance (which use the observed value in the
denominator) is not very meaningful. This is manifested in Table A-7 as very large normalized
bias and error statistics. The fractional bias and error are likely more meaningful metrics for this
pollutant. The fractional bias indicates underestimate of nitrate from April through September and
overestimation of nitrate during the colder months. The fractional errors are smaller than when
normalized directly by the observations, but are still quite large (on the order of 60 to 120 percent).
The simulated and observed values are reasonably well correlated during the winter months,
when nitrate concentrations tend to be highest, and less well to poorly correlated during the
summer months, when nitrate concentrations tend to be low.
Statistical measures of model performance for particulate nitrate concentration for the IMPROVE
sites are presented for each month in Table A-8.
A-6 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix A: Model Performance Evaluation for Non-Mercury Species
Table
A-8. Month-by-Month REMSAD Model
Performance Statistics for the 12-km Resolution Grids:
Monthly Average Nitrate Concentration
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Normalized
bias
(%)
5.7
0.3
34.5
27.9
-60.0
-35.7
-58.8
-56.5
-24.7
59.7
81.4
101.5
Normalized
gross error
(%)
52.7
53.7
68.1
70.3
70.3
86.0
87.8
85.7
93.2
116.0
114.7
140.1
Fractional
bias
(%)
-14
-20.8
6.9
-4.8
-110.5
-96.9
-126.9
-123.4
-87.8
-4.7
24
22.5
Fractional
error
(%)
52.5
58.3
52.7
60
118.2
123.2
141.4
136
122.6
86.2
71.2
80.1
(ugm 3) at IMPROVE Sites.
Mean
residual
(ngm-3)
-0.02
0.01
0.33
0.19
-0.16
-0.06
-0.15
-0.08
0.00
0.38
0.62
0.50
Mean
unsigned error
(ngm-3)
0.50
0.44
0.56
0.34
0.25
0.29
0.25
0.25
0.30
0.57
0.82
0.79
Coefficient of
Determination,
R2
0.669
0.434
0.734
0.848
0.717
0.332
0.350
0.213
0.326
0.429
0.420
0.490
The large bias and error values reflect the low nitrate concentrations for many sites (especially in the
eastern U.S.) and are likely not very meaningful. The simulated and observed values are reasonably
well correlated during the winter months, when nitrate concentrations tend to be highest, and less
well to poorly correlated during the summer months, when nitrate concentrations tend to be low.
Statistical measures of model performance for particulate ammonium concentration for the
CASTNet sites are presented for each month in Table A-9.
Table A-9. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Monthly Average Ammonium Concentration (ugm3) at CASTNet Sites.
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Normalized
bias
(%)
4.8
7.0
18.5
22.8
-13.1
-27.6
-37.1
-33.1
-14.3
39.7
31.8
69.0
Normalized
gross error
(%)
28.3
32.2
33.8
38.6
29.0
30.7
40.6
34.7
29.0
61.3
44.8
75.4
Fractional
bias
(%)
-0.9
-1.2
10.3
12
-20.6
-37.4
-50.5
-44.7
-22.8
22.4
19.9
39.9
Fractional
error
(%)
26.7
29.4
29.9
33.2
33.2
40.2
53.5
46.2
35.9
50.5
35.5
48
Mean
residual
(ngm-3)
-0.06
-0.04
0.14
0.19
-0.15
-0.38
-0.49
-0.44
-0.06
0.36
0.34
0.38
Mean
unsigned error
(ngm-3)
0.30
0.25
0.28
0.41
0.28
0.42
0.53
0.45
0.23
0.47
0.40
0.41
Coefficient of
Determination,
R2
0.819
0.781
0.821
0.352
0.781
0.629
0.760
0.834
0.826
0.638
0.882
0.830
The normalized bias ranges from approximately -40 to 40 percent, with the exception of
December for which the value is nearly 70 percent. The values are negative for the warmer
months May through September and positive for the remaining (cooler) months. Differences in the
bias and error values indicate a mix of over- and underestimation for most months. Despite the
differences between the simulated and observed values, the values are well correlated.
Summary of Monthly Model Performance for Deposition
As noted earlier in this section, there are fewer measurements for the assessment of model
performance for deposition, compared to air concentration.
A-7 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix A: Model Performance Evaluation for Non-Mercury Species
Statistical measures of model performance for sulfate deposition at NADP sites are presented for
each month in Table A-10.
Table A-10. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Monthly Sulfate Deposition (kg ha1) at NADP Sites.
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Grid
Resolution
(km)
12
12
12
12
12
12
12
12
12
12
12
12
Normalized
bias
(%)
-51.6
-51.0
-59.0
-35.2
-26.5
-47.0
-18.0
-32.6
-49.3
-38.4
-37.0
-52.4
Normalized
gross error
(%)
60.4
59.3
68.3
60.7
51.3
53.2
67.6
62.0
67.7
63.7
60.4
65.0
Fractional
bias
(%)
-82.9
-81.7
-100.3
-66.4
-49.1
-74.2
-54.4
-63.6
-89.5
-70.5
-67.7
-86.9
Fractional
error
(%)
94.4
89.9
105.5
81
66.4
79
74.7
79.8
96.5
91.7
86.7
101.8
Mean Mean
residual unsigned error
(^.gm-3) (ngm-3)
-0.25
-0.40
-0.55
-0.40
-0.34
-0.69
-0.47
-0.60
-0.60
-0.32
-0.20
-0.43
0.25
0.40
0.56
0.46
0.44
0.72
0.61
0.70
0.63
0.36
0.25
0.43
Coefficient of
Determination,
R2
0.709
0.676
0.561
0.454
0.569
0.526
0.444
0.490
0.429
0.558
0.307
0.542
The normalized bias ranges from approximately -60 to -20 percent. The bias is negative for each
month, indicating the overall underestimation of sulfate deposition by the model. There is no clear
tendency in the bias with respect to time of year. The normalized gross error ranges from
approximately 50 to 70 percent, and differences between the bias and error values indicate a mix
of over- and underestimation. The fractional bias and error measures have similar tendencies but
indicate somewhat larger errors. Despite the large relative errors, the mean residual and unsigned
error values are small, and the correlation statistics indicate that the simulation-observation pairs
are moderately correlated.
Statistical measures of model performance for nitrate deposition at NADP sites are presented for
each month in Table A-11.
Table A-11. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Monthly Nitrate Deposition (kg ha-1) at NADP Sites.
Grid
Normalized
Month Resolution bias
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
The
(km)
12
12
12
12
12
12
12
12
12
12
12
12
normalized
December, and
(%)
-7.4
4.6
-24.8
41.8
37.4
11.7
47.5
30.3
9.3
76.0
108.7
-2.0
bias ranges
Normalized
gross error
(%)
73.7
85.6
61.0
90.6
72.6
56.7
92.3
76.1
75.6
121.2
155.2
68.6
Fractional
bias
(%)
-48.2
-48.4
-59
-6.9
3.3
-11.5
-6.2
-5.9
-31.6
1
7.4
-40.4
Fractional
error
(%)
89.1
90.5
79.3
67.7
52.8
50.8
58.6
60.4
65.7
70.3
81.3
78.7
from -25 to 108 percent. The bias
positive for the remaining
months.
Mean
Mean
residual unsigned error
(ugm-3)
-0.04
-0.11
-0.29
0.03
0.11
-0.07
0.00
0.08
-0.12
0.07
0.22
-0.04
is negative
(Mam-3)
0.29
0.41
0.43
0.40
0.45
0.52
0.53
0.55
0.39
0.29
0.38
0.33
for January,
The normalized gross error ranges
Coefficient of
Determination,
R2
0.262
0.153
0.274
0.317
0.465
0.333
0.298
0.391
0.301
0.533
0.257
0.461
March, and
from
A-8 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix A: Model Performance Evaluation for Non-Mercury Species
approximately 55 to 155 percent. The large range for the relative metrics may be driven by some
low observed deposition values. Differences between the bias and error values indicate a mix of
over- and underestimation. The fractional bias reveals a greater tendency for underestimation and
suggests that much of the overestimation indicated by the normalized bias is for sites with low
nitrate deposition. The mean residual and unsigned error values are small, and the correlation
statistics indicate that the simulation-observation pairs are moderately correlated.
Statistical measures of model performance for sulfate deposition at NADP sites are presented for
each month in Table A-12.
Table A-12. Month-by-Month REMSAD Model Performance Statistics for the 12-km Resolution Grids:
Monthly Ammonia Deposition (kg ha1) at NADP Sites.
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Normalized
bias
(%)
-13.6
-21.9
-27.8
-4.2
-16.5
-26.0
-12.9
-11.1
-29.9
-1.1
-7.4
-30.6
Normalized
gross error
(%)
56.1
55.2
63.2
65.4
53.8
50.5
65.1
68.7
64.8
67.0
67.5
51.0
Fractional
bias
(%)
-28.5
-43.5
-63.1
-31.8
-38
-46.9
-42.1
-46.6
-64
-27.5
-40.2
-48.7
Fractional
error
(%)
79.3
73.5
81.2
71.7
66.1
70.7
78.6
77
85.6
79.4
76.4
78.6
Mean
residual
(ngm-3)
-0.03
-0.05
-0.08
-0.08
-0.09
-0.11
-0.09
-0.09
-0.09
-0.03
-0.03
-0.04
Mean
unsigned error
(ngm-3)
0.04
0.06
0.10
0.12
0.13
0.15
0.15
0.14
0.11
0.06
0.06
0.06
Coefficient of
Determination, R2
0.318
0.391
0.205
0.482
0.437
0.241
0.141
0.176
0.289
0.593
0.388
0.271
The normalized bias ranges from approximately -30 to -1 percent. The bias is negative for each
month, indicating an overall underestimation of ammonia deposition by the model. The normalized
gross error ranges from approximately 50 to 70 percent, and differences between the bias and
error values indicate a mix of over- and underestimation. The mean residual and unsigned error
values are small, and the correlation statistics indicate that the simulation-observation pairs are
moderately correlated.
Summary of Annual Model Performance for Particulate Species Concentration and
Deposition
Statistical measures of model performance for PM2.s and several component species are
presented for each month in Table A-13.
A-9 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix A: Model Performance Evaluation for Non-Mercury Species
Table A-13. Annual REMSAD Model Performance Statistics for the 12-km Resolution Grids
for the Various Observational Networks.
Units for Annual Average PM2.5 and Species Concentrations are |j,gm-3.
Units for Annual Deposition Totals are kg ha-1.
Network/
Species
AQS-STN
PM2.5
IMPROVE
PM2.5
CASTNet
Sulfate
IMPROVE
Sulfate
CASTNet
Nitrate
IMPROVE
Nitrate
No. Normalized
of bias
Sites (%)
-0.6
-3.4
-32.6
-26
49.5
14.4
Normalized
gross error
(%)
28.4
38.9
33.5
40.9
86.2
68
Fractional
bias
(%)
-7.9
-18.8
-41.5
-41
9.1
-15.2
Fractional
error
(%)
29.6
36.1
42.3
46.6
66.9
63.3
Mean
residual
(jigm-3)
0.2
-0.38
-0.97
-0.51
0.33
0.19
Mean
unsigned
error Qigrrv
3)
3.54
1.82
0.98
0.55
0.60
0.41
Coefficient of
Determination, R2
0.407
0.676
0.918
0.869
0.561
0.623
CASTNet
Ammonium
NADP
Sulfate
deposition
NADP
Nitrate
deposition
NADP
Ammonia
deposition
-5.8
-52.2
-3.5
-31.3
20.0
52.4
35.3
38
-9.8
-75.8
-14.4
-46.6
22.3
75.9
39.7
52.3
-0.02
-5.63
0.1
-0.74
0.16
5.63
3.28
0.86
0.914
0.738
0.523
0.497
On an annual basis, PM2.s is very well represented, but this may be a result of underestimation
of sulfate and overestimation of nitrate and other species. Deposition is underestimated,
especially for sulfate and ammonia. Overall, the annual errors tend to be smaller than the
monthly values due to a mix of over and underestimation of the various species. The results are
fairly consistent across the monitoring networks.
For PM2.s, the errors are larger for the IMPROVE sites, compared to the AQS-STN sites, but the
correlations favor the IMPROVE sitespossibly due to the lower concentrations at the Class I
area sites.
For sulfate, the metrics are similar for the CASTNet and IMPROVE sites. For nitrate, there is a
greater tendency for overestimation at the CASTNet sites, compared to the IMPROVE sites.
A-10 August 2008
-------
Appendix B: Preliminary Comparison of CMAQ
and REMSAD Tagging Simulations
Due to reliable performance in past TMDL modeling studies (e.g., Myers and Wei, 2004),
computational efficiency, and the availability of the Particle and Precursor Tagging Methodology
(PPTM) at the initiation of this project, REMSAD was used for the tagging simulations reported in
the main document. Since that time, ICF has implemented PPTM for mercury in the CMAQ model
(Douglas et al., 2006). To compare the results for the two models with PPTM, CMAQ was applied
for a three-month subset of the annual simulation period (summer 2001) using PPTM for seven
tags. This appendix compares the CMAQ and REMSAD results for this summer simulation period.
The REMSAD results were extracted from the simulation that is discussed in the main report.
Model Setup
The CMAQ modeling domain includes an outer grid with 36-km horizontal resolution and a one-
way nested (inner) grid with 12-km resolution. The nested grid covers Illinois and portions of
several surrounding states. The outer 36-km domain for CMAQ matches the 36-km domain
used for the REMSAD simulations. The 2001 meteorological input files for the 36-km domain
are those used in the CAIR and CAMR modeling and were derived from the same MM5
simulation as the inputs for the REMSAD simulation. For the 12-km domain, 2001 high
resolution meteorological input files were acquired from the EPA OAQPS. The 12-km
meteorological fields were derived from a recent MM5 simulation that covered approximately the
eastern two-thirds of the U.S. A subset of these 12-km input fields was extracted for the nested
grid simulation. The 12-km grid is depicted in the deposition plots that follow.
Emissions for CMAQ were prepared using the SMOKE processing system using the updated
2001 inventory that is discussed in detail in the main report. The emissions inputs for the two
models are therefore the same.
Seven PPTM tags were applied in the CMAQ simulation, as follows:
Powerton
Joliet 29
Joppa Steam
Other IL utilities
Remaining IL sources
All other emissions including those from other states, Canada, and Mexico, as well as from
natural and re-emission processes
Initial and boundary conditions (IC/BC).
Note that the first three tags are individual utility sources in Illinois. This set of tags overlaps the
tags prepared for REMSAD modeling and, with appropriate aggregation of some REMSAD tags,
allows a direct comparison of the simulated source contributions for the two modeling platforms.
As described in the main report (Section 5), REMSAD simulations were conducted using
boundary concentrations based on each of three different global model simulations. For the 36-
km CMAQ simulation, the boundary concentrations used were based on the GRAHM global
B-1 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix B: Preliminary Comparison of CMAQ and REMSAD Tagging Simulations
model simulation. Therefore, the CMAQ simulation results are compared to the REMSAD
simulation that also used the GRAHM-based boundary concentrations.
The CMAQ simulation was set up to be as similar as possible to the REMSAD simulation, while
utilizing the best available data. Differences in model formulation may result in somewhat
different deposition estimates from the two models (e.g., Bullock et al., 2006), but the focus here
is on comparing the estimates of relative contributions from different sources made by the two
models. In interpreting the differences in the simulation results, keep in mind that in addition to
differences in model formulation the following differences, and possibly others, are present
between the two applications:
The REMSAD simulation used 12-km resolution over the entire U.S. The CMAQ simulation
used a 12-km domain only around Illinois and parts of the surrounding states.
Because the 12-km MM5 files were not available at the initiation of this project, 12-km
meteorological fields for REMSAD were interpolated from files based on 36-km MM5 results.
The 12-km resolution meteorological fields used for CMAQ were derived from a 12-km
resolution MM5 simulation that were recently acquired from EPA OAQPS. Thus, the
meteorological fields are expected to have some differences. In particular, the REMSAD 12-
km fields may not embody some of the features contained in the CMAQ 12-km
meteorological files. These differences have not been investigated here.
The REMSAD results presented here are for the summer season (defined as June, July, and
August) of a full annual simulation of 2001. The CMAQ results are for a simulation covering
the summer season with a ten-day spin-up period prior to June 1, 2001.
REMSAD calculates re-emission of mercury dynamically (during the course of the
simulation). The CMAQ simulation includes re-emission estimates directly in the emissions
input files (as direct emissions).
The REMSAD emissions files do not include natural emissions of mercury (e.g., volcanic
emissions) within the modeling domain (roughly North America). Natural emissions are
included in the inventories used for the global simulations that provide boundary
concentrations for both the REMSAD and CMAQ simulations. The CMAQ input emissions
files do include natural emissions of mercury within the CMAQ modeling domain.
Simulation Results
The simulation results for the two models are compared in this section for the 12-km grid, as
utilized in the CMAQ simulation and for the three-month (summer 2001) simulation period.
Figure B-1 presents and compares: (a) simulated wet and dry deposition of total mercury for the
CMAQ run, and (b) simulated wet and dry deposition of total mercury for the REMSAD run.
Overall, the two distributions are similar although, in some areas, one or the other model may
produce higher deposition estimates.
The estimates of the contributions from the tagged sources are quite similar. This is examined in
further detail for the locations of the greatest impact from each of the three individually tagged
sources and for the location at which the REMSAD simulation predicts the greatest impact from
Illinois sources (the blue triangle in the REMSAD deposition plot).
Figure B-2 compares the simulated contributions to wet and dry deposition from both models at
the location of the greatest impact from the Powerton facility tag. For both models, about 50
B-2 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix B: Preliminary Comparison of CMAQ and REMSAD Tagging Simulations
percent of the deposition is attributed to background (IC/BC) and the largest source contribution
is from the Powerton plant, followed by a contribution from other Illinois utilities.
The contributions at the location of maximum contribution from Joliet 29 (Figure B-3) and at the
location of maximum contribution from Joppa Steam (Figure B-4) follow a similar pattern. The
background contribution is well over half at both the Joliet 29 and the Joppa Steam location. In
each case, though, the individual source contributes more than 10 percent of the deposition.
The two models are consistent on the ranking of contributions from the tags.
Contributions at the location of the greatest impact of Illinois sources (Figure B-5) show the
largest contribution from remaining Illinois sources (that is, sources in Illinois that were not
individually tagged). The relative contributions estimated by the two models are virtually
identical. About half of the deposition is from Illinois sources with an additional contribution from
emissions from outside the state and the remainder from background.
Summary
A comparison of CMAQ and REMSAD modeling results for a 12-km domain around Illinois and
a summer 2001 simulation period shows that the simulated spatial patterns of mercury
deposition are similar between the two models.
PPTM was used to tag seven different emissions sources (a mix of individual sources, groups of
sources, and source regions (including global background)). The estimates of source
contributions to mercury deposition are very similar between the two models.
Thus, despite some differences in the base simulation results, the relative contributions from the
tagged sources are consistent between the CMAQ and REMSAD PPTM results, for the area
and time period considered. The two models are also quite consistent on the ranking of
contributions from the tagged sources/categories. The length of the simulation period may
contribute to the similarities. Despite potential day-to-day differences in the meteorological input
files, the seasonal average conditions are likely represented in both sets of meteorological
inputs. The contribution results suggest that given the same emissions and boundary condition
inputs, PPTM is able to resolve the relative contributions from the tagged emission sources and
IC/BCs within the context of each model and that the results are effectively the same.
B-3 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix B: Preliminary Comparison of CMAQ and REMSAD Tagging Simulations
Figure B-1. Summer Total Wet and Dry Deposition of Mercury Simulated by
(a) CMAQ and (b) REMSAD.
I
22
20
18
16
14
12
10
8
6
4
2
0
g/hm2
PAUE
t>V
MCNC
90
(a)
(b)
LEVEL L THG (g/km2) -t- MAXIMUM - 7&.S s
Time: Q May 29, 20DO-0 May 28, SGOfi - MINIMUM - 3.4 g/T
1
1 72
June 1,2001 1:00:00
Min= 2 at (40,85XMax= 125 at (46,21)
B-4
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix B: Preliminary Comparison of CMAQ and REMSAD Tagging Simulations
Figure B-2. Estimates of Contributions to Total of Wet and Dry Deposition at the Location of
Greatest Impact from Powerton Based on PPTM Simulations by CMAQ and REMSAD.
CMAQ
53%
54%
24%
REMSAD
31%
D Powerton
Joliet 29
D Joppa Steam
D Other IL utilities
D Remaining IL sources
nAII other states, Canada
and Mexico
IC/BC
D Powerton
Joliet 29
D Joppa Steam
D Other IL utilities
D Remaining IL sources
nAII other states, Canada
and Mexico
IC/BC
B-5
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix B: Preliminary Comparison of CMAQ and REMSAD Tagging Simulations
Figure B-3. Estimates of Contributions to Total of Wet and Dry Deposition at the Location of
Greatest Impact from Joliet 29 Based on PPTM Simulations by CMAQ and REMSAD.
CMAQ
60%
55%
16%
8%
7%
REMSAD
16%
13%
D Powerton
Joliet 29
D Joppa Steam
D Other IL utilities
D Remaining IL sources
nAII other states, Canada
and Mexico
IC/BC
D Powerton
Joliet 29
D Joppa Steam
D Other IL utilities
D Remaining IL sources
nAII other states, Canada
and Mexico
IC/BC
B-6
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix B: Preliminary Comparison of CMAQ and REMSAD Tagging Simulations
Figure B-4. Estimates of Contributions to Total of Wet and Dry Deposition at the Location of
Greatest Impact from Joppa Steam Based on PPTM Simulations by CMAQ and REMSAD.
CMAQ
71%
63%
13%
15%
REMSAD
11%
D Powerton
Joliet 29
D Joppa Steam
D Other IL utilities
D Remaining IL sources
nAII other states, Canada
and Mexico
IC/BC
D Powerton
Joliet 29
D Joppa Steam
D Other IL utilities
D Remaining IL sources
nAII other states, Canada
and Mexico
IC/BC
B-7
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix B: Preliminary Comparison of CMAQ and REMSAD Tagging Simulations
Figure B-5. Estimates of Contributions to Total of Wet and Dry Deposition at the Location of
Greatest Impact from Illinois Sources Based on PPTM Simulations by CMAQ and REMSAD.
CMAQ
42%
51%
REMSAD
42%
51%
D Powerton
Joliet 29
D Joppa Steam
D Other IL utilities
D Remaining IL sources
nAII other states, Canada
and Mexico
IC/BC
D Powerton
Joliet 29
D Joppa Steam
D Other IL utilities
D Remaining IL sources
nAII other states, Canada
and Mexico
IC/BC
B-8
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix B: Preliminary Comparison of CMAQ and REMSAD Tagging Simulations
References
Bullock, O. R., K. Lohman, C. Seigneur, K. Vijayaraghavan, A. Dastoor, D. Davignon, N. Selin,
D. Jacob, T. Myers, K. Civerolo, C. Hogrefe, J. Ku, G. Sistla, D. Atkinson, and T. Braverman.
2006. "Results from the North American Mercury Model Inter-comparison Study (NAMMIS)."
Poster session at the Eighth International Conference on Mercury as a Global Pollutant,
Aug. 6-11, 2006, Madison, Wisconsin.
Douglas, S., T. Myers, and Y. Wei. 2006. "Implementation of Mercury Tagging in the Community
Multi-scale Air Quality Model." (Draft) Prepared for the U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards. ICF International, San Rafael,
California (06-051).
Myers, T. and Y. Wei. 2004. REMSAD Air Deposition Modeling in Support of TMDL
Development for Southern Louisiana. ICF Consulting (04-038).
B-9 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix B: Preliminary Comparison of CMAQ and REMSAD Tagging Simulations
This page deliberately left blank.
B-10 August 2008
-------
Appendix C: Details of Emissions Revisions for
Selected States
This appendix includes details of emissions revisions made for certain states that provided
extensive information on mercury emissions within the states. For a discussion of the revisions,
see Section 4 of the main report.
Also included in this section is a tabulation of sources that were included in the "Collective
Sources" tag for each state in which the in-state maximum was due to "Collective Sources."
Illinois
Table C-1. Summary of Emissions Revisions Made to IPM Sources in State of Illinois
Facility Name
Coal Fired Utilities located Inside Chicago, IL MSA
Original Emissions
Crawford
Fisk
Waukegan
Joliet29
Joliet 9
Will County
Total
Revised Emissions
Crawford
Fisk
Waukegan
Joliet 29
Joliet 9
Will County
Total
Coal Fired Utilities located Outside Chicago, IL MSA
Original Emissions
Kincaid Generation L.L.C.
Hutsonville
Duck Creek
Newton
Wood River
Havana
Joppa Steam
Coffeen
Meredosia
E. D. Edwards
Hennepin
Baldwin
Dallman
Lakeside
Powerton
HGO
(tpy)
0.076
0.057
0.235
0.298
0.097
0.168
0.931
0.104
0.059
0.310
0.222
0.057
0.139
0.892
0.103
0.003
0.016
0.087
0.009
0.008
0.205
0.018
0.007
0.019
0.012
0.040
0.012
0.001
0.389
HG2(tpy)
0.034
0.025
0.069
0.133
0.043
0.061
0.365
0.047
0.027
0.091
0.099
0.026
0.050
0.339
0.063
0.008
0.001
0.043
0.024
0.020
0.091
0.045
0.018
0.051
0.015
0.107
0.015
0.003
0.174
HGP
(tpy)
1.7E-04
1.3E-04
3.5E-04
6.7E-04
2.2E-04
3.1E-04
0.002
2.3E-04
1.3E-04
4.6E-04
5.0E-04
1.3E-04
2.5E-04
0.002
0.002
0.001
0.000
0.001
0.002
0.002
0.000
0.004
0.002
0.005
0.001
0.010
0.001
0.000
0.001
Total
(tpy)
0.110
0.082
0.304
0.432
0.140
0.229
1.298
0.151
0.086
0.401
0.322
0.083
0.189
1.232
0.167
0.012
0.017
0.131
0.035
0.030
0.297
0.066
0.026
0.075
0.028
0.156
0.028
0.004
0.564
C-1
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Facility Name
Vermilion
Southern Illinois Power Cooperative
Total
Revised Emissions
Kincaid Generation L.L.C.
Hutsonville
DuckCreek
Newton
Wood River
Havana
Joppa Steam
Coffeen
Meredosia
E. D. Edwards
Hennepin
Baldwin
Dallman
Lakeside
Powerton
Vermilion
Southern Illinois Power Cooperative
Total
Total Emissions Changes forlPM
Emissions Changes
HGO
(tpy)
0.003
0.022
0.954
0.129
0.006
0.013
0.092
0.024
0.005
0.210
0.022
0.008
0.020
0.026
0.138
0.014
0.003
0.440
0.004
0.028
1.181
0.188
HG2(tpy)
0.008
0.010
0.695
0.079
0.015
0.001
0.045
0.014
0.013
0.094
0.055
0.022
0.054
0.012
0.062
0.017
0.007
0.197
0.009
0.012
0.708
-0.014
HGP
(tpy)
0.001
0.001
0.032
0.003
0.001
0.000
0.001
0.000
0.001
0.000
0.005
0.002
0.005
0.000
0.000
0.001
0.001
0.001
0.001
0.001
0.023
-0.009
Total
(tpy)
0.012
0.033
1.681
0.210
0.021
0.014
0.138
0.039
0.019
0.305
0.082
0.033
0.079
0.038
0.200
0.032
0.010
0.638
0.014
0.041
1.913
0.165
C-2
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-2. Summary of Emissions Revisions Made to Non-IPM Sources in State of Illinois
Facility Name
Coal Fired Utilities located Inside Chicago, IL MSA
Original Emissions
Marblehead Lime Co.
Marblehead Lime Co. - South Chicago Pit.
Lone Star Industries Inc - Oglesby Plant
Dixon - Marquette Cement Inc.
Vulcan Materials - McCook Lime Pit.
Illinois Cement Co.
Lafarge Corporation
Total
Revised Emissions
Marblehead Lime Co.
Marblehead Lime Co. - South Chicago Pit.
Lone Star Industries Inc - Oglesby Plant
Dixon - Marquette Cement Inc.
Vulcan Materials - McCook Lime Pit.
Illinois Cement Co.
Lafarge Corporation
Total
Total Emissions Changes for Non-IPM
Emissions Changes
HGO
(tpy)
0.435
0.193
0.170
0.148
0.150
0.135
0.000
1.232
0.020
0.009
0.075
0.001
0.015
0.004
0.004
0.129
-1.104
HG2
(tpy)
0.054
0.024
0.030
0.026
0.019
0.023
0.000
0.176
0.003
0.001
0.013
0.000
0.002
0.001
0.001
0.020
-0.156
HGP
(tpy)
0.054
0.024
0.027
0.024
0.019
0.022
0.000
0.170
0.003
0.001
0.012
0.000
0.002
0.001
0.001
0.019
-0.151
Total
(tpy)
0.544
0.242
0.227
0.198
0.187
0.180
0.000
1.578
0.025
0.011
0.100
0.002
0.019
0.006
0.006
0.168
-1.410
C-3
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Indiana
Table C-3. Summary of Emissions Revisions Made to IPM Sources in State of Indiana
Facility Name
Original Emissions
Tanners Creek
R. Gallagher Station
Gibson Generating Station
R. M. Schahfer
Clifty Creek
State Line
E. W. Stout
Petersburg
Rockport
Merom
Cayuga (IN)
Wabash River Generating Station
Warrick Power Plant
Total
Revised Emissions
Tanners Creek
R. Gallagher Station
Gibson Generating Station
R. M. Schahfer
Clifty Creek
State Line
E. W. Stout
Petersburg
Rockport
Merom
Cayuga (IN)
Wabash River Generating Station
Warrick Power Plant
Total
Total Emissions Changes for IPM
Emissions Changes
HGO
(tpy)
0.038
0.031
0.114
0.145
0.132
0.039
0.020
0.104
0.131
0.054
0.027
0.041
0.022
0.897
0.031
0.031
0.108
0.122
0.120
0.040
0.019
0.104
0.121
0.052
0.026
0.042
0.019
0.834
-0.063
HG2
(tpy)
0.101
0.082
0.170
0.061
0.125
0.026
0.053
0.009
0.348
0.005
0.072
0.054
0.057
1.163
0.082
0.081
0.162
0.051
0.113
0.028
0.051
0.009
0.321
0.004
0.068
0.056
0.049
1.075
-0.088
HGP
(tpy)
0.009
0.007
0.015
0.001
0.007
0.000
0.005
0.000
0.031
0.000
0.006
0.005
0.005
0.092
0.007
0.007
0.014
0.001
0.007
0.000
0.005
0.000
0.029
0.000
0.006
0.005
0.004
0.086
-0.007
Total
(tpy)
0.148
0.121
0.298
0.207
0.264
0.065
0.078
0.113
0.510
0.059
0.106
0.100
0.084
2.152
0.121
0.119
0.284
0.173
0.240
0.068
0.074
0.114
0.471
0.057
0.100
0.102
0.072
1.995
-0.157
C-4
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-4. Summary of Emissions Revisions Made to Non-IPM Sources in State of Indiana
Facility Name
HGO
(tpy)
HG2
(tPV)
HGP
(tpy)
Total
(tpy)
Original Emissions
Parkview Memorial Hospital-lnc
Lone Star Industries - Inc.
Lehigh Portland Cement Company
Essroc Cement-Speed Plant
Amoco Oil Co.
Waupaca Foundry-lnc. Plant 5
Wishard Memorial Hospital
Total
Revised Emissions
Facility Name
Parkview Memorial Hospital-lnc
Lone Star Industries - Inc.
Lehigh Portland Cement Company
Essroc Cement-Speed Plant
Amoco Oil Co.
Waupaca Foundry-lnc. Plant 5
Wishard Memorial Hospital
Total
Total Emissions Changes for Non-IPM
Emissions Changes
0.0803
0.0539
0.0577
0.0540
0.0161
0.0192
1.5E-05
2.8E-01
0.054
0.070
0.059
0.059
0.069
0.057
0.037
0.404
0.123
0.0482
0.0186
0.0100
0.0094
0.0055
0.0024
9.0E-06
9.4E-02
0.032
0.024
0.010
0.010
0.024
0.007
0.022
0.129
0.035
0.0321
0.0205
0.0092
0.0086
0.0061
0.0024
6.0E-06
7.9E-02
0.021
0.026
0.009
0.009
0.026
0.007
0.015
0.115
0.036
0.1605
0.0930
0.0769
0.0720
0.0277
0.0240
3.0E-05
4.5E-01
0.107
0.120
0.079
0.079
0.118
0.071
0.073
0.648
0.194
C-5
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Iowa
FIPS
Table C-5. Summary of Emissions Revisions made to IPM Sources in State of Iowa
Facility Name
Original Emissions
19193 George Neal North
19193 George Neal South
19155 Council Bluffs
19005 Lansing
19179 Ottumwa
19115 Louisa
19045 Milton L. Kapp
19057 Burlington
19061 Dubuque
19127 Sutherland
19139 Muscatine
19163 Riverside
19113 Prairie Creek
Total
Revised Emissbns
19193
19193
19155
19005
19179
19115
19045
19057
19061
19127
19139
19163
George Neal North
George Neal South
Council Bluffs
Lansing
Ottumwa
Louisa
Milton L. Kapp
Burlington
Dubuque
Sutherland
Muscatine
Riverside
19113 Prairie Creek
Total
Total Emissions Changes
Emissions Changes
Table C-6. Summary
FIPS
Facility Name
Plant Stk
ID ID
13452
13453
13713
13530
13581
13535
13206
13689
13730
13340
13564
13294
13593
13452 1
2
3
13453 1
13713 1
2
3
13530 1
13581 1
13535 1
13206 1
13689 1
13730 1
13340 1
2
3
13564 1
2
13294 1
13593 1
2
Unit
ID
001
002
003
001
001
002
003
001
001
001
001
001
001
001
002
003
001
002
001
001
002
StkHt
(m)
121.9
143.0
167.6
152.1
182.9
185.9
74.7
93.3
155.4
75.6
67.1
105.5
61.0
68.6
91.4
121.9
143.0
76.2
76.2
167.6
93.3
182.9
185.9
74.7
93.3
48.2
75.9
75.9
75.9
68.6
91.4
105.5
61.0
26.4
of Emissions Revisions
Plant Stk
ID ID
Unit
ID
StkHt
(m)
Diam
(m)
6.10
7.86
7.62
4.69
7.62
9.14
3.96
3.57
4.45
2.90
2.59
4.08
3.96
2.87
4.65
6.10
7.62
3.66
3.66
7.62
3.58
7.62
9.14
3.96
3.58
2.74
2.90
2.90
3.05
2.60
3.20
4.08
3.96
3.79
made to
Diam
(m)
Temp
(DegK)
436
387
408
403
408
422
416
400
381
441
436
420
413
433
416
450
422
433
416
411
478
478
422
450
478
450
450
450
436
450
355
420
438
445
non-IPM
Temp
(DegK)
Vel
(m/s)
30.0
19.2
23.2
24.7
27.7
22.3
24.1
29.9
18.9
19.5
19.8
23.8
17.1
39.7
25.1
33.2
20.7
8.5
14.0
21.6
10.8
33.2
22.3
35.1
43.3
15.1
13.4
13.4
23.2
20.0
22.8
23.8
22.1
9.1
and
Vel
(m/s)
HGO
(tpy)
0.146
0.100
0.108
0.042
0.040
0.123
0.013
0.018
0.003
0.010
0.064
0.011
0.026
0.705
0.014
0.036
0.087
0.097
0.007
0.007
0.100
0.026
0.083
0.114
0.015
0.021
0.002
0.006
0.006
0.006
0.009
0.027
0.014
0.017
0.009
0.732
-0.002
HG2
(tpy)
0.055
0.045
0.043
0.021
0.018
0.018
0.013
0.008
0.007
0.007
0.006
0.005
0.004
0.250
0.005
0.014
0.033
0.043
0.003
0.003
0.040
0.013
0.037
0.016
0.014
0.009
0.005
0.004
0.004
0.004
0.001
0.003
0.006
0.003
0.001
0.338
0.012
MWI Sources in
HGO
(tpy)
HG2
(tpy)
HGP
(tpy)
0.001
2.5E-04
2.7E-04
3.9E-04
1.2E-04
8.9E-05
0.001
4.9E-05
0.001
1.1E-04
2.6E-04
2.9E-05
2.3E-05
0.004
5.5E-05
1.4E-04
3.3E-04
2.4E-04
1.8E-05
1.8E-05
2.5E-04
2.5E-04
2.4E-04
8.3E-05
0.001
5.7E-05
4.8E-04
6.5E-05
6.5E-05
6.5E-05
3.7E-05
1.1E-04
3.5E-05
1 .6E-05
7.8E-06
0.010
-1.3E-04
State of
HGP
(tpy)
Total
(tpy)
0.202
0.145
0.151
0.064
0.058
0.140
0.027
0.026
0.010
0.017
0.071
0.017
0.030
0.958
0.020
0.050
0.120
0.140
0.010
0.010
0.140
0.040
0.120
0.130
0.030
0.030
0.008
0.010
0.010
0.010
0.010
0.030
0.020
0.020
0.010
1.080
0.010
Iowa
Total
(tpy)
C-6
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
FIPS
Facility Name
Plant
ID
Stk
ID
Unit
ID
StkHt
(m)
Diam
(m)
Temp
(DegK)
Vel
(m/s)
HGO
(tpy)
HG2
(tpy)
HGP
(tpy)
Total
(tpy)
Original Emissions
City Of Cedar Rapids
WPCF
Adm Corn Processing
- Clinton
19045
19169
19041
Total
Revised Emissions
Iowa State University
Spencer Hospital
19113
19045
City Of Cedar Rapids
WPCF
Adm Corn Processing
- Clinton
Iowa State University
Spencer Hospital
19169
19041
Total
Total Emissions Changes
Emissions Changes
12181
15091
B25
M104
12181
15091
B25
M104
001
001
001
001
17.3
43.1
26.9
2.6
16.0
91.4
26.9
2.4
0.73
2.01
1.16
0.14
1.59
3.96
1.16
0.46
433
461
457
855
323
453
457
1255
10.1
10.3
9.3
1.8
14.7
19.3
9.3
14.8
0.013 0.008 0.005 0.026
0.003 0.002 0.001 0.005
2.3E-04 8.0E-05 8.8E-05 4.0E-04
9.8E-07 1.5E-05 3.9E-06 2.0E-05
0.016 0.009 0.006 0.031
0.013
0.005
0.006
0.000
0.024
0.008
0.008
0.003
0.002
0.007
0.020
0.011
0.005
0.002
0.002
0.002
0.011
0.005
0.026
0.010
0.010
0.010
0.056
0.024
C-7
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
New Jersey
Table C-7. Summary of Emissions Revisions made to IPM Sources in State of New Jersey
Facility Name
HGO
(tPV)
HG2
(tpy)
HGP
(tpy)
Total
(tpy)
Original Emissions
B L England
Hudson
Mercer
Deepwater
Logan Generating Company - L.P.
Chambers Cogeneration - L.P.
Total
Revised Emissions
B L England
Hudson
Mercer
Deepwater
Logan Generating Company - L.P.
Chambers Cogeneration - L.P.
Total
Total Emissions Changes for IPM
Emissions Changes
0.026
0.014
0.004
0.001
0.001
0.001
0.047
0.094
0.011
0.030
0.002
0.001
0.010
0.147
0.100
0.004
0.037
0.002
0.001
0.001
0.001
0.046
0.016
0.028
0.015
0.004
0.000
0.006
0.070
0.024
0.001
0.003
0.002
0.000
0.001
0.000
0.007
0.004
0.003
0.011
0.000
0.000
0.004
0.023
0.016
0.032
0.054
0.008
0.002
0.003
0.002
0.100
0.114
0.041
0.057
0.006
0.002
0.021
0.240
0.139
C-8
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-8. Summary of Emissions Revisions made to Non-IPM Sources in State of New Jersey
Facility Name
Original Emissions
Co Steel Raritan
Atlantics Tales Cast Iron Pipe
U.S. Pipe & Fndy. Co.
Co Steel Sayreville
Essex Co. RRF
Camden RRF
Union Co. RRF
Gloucester County
Warren Energy RF
Howarddown
Hoeganese
Camden Co. Muassi
Stonybrook Regional Sewerage Authority
Bayshore Regional Sewerage Authority
Somerset Raritanvalley Sewerage Authority
Northwest Bergen County Utilities Authority
Parsippany - Troyhills Township WWTP
Atlantic County Utilities Authority
Gloucester County Utilities Authority
Total
Revised Emissions
Co Steel Raritan
Atlantics Tales Cast Iron Pipe
U.S. Pipe & Fndy. Co.
Co Steel Sayreville
Essex Co. RRF
Camden RRF
Union Co. RRF
Gloucester County
Warren Energy RF
Howarddown
Hoeganese
Camden Co. Muassi
Stonybrook Regional Sewerage Authority
Bayshore Regional Sewerage Authority
Somerset Raritanvalley Sewerage Authority
Northwest Bergen County Utilities Authority
Parsippany - Troyhills Township WWTP
Atlantic County Utilities Authority
Gloucester County Utilities Authority
Total
Total Emissions Changes for Non-IPM
Emissions Changes
HGO
(tpy)
0.049
0.039
0.053
0.164
0.034
0.021
0.003
0.002
0.001
0.001
0.005
0.026
0.023
0.008
0.007
0.006
0.004
0.003
0.002
0.449
0.090
0.033
0.019
0.178
0.047
0.011
0.003
0.002
0.001
0.002
0.005
0.005
0.011
0.004
0.007
0.005
0.004
0.003
0.001
0.429
-0.020
HG2
(tpy)
0.030
0.005
0.032
0.020
0.090
0.054
0.007
0.005
0.001
0.000
0.003
0.016
0.014
0.005
0.004
0.004
0.003
0.002
0.001
0.295
0.011
0.004
0.011
0.022
0.123
0.029
0.008
0.005
0.001
0.001
0.003
0.003
0.007
0.002
0.004
0.003
0.003
0.002
0.001
0.243
-0.052
HGP
(tpy)
0.020
0.005
0.021
0.020
0.031
0.019
0.002
0.002
0.000
0.000
0.002
0.011
0.009
0.003
0.003
0.002
0.002
0.001
0.001
0.155
0.011
0.004
0.000
0.022
0.042
0.010
0.003
0.002
0.001
0.001
0.002
0.002
0.005
0.002
0.003
0.002
0.002
0.001
0.000
0.114
-0.041
Total
(tpy)
0.099
0.049
0.107
0.205
0.156
0.093
0.012
0.008
0.002
0.002
0.011
0.053
0.045
0.015
0.015
0.012
0.008
0.005
0.004
0.899
0.112
0.041
0.030
0.223
0.212
0.050
0.014
0.009
0.003
0.003
0.011
0.010
0.023
0.008
0.014
0.010
0.009
0.005
0.002
0.786
-0.113
C-9
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Nevada
Table C-9. Summary of Emissions Revisions Made to Non-IPM Sources in State of Nevada
Facility Name
Gold Mines
Original Emissions
Jerritt Canyon (Main)
Pipeline Mining Operation
Twin Creeks/Newmont Mining Corp.
Barrick Gold Strike Mine
Goldstrike Mine
Newmont Mining Corporation - Carlin South Area
Getchell Gold Corp.
Newmont Gold Co. - Lone Tree Mine
Coeur Rochester Inc.
Total
Revised Emissions
Queenstake Resources USA Inc - Jerritt Canyon Mine
Cortez Gold Mines Mill #2 (aka Pipeline Mill)
Newmont Mining Corporation - Twin Creeks Mine
Barrick Goldstrike Mines, Inc
Newmont Mining Corporation - Lone Tree Mine
Newmont Mining Corporation - Gold Quarry Operations
Newmont Midas Operations
Coeur Rochester, Inc
Florida Canyon Mine
Glamis Marigold Mine
Round Mountain Gold Corporation - Smokey Valley Common Operation
Homestoke Mining Corporation - Ruby Hill Mine
Bald Mountain Mine (including Mooney Basin Operation)
Denton - Rawhide Mine
Placer Turquoise Ridge, Inc
Battle Mountain Gold Company
Total
Cement Plant
Revised Emissions
Nevada Cement Co.
Total Emissions Changes for Non-IPM
Emissions Changes
HGO
(tpy)
6.780
2.284
1.370
0.706
0.042
0.090
0.002
0.001
0.002
11.277
0.121
0.056
0.194
0.282
0.269
0.106
0.003
0.001
0.078
0.397
0.023
7.5E-03
0.066
0.062
0.002
0.001
1.667
0.007
1.674
HG2 HGP Total
(tpy) (tpy) (tpy)
0.000
0.000
0.000
0.000
0.000
0.054
0.001
4.5E-04
0.001
0.057
0.719
0.026
0.018
0.023
0.038
0.045
0.005
0.001
0.134
0.054
0.005
0.006
0.034
0.107
0.003
1.3E-04
1.218
0.001
1.220
0.000
0.000
0.000
0.000
0.000
0.036
0.001
3.0E-04
0.001
0.038
0.018
0.002
0.006
0.007
0.005
0.008
3.1E-04
5.5E-05
8.4E-03
0.005
3.5E-04
4.0E-04
0.002
0.007
2.0E-04
8.6E-06
0.070
0.001
0.071
6.780
2.284
1.370
0.706
0.042
0.180
0.004
0.002
0.004
11.371
0.858
0.083
0.218
0.312
0.312
0.159
0.009
0.001
0.220
0.455
0.029
0.014
0.102
0.176
0.005
0.001
2.955
0.010
2.965
C-10
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-10. Summary of Origin of Data for Mercury Emissions and Speciation for Nevada Gold Mines
Facility Name Process
Newmont Lone Tree Mine Laboratory Drying Oven
Laboratory Fire Assay Furnace
Laboratory LECO Furnace
Phase I Lime Slaker
Autoclave
Carbon Kiln
Electrowinning Cells
Pregnant and Barren Solution Tanks
Newmont Gold Quarry Operations ROTP Dry Grinding Static Separator
ROTP Ore Preheaters
ROTP Ore Roasters
ROTP Calcine Quench (N & S)
AARL Carbon Kiln
AARL Carbon Kiln combustion
Zadra Carbon Kiln
Zadra Carbon Kiln combustion
AARL Carbon Stripping Tanks(solution tank 1)
AARL Carbon Stripping Tanks(solution tank 2 and 3)
Refinery Retorts
Refinery Induction & Four Furnaces
Refinery Barren Tank & Electrowinning (EW) Cells
Integrated Lab Furnaces (Lines 1 & 2)
Manual Lab Furnaces
Integrated Lab Grieve Drying Ovens
Newmont Twin Creeks Mine Aotoclave Phase 1
Aotoclave Phase 2
Juniper Carbon Kiln
Electrowinning Cells Exhaust
Juniper Hg Retort A
Speciation
Total Emissions of Hg
0.001
4.0E-06
1.5E-06
5.0E-04
0.018
0.286
0.006
0.001
3.0E-04
2.6E-02
1.7E-03
7.0E-02
3.6E-02
6.6E-04
8.6E-03
1.6E-04
3.5E-03
1.6E-03
4.6E-03
3.1E-03
2.7E-03
5.0E-05
6.5E-05
5.0E-05
0.077
0.032
0.012
0.021
0.000
HGO
35.0%
35.0%
35.0%
0.0%
78.10%
88.2%
35.0%
35.0%
0.0%
99.5%
57.2%
28.0%
99.0%
99.0%
99.0%
98.4%
95.7%
85.7%
98.2%
97.2%
97.7%
97.7%
97.7%
35.0%
97.9%
97.9%
13.3%
99.1%
99.1%
HG2
61.2%
61.2%
61.2%
0.0%
8.30%
11.0%
61.2%
61.2%
0.0%
0.3%
38.7%
61.0%
0.9%
0.9%
0.9%
0.0%
1.9%
13.4%
1.6%
2.5%
2.1%
2.1%
2.1%
61.2%
1.6%
1.6%
57.6%
0.8%
0.8%
HGP
3.8%
3.8%
3.8%
100.0%
13.60%
0.7%
3.8%
3.8%
100.0%
0.2%
4.0%
11.0%
0.1%
0.1%
0.1%
1.6%
2.4%
0.8%
0.3%
0.3%
0.3%
0.3%
0.3%
3.8%
0.5%
0.5%
29.1%
0.0%
0.0%
Notes
1
4
4
4
1
1
1
1
1
1,2
1,2
1
1,2
1,2
1,2
1,2
2
2
1,2
1,2
1
1
1
1
1
1
1
1
1
C-11
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Facility Name Process
Juniper Hg Retort B
Juniper Hg Retort C
Juniper Hg Retort D
Juniper Induction Furnaces
Pregnant/Barren solution tanks
Lab Sample Drying Ovens (2)
Lab Assay Furnaces (5)
Phase 1 Lime Verti-mill
Phase 2 Lime Verti-mill
Newmont Midas Operations Retort 1
Retort 2
Smelting Furnace
Barrick Goldstrike Mines, Inc Autoclave/Refinery
Roaster
Carbon Regeneration Kiln
Retort 1-3
West Smelting Furnace, East Smelting Furnace, Electrowinning
Retort Room
Assay & Met Laboratory
AA Machine
Coeur Rochester, Inc Smelting Furnace
Assay Lab
Florida Canyon Mine Retort
Smelting Furnace
Assay Lab
Electrowinning
Carbon Kiln
Glamis Marigold Mine Electrowinning
Speciation
Total Emissions of Hg
0.000
0.000
0.000
0.000
0.073
0.002
5.0E-04
5.0E-04
5.0E-04
4.7E-05
3.1E-04
8.2E-03
0.005
0.014
0.007
0.015
0.117
0.124
0.002
0.015
0.0036792
0.008
0.000
0.001
4.0E-06
0.001
0.025
4.3E-06
0.013
0.181
0.004
HGO
99.1%
99.1%
99.1%
99.1%
88.0%
35%
35%
0.0%
0.0%
99.1%
99.1%
35%
78.0%
78.0%
78.0%
78.0%
90.1%
98.3%
98.8%
99.1%
53.6%
35.0%
35.0%
35.0%
35.0%
99.1%
35.0%
35.0%
35.0%
35.0%
35.0%
HG2
0.8%
0.8%
0.8%
0.8%
11.0%
61.20%
61.20%
0.0%
0.0%
0.8%
0.8%
61.20%
8.3%
8.3%
8.3%
8.3%
9.5%
1.6%
0.2%
0.8%
31.0%
61.2%
61.2%
61.2%
61.2%
0.8%
61.2%
61.2%
61.2%
61.2%
61.2%
HGP
0.0%
0.0%
0.0%
0.0%
1.0%
3.80%
3.80%
100.0%
100.0%
0.0%
0.0%
3.80%
13.8%
13.8%
13.8%
13.8%
0.4%
0.2%
1.0%
0.0%
15.5%
3.8%
3.8%
3.8%
3.8%
0.0%
3.8%
3.8%
3.8%
3.8%
3.8%
Notes
1
1
1
1
1
1
4
4
4
1
1
1
1,2
1,2
1,2
1,2
1,2
1,2
1,2
1,2
2
1
1
1
1
1
1
1
1
1
1
C-12
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Facility Name
Total Annual ems (calculated)
Total Annual ems (documented)
Queenstake Resources USA Inc.
Jerritt Canyon Mine (QRJC)
Round Mountain Gold Corporation
Cortez Gold Mines Mill #2
(aka Pipeline Mill)
Process
Retort
Carbon Regeneration Kiln
Electro-Winning Circuit (3 Cells)
Smelting Furnace
Pregnant Tank
Barren Tank
Assay Lab (2 Drying Ovens)
Assay Lab (Atomic Adsorption Analytical Instrument)
Assay Lab (2 Assay Furnaces)
West Roaster
East Roaster
Carbon Regeneration Kiln
Ore Dryer
Refining Process
Large Ore Drying Ovens (5 units)
Laboratory Small Ore Dryer
Laboratory Hot Plates (2 units)
Carbon Regeneration Kiln
Electric Induction Furnace
Other Thermal Processes
Refinery Electrowining Vent
Assay Lab
High Grade Area
Carbon Regeneration Kiln (2)
Smelting Furnaces (2)
EW System #1
EW System #2
Assay Laboratory Furnace Baghouse
Gold Sluge Drying Oven
Speciation
Total Emissions of Hg
3.8E-04
0.447
0.0000011
0.001
0.001
0.001
1.1E-06
2.5E-07
7.5E-07
6.2E-01
2.3E-01
0.001
0.007
0.001
0.016
3.4E-04
1.4E-06
0.016
0.005
8.5E-06
0.007
7.0E-06
1.7E-06
0.046
0.028
0.003
0.001
0.002
0.002
HGO
35.0%
88.2%
97.7%
35.0%
35.0%
35.0%
35.0%
35.0%
35.0%
16.8%
1.8%
37.3%
99.5%
35.0%
35.0%
35.0%
35.0%
98.3%
88.0%
35.0%
35.0%
35.0%
35.0%
88.2%
35.0%
98.2%
92.9%
35.0%
35.0%
HG2
61.2%
11.0%
2.1%
61.2%
61.2%
61.2%
61.2%
61.2%
61.2%
78.9%
95.7%
62.5%
0.3%
61.2%
61.2%
61.2%
61.2%
1.5%
11.0%
61.2%
61.2%
61.2%
61.2%
11.0%
61.2%
0.5%
1.9%
61.2%
61.2%
HGP
3.8%
1.0%
0.3%
3.8%
3.8%
3.8%
3.8%
3.8%
3.8%
2.5%
0.9%
0.2%
0.2%
3.8%
3.8%
3.8%
3.8%
0.2%
1.0%
3.8%
3.8%
3.8%
3.8%
0.7%
3.8%
1.3%
5.2%
3.8%
3.8%
Notes
1
1
1
1
1
1
1
1
1,3
1
1
1
1
1
1
1
4
1
1
1
1,2
1,2
1,2
1,2
1
1
C-13
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Facility Name Process
Assay Laboratory, Atomic Absorption Spectrometers
Assay Laboratory, Drying Ovens
Assay Laboratory, Furnaces
Assay Laboratory, Hotplates
Strip Circuit Area, AA Machine
Homestake Mining Corporation- Smelting Furnace
Ruby Hill Mine Electric Carbon Kiln
Electric Mercury Retort
Electrowinning Cells
Bald Mountain Mine Carbon Regeneration Kiln
(including Mooney Basin Operation) Retort
Smelting Furnace
Fire Assay Lab
Total Annual ems (calculated)
Total Annual ems (documented)
Denton - Rawhide Mine Retort
Electrowinning Circuit
Refinery Furnace Baghouse
Fire Assay Lab Furnace Baghouse
Placer Turquoise Ridge, Inc. Lab Fire Assay/Furnaces
Drying Room
Drying Ovens
Annealing Furnaces
Hotplates
Battle Mountain Gold Company- Electric Carbon Kiln
Reona and Phoenix Projects Electrowining Cells
Pregnant & Barren Solution Vent
Notes:
Speciation
Total Emissions of Hg
6.5E-07
0.002
6.5E-06
8.0E-07
5.0E-08
0.000
0.005
0.00000025
0.010
0.048
1.4E-04
0.054
9.3E-06
0.102
0.001
0.048
0.127
2.5E-07
1.8E-04
0.004
0.001
4.5E-07
4.7E-06
0.001
1.5E-05
8.5E-07
HGO
35.0%
35.0%
35.0%
35.0%
35.0%
35.0%
88.20%
35.0%
35.0%
98.3%
99.1%
35.0%
35.0%
99.1%
35.0%
35.0%
35.0%
35.0%
35.0%
35.0%
35.0%
35.0%
88.20%
35.0%
35.0%
HG2
61.2%
61.2%
61.2%
61.2%
61.2%
61.2%
11%
61.2%
61.2%
1.5%
0.8%
61.2%
61.2%
0.8%
61.2%
61.2%
61.2%
61.2%
61.2%
61.2%
61.2%
61.2%
11%
61.2%
61.2%
HGP
3.8%
3.8%
3.8%
3.8%
3.8%
3.8%
0.70%
3.8%
3.8%
0.2%
0.0%
3.8%
3.8%
0.0%
3.8%
3.8%
3.8%
3.8%
3.8%
3.8%
3.8%
3.8%
0.70%
3.8%
3.8%
Notes
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
C-14
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
(1) Emissions from "2006 Cumulative Mercury Emissions Data Submittal Public.xls", supplied by NDEP (NDEP, 2007), which represents calendar year 2006 emissions. Only
some of the emissions in this document were based on stack testing. The rest are engineering estimates. The Nevada Control Program required all sources to do emissions
testing in 2007.
(2) Speciation from NDEP 2006 Tier-1 Test Data in document "VMRP testing Overview6.pdf supplied by NDEP (NDEP, 2007)
(3) Speciation from Ontario Hydro test results as reported in Western Environmental Services and Testing, Inc. 2006.
(4) Emissions from NDEP. 2006. Precious Metal Mining - Mercury Air Emissions Questionnaire, Nevada Division of Environmental Protection. March 2006
C-15 August 2008
-------
Table C-11. Summary of Emissions Revisions Made to IPM Sources in State of Nevada
Facility Name
Original Emissions
Mohave
Revised Emissions
Mohave
Total Emissions Changes for IPM
Emissions Changes
HGO
(tpy)
0.029
0.032
0.003
HG2
(tpy)
0.078
0.008
-0.070
HGP
(tpy)
0.007
0.000
-0.007
Total
(tpy)
0.115
0.041
-0.074
Utah
Table C-12. Summary of Emissions Revisions Made to Non-IPM Sources in State of Utah
Facility Name
Original Emissions
Nucor Steel
Wasatch Energy Systems
Holcim (US) Inc.
Tesoro West Coast
Kennecott Utah Copper Corporation - Smelter, refinery
Kennecott Utah Copper Corporation
Staker & Parson Companies
Utelite Corporation
Clean Harbors Aragonite LLC
Laidlaw Environmental Services - Inc.
Us Gypsum Sigurd Plant
Pacif Corp Gadsby
Alliant Tech Systems Inc. Bacchus Works
Alliant Tech Systems Plant 2
Hill Air Force Main Base
Total
Revised Emissions
Nucor Steel
Wasatch Energy Systems
Ash Grove Cement Company
Graymont Western US Incorporated
Holcim (US) Inc.
Tesoro West Coast
Kennecott Utah Copper Corporation - Smelter, refinery
Kennecott Utah Copper Corporation
Staker & Parson Companies
Utelite Corporation
Clean Harbors Aragonite LLC
Deseret Chemical Depot
Laidlaw Environmental Services- Inc.
Us Gypsum Sigurd Plant
HGO
(tpy)
1.25E-06
0.022
0.005
0.003
0.004
0.028
8.0E-08
2.1E-07
0.272
0.048
0.002
0.004
0.001
0.001
0.003
0.394
0.059
0.008
0.044
0.004
0.005
0.007
0.004
0.010
0.001
0.002
0.016
0.004
0
0
HG2
(tpy)
7.5E-07
0.058
0.003
0.001
0.001
0.005
5.0E-08
1.3E-07
0.094
0.016
0.001
0.003
0.001
0.001
0.002
0.184
0.007
0.021
0.015
0.001
0.002
0.001
4.7E-04
0.016
9.1E-05
0.001
0.005
0.003
0
0
HGP
(tpy)
5E-07
0.020
0.002
0.000
0.001
0.005
3.0E-08
8.0E-08
0.103
0.018
0.001
0.002
4.5E-04
3.5E-04
0.001
0.153
0.007
0.007
0.017
0.001
0.002
0.001
4.7E-04
0.001
9.1E-05
0.001
0.006
0.002
0
0
Total
(tpy)
2.5E-06
0.100
0.009
0.004
0.005
0.038
1.6E-07
4.2E-07
0.469
0.082
0.005
0.009
0.002
0.002
0.006
0.731
0.074
0.037
0.076
0.005
0.009
0.008
0.005
0.027
0.001
0.005
0.027
0.008
0
0
C-16
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Facility Name
Pacif Corp Gadsby
Alliant Tech Systems Inc. Bacchus Works
Alliant Tech Systems Plant 2
Hill Air Force Main Base
Total
Total Emissions Changes for Non-IPM
Emissions Changes
HGO
(tpy)
0
0
0
0
0.165
-0.229
HG2
(tpy)
0
0
0
0
0.073
-0.112
HGP
(tpy)
0
0
0
0
0.045
-0.108
Total
(tpy)
0
0
0
0
0.283
-0.448
Table C-13. Summary of Emissions Revisions Made to IPM Sources in State of Utah
Facility Name
Original Emissions
Pacificorp - Carbon Power Plant
Sunnyside Cogeneration Associates
Pacificorp - Hunter Power Plant
Pacificorp - Huntington Power Plant
Intermountain Power Services Corp
Total
Revised Emissions
Pacificorp - Carbon Power Plant
Sunnyside Cogeneration Associates
Pacificorp - Hunter Power Plant
Pacificorp - Huntington Power Plant
Intermountain Power Services Corp
Total
Total Emissions Changes for IPM
Emissions Changes
HGO
(tpy)
0.005
0.000
0.037
0.031
0.003
0.076
0.007
0.000
0.172
0.050
0.069
0.298
0.222
HG2
(tpy)
0.014
0.000
0.004
0.040
0.001
0.059
0.019
0.000
0.017
0.065
0.037
0.139
0.080
HGP
(tpy)
0.001
0.000
0.000
0.003
0.000
0.005
0.002
0.000
0.002
0.006
0.007
0.016
0.011
Total
(tpy)
0.020
0.000
0.041
0.074
0.005
0.140
0.028
0.001
0.191
0.121
0.113
0.453
0.313
Table C-14. Summary of Emissions Revisions Made to MWI Sources in State of Utah
Facility Name
HGO
(tpy)
HG2
(tpy)
HGP
(tpy)
Total
(tpy)
Original Emissions
Stericycle Incorporated
Revised Emissions
Stericycle Incorporated
Total Emissions Changes for MWI
Emissions Changes
3.1E-04 0.005 1.2E-03 0.006
3.4E-04 0.005 1.4E-03 0.007
3.4E-05 0.001 1.4E-04 0.001
C-17
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Breakdown of Collective Sources
In the following tables, collective sources are broken down for those states where the maximum in-state contribution was due to
collective sources. In each case, only those point sources that were within a two grid cell range of the location of the maximum
impact shown in Section 7 are included. Area sources included in the collective sources tag are presented in a separate table for
counties in the immediate vicinity of the maximum impact. States are presented in alphabetical order.
For California, which included more than 600 smaller point sources in the collective sources category in the vicinity of the maximum,
the top 34 point sources are tabulated. These 34 sources account for more than 90% of the collective source emissions in the vicinity
of the maximum.
C-18 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-15a. Point Sources Included in the Collective Sources Tag for California.
Only the top 34 point sources in the vicinity of the maximum in-state contribution to mercury deposition are included. These sources account for
more than 90% of the collective sources point source emissions in the vicinity of the maximum.
FIPS
06037
06059
06037
06037
06037
06037
06037
06037
06037
06037
06037
06037
06037
06059
06059
06037
06059
06059
06059
06037
06037
06037
06037
06037
06059
County
Los Angeles
Orange
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Orange
Orange
Los Angeles
Orange
Orange
Orange
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Orange
Facility Name
OLDQUAKERPAINTCOMPANY
COSMOTRONICCOCORP
QUEMETCOINC
SHELLOILCO(EISUSEONLY)
SHELLOILCO(EISUSEONLY)
DAICOINDINC
CommerceRefuse-to-EnergyFac.
UltramarDiamondShamrock
QUEMETCOINC
CHAPELOFTHEPINES
SULLYMILLERCONTRACTINGCO
LIFEPAINTCO
CONSOLIDATEDDRUMRECONDITIONI
BORALRESOURCESINC-IRVINEPL
LOMAVISTAMEMPARK
EVERGREENMEMORIALCAREINC-EV
UNIONOILCO-SCIENCE&TECHD
SENTINELCREMATIONSOCIETIESI
SENTINELCREMATIONSOCIETIESI
BLUEDIAMONDMATERIALS-SUNVA
EquilonfformerlyTexacoRefining&Marketing
INDUSTRIALASPHALT
BP(formerlyAtlanticRichfieldCo.(ARCO))
AshlandChemicalCompany
UNIONOILCO-SCIENCE&TECHD
Plant
ID
19102621976
30102612583
1910268547
191026800116
191026800116
19102644023
LMWC-2
ESD032
1910268547
19102621472
19102634055
19102618990
19102615490
30102650079
3010266999
191026100800
30102613979
30102610472
30102610472
19102619390
ESD019
19102621395
ESD014
704
30102613979
Point
ID
1
70102
70001
70120
70122
70002
LMWC-1
CCUJ
70001
70001
70001
70001
70001
70001
1
1
70018
70001
70003
70002
CCUJ
70002
CCUJ
1
70019
Stack
ID
3560
4577
3722
3639
3640
3583
4253
3918
3723
3559
3577
3555
3547
4613
4616
3537
4585
4574
4575
3557
3814
3558
3737
3732
4586
Latitude
33.8358
33.6936
34.0199
33.8092
33.8092
33.8535
33.9966
33.7836
34.0199
34.0440
34.0999
33.9209
33.9799
33.6936
33.8985
34.0411
33.9090
33.7924
33.7924
34.0999
33.7948
34.1303
33.8163
33.9875
33.9090
Longitude
-118.2630
-117.8307
-117.9855
-118.2423
-118.2423
-118.2322
-118.1509
-118.2315
-117.9855
-118.2999
-117.9322
-118.0601
-118.1258
-117.8307
-117.9300
-118.2009
-117.8544
-117.8965
-117.8965
-117.9322
-118.2291
-117.9326
-118.2449
-118.1389
-117.8544
StkHt
(m)
3.0
3.0
3.0
3.0
3.0
3.0
34.7
42.3
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
42.3
3.0
42.3
26.9
3.0
Diam
(m)
0.00
0.00
0.00
0.00
0.00
0.00
1.98
2.16
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
2.16
0.00
2.16
1.16
0.00
Temp
(K)
295
295
295
295
295
295
422
505
295
295
295
295
295
295
295
295
295
295
295
295
505
295
505
457
295
Vel
(m/s)
0.0
0.0
0.0
0.0
0.0
0.0
12.3
16.2
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
16.2
0.0
16.2
9.3
0.0
HGO (tpy)
5.67E-02
1.80E-02
1.80E-02
8.66E-03
8.57E-03
O.OOE+00
3.52E-03
8.40E-03
7.60E-03
3.73E-04
4.37E-03
4.28E-03
2.37E-03
3.60E-03
1.91E-03
1.81E-03
1.75E-03
1.74E-04
1.74E-04
2.35E-03
1.95E-03
1.60E-03
1.39E-03
9.30E-04
7.50E-04
HG2(tpy)
7.09E-03
1.08E-02
2.25E-03
5.20E-03
5.14E-03
1.68E-02
9.28E-03
1.05E-03
9.50E-04
5.59E-03
5.46E-04
5.36E-04
1.42E-03
4.50E-04
1.15E-03
1.09E-03
1.05E-03
2.61 E-03
2.61 E-03
2.94E-04
2.43E-04
2.00E-04
1.74E-04
3.21 E-04
4.50E-04
HGP
(tpy)
7.09E-03
7.19E-03
2.25E-03
3.46E-03
3.43E-03
O.OOE+00
3.20E-03
1.05E-03
9.50E-04
1.49E-03
5.46E-04
5.36E-04
9.46E-04
4.50E-04
7.65E-04
7.26E-04
7.00E-04
6.95E-04
6.95E-04
2.94E-04
2.43E-04
2.00E-04
1.74E-04
3.53E-04
3.00E-04
Total Hg
»y)
7.09E-02
3.60E-02
2.25E-02
1.73E-02
1.71E-02
1.68E-02
1.60E-02
1.05E-02
9.50E-03
7.45E-03
5.46E-03
5.36E-03
4.73E-03
4.50E-03
3.82E-03
3.63E-03
3.50E-03
3.47E-03
3.47E-03
2.94E-03
2.43E-03
2.00E-03
1.74E-03
1.60E-03
1.50E-03
C-19
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
06037
06037
06037
06037
06037
06037
06037
06037
06037
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
Los Angeles
BLUEDIAMONDMATERIALS-GARDEN
BLUEDIAMONDMATERIALS-SUNVA
LA.COUNTYDEPT.OFPUBLICWO
ARCOCQCKILN
LACODEPTHEALTHSRV-UCLAHAR
LIVEOAKMEMORIALPARK
LIVEOAKMEMORIALPARK
LIVEOAKMEMORIALPARK
UNIVERSITYSOCALIFORNIA-HEALT
1910266578
19102619390
060370354
19102647232
191026457
19102622839
19102622839
19102622839
19102656
70001
70001
A185
70001
70010
70001
70002
70003
72011
3621
3556
3523
3595
3594
3566
3567
3568
3611
34.0999
34.0999
34.0851
33.7699
33.8312
34.1316
34.1316
34.1316
34.0599
-117.9322
-117.9322
-118.1510
-118.2202
-118.2967
-117.9976
-117.9976
-117.9976
-118.2027
3.0
3.0
12.3
3.0
3.0
3.0
3.0
3.0
3.0
0.00
0.00
1.04
0.00
0.00
0.00
0.00
0.00
0.00
295
295
595
295
295
295
295
295
295
0.0
0.0
25.6
0.0
0.0
0.0
0.0
0.0
0.0
1.20E-03
1.17E-03
6.70E-04
9.60E-04
O.OOE+00
4.25E-05
4.25E-05
4.25E-05
3.46E-04
1.50E-04
1.47E-04
4.02E-04
1.20E-04
1.11E-03
6.38E-04
6.38E-04
6.38E-04
2.08E-04
1.50E-04
1.47E-04
2.68E-04
1.20E-04
O.OOE+00
1.70E-04
1.70E-04
1.70E-04
1.39E-04
1.50E-03
1.47E-03
1.34E-03
1.20E-03
1.11E-03
8.50E-04
8.50E-04
8.50E-04
6.93E-04
C-20
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-15b. Non-point Sources Included in the Collective Sources Tag for California.
Counties in the vicinity of the maximum in-state contribution to mercury deposition with significant non-point emissions are included.
FIPS
06037
06037
06059
06059
06037
06037
06059
06059
06037
06059
06059
06059
06037
06059
06037
06059
06037
06059
County
LOS
ANGELES
LOS
ANGELES
ORANGE
ORANGE
LOS
ANGELES
LOS
ANGELES
ORANGE
ORANGE
LOS
ANGELES
ORANGE
ORANGE
ORANGE
LOS
ANGELES
ORANGE
LOS
ANGELES
ORANGE
LOS
ANGELES
ORANGE
Source Description
Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, Dental Alloy (Mercury
Amalgams) Production
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, Dental Alloy (Mercury
Amalgams) Production
Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
Miscellaneous Area Sources, Other Combustion, Cremation
External Combustion Boilers, Industrial, Liquid Waste
Miscellaneous Area Sources, Other Combustion, Cremation
External Combustion Boilers, Industrial, Liquid Waste
Stationary Source Fuel Combustion, Commercial/Institutional, Residual Oil
Stationary Source Fuel Combustion, Industrial, Residual Oil
External Combustion Boilers, Industrial, Distillate Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Residual Oil
Miscellaneous Area Sources, Other Combustion, Motor Vehicle Fires
Miscellaneous Area Sources, Other Combustion, Motor Vehicle Fires
Industrial Processes, Chemical Manufacturing, Other Not Classified
Industrial Processes, Chemical Manufacturing, Other Not Classified
Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling
Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling
HGO
(tpy)
3.25E-02
1.54E-02
1.40E-02
9.62E-03
1.28E-03
2.09E-03
3.80E-04
7.40E-04
6.10E-04
5.41 E-04
3.52E-04
2.12E-04
9.07E-05
1.85E-05
1.84E-05
4.18E-06
3.08E-06
9.10E-07
HG2
(tpy)
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
3.38E-03
1.25E-03
1.00E-03
4.44E-04
3.66E-04
3.24E-04
2.11 E-04
1.27E-04
2.39E-04
4.87E-05
2.30E-06
5.20E-07
O.OOE+00
O.OOE+00
HGP
(tpy)
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
1.17E-03
8.35E-04
3.45E-04
2.96E-04
2.44E-04
2.16E-04
1.41 E-04
8.49E-05
8.24E-05
1.68E-05
2.30E-06
5.20E-07
O.OOE+00
O.OOE+00
Total Hg
3.25E-02
1.54E-02
1.40E-02
9.62E-03
5.83E-03
4.17E-03
1.73E-03
1.48E-03
1.22E-03
1.08E-03
7.04E-04
4.24E-04
4.12E-04
8.40E-05
2.30E-05
5.22E-06
3.08E-06
9.10E-07
C-21
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-16a. Point Sources Included in the Collective Sources Tag for District
of Columbia.
Only sources in the vicinity of the maximum in-state contribution to mercury deposition are included.
FIPS
11001
11001
11001
11001
11001
County
Dist.
Columbia
Dist.
Columbia
Dist.
Columbia
Dist.
Columbia
Dist.
Columbia
Facility Name Plant Point Stack Latitude Longitude StkHt Diam Temp
ID ID ID (m) (m) (K)
PWCWashington 11001F001 A812 9339 38.8834 -77.0147 8.7 0.46 493
BollingAFB 11001F002 A809 9340 38.8834 -77.0147 12.3 1.04 595
WALTERREEDARMYMEDICALCENTER 11001F003 A813 9341 38.8834 -77.0147 13.0 0.82 404
FORTLESLEYJ.MCNAIR 11001F004 A810 9342 38.8834 -77.0147 13.0 0.82 404
NavalResearchLaboratory 110010033 A811 9338 38.8194 -77.0214 8.7 0.46 493
Vel HGO HG2
(m/s) (tpy) (tpy)
19.2 5.70E-06 3.42E-06
25.6 1.73E-05 1.04E-05
8.8 1.11E-05 6.64E-06
8.8 3.44E-06 2.06E-06
19.2 2.74E-05 1.64E-05
HGP
(tpy)
2.28E-06
6.93E-06
4.43E-06
1.38E-06
1.10E-05
Total Hg
(tpy)
1.14E-05
3.46E-05
2.21 E-05
6.88E-06
5.48E-05
Table C-16b. Non-point Sources Included in the Collective Sources Tag for District of Columbia.
FIPS
11001
11001
11001
11001
11001
11001
11001
11001
11001
11001
11001
11001
County
DIST. COLUMBIA
DIST. COLUMBIA
DIST. COLUMBIA
DIST. COLUMBIA
DIST. COLUMBIA
DIST. COLUMBIA
DIST. COLUMBIA
DIST. COLUMBIA
DIST. COLUMBIA
DIST. COLUMBIA
DIST. COLUMBIA
DIST. COLUMBIA
Source Description
Stationary Source Fuel Combustion, Residential, Distillate Oil
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, Laboratories
Miscellaneous Area Sources, Other Combustion, Cremation
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, Dental Alloy (Mercury Amalgams)
Production
Stationary Source Fuel Combustion, Commercial/Institutional, Bituminous/Subbituminous Coal
External Combustion Boilers, Industrial, Liquid Waste
Stationary Source Fuel Combustion, Industrial, Residual Oil
Stationary Source Fuel Combustion, Residential, Bituminous/Subbituminous Coal
External Combustion Boilers, Industrial, Distillate Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite Coal
Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling
Industrial Processes, Chemical Manufacturing, Other Not Classified
HGO HG2
(tpy) (tpy)
9.17E-04 5.50E-04
1.71E-03 O.OOE+00
4.79E-05 1.26E-04
2.05E-04 O.OOE+00
6.32E-05 3.79E-05
5.33E-05 3.20E-05
3.90E-05 2.34E-05
3.89E-05 2.33E-05
2.54E-05 1.52E-05
8.85E-06 5.31 E-06
1.70E-07 O.OOE+00
2.00E-08 O.OOE+00
HGP
(tpy)
3.67E-04
O.OOE+00
4.36E-05
O.OOE+00
2.53E-05
2.13E-05
1.56E-05
1.56E-05
1.01E-05
3.54E-06
O.OOE+00
O.OOE+00
Total Hg
(tpy)
1.83E-03
1.71E-03
2.18E-04
2.05E-04
1.26E-04
1.07E-04
7.79E-05
7.77E-05
5.07E-05
1.77E-05
1.70E-07
2.00E-08
C-22
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-17a. Point Sources Included in the Collective Sources Tag for Illinois
Only sources in the vicinity of the maximum in-state contribution to mercury deposition are
FIPS
17001
17001
17001
17001
17169
17001
17001
17001
17001
17001
17001
17001
17001
17001
17001
17001
17001
17001
17001
17067
17001
17001
17001
17001
17001
17001
17001
17001
17001
17001
17067
17001
County
Adams
Adams
Adams
Adams
Schuyler
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Hancock
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Hancock
Adams
Facility Name
Blessing Hospital
Diamond Construction Co.
Quincy soy bean co.
Quincy soy bean co.
Freesen incorporated
Blessing hospital
Quincy humane society
Titan wheel international Inc.
Quincy Wilbert vault company
ADM quincy
Blessinghospital
Blessing hospital
Adm quincy
Trinity industries Inc.
Blessing hospital
Blessing hospital
Titan wheel international Inc.
Prairie farms dairy-Quincy milk div.
J.M.Huber corporation
Memorial hospital
J.M.Huber corporation
Adm quincy
Klingele veterinary clinic
J.M.Huber corporation
Titan wheel international Inc.
Adm quincy
Adm quincy
Archer daniels midland company
Adm quincy
J.M.Huber corporation
Memorial hospital
Quincy municipal #4 Landfill
Plant
ID
001065ADQ
001820AAB
15169
15169
169802AAA
001065AAJ
001815ABK
001806AAB
001806AAR
001815AAF
001065AAJ
001065ADQ
001815AAF
001815AAD
001065AAJ
001065ADQ
001806AAB
001065ACS
001815AAS
067025AAL
001815AAS
001815AAF
001065AGA
001815AAS
001806AAB
001815AAF
001815AAF
001806AAM
001815AAF
001815AAS
067025AAL
LF10363
Point
ID
0001
0001
60541
60542
0001
0004
0001
0010
0001
0045
0003
0002
0046
0005
0003
0002
0013
0002
0103
0001
0116
0061
0001
0091
0013
0063
0044
0001
0044
0106
0002
A302
Stack
ID
10671
10713
10715
10716
17522
10668
10712
10679
10693
10699
10667
10673
10700
10695
10666
10672
10680
10670
10707
14787
10708
10702
10677
10706
10681
10701
10697
10692
10698
10705
14788
10717
Latitude
39.9492
39.9583
39.9056
39.9056
40.2251
39.9364
39.8903
39.9492
39.9564
39.9056
39.9364
39.9492
39.9056
39.9019
39.9364
39.9492
39.9492
39.9358
39.8939
40.4078
39.8939
39.9056
39.9352
39.8939
39.9492
39.9056
39.9056
39.9917
39.9056
39.8939
40.4078
39.9786
Longitude
-91.4071
-91.3747
-91.4100
-91.4100
-90.8605
-91.3990
-91.3964
-91.3692
-91.3491
-91.4100
-91.3990
-91.4071
-91.4100
-91.4089
-91.3990
-91.4071
-91.3692
-91.3869
-91.4103
-91.1306
-91.4103
-91.4100
-91.3415
-91.4103
-91.3692
-91.4100
-91.4100
-91.2335
-91.4100
-91.4103
-91.1306
-91.2111
StkHt
(m)
9.1
9.8
43.1
43.1
9.1
9.1
10.1
63.1
6.1
12.2
10.7
33.5
12.2
6.1
10.7
33.5
63.1
9.4
10.1
7.0
10.1
10.1
5.5
10.1
63.1
17.1
12.2
10.1
12.2
16.8
9.1
3.0
Diam
(m)
1.22
0.10
2.01
2.01
3.05
0.40
0.10
1.22
0.52
1.52
0.10
2.29
1.52
0.10
0.10
2.29
1.22
0.10
0.10
0.24
0.10
0.10
0.10
0.10
1.22
0.10
1.46
0.10
1.46
0.91
0.76
0.00
Temp
(K)
308
294
461
461
339
432
294
585
866
561
294
366
561
294
294
366
585
294
294
700
294
294
294
294
585
294
411
294
411
294
433
295
Vel
(m/s)
0.5
0.1
10.3
10.3
2.1
10.3
0.1
8.1
5.7
2.4
0.1
1.1
2.4
0.1
0.1
1.1
8.1
0.1
0.1
9.3
0.1
0.1
0.1
0.1
8.1
0.1
9.8
0.1
9.8
8.6
4.1
0.0
included.
HGO
(tpy)
0.0481
0.0022
0.0003
0.0003
0.0005
0.0001
6.01 E-05
4.52E-05
1.91 E-05
1.75E-05
1.23E-05
1.23E-05
1.12E-05
8.96E-06
8.35E-06
8.35E-06
8.12E-06
8.11E-06
7.41 E-06
7.01 E-06
6.53E-06
6.27E-06
5.70E-07
5.39E-06
4.91 E-06
4.26E-06
3.32E-06
8.02E-06
2.14E-06
1.87E-06
1.79E-06
4.58E-06
HG2
(tpy)
0.0289
0.0003
0.0002
0.0002
0.0001
0.0001
3.61 E-05
2.71 E-05
1.15E-05
1.05E-05
7.38E-06
7.38E-06
6.75E-06
5.38E-06
5.01 E-06
5.01 E-06
4.87E-06
4.86E-06
4.45E-06
4.21 E-06
3.92E-06
3.76E-06
8.59E-06
3.24E-06
2.95E-06
2.56E-06
1.99E-06
1.00E-06
1.28E-06
1.12E-06
1.07E-06
5.70E-07
HGP
(tpy)
0.0192
0.0003
0.0001
0.0001
0.0001
0.0001
2.41 E-05
1.81 E-05
7.63E-06
6.99E-06
4.92E-06
4.92E-06
4.50E-06
3.58E-06
3.34E-06
3.34E-06
3.25E-06
3.24E-06
2.96E-06
2.81 E-06
2.61 E-06
2.51 E-06
2.29E-06
2.16E-06
1.97E-06
1.70E-06
1.33E-06
1.00E-06
8.60E-07
7.50E-07
7.20E-07
5.70E -07
Total Hg
(tpy)
0.0962
0.0027
0.0005
0.0005
0.0006
0.0003
1.20E-04
9.04E-05
3.82E-05
3.50E-05
2.46E-05
2.46E-05
2.25E-05
1.79E-05
1.67E-05
1.67E-05
1.62E-05
1.62E-05
1.48E-05
1.40E-05
1.31 E-05
1.25E-05
1.15E-05
1.08E-05
9.83E-06
8.52E-06
6.64E-06
1.00E-05
4.28E-06
3.74E-06
3.58E-06
5.72E-06
C-23
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
FIPS
17001
17001
17001
17001
17001
17001
17001
17001
17001
17001
17001
17001
17001
17001
17001
17067
17001
17001
17001
17001
17001
17001
17001
17001
County
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Hancock
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Adams
Facility Name
Titan wheel international Inc.
Gemcity concrete Lie.
Prairie farms dairy-Quincy milk div.
Foam productscorp
Gemcity concrete Lie
Trinity industries-lnc.
Amerencips
Prairie farms dairy inc.-Durst div.
Gardner Denver machinery Inc.
Titan wheel international Inc.
Gardner Denver machinery Inc.
Huckstore fixture inc.
Chesterbross construction co.
Amerencips
Midwest patterns Inc.
Wlmiller company
Archer daniels midland company
Archer daniels midland company
Kuester tool and die inc.
Archer daniels midland company
Archer daniels midland company
Archer daniels midland company
Titan wheel international Inc.
Archer daniels midland company
Plant
ID
001806AAB
001065AEK
001065ACS
001815ABG
001806AAC
001815AAD
001815ABD
001065AEN
001815AAK
001806AAB
001815AAK
001065AFU
001811AAA
001815ABD
001820AAG
067040AAB
001806AAF
001806AAF
001065AGH
001806AAF
001806AAF
001806AAF
001806AAB
001806AAF
Point
ID
0026
0003
0001
0005
0004
0009
0001
0001
0018
0027
0019
0008
0001
0002
0004
0001
0015
0019
0001
0014
0001
0001
0028
0016
Stack
ID
10682
10674
10669
10711
10685
10696
10709
10675
10703
10683
10704
10676
10694
10710
10714
14791
10689
10691
10678
10688
10687
10686
10684
10690
Latitude
39.9492
39.9364
39.9358
39.8784
39.9572
39.9019
39.9459
39.9358
39.9083
39.9492
39.9083
39.9439
40.0000
39.9459
39.9919
40.3842
39.9447
39.9447
39.9597
39.9447
39.9447
39.9447
39.9492
39.9447
Longitude
-91.3692
-91.4150
-91.3869
-91.3983
-91.3653
-91.4089
-91.2913
-91.3777
-91.4111
-91.3692
-91.4111
-91.3703
-91.0000
-91.2913
-91.3961
-91.3575
-91.3669
-91.3669
-91.3774
-91.3669
-91.3669
-91.3669
-91.3692
-91.3669
StkHt
(m)
13.7
6.1
15.8
10.1
6.1
7.6
1.5
3.4
10.1
13.7
10.1
15.1
10.1
1.5
2.1
8.8
15.1
18.3
6.7
16.2
10.4
10.4
13.7
15.1
Diam
(m)
0.67
0.40
0.10
0.10
0.40
0.46
0.10
0.40
0.10
0.67
0.10
0.82
0.10
0.10
0.30
0.85
0.82
0.10
0.30
0.70
0.76
0.76
0.67
0.82
Temp
(K)
383
422
294
294
366
478
294
475
294
383
294
445
294
294
811
450
445
294
978
561
561
561
383
445
Vel
(m/s)
10.1
11.5
0.1
0.1
3.8
2.9
0.1
28.7
0.1
10.1
0.1
7.9
0.1
0.1
1.2
44.5
7.9
0.1
3.7
91.7
77.6
77.6
10.1
7.9
HGO
(tpy)
1.12E-06
9.10E-07
8.80E-07
8.70E-07
7.50E-07
5.90E-07
6.10E-07
4.60E-07
4.00E-07
3.20E-07
3.30E-07
2.80E-07
7.10E-07
1.20E-07
7.00E-08
2.60E-07
5.00E-08
1.28E-09
1.17E-09
3.90E-10
3.77E-10
3.25E-10
1.95E-11
9.75E-12
HG2
(tpy)
6.70E-07
5.50E-07
5.30E-07
5.20E-07
4.50E-07
3.50E-07
3.70E-07
2.80E-07
2.40E-07
1.90E-07
2.00E-07
1.70E-07
9.00E-08
7.00E-08
4.00E-08
3.00E-08
3.00E-08
7.68E-10
7.02E-10
2.34E-10
2.26E-10
1.95E-10
1.17E-11
5.85E-12
HGP
(tpy)
4.50E-07
3.70E-07
3.50E-07
3.50E-07
3.00E-07
2.40E-07
2.40E-07
1.90E-07
1.60E-07
1.30E-07
1.30E-07
1.10E-07
9.00E-08
5.00E-08
3.00E-08
3.00E-08
2.00E-08
5.12E-10
4.68E-10
1.56E-10
1.51E-10
1.30E-10
7.80E-12
3.90E-12
Total Hg
(tpy)
2.24E-06
1.83E-06
1.76E-06
1.74E-06
1.50E-06
1.18E-06
1.22E-06
9.30E-07
8.00E-07
6.40E-07
6.60E-07
5.60E-07
8.90E-07
2.40E-07
1.40E-07
3.20E-07
1.00E-07
2.56E-09
2.34E-09
7.80E-10
7.55E-10
6.50E-10
3.90E-11
1.95E-11
C-24
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-17b. Non-point Sources Included in the Collective Sources Tag for Illinois.
The county in the vicinity of the maximum in-state contribution to mercury deposition with significant non-point emissions is included.
FIPS
County
Source Description
HGO
(tpy)
HG2
(fry)
HGP
(tpy)
Total Hg
(fry)
17001
17001
17001
17001
17001
17001
17001
17001
17001
17001
17001
17001
17001
ADAMS Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
ADAMS Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools,
Laboratories
ADAMS Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, Dental Alloy
(Mercury Amalgams) Production
ADAMS External Combustion Boilers, Industrial, Liquid Waste
ADAMS Stationary Source Fuel Combustion, Industrial, Residual Oil
ADAMS Stationary Source Fuel Combustion, Commercial/Institutional,
Bituminous/Subbituminous Coal
ADAMS Miscellaneous Area Sources, Other Combustion, Cremation
ADAMS Stationary Source Fuel Combustion, Residential, Distillate Oil
ADAMS Stationary Source Fuel Combustion, Commercial/Institutional, Residual Oil
ADAMS Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite Coal
ADAMS Stationary Source Fuel Combustion, Residential, Bituminous/Subbituminous Coal
ADAMS Industrial Processes, Chemical Manufacturing, Other Not Classified
ADAMS Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling
2.33E-04
2.21 E-04
2.05E-04
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00 O.OOE+00
2.33E-04
2.21 E-04
2.05E-04
1.90E-05
1.39E-05
1.33E-05
5.06E-06
9.41 E-06
4.61 E-06
1.86E-06
1.04E-06
2.00E-08
2.00E-08
1.14E-05
8.35E-06
7.96E-06
1.33E-05
5.65E-06
2.77E-06
1.12E-06
6.20E-07
O.OOE+00
O.OOE+00
7.62E-06
5.57E-06
5.31 E-06
4.60E-06
3.76E-06
1.84E-06
7.40E-07
4.20E-07
O.OOE+00
O.OOE+00
3.81 E-05
2.78E-05
2.66E-05
2.30E-05
1.88E-05
9.22E-06
3.72E-06
2.08E-06
2.00E-08
2.00E-08
C-25
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-18a. Point Sources Included in the Collective Sources Tag for Kansas.
Only sources in the vicinity of the maximum in-state contribution to mercury deposition are
FIPS
20205
20205
20001
20001
20001
20001
20133
20205
20133
20207
20205
20133
County
WILSON
WILSON
ALLEN
ALLEN
ALLEN
ALLEN
NEOSHO
WILSON
NEOSHO
WOODSON
WILSON
NEOSHO
Facility Name
Lafarge
Lafarge
MONARCHCEMENTCOMPANY(THE)
MONARCHCEMENTCOMPANY(THE)
MONARCHCEMENTCOMPANY(THE)
AllenCounty Landfill
CityofChanuteLandfill
WilsonCountyLandfill
NeoshoCountyLandfill
WoodsonCountyTransferStation
ARCHERDANIELSMIDLANDCO.
CITYOFCHANUTEELEC.DEPT.-
Plant
ID
323
322
00009
00009
00009
LF10150
LF4022
LF254
LF4021
LF204
202050007
201330002
Point
ID
1
1
000006
000007
000005
A36
A874
A1510
A2652
A1106
A1224
A1213
Stack
ID
20091
20090
19903
19904
19902
19908
20050
20092
20049
20093
20089
20046
Latitude
37.5094
37.5094
37.7983
37.7983
37.7983
37.9155
37.6677
37.5592
37.5588
37.8865
37.5356
37.6883
Longitude
-95.8217
-95.8217
-95.4244
-95.4244
-95.4244
-95.2979
-95.4413
-95.7436
-95.3066
-95.7401
-95.7603
-95.4006
StkHt
(m)
26.9
26.9
42.7
42.7
43.1
3.0
3.0
3.0
3.0
3.0
8.7
12.3
Diam
(m)
1.16
1.16
3.35
3.35
2.68
0.00
0.00
0.00
0.00
0.00
0.46
1.04
Temp
(K)
457
457
372
372
434
295
295
295
295
295
493
595
Vel
(m/s)
9.3
9.3
19.8
19.8
15.7
0.0
0.0
0.0
0.0
0.0
19.2
25.6
included.
HGO
(tpy)
3.25E-02
1.06E-02
2.65E-03
2.50E-03
1.00E-03
9.51 E-06
1.21E-06
9.30E-07
4.90E-07
4.20E-07
1.70E-07
1.00E-08
HG2
(tpy)
1.12E-02
3.65E-03
4.59E-04
4.33E-04
1.74E-04
1.19E-06
1.50E-07
1.20E-07
6.00E-08
5.00E-08
1.00E-07
1.00E-08
HGP
(tpy)
1.23E-02
4.01 E-03
4.24E-04
4.00E-04
1.60E-04
1.19E-06
1.50E-07
1.20E-07
6.00E-08
5.00E-08
7.00E-08
1.00E-08
Total Hg
(tpy)
5.61 E-02
1.82E-02
3.53E-03
3.33E-03
1.34E-03
1.19E-05
1.51 E-06
1.17E-06
6.10E-07
5.20E-07
3.40E-07
3.00E-08
S.SANTAFE
Table C-18b. Non-point Sources Included in the Collective Sources Tag for Kansas.
Counties in the vicinity of the maximum in-state contribution to mercury deposition with significant non-point emissions are included.
C-26
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
FIPS
20133
20133
20001
20001
20205
20205
20133
20001
20133
20001
20205
20133
20001
20205
20133
20205
20001
20133
20001
20205
20205
20133
20133
20001
20205
20133
20205
20133
20001
20001
20205
20205
County
NEOSHO
NEOSHO
ALLEN
ALLEN
WILSON
WILSON
NEOSHO
ALLEN
NEOSHO
ALLEN
WILSON
NEOSHO
ALLEN
WILSON
NEOSHO
WILSON
ALLEN
NEOSHO
ALLEN
WILSON
WILSON
NEOSHO
NEOSHO
ALLEN
WILSON
NEOSHO
WILSON
NEOSHO
ALLEN
ALLEN
WILSON
WILSON
Source Description
Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools,
Laboratories
Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools,
Laboratories
Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools,
Laboratories
Stationary Source Fuel Combustion, Commercial/Institutional, Distillate Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Distillate Oil
External Combustion Boilers, Industrial, Liquid Waste
External Combustion Boilers, Industrial, Liquid Waste
Stationary Source Fuel Combustion, Commercial/Institutional, Distillate Oil
Stationary Source Fuel Combustion, Industrial, Residual Oil
Stationary Source Fuel Combustion, Industrial, Residual Oil
External Combustion Boilers, Industrial, Liquid Waste
External Combustion Boilers, Industrial, Distillate Oil
Stationary Source Fuel Combustion, Industrial, Residual Oil
External Combustion Boilers, Industrial, Distillate Oil
Miscellaneous Area Sources, Other Combustion, Cremation
Miscellaneous Area Sources, Other Combustion, Cremation
External Combustion Boilers, Industrial, Distillate Oil
Miscellaneous Area Sources, Other Combustion, Cremation
Stationary Source Fuel Combustion, Residential,
Bituminous/Subbituminous Coal
Stationary Source Fuel Combustion, Commercial/Institutional, Residual Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Residual Oil
Stationary Source Fuel Combustion, Residential, Distillate Oil
Stationary Source Fuel Combustion, Residential, Distillate Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Residual Oil
Stationary Source Fuel Combustion, Commercial/Institutional,
Bituminous/Subbituminous Coal
Stationary Source Fuel Combustion, Commercial/Institutional,
Bituminous/Subbituminous Coal
Stationary Source Fuel Combustion, Residential, Distillate Oil
Stationary Source Fuel Combustion, Commercial/Institutional,
Bituminous/Subbituminous Coal
Industrial Processes, Chemical Manufacturing, Other Not Classified
HGO (tpy)
5.80E-05
5.49E-05
5.03E-05
4.76E-05
3.60E-05
3.41 E-05
1.53E-05
1.10E-05
6.76E-06
6.25E-06
5.20E-06
4.94E-06
4.57E-06
4.12E-06
3.22E-06
3.01 E-06
2.97E-06
1.18E-06
1.03E-06
1.96E-06
7.40E-07
1.14E-06
8.70E-07
6.20E-07
4.00E-07
3.40E-07
3.00E-07
2.20E-07
1.50E-07
1.00E-07
8.00E-08
7.00E-08
HG2 (Ipy)
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
9.19E-06
6.57E-06
4.06E-06
3.75E-06
3.12E-06
2.96E-06
2.74E-06
2.47E-06
1.93E-06
1.81 E-06
1.78E-06
3.12E-06
2.71 E-06
1.18E-06
1.94E-06
6.80E-07
5.20E-07
3.80E-07
2.40E-07
2.00E-07
1.80E-07
1.30E-07
9.00E-08
6.00E-08
5.00E-08
1.00E-08
HGP (tpy)
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
6.13E-06
4.38E-06
2.70E-06
2.50E-06
2.08E-06
1.98E-06
1.83E-06
1.65E-06
1.29E-06
1.20E-06
1.19E-06
1.08E-06
9.30E-07
7.80E-07
6.70E-07
4.50E-07
3.50E-07
2.50E-07
1.60E-07
1.30E-07
1.20E-07
9.00E-08
6.00E-08
4.00E-08
3.00E-08
1.00E-08
Hg total
5.80E-05
5.49E-05
5.03E-05
4.76E-05
3.60E-05
3.41 E-05
3.07E-05
2.19E-05
1.35E-05
1.25E-05
1.04E-05
9.88E-06
9.14E-06
8.24E-06
6.44E-06
6.02E-06
5.94E-06
5.38E-06
4.67E-06
3.92E-06
3.35E-06
2.27E-06
1.74E-06
1.25E-06
8.00E-07
6.70E-07
6.00E-07
4.40E-07
3.00E-07
2.00E-07
1.60E-07
9.00E-08
C-27
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
FIPS
20133
20001
20205
20133
County
Source Description
HGO (tpy)
HG2 (Ipy)
NEOSHO Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite 3.00E-08 2.00E-08
Coal
ALLEN Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite 2.00E-08 1.00E-08
Coal
WILSON Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite 1.00E-08 1.00E-08
Coal
NEOSHO Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling 1.00E-08 O.OOE+00
Table C-19a. Point Sources Included in the Collective Sources Tag for Kentucky.
HGP (tpy) Hg total
1.00E-08 6.00E-08
1.00E-08 4.00E-08
O.OOE+00 2.00E-08
O.OOE+00 1.00E-08
Only sources in the vicinity of the maximum in-state contribution to mercury deposition are included.
FIPS
21157
21157
21157
21157
21157
21157
21157
21157
21157
21157
21033
County
Marshall
Marshall
Marshall
Marshall
Marshall
Marshall
Marshall
Marshall
Marshall
Marshall
Caldwell
Facility Name
LWD-lnc.
WESTLAKECA&OCORP.
LWD-lnc.
WESTLAKEMONOMERS-INC
AIRPRODUCTS&CHEMICALS
AIRPRODUCTS&CHEMICALS
BFGOODRICHCO
EIFAtochemNorthAmerica-lnc.
LWD-lnc.
BFGOODRICHCO
Crider&Rogers Landfill
Plant
ID
210
T$5563
211
15267
15113
15113
14082
A27
212
14082
LF9348
Point
ID
1
1
1
59625
60360
60371
58274
1
1
58264
A2028
Stack ID
20248
20253
20249
20247
20245
20246
20244
20251
20250
20243
20148
Latitude
37.0475
37.0553
37.0475
37.0481
37.0464
37.0464
37.0517
37.0542
37.0475
37.0517
37.1191
Longitude
-88.3386
-88.3308
-88.3386
-88.3569
-88.3503
-88.3503
-88.3322
-88.3650
-88.3386
-88.3322
-87.8752
StkHt
(m)
26.9
17.3
26.9
68.6
43.1
43.1
43.1
26.9
26.9
43.1
3.0
Diam
(m)
1.16
0.76
1.16
0.18
2.01
2.01
2.01
1.16
1.16
2.01
0.00
Temp
(K)
457
359
457
293
461
461
461
457
457
461
295
Vel
(m/s)
9.3
7.1
9.3
0.7
10.3
10.3
10.3
9.3
9.3
10.3
0.0
HGO
(tpy)
3.39E-01
2.75E-01
1.37E-01
6.85E-04
6.85E-04
6.85E-04
6.85E-04
2.48E-04
2.12E-04
1.19E-04
1.55E-06
HG2
(tpy)
1.17E-01
1.65E-01
4.71 E-02
4.11E-04
4.11E-04
4.11E-04
4.11E-04
8.54E-05
7.30E-05
7.13E-05
1.90E-07
HGP
(tpy)
1.29E-01
1.10E-01
5.18E-02
2.74E-04
2.74E-04
2.74E-04
2.74E-04
9.40E-05
8.03E-05
4.75E-05
1.90E-07
Total Hg
(tpy)
5.84E-01
5.50E-01
2.35E-01
1.37E-03
1.37E-03
1.37E-03
1.37E-03
4.27E-04
3.65E-04
2.38E-04
1.93E-06
C-28
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-19b. Non-point Sources Included in the Collective Sources Tag for Kentucky.
The county in the vicinity of the maximum in-state contribution to mercury deposition with significant non-point emissions is included.
FIPS
21157
21157
21157
21157
21157
21157
21157
21157
21157
21157
County
Source Description
HGO
(tpy)
HG2 HGP
fry) (tpy)
MARSHALL Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total 1.05E-04 O.OOE+00 O.OOE+00
MARSHALL Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, 9.98E-05 O.OOE+00 O.OOE+00
Laboratories
MARSHALL Stationary Source Fuel Combustion, Commercial/Institutional, 1.10E-05 6.60E-06 4.40E-06
Bituminous/Subbituminous Coal
MARSHALL External Combustion Boilers, Industrial, Liquid Waste 9.99E-06 5.99E-06 4.00E-06
MARSHALL Stationary Source Fuel Combustion, Residential, Distillate Oil 8.68E-06 5.21E-06 3.47E-06
MARSHALL Miscellaneous Area Sources, Other Combustion, Cremation 8.70E-07 2.29E-06 7.90E-07
MARSHALL Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite Coal 1.54E-06 9.20E-07 6.20E-07
MARSHALL Stationary Source Fuel Combustion, Commercial/Institutional, Residual Oil 1.02E-06 6.10E-07 4.10E-07
MARSHALL Stationary Source Fuel Combustion, Residential, Bituminous/Subbituminous 7.10E-07 4.30E-07 2.90E-07
Coal
MARSHALL Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling 1.00E-08 O.OOE+00 O.OOE+00
Table C-20a. Point Sources Included in the Collective Sources Tag for Mississippi.
Total Hg
fry)
1.05E-04
9.98E-05
2.20E-05
2.00E-05
1.74E-05
3.95E-06
3.08E-06
2.04E-06
1.43E-06
1.00E-08
Only sources in the vicinity of the maximum in-state contribution to mercury deposition are included.
FIPS
28043
28013
28013
28013
28013
28013
28013
28013
28013
28043
28161
County
Grenada
Calhoun
Calhoun
Calhoun
Calhoun
Calhoun
Calhoun
Calhoun
Calhoun
Grenada
Yalobusha
Facility Name
NEWSPRINTSOUTH
MEMPHISHARDWOOD
MEMPHISHARDWOOD
WEYERHAEUSERCO
WEYERHAEUSERCO
WEYERHAEUSERCO
WEYERHAEUSERCO
WEYERHAEUSERCO
WEYERHAEUSERCO
KOPPERSINDUSTRIESINC
WaterValleySanitai-yLandfill
Plant
ID
2804300015
14606
14606
15804
15804
15804
280130032
280130032
280130032
2804300012
LF107
Point
ID
011
59254
59255
60734
60735
60736
001
002
003
001
A367
Stack ID
26623
26602
26601
26603
26604
26605
26606
26607
26608
26622
26828
Latitude
33.8345
33.9419
33.9419
33.9419
33.9419
33.9419
33.9419
33.9419
33.9419
33.7300
34.0295
Longitude
-89.8169
-89.3236
-89.3236
-89.3236
-89.3236
-89.3236
-89.3236
-89.3236
-89.3236
-89.7814
-89.7194
StkHt
(m)
3.0
23.8
23.8
23.8
23.8
23.8
23.8
23.8
23.8
24.4
3.0
Diam
(m)
0.00
1.01
1.01
1.01
1.01
1.01
1.01
1.01
1.01
0.00
0.00
Temp
(K)
295
476
476
476
476
476
476
476
476
295
295
Vel
(m/s)
0.0
10.5
10.5
10.5
10.5
10.5
10.5
10.5
10.5
0.0
0.0
HGO
(tpy)
3.72E-02
1.13E-04
1.13E-04
1.13E-04
1.13E-04
1.13E-04
8.54E-05
8.54E-05
8.54E-05
5.00E-05
3.30E-07
HG2
(tpy)
2.23E-02
6.75E-05
6.75E-05
6.75E-05
6.75E-05
6.75E-05
5.12E-05
5.12E-05
5.12E-05
3.00E-05
4.00E-08
HGP
(tpy)
1.49E-02
4.50E-05
4.50E-05
4.50E-05
4.50E-05
4.50E-05
3.42E-05
3.42E-05
3.42E-05
2.00E-05
4.00E-08
Total Hg
(tpy)
7.45E-02
2.25E-04
2.25E-04
2.25E-04
2.25E-04
2.25E-04
1.71E-04
1.71E-04
1.71E-04
1.00E-04
4.10E-07
C-29
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-20b. Non-point Sources Included in the Collective Sources Tag for Mississippi.
Counties in the vicinity of the maximum in-state contribution to mercury deposition with significant non-point emissions are included.
FIPS
28043
28013
28043
28043
28043
28013
28043
28013
28013
28013
28043
28043
28013
28013
28013
28043
28043
County
GRENADA
CALHOUN
GRENADA
GRENADA
GRENADA
CALHOUN
GRENADA
CALHOUN
CALHOUN
CALHOUN
GRENADA
GRENADA
CALHOUN
CALHOUN
CALHOUN
GRENADA
GRENADA
Source Description
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, Laboratories
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, Laboratories
Stationary Source Fuel Combustion, Commercial/Institutional, Distillate Oil
External Combustion Boilers, Industrial, Liquid Waste
Stationary Source Fuel Combustion, Industrial, Residual Oil
External Combustion Boilers, Industrial, Liquid Waste
External Combustion Boilers, Industrial, Distillate Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Distillate Oil
Stationary Source Fuel Combustion, Industrial, Residual Oil
External Combustion Boilers, Industrial, Distillate Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Residual Oil
Miscellaneous Area Sources, Other Combustion, Cremation
Miscellaneous Area Sources, Other Combustion, Cremation
Stationary Source Fuel Combustion, Commercial/Institutional, Residual Oil
Stationary Source Fuel Combustion, Residential, Distillate Oil
Stationary Source Fuel Combustion, Residential, Distillate Oil
Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling
HGO
(tpv)
7.41 E-05
4.92E-05
2.23E-05
1.30E-05
9.50E-06
6.82E-06
6.19E-06
5.20E-06
4.99E-06
3.25E-06
1.27E-06
5.30E-07
3.50E-07
3.00E-07
9.00E-08
8.00E-08
1.00E-08
HG2
(tpv)
O.OOE+00
O.OOE+00
1.34E-05
7.80E-06
5.70E-06
4.09E-06
3.71 E-06
3.12E-06
2.99E-06
1.95E-06
7.60E-07
1.41 E-06
9.30E-07
1.80E-07
5.00E-08
5.00E-08
O.OOE+00
HGP
(tpv)
O.OOE+00
O.OOE+00
8.93E-06
5.20E-06
3.80E-06
2.73E-06
2.47E-06
2.08E-06
1.99E-06
1.30E-06
5.10E-07
4.90E-07
3.20E-07
1.20E-07
3.00E-08
3.00E-08
O.OOE+00
Total Hg
(tpy)
7.41 E-05
4.92E-05
4.46E-05
2.60E-05
1.90E-05
1.36E-05
1.24E-05
1.04E-05
9.97E-06
6.50E-06
2.54E-06
2.43E-06
1.60E-06
6.00E-07
1.70E-07
1.60E-07
1.00E-08
Table C-21. Non-point Sources Included in the Collective Sources Tag for New Jersey.
The county in the vicinity of the maximum in-state contribution to mercury deposition with significant non-point emissions is included.
(Note: this is an area source. No point sources in the collective source tag were located near the simulated peak.)
FIPS
County
Source Description
HGO (tpy)
HG2 (tpy)
HGP (tpy)
Total Hg (tpy)
34031
PASSAIC
Stationary Source Fuel Combustion, Industrial, Residual Oil
9.94E-05
5.97E-05
3.98E-05
1.99E-04
C-30
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-22a
. Point Sources Included in
Only sources in the vicinity of the maximum
FIPS
39085
39035
39085
39085
39085
39085
39007
39085
39085
County
Lake
Cuyahoga
Lake
Lake
Lake
Lake
Ashtabula
Lake
Lake
Facility Name
LubrizolCorp.
NEORSDEASTERLYWASTEWATERTREATMENTPLA
EAGLEPICHERIND.-MATSDIV.
LubrizolCorporation
LakeCountyRecycling&DisposalLF
LINCOLN ELECTRIC
WMI-GenevaLandfill
LakeCountySolidWasteLandfill
CityofWilloughbyLandfill
Plant
ID
B30
12213
13778
A36
LF5858
13897
LF10003
LF5859
LF5860
Point
ID
1
53881
57654
1
A1289
57958
A3
A1290
A1291
Stack ID
30477
30304
30472
30476
30478
30473
30255
30479
30480
in-state
Latitude
41.6111
41.5694
41.6444
41.7203
41.6323
41.6911
41.7940
41.7550
41.6456
the Collective Sources Tag for Ohio.
contribution to mercury deposition are included.
Longitude
-81.4794
-81.5850
-81.4156
-81.2736
-81.4095
-81.3097
-80.9054
-81.2047
-81.3963
StkHt (m)
26.9
17.3
43.1
26.9
3.0
35.1
3.0
3.0
3.0
Diam (m)
1.16
0.73
2.01
1.16
0.00
1.62
0.00
0.00
0.00
Temp (K)
457
433
461
457
295
464
295
295
295
Vel (m/s)
9.3
10.1
10.3
9.3
0.0
8.4
0.0
0.0
0.0
HGO
(tpy)
1.17E-01
2.82E-04
1.37E-04
3.18E-05
3.16E-05
1.95E-05
3.04E-05
1.62E-05
1.67E-06
HG2
(tpy)
4.05E-02
1.69E-04
8.25E-05
1.10E-05
3.95E-06
1.17E-05
3.80E-06
2.03E-06
2.10E-07
HGP
(tpy)
4.45E-02
1.13E-04
5.50E-05
1.21E-05
3.95E-06
7.79E-06
3.80E-06
2.03E-06
2.10E-07
Total Hg
(tpy)
2.02E-01
5.64E-04
2.75E-04
5.48E-05
3.95E-05
3.89E-05
3.80E-05
2.03E-05
2.09E-06
C-31
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-22b. Non-point Sources Included in the Collective Sources Tag for Ohio.
Counties in the vicinity of the maximum in-state contribution to mercury deposition with significant non-point emissions are included.
FIPS
39035
39035
39035
39035
39035
39085
39085
39035
39007
39085
39035
39035
39085
39035
39007
39007
39035
39035
39085
39085
39007
39085
39085
39085
39007
39007
39007
39035
39007
39085
39085
County
CUYAHOGA
CUYAHOGA
CUYAHOGA
CUYAHOGA
CUYAHOGA
LAKE
LAKE
CUYAHOGA
ASHTABULA
LAKE
CUYAHOGA
CUYAHOGA
LAKE
CUYAHOGA
ASHTABULA
ASHTABULA
CUYAHOGA
CUYAHOGA
LAKE
LAKE
ASHTABULA
LAKE
LAKE
LAKE
ASHTABULA
ASHTABULA
ASHTABULA
CUYAHOGA
ASHTABULA
LAKE
LAKE
Source Description
Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, Dental Alloy
(Mercury Amalgams) Production
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, Laboratories
Stationary Source Fuel Combustion, Commercial/Institutional, Distillate Oil
External Combustion Boilers, Industrial, Liquid Waste
Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, Laboratories
Stationary Source Fuel Combustion, Industrial, Residual Oil
Stationary Source Fuel Combustion, Residential, Distillate Oil
Stationary Source Fuel Combustion, Residential, Distillate Oil
External Combustion Boilers, Industrial, Distillate Oil
Miscellaneous Area Sources, Other Combustion, Cremation
Stationary Source Fuel Combustion, Commercial/Institutional, Distillate Oil
Stationary Source Fuel Combustion, Residential, Distillate Oil
Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, Laboratories
Stationary Source Fuel Combustion, Commercial/Institutional, Residual Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Bituminous/Subbituminous
Coal
External Combustion Boilers, Industrial, Liquid Waste
Stationary Source Fuel Combustion, Industrial, Residual Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Distillate Oil
External Combustion Boilers, Industrial, Distillate Oil
Miscellaneous Area Sources, Other Combustion, Cremation
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, Dental Alloy
(Mercury Amalgams) Production
External Combustion Boilers, Industrial, Liquid Waste
Stationary Source Fuel Combustion, Industrial, Residual Oil
Miscellaneous Area Sources, Other Combustion, Cremation
Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite Coal
External Combustion Boilers, Industrial, Distillate Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Residual Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Bituminous/Subbituminous
Coal
HGO
(tpy)
4.78E-03
4.68E-03
4.53E-03
2.23E-03
5.11E-04
7.91 E-04
7.50E-04
3.73E-04
2.54E-04
2.44E-04
2.43E-04
1.07E-04
2.36E-04
1.91 E-04
3.60E-04
3.41 E-04
1.27E-04
1.22E-04
9.59E-05
7.01 E-05
6.74E-05
4.56E-05
1.77E-05
7.56E-05
3.16E-05
2.31 E-05
8.05E-06
1.71 E-05
1.50E-05
1.34E-05
1.30E-05
HG2
(tpy)
O.OOE+OO
O.OOE+00
O.OOE+OO
1.34E-03
3.07E-04
O.OOE+OO
O.OOE+OO
2.24E-04
1.53E-04
1.46E-04
1.46E-04
2.82E-04
1.42E-04
1.14E-04
O.OOE+OO
O.OOE+OO
7.60E-05
7.33E-05
5.75E-05
4.20E-05
4.04E-05
2.74E-05
4.67E-05
O.OOE+OO
1.90E-05
1.38E-05
2.12E-05
1.03E-05
9.01 E-06
8.05E-06
7.79E-06
HGP
(tpy)
O.OOE+OO
O.OOE+OO
O.OOE+OO
8.91 E-04
2.04E-04
O.OOE+OO
O.OOE+OO
1.49E-04
1.02E-04
9.75E-05
9.72E-05
9.72E-05
9.44E-05
7.63E-05
O.OOE+OO
O.OOE+OO
5.07E-05
4.89E-05
3.84E-05
2.80E-05
2.69E-05
1.82E-05
1.61 E-05
O.OOE+OO
1.26E-05
9.23E-06
7.32E-06
6.84E-06
6.01 E-06
5.37E-06
5.19E-06
Total Hg (tpy)
4.78E-03
4.68E-03
4.53E-03
4.46E-03
1.02E-03
7.91 E-04
7.50E-04
7.47E-04
5.09E-04
4.88E-04
4.86E-04
4.86E-04
4.72E-04
3.82E-04
3.60E-04
3.41 E-04
2.53E-04
2.44E-04
1.92E-04
1.40E-04
1.35E-04
9.12E-05
8.05E-05
7.56E-05
6.32E-05
4.61 E-05
3.66E-05
3.42E-05
3.00E-05
2.68E-05
2.60E-05
C-32
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
FIPS
39007
39035
39085
39035
39007
39007
39085
39085
39007
39007
39035
39035
39007
39085
39085
39007
FIPS
44007
44007
44007
44007
44007
44007
44007
44007
44007
44007
44007
44007
44007
44007
44007
44007
County
Providence
Providence
Providence
Providence
Providence
Providence
Providence
Providence
Providence
Providence
Providence
Providence
Providence
Providence
Providence
Providence
County
Source Description
HGO
(tpy)
HG2
(tpy)
ASHTABULA Stationary Source Fuel Combustion, Residential, Bituminous/Subbituminous Coal 1.14E-05 6.83E-06
CUYAHOGA Stationary Source Fuel Combustion, Residential, Bituminous/Subbituminous Coal 5.51E-06 3.31E-06
LAKE Stationary Source Fuel Combustion, Residential, Bituminous/Subbituminous Coal 4.98E-06 2.99E-06
CUYAHOGA Industrial Processes, Chemical Manufacturing, Other Not Classified 7.22E-06 9.00E-07
ASHTABULA Stationary Source Fuel Combustion, Commercial/Institutional, Residual Oil 3.83E-06 2.30E-06
ASHTABULA Stationary Source Fuel Combustion, Commercial/Institutional, Bituminous/Subbituminous 3.78E-06 2.26E-06
Coal
LAKE Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite Coal 1.82E-06 1.09E-06
LAKE Industrial Processes, Chemical Manufacturing, Other Not Classified 1.76E-06 2.20E-07
ASHTABULA Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite Coal 5.30E-07 3.20E-07
ASHTABULA Stationary Source Fuel Combustion, Residential, Anthracite Coal 3.00E-07 1.80E-07
CUYAHOGA Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling 4.50E-07 O.OOE+00
CUYAHOGA Stationary Source Fuel Combustion, Residential, Anthracite Coal 1 .50E-07 9.00E-08
ASHTABULA Industrial Processes, Chemical Manufacturing, Other Not Classified 2.10E-07 3.00E-08
LAKE Stationary Source Fuel Combustion, Residential, Anthracite Coal 1.30E-07 8.00E-08
LAKE Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling 7.00E-08 O.OOE+00
ASHTABULA Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling 3.00E-08 O.OOE+00
Table C-23a. Point Sources Included in the Collective Sources Tag for Rhode Island.
Only sources in the vicinity of the maximum in-state contribution to mercury deposition are included.
Facility Name
ALGONQUINGASTRANSMISSIONCO.(B)
BROWNUNIVERSITY
BROWNUNIVERSITY
SEVILLEDYEINGCO.-INC.
DORADOPROCESSINGCO.-INC.
SLATERDYEWORKS&SLATERSCREENPRINT
TEKNORAPEXCO.(CENTRALAVENUE)
CCLCUSTOMMFG.
OSRAMSYLVANIAPRODUCTSINC.
BROWNUNIVERSITY
PROVIDENCEMETALLIZINGCO.-INC.
ELIZABETHWEBBINGMILLSCO.-INC.
AMERICANINSULATEDWIRECORP.PAWTUCKET
SEVILLEDYEINGCO.-INC.
ELIZABETHWEBBINGMILLSCO.-INC.
BICCGENERAL
Plant
ID
00500
00033
00033
00013
00034
00030
00029
00453
00003
00033
06011
09008
80000
00013
09008
09046
Point ID
003
002
001
001
001
001
001
001
001
001
001
001
001
001
001
001
Stack
ID
32028
32018
32016
32011
32020
32015
32013
32026
32010
32017
32032
32033
32040
32012
32034
32038
Latitude
41.9875
41.8271
41.8271
42.0029
42.0053
41.8587
41.8856
41.9276
41.8962
41.8271
41.8736
41.8860
41.8856
42.0029
41.8860
41.9231
Longitude
-71.7597
-71.3991
-71.3992
-71.5267
-71.5242
-71.3742
-71.3630
-71.4272
-71.3887
-71.3992
-71.4101
-71.3821
-71.3668
-71.5267
-71.3821
-71.4807
StkHt
(m)
16.6
3.0
18.3
15.2
10.7
32.0
24.4
9.4
24.4
18.3
17.7
24.4
53.3
15.2
24.4
10.7
Diam
(m)
0.76
0.00
1.16
0.61
0.61
3.05
0.61
0.61
1.52
1.16
0.76
0.30
3.05
0.61
0.30
1.22
Temp
(K)
605
295
454
464
456
468
466
468
439
454
589
451
468
464
451
468
Vel
(m/s)
23.8
0.0
7.6
8.2
8.2
9.4
8.2
9.4
9.4
7.6
9.1
8.8
9.4
8.2
8.8
9.4
HGO
(tpy)
2.00E-03
5.00E-04
4.75E-05
3.25E-05
2.83E-05
2.25E-05
1.25E-05
1.00E-05
1.00E-05
8.75E-06
7.50E-06
7.50E-06
7.50E-06
7.45E-06
6.00E-06
5.50E-06
HGP Total Hg (tpy)
(tpy)
4.56E-06 2.28E-05
2.21E-06 1.10E-05
1.99E-06 9.96E-06
9.00E-07 9.02E-06
1.53E-06 7.66E-06
1.51E-06 7.55E-06
7.30E-07 3.64E-06
2.20E-07 2.20E-06
2.10E-07 1.06E-06
1.20E-07 6.00E-07
O.OOE+00 4.50E-07
6.00E-08 3.00E-07
3.00E-08 2.70E-07
5.00E-08 2.60E-07
O.OOE+00 7.00E-08
O.OOE+00 3.00E-08
HG2
(tpy)
1.20E-03
3.00E-04
2.85E-05
1.95E-05
1.70E-05
1.35E-05
7.50E-06
6.00E-06
6.00E-06
5.25E-06
4.50E-06
4.50E-06
4.50E-06
4.47E-06
3.60E-06
3.30E-06
HGP
(tpy)
8.00E-04
2.00E-04
1.90E-05
1.30E-05
1.13E-05
9.00E-06
5.00E-06
4.00E-06
4.00E-06
3.50E-06
3.00E-06
3.00E-06
3.00E-06
2.98E-06
2.40E-06
2.20E-06
Total Hg
(tpy)
4.00E-03
1.00E-03
9.50E-05
6.50E-05
5.65E-05
4.50E-05
2.50E-05
2.00E-05
2.00E-05
1.75E-05
1.50E-05
1.50E-05
1.50E-05
1.49E-05
1.20E-05
1.10E-05
C-33
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
44007
44007
44007
44007
44007
44007
44007
44007
44007
44007
Providence
Providence
Providence
Providence
Providence
Providence
Providence
Providence
Providence
Providence
DORADOPROCESSINGCO.-INC.
OCEANSTATEPOWER
SLATERDYEWORKSCUMBERLAND
FOAMTECHNOLOGY
TEKNORAPEXCO.(CENTRALAVENUE)
NorthProvidenceSanitaryLandfill
ALGONQUINGASTRANSMISSIONCO.(B)
BICCGENERAL
BurriNvilleSanitaryLandfill
PAWTUCKETPOWERASSOCIATED
00034
00545
00706
00716
00029
LF2954
00500
09046
LF2956
90069
001
001
001
001
001
A1795
001
002
A1796
002
32019
32029
32030
32031
32014
32043
32027
32039
32044
32042
42.0053
41.9918
41.9531
41.8812
41.8856
41.8716
41.9875
41.9231
41.8716
41.8601
-71.5242
-71.6829
-71.4018
-71.4050
-71.3630
-71.5590
-71.7597
-71.4807
-71.5590
-71.4083
10.7
45.7
5.5
18.3
24.4
3.0
3.0
3.0
3.0
54.1
0.61
4.80
0.91
0.61
0.61
0.00
0.00
0.00
0.00
3.58
456
370
468
451
466
295
295
295
295
389
8.2
18.7
9.4
8.8
8.2
0.0
0.0
0.0
0.0
21.9
5.00E-06
3.50E-06
1.75E-06
1.25E-06
1.25E-06
1.53E-06
5.00E-07
4.00E-07
2.20E-07
3.00E-08
3.00E-06
2.10E-06
1.05E-06
7.50E-07
7.50E-07
1.90E-07
3.00E-07
2.40E-07
3.00E-08
2.00E-08
2.00E-06
1.40E-06
7.00E-07
5.00E-07
5.00E-07
1.90E-07
2.00E-07
1.60E-07
3.00E-08
1.00E-08
1.00E-05
7.00E-06
3.50E-06
2.50E-06
2.50E-06
1.91E-06
1.00E-06
8.00E-07
2.80E-07
6.00E-08
Table C-23b. Non-point Sources Included in the Collective Sources Tag for Rhode Island.
The county in the vicinity of the maximum in-state contribution to mercury deposition with significant non-point emissions is included.
FIPS
44007
44007
44007
44007
44007
44007
44007
44007
44007
44007
44007
44007
County
PROVIDENCE
PROVIDENCE
PROVIDENCE
PROVIDENCE
PROVIDENCE
PROVIDENCE
PROVIDENCE
PROVIDENCE
PROVIDENCE
PROVIDENCE
PROVIDENCE
PROVIDENCE
Source Description
Stationary Source Fuel Combustion, Residential, Distillate Oil
Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools,
Laboratories
Stationary Source Fuel Combustion, Commercial/Institutional, Distillate Oil
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools,
Dental Alloy (Mercury Amalgams) Production
External Combustion Boilers, Industrial, Liquid Waste
Miscellaneous Area Sources, Other Combustion, Cremation
Stationary Source Fuel Combustion, Commercial/Institutional,
Bituminous/Subbituminous Coal
Stationary Source Fuel Combustion, Residential, Anthracite Coal
Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite Coal
Industrial Processes, Chemical Manufacturing, Other Not Classified
Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling
HGO
(tpy)
7.62E-03
2.00E-03
1.89E-03
7.44E-04
8.10E-04
1.96E-04
8.34E-05
7.44E-05
1.25E-05
1.04E-05
1.24E-06
1.90E-07
HG2
(tpy)
4.57E-03
O.OOE+00
O.OOE+00
4.47E-04
O.OOE+00
1.17E-04
2.20E-04
4.46E-05
7.52E-06
6.25E-06
1.50E-07
O.OOE+00
HGP
(tpy)
3.05E-03
O.OOE+00
O.OOE+00
2.98E-04
O.OOE+00
7.82E-05
7.59E-05
2.97E-05
5.01 E-06
4.16E-06
1.50E-07
O.OOE+00
Total Hg
»y)
1.52E-02
2.00E-03
1.89E-03
1.49E-03
8.10E-04
3.91 E-04
3.79E-04
1.49E-04
2.51 E-05
2.08E-05
1.54E-06
1.90E-07
C-34
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
FIPS County
Table C-24a. Point Sources Included in the Collective Sources Tag for South Dakota.
Only sources in the vicinity of the maximum in-state contribution to mercury deposition are included.
Facility Name
Plant ID
Point
ID
Stack ID
Latitude
Longitude
StkHt
(m)
Diam
(m)
Temp
(K)
Vel
(m/s)
HGO
(tpy)
HG2
(tpy)
HGP
(tpy)
Total Hg
(tpy)
46019 Butte
AMERICANCOLLOIDCO
13876
57859
32793
44.8927
-103.5066
40.1
1.92
459
7.5
2.78E-03
1.67E-03
1.11E-03
5.56E-03
Table C-24b. Non-point Sources Included in the Collective Sources Tag for South Dakota.
The county in the vicinity of the maximum in-state contribution to mercury deposition with significant non-point emissions is included.
FIPS
46019
46019
46019
46019
46019
46019
46019
46019
46019
46019
46019
46019
County
BUTTE
BUTTE
BUTTE
BUTTE
BUTTE
BUTTE
BUTTE
BUTTE
BUTTE
BUTTE
BUTTE
BUTTE
Source Description
Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools,
Laboratories
Stationary Source Fuel Combustion, Commercial/Institutional, Distillate Oil
Stationary Source Fuel Combustion, Residential, Distillate Oil
Miscellaneous Area Sources, Other Combustion, Cremation
External Combustion Boilers, Industrial, Liquid Waste
Stationary Source Fuel Combustion, Industrial, Residual Oil
Stationary Source Fuel Combustion, Commercial/Institutional,
Bituminous/Subbituminous Coal
Stationary Source Fuel Combustion, Residential, Bituminous/Subbituminous
Coal
Stationary Source Fuel Combustion, Commercial/Institutional, Residual Oil
External Combustion Boilers, Industrial, Distillate Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite Coal
HGO
»y)
3.05E-05
2.89E-05
5.52E-06
5.11E-06
4.70E-07
6.10E-07
4.50E-07
3.80E-07
3.30E-07
3.10E-07
3.00E-07
6.00E-08
HG2
»y)
O.OOE+00
O.OOE+00
3.31 E-06
3.07E-06
1.24E-06
3.70E-07
2.70E-07
2.30E-07
2.00E-07
1.90E-07
1.80E-07
3.00E-08
HGP
»y)
O.OOE+00
O.OOE+00
2.21 E-06
2.04E-06
4.30E-07
2.50E-07
1.80E-07
1.50E-07
1.30E-07
1.30E-07
1.20E-07
2.00E-08
Total Hg
(ipy)
3.05E-05
2.89E-05
1.10E-05
1.02E-05
2.14E-06
1.23E-06
9.00E-07
7.60E-07
6.60E-07
6.30E-07
6.00E-07
1.10E-07
C-35
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-25a. Point Sources Included in
the Collective Sources
Only sources in the vicinity of the maximum in-state contribution to
FIPS
47165
47165
47165
47165
47165
47111
47189
47165
47111
47111
47111
47111
47165
47165
47169
47111
47165
County
Sumner
Sumner
Sumner
Sumner
Sumner
Macon
Wilson
Sumner
Macon
Macon
Macon
Macon
Sumner
Sumner
Trousdale
Macon
Sumner
Facility Name
ResourceAuthorityinSumnerCo.
ResourceAuthorityinSumnerCo.
CRESENTMANUFACTURINGCOMPANY
GALLATINBLOCKCOMPANY
CRESENTMANUFACTURINGCOMPANY
LAFAYETTEMANUFACTURINGCOMPANY
WilsonCountyLandfill
CRESENTMANUFACTURINGCOMPANY
LAFAYETTEMANUFACTURINGCOMPANY
LafayetteManufacturingCompany
LAFAYETTEMANUFACTURINGCOMPANY
LAFAYETTEMANUFACTURINGCOMPANY
HendersonvilleLandfill
SumnerCounty Landfill
Hartsville/TrousdaleLandfill
LafayetteCityLandfil
GallatinCity Landfill
Plant
ID
SMWC-29
SMWC-29
15352
14981
15352
14549
LF252
15352
14549
470560002
14549
14549
LF1181
LF1180
LF984
LF5073
LF1179
Point
ID
SMWC-1
SMWC-2
60975
60054
60973
59152
A501
60974
59149
001
59150
59151
A457
A456
A7275
A3510
A454
Stack
ID
33263
33264
33259
33256
33257
33045
33306
33258
33042
33046
33043
33044
33262
33261
33265
33047
33260
Latitude
36.3832
36.3832
36.3832
36.3832
36.3832
36.5404
36.2626
36.3832
36.5404
36.5388
36.5404
36.5404
36.4487
36.4487
36.3926
36.5260
36.4487
Longitude
-86.4548
-86.4548
-86.4548
-86.4548
-86.4548
-86.0122
-86.2944
-86.4548
-86.0122
-86.0119
-86.0122
-86.0122
-86.4799
-86.4799
-86.1311
-86.0088
-86.4799
Tag for Tennessee.
mercury deposition are
StkHt
(m)
31.1
31.1
25.5
45.6
25.5
4.7
3.0
25.5
31.6
31.6
4.7
4.7
3.0
3.0
3.0
3.0
3.0
Diam
(m)
1.89
1.89
1.25
2.07
1.25
0.65
0.00
1.25
1.40
1.40
0.65
0.65
0.00
0.00
0.00
0.00
0.00
Temp
(K)
505
505
452
482
452
295
295
452
430
430
295
295
295
295
295
295
295
Vel
(mis)
13.3
13.3
10.5
9.7
10.5
9.0
0.0
10.5
11.5
11.5
9.0
9.0
0.0
0.0
0.0
0.0
0.0
included.
HGO
fry)
1.25E-02
1.25E-02
2.51 E-05
1.02E-05
8.95E-06
6.98E-06
8.15E-06
4.47E-06
4.13E-06
3.13E-06
3.60E-07
3.60E-07
3.90E-07
3.40E-07
3.30E-07
3.30E-07
2.60E-07
HG2
(tpy)
3.30E-02
3.30E-02
1.51 E-05
6.10E-06
5.37E-06
4.19E-06
1.02E-06
2.68E-06
2.48E-06
1.88E-06
2.20E-07
2.20E-07
5.00E-08
4.00E-08
4.00E-08
4.00E-08
3.00E-08
HGP
»y)
1.14E-02
1.14E-02
1.00E-05
4.07E-06
3.58E-06
2.79E-06
1.02E-06
1.79E-06
1.65E-06
1.25E-06
1.40E-07
1.40E-07
5.00E-08
4.00E-08
4.00E-08
4.00E-08
3.00E-08
Total Hg
»y)
5.69E-02
5.69E-02
5.02E-05
2.03E-05
1.79E-05
1.40E-05
1.02E-05
8.94E-06
8.26E-06
6.26E-06
7.20E-07
7.20E-07
4.90E-07
4.20E-07
4.10E-07
4.10E-07
3.20E-07
C-36
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-25b. Non-point Sources Included in the Collective Sources Tag for Tennessee.
Counties in the vicinity of the maximum in-state contribution to mercury deposition with significant non-point emissions are included.
FIPS
47165
47165
47189
47189
47165
47189
47165
47189
47165
47111
47111
47165
47165
47189
47189
47165
47165
47189
47189
47165
47111
47189
47111
47189
47165
47111
47189
47111
47111
47165
47165
47189
47111
47189
47111
47111
47111
County
SUMMER
SUMMER
WILSON
WILSON
SUMNER
WILSON
SUMNER
WILSON
SUMNER
MACON
MACON
SUMNER
SUMNER
WILSON
WILSON
SUMNER
SUMNER
WILSON
WILSON
SUMNER
MACON
WILSON
MACON
WILSON
SUMNER
MACON
WILSON
MACON
MACON
SUMNER
SUMNER
WILSON
MACON
WILSON
MACON
MACON
MACON
Source Description
Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, Laboratories
Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, Laboratories
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, Dental Alloy (Mercury
Amalgams) Production
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, Dental Alloy (Mercury
Amalgams) Production
Stationary Source Fuel Combustion, Commercial/Institutional, Distillate Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Distillate Oil
External Combustion Boilers, Industrial, Liquid Waste
Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, Laboratories
Stationary Source Fuel Combustion, Industrial, Residual Oil
External Combustion Boilers, Industrial, Distillate Oil
External Combustion Boilers, Industrial, Liquid Waste
Stationary Source Fuel Combustion, Industrial, Residual Oil
Stationary Source Fuel Combustion, Residential, Distillate Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Bituminous/Subbituminous Coal
External Combustion Boilers, Industrial, Distillate Oil
Stationary Source Fuel Combustion, Residential, Distillate Oil
Miscellaneous Area Sources, Other Combustion, Cremation
Stationary Source Fuel Combustion, Commercial/Institutional, Distillate Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Bituminous/Subbituminous Coal
External Combustion Boilers, Industrial, Liquid Waste
Miscellaneous Area Sources, Other Combustion, Cremation
Stationary Source Fuel Combustion, Commercial/Institutional, Residual Oil
Stationary Source Fuel Combustion, Industrial, Residual Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Residual Oil
Stationary Source Fuel Combustion, Residential, Distillate Oil
External Combustion Boilers, Industrial, Distillate Oil
Stationary Source Fuel Combustion, Residential, Bituminous/Subbituminous Coal
Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite Coal
Stationary Source Fuel Combustion, Residential, Bituminous/Subbituminous Coal
Miscellaneous Area Sources, Other Combustion, Cremation
Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite Coal
Stationary Source Fuel Combustion, Commercial/Institutional, Bituminous/Subbituminous Coal
Stationary Source Fuel Combustion, Commercial/Institutional, Residual Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite Coal
HGO
(tpy)
4.39E-04
4.16E-04
3.01 E-04
2.85E-04
2.05E-04
2.05E-04
7.57E-05
6.25E-05
3.84E-05
6.46E-05
6.12E-05
2.81 E-05
1.83E-05
1.79E-05
1.31 E-05
1.16E-05
1.12E-05
8.49E-06
8.14E-06
3.45E-06
7.36E-06
7.36E-06
5.46E-06
2.37E-06
4.30E-06
3.99E-06
3.55E-06
3.27E-06
2.60E-06
2.40E-06
1.57E-06
1.31E-06
5.10E-07
1.03E-06
8.30E-07
4.20E-07
1.10E-07
HG2
(tpy)
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
4.54E-05
3.75E-05
2.30E-05
O.OOE+00
O.OOE+00
1.68E-05
1.10E-05
1.07E-05
7.83E-06
6.95E-06
6.70E-06
5.10E-06
4.88E-06
9.08E-06
4.42E-06
4.42E-06
3.28E-06
6.24E-06
2.58E-06
2.39E-06
2.13E-06
1.96E-06
1.56E-06
1.44E-06
9.40E-07
7.90E-07
1.33E-06
6.20E-07
5.00E-07
2.50E-07
7.00E-08
HGP
(tpy)
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
3.03E-05
2.50E-05
1.54E-05
O.OOE+00
O.OOE+00
1.12E-05
7.30E-06
7.15E-06
5.22E-06
4.64E-06
4.46E-06
3.40E-06
3.26E-06
3.13E-06
2.95E-06
2.94E-06
2.19E-06
2.15E-06
1.72E-06
1.60E-06
1.42E-06
1.31E-06
1.04E-06
9.60E-07
6.30E-07
5.20E-07
4.60E-07
4.10E-07
3.30E-07
1.70E-07
5.00E-08
Total Hg (tpy)
4.39E-04
4.16E-04
3.01 E-04
2.85E-04
2.05E-04
2.05E-04
1.51 E-04
1.25E-04
7.68E-05
6.46E-05
6.12E-05
5.61 E-05
3.65E-05
3.57E-05
2.61 E-05
2.32E-05
2.23E-05
1.70E-05
1.63E-05
1.57E-05
1.47E-05
1.47E-05
1.09E-05
1.08E-05
8.60E-06
7.98E-06
7.10E-06
6.54E-06
5.20E-06
4.80E-06
3.14E-06
2.62E-06
2.30E-06
2.06E-06
1.66E-06
8.40E-07
2.30E-07
C-37
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
FIPS
47165
47189
47165
47189
47111
FIPS County
County
SUMMER
WILSON
SUMMER
WILSON
MACON
Source Description
Industrial Processes, Chemical Manufacturing, Other Not Classified
Industrial Processes, Chemical Manufacturing, Other Not Classified
Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling
Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling
Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling
Table C-26a. Point Sources Included in the Collective
Only sources in the vicinity of the maximum in-state contribution to
Facility Name Plant Point Stack Latitude Longitude
ID ID ID
HGO
(tpy)
5.00E-08
5.00E-08
4.00E-08
3.00E-08
1.00E-08
Sources Tag for Utah.
mercury deposition are
StkHt Diam Temp Vel
(m) (m) (K) (m/s)
HG2
(tpy)
1.00E-08
1.00E-08
O.OOE+00
O.OOE+00
O.OOE+00
included.
HGO
»y)
HGP
(tpy)
1.00E-08
1.00E-08
O.OOE+00
O.OOE+00
O.OOE+00
HG2 HGP
(tpy) (tpy)
Total Hg (tpy)
7.00E-08
7.00E-08
4.00E-08
3.00E-08
1.00E-08
Total Hg
(tpy)
49029
MORGAN
Holcim (US) Inc.
N/A
N;A
N;A
41.0614
-111.5296
89.0
3.43
533
8.4
5.42E-03
1.87E-03
2.06E-03
9.35E-03
Table C-26b. Non-point Sources Included in the Collective Sources Tag for Utah.
The county in the vicinity of the maximum in-state contribution to mercury deposition with significant non-point emissions is included.
FIPS
49029
49029
49029
49029
49029
49029
49029
49029
49029
49029
County
MORGAN
MORGAN
MORGAN
MORGAN
MORGAN
MORGAN
MORGAN
MORGAN
MORGAN
MORGAN
Source Description
Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools,
Laboratories
Stationary Source Fuel Combustion, Commercial/Institutional, Distillate Oil
External Combustion Boilers, Industrial, Liquid Waste
Stationary Source Fuel Combustion, Residential, Bituminous/Subbituminous Coal
Stationary Source Fuel Combustion, Industrial, Residual Oil
Miscellaneous Area Sources, Other Combustion, Cremation
Stationary Source Fuel Combustion, Commercial/Institutional,
Bituminous/Subbituminous Coal
External Combustion Boilers, Industrial, Distillate Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Residual Oil
Stationary Source Fuel Combustion, Residential, Distillate Oil
HGO
(ipy)
2.38E-05
3.25E-06
1.02E-06
8.20E-07
7.50E-07
3.10E-07
5.50E-07
4.80E-07
1.90E-07
1.10E-07
HG2
»y)
O.OOE+00
1.95E-06
6.10E-07
4.90E-07
4.50E-07
8.30E-07
3.30E-07
2.90E-07
1.10E-07
7.00E-08
HGP
(ipy)
O.OOE+00
1.30E-06
4.10E-07
3.30E-07
3.00E-07
2.90E-07
2.20E-07
1.90E-07
7.00E-08
4.00E-08
Total Hg
(ipy)
2.38E-05
6.50E-06
2.04E-06
1.64E-06
1.50E-06
1.43E-06
1.10E-06
9.60E-07
3.70E-07
2.20E-07
C-38
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-27a. Point Sources Included in the Collective Sources Tag for Virginia.
FIPS
51710
51710
51740
51740
51740
51740
51550
51810
51710
51650
51710
51650
51810
51810
51710
51650
51550
51740
51740
51650
51810
51810
51650
51810
51710
County
Norfolk
Norfolk
Portsmouth
Portsmouth
Portsmouth
Portsmouth
Chesapeake
Virginia
Beach
Norfolk
Hampton
Norfolk
Hampton
Virginia
Beach
Virginia
Beach
Norfolk
Hampton
Chesapeake
Portsmouth
Portsmouth
Hampton
Virginia
Beach
Virginia
Beach
Hampton
Virginia
Beach
Norfolk
Only sources in the vicinity
Facility Name
STERICYCLEJNC.fFORMERLYAMERICANWASTEIND
STERICYCLEJNC.fFORMERLYAMERICANWASTEIND
SPSAWastetoEnergySteam&PowerPlant
SPSAWastetoEnergySteam&PowerPlant
SPSAWastetoEnergySteam&PowerPlant
SPSAWastetoEnergySteam&PowerPlant
CARGILLINC
HamptonRoadsSanitationDist.-AtlanticPlan
ThrasherDebrisLandfill
HamptonCityLF
HamptonRoadsSanitationDist.-VIP
WilliamsPavingCo-bigBethel
VirginiaBeachLandfillWasteManagement
CHRISTIANBROADCASTINGNETWORK
NavalBaseNorfolk
BethelLandfill
ChesapeakeCity/CivicCtrLF
NavalMedicalCenterPortsmouthfPWCN.)
NORFOLKVENEERMILLS
LangleyAFB
LittleCreekNavalAmphibiousBase
FCTCLANTDAMNECK
BigBethelRdLandfill/WilliamsPavingCo
OceanaNavalAirStation
NCTAMSLANTNorfolk
of the maximum
Plant
ID
MWI-39
MWI-39
LMWC-64
LMWC-64
LMWC-64
LMWC-64
15334
51810M001
LF3912
LF7087
517100197
LF7088
LF642
518100030
51710F002
LF7089
LF9070
517400007
14647
516500007
518100013
518100006
LF7085
518100004
51710F001
Point
ID
39-1
39-2
LMWC-
3
LMWC-
1
LMWC-
4
LMWC-
2
59766
A3326
A848
A1564
A3300
A1565
A1447
A3323
A3301
A1566
A6685
A3306
59329
A3290
A3328
A3324
A1562
A3330
A3304
in-state contribution to mercury
Stack
ID
89
90
35413
35411
35414
35412
35357
35454
35402
35371
35397
35372
35457
35450
35400
35373
35364
35408
35406
35367
35449
35447
35369
35445
35399
Latitude
36.8345
36.8345
36.8676
36.8676
36.8676
36.8676
36.8039
36.8369
36.7379
37.0411
36.8819
37.0764
36.7891
36.8031
36.8950
37.0411
36.7650
36.8442
36.8083
37.0706
36.9089
36.7717
37.0764
36.8092
36.8950
Longitude
-76.2727
-76.2727
-76.3770
-76.3770
-76.3770
-76.3770
-76.2856
-76.0256
-76.1955
-76.3623
-76.3172
-76.3833
-76.2020
-76.1936
-76.2590
-76.3623
-76.2860
-76.3064
-76.2956
-76.3608
-76.1753
-75.9644
-76.3833
-76.0381
-76.2590
StkHt
(m)
4.6
4.6
3.0
3.0
3.0
3.0
43.1
12.3
3.0
3.0
8.7
3.0
3.0
12.3
12.3
3.0
3.0
8.7
20.7
12.3
12.3
12.3
3.0
12.3
8.7
deposition
Diam
(m)
0.29
0.29
0.30
0.30
0.30
0.30
2.01
1.04
0.00
0.00
0.46
0.00
0.00
1.04
1.04
0.00
0.00
0.46
0.98
1.04
1.04
1.04
0.00
1.04
0.46
Temp
(K)
512
512
295
295
295
295
461
595
295
295
493
295
295
595
595
295
295
493
476
595
595
595
295
595
493
are included.
Vel
(mis)
3.2
3.2
4.6
4.6
4.6
4.6
10.3
25.6
0.0
0.0
19.2
0.0
0.0
25.6
25.6
0.0
0.0
19.2
10.4
25.6
25.6
25.6
0.0
25.6
19.2
HGO
(tpy)
USE-
OS
USE-
OS
7.1 1E-
04
3.81 E-
04
3.23E-
04
2.60E-
04
7.67E-
05
3.93E-
05
2.40E-
05
1.47E-
05
9.05E-
06
1.40E-
05
1.25E-
05
6.98E-
06
6.91 E-
06
6.66E-
06
5.79E-
06
3.58E-
06
3.27E-
06
2.86E-
06
1.89E-
06
1.37E-
06
2.00E-
06
1.09E-
06
1.01E-
06
HG2
»y)
2.18E-02
2.18E-02
1.87E-03
1.00E-03
8.53E-04
6.84E-04
4.60E-05
2.36E-05
3.01 E-06
1.84E-06
5.43E-06
1.75E-06
1.56E-06
4.19E-06
4.15E-06
8.30E-07
7.20E-07
2.15E-06
1.96E-06
1.72E-06
1.13E-06
8.20E-07
2.50E-07
6.50E-07
6.10E-07
HGP
»y)
5.81 E-03
5.81 E-03
6.46E-04
3.46E-04
2.94E-04
2.36E-04
3.07E-05
1.57E-05
3.01 E-06
1.84E-06
3.62E-06
1.75E-06
1.56E-06
2.79E-06
2.76E-06
8.30E-07
7.20E-07
1.43E-06
1.31 E-06
1.15E-06
7.60E-07
5.50E-07
2.50E-07
4.40E-07
4.00E-07
Total Hg
(tpy)
2.91 E-02
2.91 E-02
3.23E-03
1.73E-03
1.47E-03
1.18E-03
1.53E-04
7.87E-05
3.01 E-05
1.84E-05
1.81 E-05
1.75E-05
1.56E-05
1.40E-05
1.38E-05
8.32E-06
7.23E-06
7.16E-06
6.54E-06
5.73E-06
3.78E-06
2.74E-06
2.50E-06
2.18E-06
2.02E-06
C-39
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
FIPS
51810
51710
51810
51810
51810
51740
51650
51810
51550
51810
51740
51710
51800
51650
51550
51810
51700
51710
51550
51810
51740
County
Virginia
Beach
Norfolk
Virginia
Beach
Virginia
Beach
Virginia
Beach
Portsmouth
Hampton
Virginia
Beach
Chesapeake
Virginia
Beach
Portsmouth
Norfolk
Suffolk
Hampton
Chesapeake
Virginia
Beach
Newport
News
Norfolk
Chesapeake
Virginia
Beach
Portsmouth
Facility Name
FewlncDebrisLandfill
FISCNorfolk
LakesideConstruction
CentervilleTpWMtTrashmoreli
FORTSTORY
NorfolkNavalShipyard(PWCNorfolk)
FORTMONROE
NSGANorthwest
CHESAPEAKEGENERALHOSPITAL
VABEACHGENHOSPITAL
MARYVIEWHOSPITAL
NAVYPUBLICWKSCTR
HollandLandfill/Suffolk
WilliamsBigBethelRdLF
SOUTH EASTERNVATRAININGCENTER
WESTMINSTER-CANTERBURY
HamptonRoadsSanitationDist.-BoatH arbor
HamptonRoadsSanitationDist.-ArmyBase
MORSONINC
HamptonRoadsSanitationDist.-Chesapeake-E
St.Julien'sCreekAnnex(PWCNorfolk)
Plant
ID
LF640
517100204
LF641
LF643
518100005
517400006
516500052
51810F001
515500051
518100008
517400012
517100010
LF1254
LF7086
515500086
518100042
517000068
517100196
14633
518100034
51740F001
Point
ID
A1435
A3298
A4529
A4541
A3325
A3307
A3289
A3329
A3283
A3331
A3305
A3303
A156
A1563
A3287
A3332
A3296
A3299
59303
A3327
A3308
Stack
ID
35455
35398
35456
35458
35446
35407
35368
35453
35358
35448
35409
35395
35441
35370
35359
35452
35388
35396
35356
35451
35410
Latitude
36.8857
36.9406
36.7411
36.7411
36.9244
36.8156
37.0050
36.7411
36.7506
36.8650
36.8364
36.9469
36.8253
37.0764
36.7933
36.9094
36.9885
36.9214
36.7622
36.9067
36.8676
Longitude
-76.1836
-76.3258
-76.0480
-76.0480
-76.0081
-76.3006
-76.3036
-76.0480
-76.2411
-76.0264
-76.3492
-76.3194
-76.3196
-76.3833
-76.2247
-76.0800
-76.4212
-76.3258
-76.2281
-76.1664
-76.3770
StkHt
(m)
3.0
8.7
3.0
3.0
12.3
12.3
12.3
3.0
12.3
12.3
12.3
12.3
3.0
3.0
8.7
12.3
8.7
8.7
19.9
8.7
12.3
Diam
(m)
0.00
0.46
0.00
0.00
1.04
1.04
1.04
0.30
1.04
1.04
1.04
1.04
0.00
0.00
0.46
1.04
0.46
0.46
0.98
0.46
1.04
Temp
(K)
295
493
295
295
595
595
595
295
595
595
595
595
295
295
493
595
493
493
418
493
595
Vel
(m/s)
0.0
19.2
0.0
0.0
25.6
25.6
25.6
4.6
25.6
25.6
25.6
25.6
0.0
0.0
19.2
25.6
19.2
19.2
10.1
19.2
25.6
HGO
(tpy)
1.30E-
06
7.30E-
07
1.16E-
06
1.08E-
06
6.30E-
07
5.20E-
07
4.60E-
07
4.1 OE-
07
3.30E-
07
3.30E-
07
3.20E-
07
2.30E-
07
3.20E-
07
3.1 OE-
07
1.90E-
07
1.80E-
07
1.40E-
07
1.40E-
07
1.40E-
07
1.40E-
07
1.30E-
07
HG2
(tpy)
1.60E-07
4.40E-07
1.50E-07
1.40E-07
3.80E-07
3.10E-07
2.80E-07
2.50E-07
2.00E-07
2.00E-07
1.90E-07
1.40E-07
4.00E-08
4.00E-08
1.20E-07
1.10E-07
8.00E-08
8.00E-08
8.00E-08
8.00E-08
8.00E-08
HGP
(tpy)
1.60E-07
2.90E-07
1.50E-07
1.40E-07
2.50E-07
2.10E-07
1.80E-07
1.70E-07
1.30E-07
1.30E-07
1.30E-07
9.00E-08
4.00E-08
4.00E-08
8.00E-08
7.00E-08
6.00E-08
6.00E-08
6.00E-08
6.00E-08
5.00E-08
Total Hg
(tpy)
1.62E-06
1.46E-06
1.46E-06
1.36E-06
1.26E-06
1.04E-06
9.20E-07
8.30E-07
6.60E-07
6.60E-07
6.40E-07
4.60E-07
4.00E-07
3.90E-07
3.90E-07
3.60E-07
2.80E-07
2.80E-07
2.80E-07
2.80E-07
2.60E-07
C-40
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-27b. Non-point Sources Included in the Collective Sources Tag for Virginia.
Counties in the vicinity of the maximum in-state contribution to mercury deposition with significant non-point emissions are included.
FIPS
51810
51810
51810
51710
51710
51700
51810
51550
51710
51700
51710
51550
51550
51700
51700
51810
51650
51650
51650
51700
51550
51550
51740
51740
County
VIRGINIA BEACH
VIRGINIA BEACH
VIRGINIA BEACH
NORFOLK
NORFOLK
NEWPORT NEWS
VIRGINIA BEACH
CHESAPEAKE
NORFOLK
NEWPORT NEWS
NORFOLK
CHESAPEAKE
CHESAPEAKE
NEWPORT NEWS
NEWPORT NEWS
VIRGINIA BEACH
HAMPTON
HAMPTON
HAMPTON
NEWPORT NEWS
CHESAPEAKE
CHESAPEAKE
PORTSMOUTH
PORTSMOUTH
Source Description
Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
Stationary Source Fuel Combustion, Residential, Distillate Oil
Industrial Processes, Photo Equip/Health Care/Labs/Air
Condit/SwimPools, Laboratories
Stationary Source Fuel Combustion, Residential, Distillate Oil
Industrial Processes, Photo Equip/Health Care/Labs/Air
Condit/SwimPools, Dental Alloy (Mercury Amalgams) Production
Industrial Processes, Photo Equip/Health Care/Labs/Air
Condit/SwimPools, Dental Alloy (Mercury Amalgams) Production
Stationary Source Fuel Combustion, Commercial/Institutional, Distillate
Oil
Stationary Source Fuel Combustion, Residential, Distillate Oil
Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
Stationary Source Fuel Combustion, Residential, Distillate Oil
Industrial Processes, Photo Equip/Health Care/Labs/Air
Condit/SwimPools, Laboratories
Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
Industrial Processes, Photo Equip/Health Care/Labs/Air
Condit/SwimPools, Laboratories
Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
Industrial Processes, Photo Equip/Health Care/Labs/Air
Condit/SwimPools, Laboratories
Industrial Processes, Photo Equip/Health Care/Labs/Air
Condit/SwimPools, Dental Alloy (Mercury Amalgams) Production
Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total
Industrial Processes, Photo Equip/Health Care/Labs/Air
Condit/SwimPools, Laboratories
Stationary Source Fuel Combustion, Residential, Distillate Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Distillate
Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Distillate
Oil
Industrial Processes, Photo Equip/Health Care/Labs/Air
Condit/SwimPools, Dental Alloy (Mercury Amalgams) Production
Stationary Source Fuel Combustion, Residential, Distillate Oil
Industrial Processes, Photo Equip/Health Care/Labs/Air
Condit/SwimPools, Laboratories
HGO
(tpy)
1.51E-03
7.55E-04
1.43E-03
6.71 E-04
1.17E-03
1.07E-03
4.47E-04
4.17E-04
7.87E-04
3.81 E-04
7.45E-04
7.06E-04
6.69E-04
6.24E-04
5.91 E-04
4.86E-04
4.78E-04
4.53E-04
2.13E-04
2.05E-04
2.00E-04
3.89E-04
1.67E-04
3.24E-04
HG2
(tpy)
O.OOE+00
4.53E-04
O.OOE+00
4.02E-04
O.OOE+00
O.OOE+00
2.68E-04
2.50E-04
O.OOE+00
2.29E-04
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
1.28E-04
1.23E-04
1.20E-04
O.OOE+00
1.00E-04
O.OOE+00
HGP
(tpy)
O.OOE+00
3.02E-04
O.OOE+00
2.68E-04
O.OOE+00
O.OOE+00
1.79E-04
1.67E-04
O.OOE+00
1.52E-04
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
8.50E-05
8.21 E-05
8.01 E-05
O.OOE+00
6.68E-05
O.OOE+00
Total Hg
(tpy)
1.51E-03
1.51E-03
1.43E-03
1.34E-03
1.17E-03
1.07E-03
8.93E-04
8.33E-04
7.87E-04
7.62E-04
7.45E-04
7.06E-04
6.69E-04
6.24E-04
5.91 E-04
4.86E-04
4.78E-04
4.53E-04
4.25E-04
4.11 E-04
4.01 E-04
3.89E-04
3.34E-04
3.24E-04
C-41
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
FIPS
51650
51740
51740
51700
51810
51710
51700
51710
51550
51810
51700
51810
51710
51710
51650
51810
51710
51550
51810
51740
51650
51700
51550
51700
51550
51810
51650
51650
County
HAMPTON
PORTSMOUTH
PORTSMOUTH
NEWPORT NEWS
VIRGINIA BEACH
NORFOLK
NEWPORT NEWS
NORFOLK
CHESAPEAKE
VIRGINIA BEACH
NEWPORT NEWS
VIRGINIA BEACH
NORFOLK
NORFOLK
HAMPTON
VIRGINIA BEACH
NORFOLK
CHESAPEAKE
VIRGINIA BEACH
PORTSMOUTH
HAMPTON
NEWPORT NEWS
CHESAPEAKE
NEWPORT NEWS
CHESAPEAKE
VIRGINIA BEACH
HAMPTON
HAMPTON
Source Description
Industrial Processes, Photo Equip/Health Care/Labs/Air
Condit/SwimPools, Dental Alloy (Mercury Amalgams) Production
Industrial Processes, Photo Equip/Health Care/Labs/Air
Condit/SwimPools, Dental Alloy (Mercury Amalgams) Production
Stationary Source Fuel Combustion, Commercial/Institutional, Distillate
Oil
External Combustion Boilers, Industrial, Liquid Waste
Miscellaneous Area Sources, Other Combustion, Cremation
External Combustion Boilers, Industrial, Liquid Waste
External Combustion Boilers, Industrial, Distillate Oil
Miscellaneous Area Sources, Other Combustion, Cremation
Miscellaneous Area Sources, Other Combustion, Cremation
Stationary Source Fuel Combustion, Commercial/Institutional, Residual
Oil
Miscellaneous Area Sources, Other Combustion, Cremation
Stationary Source Fuel Combustion, Commercial/Institutional,
Bituminous/Subbituminous Coal
External Combustion Boilers, Industrial, Distillate Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Residual
Oil
Miscellaneous Area Sources, Other Combustion, Cremation
External Combustion Boilers, Industrial, Liquid Waste
Stationary Source Fuel Combustion, Commercial/Institutional,
Bituminous/Subbituminous Coal
External Combustion Boilers, Industrial, Liquid Waste
Stationary Source Fuel Combustion, Industrial, Residual Oil
Miscellaneous Area Sources, Other Combustion, Cremation
External Combustion Boilers, Industrial, Liquid Waste
Stationary Source Fuel Combustion, Commercial/Institutional, Residual
Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Residual
Oil
Stationary Source Fuel Combustion, Commercial/Institutional,
Bituminous/Subbituminous Coal
Stationary Source Fuel Combustion, Commercial/Institutional,
Bituminous/Subbituminous Coal
External Combustion Boilers, Industrial, Distillate Oil
Stationary Source Fuel Combustion, Industrial, Residual Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Residual
Oil
HGO
(tpy)
2.05E-04
2.05E-04
8.16E-05
7.55E-05
2.70E-05
4.26E-05
3.59E-05
1.41E-05
1.26E-05
2.54E-05
1.12E-05
2.25E-05
2.03E-05
1.99E-05
8.55E-06
1.91E-05
1.78E-05
1.59E-05
1.40E-05
6.13E-06
1.23E-05
1.17E-05
1.14E-05
9.79E-06
9.42E-06
9.07E-06
8.97E-06
8.96E-06
HG2
(tpy)
O.OOE+00
O.OOE+00
4.90E-05
4.53E-05
7.13E-05
2.56E-05
2.15E-05
3.71 E-05
3.33E-05
1.52E-05
2.94E-05
1.35E-05
1.22E-05
1.20E-05
2.26E-05
1.15E-05
1.07E-05
9.51 E-06
8.37E-06
1.62E-05
7.36E-06
7.01 E-06
6.83E-06
5.88E-06
5.65E-06
5.45E-06
5.38E-06
5.38E-06
HGP
(tpy)
O.OOE+00
O.OOE+00
3.27E-05
3.02E-05
2.46E-05
1.71 E-05
1.44E-05
1.28E-05
1.15E-05
1.02E-05
1.02E-05
9.01 E-06
8.11 E-06
7.97E-06
7.78E-06
7.64E-06
7.12E-06
6.34E-06
5.58E-06
5.57E-06
4.91 E-06
4.67E-06
4.56E-06
3.92E-06
3.77E-06
3.63E-06
3.59E-06
3.58E-06
Total Hg
(tpy)
2.05E-04
2.05E-04
1.63E-04
1.51E-04
1.23E-04
8.52E-05
7.18E-05
6.40E-05
5.75E-05
5.08E-05
5.08E-05
4.51 E-05
4.05E-05
3.99E-05
3.89E-05
3.82E-05
3.56E-05
3.17E-05
2.79E-05
2.79E-05
2.45E-05
2.34E-05
2.28E-05
1.96E-05
1.88E-05
1.82E-05
1.79E-05
1.79E-05
C-42
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
FIPS
51650
51550
51740
51740
51740
51710
51740
51740
51810
51710
51700
51550
51700
51650
51810
51740
51550
51550
51650
51740
51710
51810
51810
51550
51710
51700
County
HAMPTON
CHESAPEAKE
PORTSMOUTH
PORTSMOUTH
PORTSMOUTH
NORFOLK
PORTSMOUTH
PORTSMOUTH
VIRGINIA BEACH
NORFOLK
NEWPORT NEWS
CHESAPEAKE
NEWPORT NEWS
HAMPTON
VIRGINIA BEACH
PORTSMOUTH
CHESAPEAKE
CHESAPEAKE
HAMPTON
PORTSMOUTH
NORFOLK
VIRGINIA BEACH
VIRGINIA BEACH
CHESAPEAKE
NORFOLK
NEWPORT NEWS
Source Description
Stationary Source Fuel Combustion, Commercial/Institutional,
Bituminous/Subbituminous Coal
External Combustion Boilers, Industrial, Distillate Oil
External Combustion Boilers, Industrial, Liquid Waste
Stationary Source Fuel Combustion, Industrial, Residual Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Residual
Oil
Stationary Source Fuel Combustion, Residential,
Bituminous/Subbituminous Coal
Stationary Source Fuel Combustion, Commercial/Institutional,
Bituminous/Subbituminous Coal
External Combustion Boilers, Industrial, Distillate Oil
Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite
Coal
Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite
Coal
Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite
Coal
Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite
Coal
Stationary Source Fuel Combustion, Residential,
Bituminous/Subbituminous Coal
Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite
Coal
Stationary Source Fuel Combustion, Residential,
Bituminous/Subbituminous Coal
Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite
Coal
Industrial Processes, Chemical Manufacturing, Other Not Classified
Stationary Source Fuel Combustion, Residential,
Bituminous/Subbituminous Coal
Stationary Source Fuel Combustion, Residential,
Bituminous/Subbituminous Coal
Industrial Processes, Chemical Manufacturing, Other Not Classified
Stationary Source Fuel Combustion, Residential, Anthracite Coal
Industrial Processes, Chemical Manufacturing, Other Not Classified
Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling
Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling
Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling
Stationary Source Fuel Combustion, Residential, Anthracite Coal
HGO
(tpy)
7.80E-06
7.53E-06
6.72E-06
4.91 E-06
4.64E-06
4.33E-06
4.24E-06
3.20E-06
3.16E-06
2.49E-06
1.37E-06
1.32E-06
1.28E-06
1.09E-06
7.80E-07
6.00E-07
8.40E-07
4.50E-07
3.70E-07
2.10E-07
1.00E-07
1.40E-07
1.40E-07
7.00E-08
7.00E-08
3.00E-08
HG2
(tpy)
4.68E-06
4.52E-06
4.04E-06
2.95E-06
2.79E-06
2.60E-06
2.54E-06
1.92E-06
1.89E-06
1.49E-06
8.20E-07
7.90E-07
7.60E-07
6.50E-07
4.70E-07
3.60E-07
1.00E-07
2.70E-07
2.20E-07
3.00E-08
6.00E-08
2.00E-08
O.OOE+00
O.OOE+00
O.OOE+00
2.00E-08
HGP
(tpy)
3.12E-06
3.01 E-06
2.69E-06
1.97E-06
1.86E-06
1.73E-06
1.70E-06
1.28E-06
1.26E-06
1.00E-06
5.50E-07
5.30E-07
5.10E-07
4.40E-07
3.10E-07
2.40E-07
1.00E-07
1.80E-07
1.50E-07
3.00E-08
4.00E-08
2.00E-08
O.OOE+00
O.OOE+00
O.OOE+00
1.00E-08
Total Hg
(tpy)
1.56E-05
1.51E-05
1.35E-05
9.83E-06
9.29E-06
8.66E-06
8.48E-06
6.40E-06
6.31 E-06
4.98E-06
2.74E-06
2.64E-06
2.55E-06
2.18E-06
1.56E-06
1.20E-06
1.04E-06
9.00E-07
7.40E-07
2.70E-07
2.00E-07
1.80E-07
1.40E-07
7.00E-08
7.00E-08
6.00E-08
C-43
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
FIPS
51700
51650
51810
51710
51740
51550
51650
51650
51700
FIPS County
County
NEWPORT NEWS
HAMPTON
VIRGINIA BEACH
NORFOLK
PORTSMOUTH
CHESAPEAKE
HAMPTON
HAMPTON
NEWPORT NEWS
Table
Only sources
Facility Name Plant
ID
Source Description HGO
(tpy)
Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling 6.00E-08
Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling 5.00E-08
Stationary Source Fuel Combustion, Residential, Anthracite Coal 2.00E-08
Industrial Processes, Chemical Manufacturing, Other Not Classified 3.00E-08
Industrial Processes, Electrical Equipment, Fluorescent Lamp Recycling 3.00E-08
Stationary Source Fuel Combustion, Residential, Anthracite Coal 1.00E-08
Stationary Source Fuel Combustion, Residential, Anthracite Coal 1.00E-08
Industrial Processes, Chemical Manufacturing, Other Not Classified 2.00E-08
Industrial Processes, Chemical Manufacturing, Other Not Classified 1.00E-08
C-28a. Point Sources Included in the Collective Sources Tag for Wyoming.
HG2
(tpy)
O.OOE+00
O.OOE+00
1.00E-08
O.OOE+00
O.OOE+00
1.00E-08
1.00E-08
O.OOE+00
O.OOE+00
HGP
(tpy)
O.OOE+00
O.OOE+00
1.00E-08
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
Total Hg
(tpy)
6.00E-08
5.00E-08
4.00E-08
3.00E-08
3.00E-08
2.00E-08
2.00E-08
2.00E-08
1.00E-08
in the vicinity of the maximum in-state contribution to mercury deposition are included.
Point Stack ID Latitude Longitude StkHt Diam (m) Temp(K) Vel(m/s) HGO (tpy)
ID (m)
HG2(tpy)
HGP (tpy)
Total Hg (tpy)
56045
Weston BLACKHILLSOSAGE
58182
135277
4568
43.9600
-104.4000
38.1
2.96
446
27.1
1.24E-02
7.43E-03
C-44
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
Table C-28b. Non-point Sources Included in the Collective Sources Tag for Wyoming.
The county in the vicinity of the maximum in-state contribution to mercury deposition with significant non-point emissions is included.
FIPS
County
Source Description
HGO
(tpy)
HG2
(tpy)
HGP
(fry)
Total Hg
(fry)
56045
56045
56045
56045
56045
56045
56045
56045
56045
56045
56045
56045
WESTON Stationary Source Fuel Combustion, Residential, Bituminous/Subbituminous 1.59E-05 9.52E-06 6.35E-06 3.17E-05
Coal
WESTON Miscellaneous Area Sources, Fluorescent Lamp Breakage, Total 2.23E-05 O.OOE+00 O.OOE+00 2.23E-05
WESTON Industrial Processes, Photo Equip/Health Care/Labs/Air Condit/SwimPools, 2.11E-05 O.OOE+00 O.OOE+00 2.11E-05
Laboratories
WESTON Stationary Source Fuel Combustion, Residential, Distillate Oil 5.35E-06 3.21E-06 2.14E-06 1.07E-05
WESTON Stationary Source Fuel Combustion, Commercial/Institutional, Distillate Oil 3.82E-06 2.29E-06 1.53E-06 7.64E-06
WESTON Miscellaneous Area Sources, Other Combustion, Cremation 6.00E-07 1.59E-06 5.50E-07 2.74E-06
WESTON External Combustion Boilers, Industrial, Liquid Waste 8.20E-07 4.90E-07 3.30E-07 1.64E-06
WESTON Stationary Source Fuel Combustion, Industrial, Residual Oil 6.00E-07 3.60E-07 2.40E-07 1.20E-06
WESTON External Combustion Boilers, Industrial, Distillate Oil 3.90E-07 2.30E-07 1.60E-07 7.80E-07
WESTON Stationary Source Fuel Combustion, Commercial/Institutional, Residual Oil 2.20E-07 1.30E-07 9.00E-08 4.40E-07
WESTON Stationary Source Fuel Combustion, Commercial/Institutional, 1.80E-07 1.10E-07 7.00E-08 3.60E-07
Bituminous/Subbituminous Coal
WESTON Stationary Source Fuel Combustion, Commercial/Institutional, Anthracite 3.00E-08 1.00E-08 1.00E-08 5.00E-08
Coal
C-45
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix C: Details of Emissions Revisions for Selected States
References
EPA Region 9. 2007. E-mail correspondence from S. Chilingaryan with attachment
"speciationformodelingoct31.xls" dated October 31, 2007.
NDEP. 2007. E-mail correspondence from F. Forsgren with attachments '2006 Cumulative
Mercury Emissions Data Submittal Public.xls' and 'VMRP testing Overview6.pdf dated July 18,
2007
Western Environmental Services and Testing, Inc. 2006. "Mercury Emissions Test: Ontario
Hydro Method." (http://ndep.nv.gov/bapc/mercury/2006 Queenstake OHM.pdf.
C-46 August 2008
-------
Appendix D: Sensitivity of REMSAD Results
for Northern Utah to
Precipitation Amount
Introduction
This sensitivity analysis examines and quantifies the effects of rainfall amount on the REMSAD
simulation results. High amounts of rainfall over northern Utah, southwestern Wyoming, and
northwestern Colorado in the MM5-based files used as input to the REMSAD simulation result in
high mercury wet deposition in this area. The MM5-derived precipitation input files were reviewed
and it was found the simulated rainfall amounts to be higher than observed in the area of the
country referenced above. To examine the effect of this overestimation of precipitation, a
sensitivity simulation was conducted in which the rainfall amounts were reduced. Specifically, the
rainfall amount was reduced based on the monthly ratio of observed to simulated rainfall at six
meteorological monitoring sites. Adjustments were made to all grid cells within a rectangular area
including the area of interest. For consistency, rain liquid water content was also adjusted. (Rain
liquid water content is the amount of precipitable moisture in the air expressed as a mixing ratio.)
The sensitivity simulation was run for a three-month period that includes June, July, and August of
2001. The effects of reducing the rainfall and other moisture-related parameters were quantified
by comparing the simulated mercury deposition amounts to the original REMSAD simulation
results.
The methods and results of the sensitivity analysis are presented in this appendix.
Evaluation of the MM5-Based Meteorological Input Fields for Utah
The first step in this analysis was to examine how well the MM5-derived meteorological fields for
2001 represent the observed rainfall and other key meteorological parameters in northwestern
Utah and the surrounding areas. The MM5 inputs were used in the recent REMSAD-based
mercury deposition modeling analysis discussed in the main report. The evaluation
methodology for the meteorological inputs was based on that used for the entire U.S., as
discussed by Douglas et al. (2005).
Surface and upper-air meteorological monitoring sites were identified in the area of interest
(which includes Salt Lake City and the surrounding areas) and compared the simulated and
observed monthly average values of temperature, dew-point temperature, wind speed, wind
direction, rainfall amount, and several other derived parameters.
Statistical measures (bias, unsigned error, and RMS error) were also used to compare the daily values
of these parameters with observed data. These include:
Bias = 1/A/S(S;-O;)
Unsigned Error = 1/A/Z |S/-O/|
RMS Error = J| |
i=\
The surface and upper-air meteorological monitoring sites are listed in Table D-1.
D-1 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
Table D-1. Locations and Identifiers of Surface and Upper-Air Monitoring Sites
Used for the Statistical Evaluation of MM5 for Utah.
Site Name (ID)
Salt Lake City, UT (24127)
Logan, UT (94128)
Ogden, UT (24126)
Vernal, UT (94030)
Evanston, WY(04111)
Rock Springs, WY (24027)
Salt Lake City, UT (241 27)
Riverton,WY(24127)
Latitude
Surface Sites
40°47'N
41°47'N
41°12'N
40°26'N
41°16'N
41°36'N
Upper-Air Sites
40°47'N
43°04'N
Longitude
111°58'W
111°51'W
112° 01'W
109°33'W
111°02'W
109°04'W
111°58'W
108°28'W
Note that for Rock Springs, data are available for May through December only.
A focal point of this comparison is rainfall amount, since this parameter is important in the
simulation of mercury wet deposition. A visual comparison of plots of measured annual rainfall
amount (Figure D-1 a) and MM5 derived precipitation fields (Figure D-1b) shows relatively high
amounts of precipitation estimated by MM5 over northwestern Utah and along the
Utah/Colorado boundary. Similar high values of precipitation are not indicated by the
observations (Figure D-1 a). The observed precipitation plot was obtained from the National
Atmospheric Deposition Program (NADP) web site (http://nadp.sws.uius.edu) (NADP 2005).
Figure D-1 a. Observed Annual Rainfall Totals (cm) for 2001.
Total precipitation, 2001
Sftes not pictured:
AK01 35 cm
AK03 24 cm
H199 374 cm
VID1 105 cm
National Atmospheric Deposition ProgranVNational Trends Network
httpy/nadp, swsuiuc.edu
D-2
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
Figure D-1b. MM5-Derived Annual Rainfall Totals (cm) for 2001.
IflVEl. 1 RAIN (cm)
Time: 0 Apr II. 20!O 0 Apr 10, 2020
-2738. -8018 -1Z96
MAXIMUM - 514.1 cm (145,5)
MINIMUM 0.4 <;m (30.32)
an
Total Precipitation lor 2001 MM5 RCMSAD input file (cm)
For each of the meteorological monitoring sites listed in Table D-1, MM5-derived monthly and
annual mean values of a variety of surface and upper-air meteorological parameters were
calculated and compared with the observed mean values. Monthly and annual values of the
bias, unsigned error, and RMS error were also calculated. Table D-2 summarizes selected
annual metrics and statistical measures for each surface monitoring site.
Table D-2a. Summary of Annual Average Metrics and Statistical Measures for MM5
for the 2001 Annual Simulation Period: Salt Lake City, UT (Surface).
Parameter
Maximum Temperature (°C)
Minimum Temperature (°C)
Dew Point Temperature (°C)
Surface Wind Speed (ms-1)
Surface Wind Direction (degrees)
Total Daily Rainfall (in)
Mean Observed
17.78
6.89
0.73
0.78
166.81
0.04
Mean Simulated
11.91
2.34
-2.49
2.02
149.39
0.07
Bias
-5.87
-4.56
-3.21
0.9
2.65
0.03
Unsigned Error
6.08
4.70
3.73
1.52
59.58
0.06
D-3
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
Table D-2b. Summary of Annual Average Metrics and Statistical Measures
for the 2001
Parameter
Maximum Temperature (°C)
Minimum Temperature (°C)
Dew Point Temperature (°C)
Surface Wind Speed (ms-1)
Surface Wind Direction (degrees)
Total Daily Rainfall (in)
Annual Simulation
Mean Observed
15.47
0.04
-0.67
0.27
257.08
0.04
for MM5
Period: Logan, UT (Surface).
Mean Simulated
11.08
0.92
-2.94
1.21
166.8
0.06
Bias
-4.39
0.88
-2.27
1.83
1.51
0.02
Table D-2c. Summary of Annual Average Metrics and Statistical Measures
for the 2001
Parameter
Maximum Temperature (°C)
Minimum Temperature (°C)
Dew Point Temperature (°C)
Surface Wind Speed (ms-1)
Surface Wind Direction (degrees)
Total Daily Rainfall (in)
Annual Simulation
Mean Observed
16.85
6.03
-0.39
1.31
185.92
0.04
Period: Ogden, UT
Mean Simulated
11.99
2.06
-2.19
1.48
149.26
0.08
(Surface).
Bias
-4.86
-3.97
-1.8
0.65
0.08
0.04
Table D-2d. Summary of Annual Average Metrics and Statistical Measures
for the 2001
Parameter
Maximum Temperature (°C)
Minimum Temperature (°C)
Dew Point Temperature (°C)
Surface Wind Speed (ms-1)
Surface Wind Direction (degrees)
Total Daily Rainfall (in)
Annual Simulation
Mean Observed
16.23
1.79
-1.1
0.82
273.04
0.03
Period: Vernal, UT
Mean Simulated
10.92
1.14
-3.9
1.92
294.81
0.05
(Surface).
Bias
-5.32
-0.65
-2.8
1.87
0.91
0.02
Unsigned Error
5.44
3.32
3.35
1.90
80.26
0.05
for MM5
Unsigned Error
5.02
4.04
2.80
1.23
50.98
0.06
for MM5
Unsigned Error
5.55
2.72
3.32
1.95
50.66
0.05
D-4
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
Table D-2e. Summary of Annual Average Metrics and Statistical Measures for MM5
for the 2001 Annual Simulation Period: Evanston, WY (Surface).
Parameter
Maximum Temperature (°C)
Minimum Temperature (°C)
Dew Point Temperature (°C)
Surface Wind Speed (ms-1)
Surface Wind Direction (degrees)
Total Daily Rainfall (in)
Mean Observed
11.87
0.73
-4.03
2.22
233.81
0.03
Mean Simulated
11.43
1.01
-3.55
2.43
235.7
0.05
Bias
-0.44
0.28
0.48
0.4
0.32
0.02
Unsigned Error
1.74
1.66
2.21
0.89
33.04
0.05
Table D-2f. Summary of Annual Average Metrics and Statistical Measures for MM5
for the 2001 Annual Simulation Period: Rock Springs, WY (Surface).
Parameter Mean Observed Mean Simulated Bias Unsigned Error
Maximum Temperature (°C) 18 16.15 -1.86 2.43
Minimum Temperature (°C) 4.3 3.82 -0.48 2.11
Dew Point Temperature (°C) -3.3 -1.62 1.68 2.96
Surface Wind Speed (ms-i) 2.85 2.71 0.06 1.33
Surface Wind Direction (degrees) 243.62 252.15 0.12 29.91
Total Daily Rainfall (in) 0.02 0.06 0.04 0.05
On an annual basis, the comparison of the simulated and observed values show that MM5
tends to underestimate surface temperature and dew point temperature and overestimate
rainfall amounts. Surface wind speeds are also overestimated. Surface wind direction errors are
on the order of 30 to 80 degrees considering all sites.
Table D-3 summarizes the annual metrics and statistical measures for the two upper-air
monitoring sites for both sets of meteorological fields. For this analysis, the 700 mb level which
is approximately 3000 meters above sea level was examined. This level was chosen to account
for the high elevations in the area of interest.
D-5 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
Table D-3a. Summary of Annual Metrics and Statistical Measures for MM5
for the 2001 Annual Simulation Period: Salt Lake City, UT (Upper-Air).
Parameter
700 mb Temperature (am) (°C)
700 mb Temperature (pm) (°C)
700 mb Dew Point Temperature (am) (°C)
700 mb Dew Point Temperature (pm) (°C)
700 mb Wind Speed (am) (ms-1)
700 mb Wind Speed (pm) (degrees)
700 mb Wind Direction (am) (ms-1)
700 mb Wind Direction (pm) (degrees)
Mean Observed
1.62
2.16
-9.7
-10.19
4.87
4.62
260.85
245.84
Mean Simulated
2.33
2.37
-8.68
-7.51
3.79
4.01
252.12
241 .74
Bias
0.79
0.27
1.16
2.7
-0.73
-1.18
0.18
0.67
Unsigned Error
1.30
1.03
3.25
4.13
2.50
2.48
28.63
24.27
Table D-3b. Summary of Annual Metrics and Statistical Measures for MM5
for the 2001 Annual Simulation Period: Riverton, WY (Upper-Air).
Parameter Mean Observed Mean Simulated Bias Unsigned Error
700 mb Temperature (am) (°C)
700 mb Temperature (pm) (°C)
700 mb Dew Point Temperature (am) (°C)
700 mb Dew Point Temperature (pm) (°C)
700 mb Wind Speed (am) (ms-1)
700 mb Wind Speed (pm) (degrees)
700 mb Wind Direction (am) (ms-1)
700 mb Wind Direction (pm) (degrees)
0.92
2.49
-10.97
-11.14
5.33
4.98
276.74
269.68
1.34
1.95
-9.65
-9.78
4.65
4.27
288.36
281.74
0.44
-0.51
1.31
1.35
-0.39
-1.06
0.15
0.2
1.44
1.50
3.47
3.14
2.72
2.60
32.10
33.19
At the 700 mb level, temperatures and dew point temperatures are well represented and slightly
overestimated for both sites in the MM5 estimates. Wind speed errors are on the order of 2 to 3
ms"1 and wind speeds tend to be underestimated by MM5. Wind directions aloft are well
represented, with errors on the order of 25 to 35 degrees.
Thus, on an annual average basis, and for the sites considered here, MM5 has difficulty
representing the surface conditions and overestimates precipitation. The model appears to do a
good job representing upper-air temperature, moisture, and wind conditions.
Focusing in on rainfall, Figure D-2 compares observed and simulated annual rainfall totals for
the six surface sites considered in this analysis.
D-6 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
Figure D-2. Comparison of Observed and Simulated Annual Rainfall Amount (in).
SLC
Logan Ogden Vernal
Exranston Rock
Springs
This plot illustrates the overestimation of annual rainfall by MM5 in northern Utah and the
surrounding area.
Figure D-3 shows the variation in monthly mean rainfall amounts, observed and simulated by MM5.
Figure D-3a. Daily Average Rainfall Amount (in)
Based on Observed and Simulated Daily Precipitation Values: Salt Lake City, UT.
-P 0.2
Salt Lake City, UT
-Obs
-MM5
10 11 12
D-7
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
Figure D-3b. Daily Average Rainfall Amount (in)
Based on Observed and Simulated Daily Precipitation Values: Logan, UT.
-c 0.
Logan, UT
Obs
MM5
12
Figure D-3c. Daily Average Rainfall Amount (in)
Based on Observed and Simulated Daily Precipitation Values: Ogden, UT.
Ogden, UT
0bs
MM5
10 11 12
D-8
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
Figure D-3d. Daily Average Rainfall Amount (in)
Based on Observed and Simulated Daily Precipitation Values: Vernal, UT.
-? 0.2
Vernal, UT
-Obs
-MM5
10 11 12
Figure D-3e. Daily Average Rainfall Amount (in)
Based on Observed and Simulated Daily Precipitation Values: Evanston, WY.
Ev
fl 9
c v.^
n ifi
.E n 1?
2
o> n ns
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
Figure D-3f. Daily Average Rainfall Amount (in)
Based on Observed and Simulated Daily Precipitation Values: Rock Springs, WY.
= 0.16
I 0.12
o> 0.08
> 0.04
Rock Springs, WY
-Obs
-MM5
6 7
Month
10 11 12
Figure D-3 illustrates that most of the overestimation of rainfall occurs during the summer months,
but that there is also some higher than observed rainfall during the spring and fall transition
periods. MM5 most severely overestimates precipitation for Ogden, Vernal, and Rock Springs.
Adjustment Methodology
Based on these findings, a sensitivity simulation was conducted to examine and quantify the
effects of the overabundant rainfall on the REMSAD simulation results. Specifically, the
precipitation amounts were reduced over northern Utah, southwestern Wyoming, and
northwestern Colorado. The amount of reduction was based on the monthly ratio of observed to
simulated rainfall at the sites considered in this analysis. The rain liquid water content was also
reduced by a percentage amount equal to that used for the precipitation.
The area over which the moisture parameters were adjusted is shown in Figure D-4.
D-10
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
Figure D-4. REMSAD Grid Cells Corresponding to Northern Utah, Southwestern Wyoming, and
Northwestern Colorado.
I£VEL 1 RAIN (era)
Time; 0 Apr 11. 2010-0 Apr 10, 3030
-I MAXIMUM = &4A.4 cm (I4&.S)
- MINIMUM = 0.4 cm (30,33)
*a % aa'°o
^vvvvvvv
Total Precipitation for KO01 MM5 REMSAD input file (cm)
Adjustment factors, calculated as the ratio of the average observed to average MM5 rainfall for
all sites, are shown below. The factors were applied by month for June, July, and August to
precipitation and rain liquid water mixing ratio uniformly in all grid cells in this area in order to
compensate for overestimation by MM5.
Rock Avg. Adj.
Springs Factor
0.02 0.03 0.44
SLC Logan Ogden Vernal Evanston
June
July
August
Obs.
MM5
Obs.
MM5
Obs.
MM5
0.04
0.07
0.01
0.08
0.03
0.07
0.03
0.06
0.01
0.05
0.00
0.03
0.03
0.06
0.03
0.16
0.01
0.10
0.01
0.03
0.04
0.07
0.03
0.17
0.02
0.06
0.03
0.07
0.03
0.10
0.06
0.02
0.12
0.03
0.13
0.03
0.06
0.02
0.09
0.02
0.10
0.25
0.22
After applying the factors, minimum values were assigned as follows:
Water vapor (mixing ratio): 0.0001 kg/kg
Rain liquid water mixing ratio:
0.0001 kg/kg
Figure D-5 illustrates the results of applying the reduction factors.
D-11
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
Figure D-5. Daily Average Rainfall Amount (in):
Observed, Simulated (MM5), and Adjusted (Adj).
o:
rf
£ n
Salt Lake City, UT
-*-Obs -B-MM5 Adj
678
Month
Vernal, UT
o:
fi
ro n
Q OH
Logan, UT
--Obs -H-MM5 Adj
m
0
678
Month
Evanston, WY
-Obs
-MM5
Adj
£ U. 10
o:
fi
Tin 04 -
n
Ogden, UT
-»-Obs -B-MM5 Adj
^^^*~~~~~~~---~^
^^ ^~~~~B
w^
"~ * ^
6
7
Month
8
Rock Springs, WY
REMSAD Simulation Results
The summer season consisting of the months June, July, and August 2001 was simulated with
REMSAD using the adjusted meteorological fields and the base emissions files. A limited
number of tags were simulated including the CTM background tag, re-emissions, and four Utah
tags: Intermountain Power, Ash Grove, Hunter, Nucor Steel, and UT Collective Sources.
Dry, wet, and total mercury deposition for the summer period for both the base simulation using
unadjusted meteorology and the sensitivity simulation using adjusted meteorology are
presented in Figures D-6 through D-8. For the sensitivity case, the area over which rainfall was
adjusted in outlined in blue. Small differences are present in the dry deposition distributions,
presumably due to the changes in overall air concentration resulting from the reduced wet
deposition. The largest changes in wet deposition are several g/km2, and these changes carry
over into the change in total deposition. The change in wet deposition can be more clearly seen
in the plot in Figure D-9, which shows the calculated difference of the summer deposition from
D-12
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
the sensitivity simulation minus the summer deposition in the base simulation. (The adjusted
area is again outlined in blue.) The largest reduction in wet deposition occurs in eastern Nevada
near the location of goldmine operations in that state. Reductions of 1 or 2 g/km2 are present
throughout most of the area of Utah in which the rainfall was adjusted. Interestingly, there is an
increase in wet deposition just east of the area in which the rainfall was adjusted, implying that
reduced removal of mercury in the target area might lead to more mercury in the atmosphere
immediately downwind of the target area compared to the baseline and hence higher deposition
in that downwind area.
Figure D-6. Summer Dry Deposition of Mercury: (a) Base Simulation; (b) Sensitivity Simulation with
Reduced Rainfall.
LKVKI. I THU (4£/ltm2)
(a)
MAXIMUM - B2.&a R/km2 ! JU.HiJJ
- UIMUUU - O.SB ft/kml! (100.47)
LEVEL 1 TUG (
(b)
. UAXtMLM - KM I «/hmii I-SU.BU)
- UIMUUU - 050 g/kmZ (flM IHIJ
l dry deE^oail iari «f THG I Summer 2001
rtl tJry t
nf TUG I Summer
D-13
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
Figure D-7. Summer Wet Deposition of Mercury: (a) Base Simulation; (b) Sensitivity Simulation with
Reduced Rainfall.
(a)
UAXIWI:M - I4Z4 n/xms (4i.no)
UIMMUU ^ QUO K/km2 (1 II)
(b)
LEVEL I TUG (
12 10 |t/kmK (39.110)
MIMMLIM - 0.00 /kn^ 11.1 I)
1300 lose me
i _ri ,
too ixo
To till wet deposition of THG_1 - - Summer 2OQ]
Total wel deposition of T!IG_1 - - Summer 2001
D-14
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
Figure D-8. Summer Total Deposition of Mercury: (a) Base Simulation; (b) Sensitivity Simulation with
Reduced Rainfall.
(a)
LbiVKL 1 TUG (fi/km2>
1-tQ.noj
km^ (05.4OJ
(b)
MAXIMUM - 60.09 t/kmS (40.130)
- MINIMI >u * I ttZ s/km2 {95.44)
Total weLtdry deposition of THG_I -- Summer 2001
Ml w*1 i ttry depoxit KJII nf TMt! I - - Summer 2001
D-15
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
Figure D-9. Difference in Summer Wet Deposition of Mercury: Sensitivity Simulation minus Base
Simulation.
LEVEL 1 THG (g/km3)
( MAXIMUM = 2.9 g/km2 (104,77)
- MINIMUM - -8.7 g/km2 (41.80)
Km
The above displays show that reduced rainfall, as expected, reduces wet deposition. It is also of
interest to know whether the contributors to deposition also change due to alterations in the wet
deposition patterns. Figure D-10a and D-10b present a summary of contributions to mercury
deposition for summer 2001 in the base run and in the sensitivity simulation. These summaries
are for the same grid cell, the location of the maximum in-state contribution to mercury
deposition for Utah that was presented in the main report (see Figures 7-42a and 7-42b; Figure
7-42a is repeated here as Figure D-10c).
D-16
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation Amount
Figure D-10a. Utah. Analysis of Summer Deposition with Base Rainfall for the Single Grid Cell (the Blue Triangle in the Figure 7-42a)
Where In-State Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (Summer deposition: 9.3 g/km2).
Utah
Emissions vs. Background Contributions (g/km2) (1,2,3)
Wet vs. Dry Contributions (g/km2) (1,2)
8 -,
7 -
6
5
4 -
-a
2
1 -
3.0
Emiss
6.3
CTM
REMSAD
CTM Bckgnd
(REMSAD)
Total (REMSAD)
Contributions to Total Deposition (2)
Contributions from Utah Sources without Background (4,5)
DCTM Background
(REMSAD) 67.6%
I Utah 26.1%
D Other 4.4%
IReemission 1.8%
JV
2nd pie
Other sources (19.3%)
Tagged UT sources
(80.7%)
DUT Collective Sources
96.3%
DUT Intermountain Power
1.7%
DUT Ash Grove 1.5%
UTNucor Steel 0.3%
OUT Hunter 0.2%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
5) This sensitivity simulation used a limited number of Utah tags. Sources such as Davis/Wasatch are included in "Other sources".
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition from sources located within
D-17
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation Amount
that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
D-18 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation Amount
Figure D-1 Ob. Utah. Analysis of Summer Deposition with Reduced Rainfall for the Single Grid Cell (the Blue Triangle in the Figure 7-42a)
Where In-State Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (Summer deposition: 5.7 g/km2).
Utah
Emissions vs. Background Contributions (g/km2) (1,2,3)
7 -
6 -
5 -
4 -
3
2 -
1 -
0 -
2.6
Emiss
CTM
REMSAD
Wet vs. Dry Contributions (g/km2) (1,2)
CTM Bckgnd
(REMSAD)
4.0
Total (REMSAD)
Contributions to Total Deposition (2)
Contributions from Utah Sources without Background (4,5)
DCTM Background
(REMSAD) 53.3%
I Utah 38.8%
D Other 6.4%
IReemission 1.4%
2nd pie
Other sources (16.8%)
Tagged UT sources
(83.2%)
~[IUT"ColTective~Sburces~
98.0%
UT Intermountain Power
0.8%
OUT Ash Grove 0.8%
UT Hunter 0.2%
DUTNucor Steel 0.2%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
5) This sensitivity simulation used a limited number of Utah tags. Sources such as Davis/Wasatch are included in "Other sources".
Contributions to mercury deposition are displayed only for one grid cell within the state, i.e., the grid cell of greatest deposition
from sources located within that same state. Results should not be extrapolated to indicate source contributions on a statewide basis.
D-19
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
Figure D-10c. REMSAD-simulated Total (Wet and Dry) Annual Mercury Deposition (g km ) for Utah.
i.KVKL 1 T1IC (g/kinH)
Time: 0 ian 1. S001-0 Dec 31, 3001
t MAXIMUM - S55A f/kmZ (8,56)
- MINIMUM - 5.0 g/km2 (-17,62)
Annim! l.otal wet. (dry deposition of TUG
within Utah
2001
Differences are clearly visible between the two sets of charts. The change in the contribution
from background concentrations is greater than change in the contribution from emissions
sources. Consequently, the relative proportions of background and emissions contributions are
different between the base and sensitivity simulations. The most important emissions contributor
(UT collective sources) is the same in both cases.
The particular grid cell examined in Figure D-10 includes a large contribution from emissions
sources. In order to see how contributions are affected in an area not so strongly affected by
emissions, another grid cell located to the southwest of the peak was also examined. The
location of this grid cell is denoted by the blue triangle in Figure D-11.
D-20
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
Figure D-11. Alternate Location for Analysis of Deposition Contributions.
LEVEL 1 THG
+ MAXIMUM = 73.B4 g/km2 (3,56)
- MINIMUM = 1-57 K/km2 (39,5)
V ty,°Oe
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation Amount
Figure D-12a. Utah. Analysis of Summer Deposition with Base Rainfall for the Single Grid Cell (the Blue Triangle in the Figure 7-42a)
Where In-State Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (Summer deposition: 8.8 g/km2).
Utah
Emissions vs. Background Contributions (g/km2) (1,2,3)
7 -
6
5
4 -
3
2
1 -
0.8
I I
Emiss
CTM
REMSAD
Wet vs. Dry Contributions (g/km2) (1,2)
CTM Bckgnd
(REMSAD)
Total (REMSAD)
Contributions to Total Deposition (2)
Contributions from Utah Sources without Background (4,5)
DCTM Background
(REMSAD) 91.1%
I Utah 2.7%
DOther 3.0%
DReemission 3.2%
1st pie
2nd pie
Other sources (69.7%)
Tagged UT sources
(30.3%)
"tTuTl ntermou ntai n" Power
43.0%
UT Collective Sources
30.5%
OUT Ash Grove 24.0%
UTNucor Steel 2.2%
OUT Hunter 0.3%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
5) This sensitivity simulation used a limited number of Utah tags. Sources such as Davis/Wasatch are included in "Other sources".
Contributions to mercury deposition are displayed only for a single, arbitrarily chosen grid cell within the state. Results should not be extrapolated to
indicate source contributions on a statewide basis.
D-22
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation Amount
Figure D-12b. Utah. Analysis of Summer Deposition with Reduced Rainfall for the Single Grid Cell (the Blue Triangle in the Figure 7-42a)
Where In-State Sources Contributed the Most to Simulated Annual Total Mercury Deposition for 2001 (Summer deposition: 5.7 g/km2).
Utah
Emissions vs. Background Contributions (g/km2) (1,2,3) Wet vs. Dry Contributions (g/km2) (1,2)
8 -,
7 -
6
5
4 -
3
2
1 -
0.6
i |
Emiss
5.1
CTM
REMSAD
Emissions
CTM Bckgnd
(REMSAD)
3.9
1.9
Total (REMSAD)
Contributions to Total Deposition (2)
Contributions from Utah Sources without Background (4,5)
DCTM Background
(REMSAD) 89.7%
I Utah 3.3%
DOther 3.7%
IReemission 3.4%
Jf
2nd pie
Other sources (68.3%)
Tagged UT sources
(31.7%)
[JUT Intermountain Power
45.1%
UT Collective Sources
26.7%
OUT Ash Grove 26.1%
UTNucor Steel 1.6%
OUT Hunter 0.5%
Notes: 1) "Emissions" refers to emissions contributions from the U.S., Canada, and Mexico as simulated by REMSAD.
2) "Background" refers to the effects of initial and boundary concentrations and embodies the effects of global emissions.
3) The CTM, GRAHM (GRHM), and GEOS-Chem (G-C) global models provided boundary conditions for REMSAD and CMAQ, as discussed in Section 5.2.
4) Percentages of cut-out pie segments are calculated based on total represented in only the cut-out pie.
5) This sensitivity simulation used a limited number of Utah tags. Sources such as Davis/Wasatch are included in "Other sources".
Contributions to mercury deposition are displayed only for a single, arbitrarily chosen grid cell within the state. Results should not be extrapolated to
indicate source contributions on a statewide basis.
D-23
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
The modeling results suggest that, even for national-scale applications, accurate representation
of regional meteorological conditions, especially the timing and amount of rainfall, is key to
reliable mercury deposition modeling.
Future modeling efforts for this simulation period, should consider the use of improved or
alternative meteorological inputs. A diagnostic evaluation of MM5 for the western U.S. should be
performed and the MM5 options pertaining to grid resolution, cloud parameterizations, and
moist physics, radiation, and soil temperature schemes should be examined. It is possible that
the options selected for the full (continental-scale) application of MM5 could be refined for the
west and specifically for the Utah/Colorado/Wyoming area. Alternatively, the inputs could be
prepared using other meteorological modeling tools such as WRF or RUC.
References
Douglas, S., B. Hudischewskyj, S. Beckmann, and T. Myers. 2005. "Comparison of MM5- and
RUC-Based Meteorological Input Fields for REMSAD Mercury Modeling (Revised)",
Memorandum to Ruth Chemerys and Dwight Atkinson, EPA Office of Water. Prepared by
ICF International, San Rafael, California.
National Atmospheric Deposition Program (NRSP-3). 2005. NADP Program Office, Illinois State
Water Survey, 2204 Griffith Dr., Champaign, Illinois, 61820.
D-24 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix D: Sensitivity of REMSAD Results for Northern Utah to Precipitation
Amount
This page deliberately left blank.
D-25 August 2008
-------
Appendix E: Use of Classification and
Regression Tree (CART) Analysis to
Examine the Impact of Year-to-Year
Meteorological Variability on
Mercury Deposition for Several
Locations throughout the U.S.
Introduction
The REMSAD simulations reported in the main body of this report are limited to a single year (2001),
largely due to the level of effort involved in preparing the necessary inputs and conducting regional-
scale mercury deposition modeling. It is of interest, however, to have some estimate of how much the
simulation results might vary due to differing meteorological conditions that are present in other
years. In this companion study, statistical analysis is used to estimate the potential year-to-year
variability in the simulation results due to variations in meteorology.
The technique examines the annual variability in wet and dry mercury (Hg) deposition for several
locations throughout the U.S., based on the integrated analysis of observed meteorological data
and mercury deposition modeling results. The objective of this study was to use the meteorological
and deposition information available from the mercury deposition modeling study described in the
main body of this report to estimate the variability in deposition for a ten-year period for selected
locations of interest. Since mercury wet deposition data are typically not available for a full ten-year
period for most locations, it was of interest to use the available simulation results and a longer
period of record of meteorological data to estimate variability in wet deposition. Dry deposition
measurements are also typically not available, except during short-term special monitoring
programs, so this technique also provides a means to estimate the amount and variability of dry
deposition over a multi-year period.
The modeling results used for this analysis are from the national-scale application of the Regional
Modeling System for Aerosols and Deposition (REMSAD) discussed in the main body of this report.
The modeling was conducted for the EPA Office of Water (OW). REMSAD was applied to the study
of mercury transport and deposition and to quantify the contributions of specific sources and source
categories to mercury deposition within the contiguous 48 states. REMSAD simulates wet and dry
mercury deposition. Wet deposition occurs as a result of precipitation scavenging. Dry deposition is
calculated for each species based on land-use characteristics and meteorological parameters.
REMSAD also includes algorithms for the re-emission of mercury into the atmosphere from land and
water surfaces, due to naturally occurring (e.g., microbial) processes. REMSAD provides estimates
of the concentrations and deposition of mercury and all other simulated pollutants at each grid
location in the modeling domain. For the OW study, the REMSAD modeling domain encompasses
the contiguous 48 states with a horizontal grid spacing of approximately 12 km. REMSAD was
applied for the annual simulation period 2001.
The modeling results were used in conjunction with observed data and the Classification and
Regression Tree (CART) analysis technique to estimate wet and dry mercury deposition for fifteen
selected locations for each year of the ten-year period 1997-2006. CART was used in combination
with the REMSAD meteorological input fields and simulated deposition values to match the
deposition values to specific meteorological parameters. The CART results were used to identify
E-1 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
the meteorological conditions associated with certain ranges of deposition. Each day for the one-
year simulation period was placed into a classification bin based on the values of the
meteorological input parameters and the simulated deposition. The resulting bins contained days
with similar deposition values and meteorological characteristics.
The CART results then provided the framework for estimating the deposition characteristics for all
days within the ten-year period of record. Actual meteorological data were obtained for each day
and based on these data the days were placed in the CART bins (as derived using the REMSAD
inputs and outputs as described above). Each day for the ten-year period was then assigned a
deposition value, which was equal to the calculated average for the similar (REMSAD-simulated)
days. In this manner, estimated daily totals of Hg deposition for each day of the ten-year periods
were obtained. These were summed to provide annual totals, from which the variance could be
calculated.
The deposition analysis was conducted for 15 locations throughout the continental U.S,
representing different climate zones and different areas within the climate zone. For each location,
the analysis was conducted separately for wet and dry Hg deposition.
The results are intended to provide perspective to the analysis and use of the REMSAD simulation
results for the 2001 annual simulation period and for use in water quality modeling. When
examining the results, the reader should keep in mind that the purpose of this analysis was only to
investigate the impact on deposition of year-to-year variability due to meteorological influences.
Other important factors that would affect changes in deposition from one year to another were
beyond the scope of this analysis. Most notably, the year-to-year changes in emissions from U.S.
and other global sources were not accounted for and were beyond the scope. Instead, the analysis
includes the implicit assumption that all emissions are constant throughout the 10-year CART study
period.
The methods and results of the CART-based mercury deposition analysis are presented in the
remainder of this appendix. The following three sections address site selection, CART analysis
methods and results, and use of the CART results to estimate mercury deposition. A summary and
some conclusions are presented in the final section.
The CART analysis method used here was first tested in the context of mercury deposition variability
due to meteorological parameters as part of the Devil's Lake TMDL Pilot (Myers et al., 2003).
Application of CART in this report is consistent with that earlier peer reviewed methodology.
Site Selection
The mercury deposition analysis was conducted for 15 sites, located throughout the continental
U.S, representing different climate zones and different regions within each climate zone. The site
selection procedures and the selected locations are summarized in this section.
SITE SELECTION PROCEDURES
The National Climatic Data Center (NCDC) divides the contiguous U.S. into nine regions in order to
summarize climate data and information and to analyze climate anomalies in the context of an
historical perspective. These regions have been identified through climate analysis as climatically
consistent, especially with respect to temperature and precipitation (NCDC, 2007; Karl and Koss,
E-2 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
1984). The regions include Northwest, West, Southwest, West North Central, East North Central,
Central, South, Southeast, and Northeast and are shown in Figure E-1.
Figure E-1. U.S. Climate Regions for Temperature and Precipitation (Source: NCDC, 2008).
U.S. Standard Regions
for Temperature & Precipitation
National Climatic Data Center, NOAA
The normal temperature and precipitation amounts for these regions (based on thirty years of
observed data for 1961-1990) are listed in Table E-1, along with the national average values.
Table E-1. Normal Precipitation and Temperature for the Nine NCDC Climate Regions and the
Contiguous U.S., Based on Observed Data for 1961-1990.
Zone/Region
Northwest
West
Southwest
West North Central
East North Central
Central
South
Southeast
Northeast
National
Normal Precipitation
(in)
27.5
16.5
13.6
16.9
30.5
43.1
35.7
51.0
41.6
29.5
Normal Temperature
(°F)
46.7
55.0
51.8
43.3
43.5
53.2
62.0
62.4
46.1
52.4
E-3
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
For the mercury deposition study, the goal was to identify one or more locations within each of the
climate zones for further analysis, while keeping the total number of locations to within 15. In order
to have some mercury deposition data to verify the reasonableness of the results, the focus was on
the locations of Mercury Deposition Network (MDN) sites and identified several MDN sites to
represent each climate zone. To refine this list, the following were also considered:
Variability in geography within each climate zone,
Observed and/or simulated (REMSAD) mercury deposition (specifically, locations with high
deposition), and
Data availability for each MDN site.
Data availability during the analysis period was considered so that there would be some observed
data to compare with the results. This was done to ensure that the estimated values were within a
reasonable range, compared to the observed data. The selected sites/locations are listed in Table
E-2 and plotted in Figure E-2.
E-4 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
Table E-2. Mercury Deposition Analysis Locations, by Climate Region.
Site# MDN
Site ID
1 WA03
2 ID03
3 CA75
4 NV99
5 UT97
6 C099
Site Name
Makah National Fish Hatchery
Craters of the Moon National
Monument
Sequoia National Park-Giant
Forest
Gibb's Ranch
Salt Lake City
Mesa Verde National Park-
Chapin Mesa
Start Date Status Elev
(m)
Region: Northwest
3/2/2007 Active 6
10/20/2006 Active 1807
Region: West
7/22/2003 Active 1902
2/13/2003 Active 1805
Region: Southwest
5/16/2007 Active 1297
12/26/2001 Active 2172
Lat (deg) Lon (deg)
48.2892 -124.6519
43.4605 -113.5551
36.5661 -118.7776
41.5516 -115.2132
40.7118 111.9609
37.1981 -108.4903
Region: West North Central
7 SD18
Eagle Butte
3/21/2007 Active 742
44.9931 -101.2403
Region: East North Central
8 MN23
9 IN20
10 TN11
11 OK99
12 LA28
13 FLOS
14 NCOS
15 NJ30
Camp Ripley
Roush Lake
Great Smoky Mountains
National Park-Elkmont
Stilwell
Hammond
Chassahowitzka National
Wildlife Refuge
Waccamaw State Park
New Brunswick
7/2/1996 Active 410
Region: Central
10/26/2000 Active 244
1/30/2002 Active 640
Region: South
4/29/2003 Active 304
10/7/1998 Active 9
Region: Southeast
7/1/1997 Active 3
2/27/1996 Active 10
Region: Northeast
1/17/2006 Active 21
46.2494 -94.4972
40.8400 -85.4639
35.6645 -83.5903
35.7514 -94.6717
30.5031 -90.3769
28.7486 -82.5551
34.2592 -78.4777
40.4728 -74.4226
E-5
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
Figure E-2. Mercury Deposition Analysis Locations, by Climate Region.
U.S. Standard Regions
for Temperature & Precipitation
National Climatic Data Center, NOAA
DEPOSITION SITES, CHARACTERISTICS AND METEOROLOGICAL SITE PAIRS
Two sites represent the Northwest climate zone. The Makah National Fish Hatchery site (WA03)
became operational in March 2007. This site represents the coastal portion of the Northwest
climate zone. The Craters of the Moon National Monument site (I DOS) became operational in
October 2006. This site represents the interior portion of the Northwest climate zone.
Two sites represent the West climate zone. The Sequoia National Park site (CA75) became active
in July 2003 and the Gibb's Ranch site (NV99) was established in February 2003. Both sites are
high elevation sites. NV99 is near several potential sources of mercury (mines) and also near an
area of high simulated mercury deposition in the REMSAD modeling.
Two sites represent the Southwest climate zone. The Salt Lake City site (UT97) is a new site and
was established in July 2007. This site is of interest because it is near several potential sources of
mercury (mines) and also near an area of high simulated mercury deposition in the REMSAD
modeling. The Mesa Verde National Park site (CO99) has a longer period of record with data going
back to December 2001. It is a high elevation site and is centrally located within the Southwest
climate zone.
The Eagle Butte site (SD18) is nearly centrally located in the West North Central climate zone. It is
a new site (March 2007) and one of only a few sites in this region.
The Camp Ripley (MN23) site represents the East North Central climate zone. It was established in
July 1996 and thus the data for this site fully overlap with the analysis period. This site is located
within a region characterized by mercury-sensitive water bodies (especially in Minnesota and
Wsconsin).
E-6
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
Two sites represent the Central climate zone. The Roush Lake site (IN20) was established in
October 2000. Further south within this climate zone and at a higher elevation is the Great Smoky
Mountains National Park site (TN11). This site was established in January 2002.
Two sites represent the South climate zone. The Stilwell site (OK99) became operational in April
2003. This site represents the interior (plains) portion of the South climate zone. The Hammond
site (LA28), established in October 1998, represents the coastal Gulf of Mexico region (a mercury
sensitive area).
Two sites represent the Southeast climate zone. The Chassahowitzka site (FLOS) is also along the
Gulf of Mexico (to the east of the Gulf) and represents Florida, another mercury sensitive area.
Further north in the Southeast climate zone is Waccamaw State Park (NCOS) which is located
along the Atlantic coast. Both sites have long periods of record that nearly fully overlap the analysis
period.
Finally, the New Brunswick site (NJ30) is located in the Northeast climate zone. It was established
in January 2006. As the representative site for the northeastern U.S, it is expected to represent
sites that are potentially influenced by local and regional sources of mercury as well as long-range
transport.
For the purposes of conducting the CART analysis, each site was paired with one surface and one or
more upper-air meteorological monitoring sites. The meteorological data for these sites were used to
match each day for the period 1997-2006 to the CART classification bins (one wet deposition and
one dry deposition bin) to obtain estimates of mercury deposition for that day. The meteorological site
pairs are listed in Table E-3, and for each site the elevation and distance from the MDN site are
given. It is assumed that upper-air data within approximately 300 km of a site can be used to
represent wind, temperature, and moisture conditions aloft. This assumption is consistent with the
spacing of the National Weather Service (NWS) upper-air sites.
E-7 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
Table E-3. MDN and Meteorological Site Pairings for Use in Estimating Mercury Deposition
Based on Meteorological Parameters.
Site# MDN Site
ID
1 WA03
2 ID03
3 CA75
4 NV99
5 UT97
6 C099
7 SD18
8 MN23
9 IN20
10 TN11
11 OK99
12 LA28
13 FLOS
14 NCOS
15 NJ30
Surface Meteorological
Site WBAN No.
94240
24156
23157
4114
24127
93069
94056
94938
14827
13891
93993
12916
12818
13748
14734
Elev Distance Upper-Air Meteorological Site Name Distance
(m) (km) (WBAN No.) (km)
Region: Northwest
59 38.5
1353 100.7
Region: West
1264 94.7
1650 83.1
Region: Southwest
1287 7.9
1803 17.0
Region: West North Central
787 61.3
Region: East North Central
374 32.6
Region: Central
241 28.4
293 39.3
Region: South
381 53.2
1 57.3
Region: Southeast
23 33.0
14 48.7
Region: Northeast
2 33.2
Quillayute, WA (94240)
Boise, ID (24131)
Oakland, CA (23230)
Elko, NV (4105)
Salt Lake City, UT (24127)
Grand Junction, CO (23066)
Rapid City, SD (94043)
Minneapolis, MN (94983)
Wilmington, OH (13841)
Atlanta, G A (5381 9)
Nashville, TN (13897)
Greensboro, NC (13723)
Springfield, MO (13995)
Little Rock, AR (3952)
Norman, OK (3948)
Slidell, LA (53813)
Tampa Bay, FL (12842)
Moorehead City, NC (93768)
Brookhaven, NY (94703)
38.5
215.2
330.4
87.4
6.5
213.9
188.1
174.8
212.0
272.9
275.3
330.7
199.5
242.2
260.9
56.9
117.7
161.2
137.9
Note that for all sites, the MDN data consist of wet deposition measurements only. The available
data indicate that there is considerable year-to-year variability in wet mercury deposition. For
example, annual wet deposition for five of the selected MDN sites is plotted in Figure E-3, for the
E-8
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
period 2002-2005. For these five sites, the quarterly standard deviation from the mean represents
from 7 to 26 percent of the mean for each site, which indicates significant variability. Note that this
variability in the observed data is attributable to a combination of meteorological and emissions
factors. In this study, only the meteorological variability is addressed, and the goal was to extend
the analysis of variability to cover the ten-year period and both wet and dry mercury deposition for
all of the selected locations. When the data collection periods overlap with the analysis period, the
wet deposition measurements are used as a basis for assessing the reasonableness of the CART-
based deposition estimates (both in terms of the overall magnitude of wet deposition and that
portion of the year-to-year variability of the deposition estimates attributable to meteorological
influences).
Figure E-3. Annual Mercury Wet Deposition (g km"2) for 2002-2005 for the Mesa Verde (CO99), Camp
Ripley (MN23), Great Smoky Mountains (TN11), Hammond (LA28) and Waccamaw State Park (NCOS)
MDN Monitoring Sites.
2002
2003
D2004
2005
CO99
MN23
TN11
LA28
NCOS
CART Application
The CART analysis software is a statistical analysis tool developed by Breiman, et al. (1984) and
enhanced by Steinberg, et al. (1997) and Salford Systems (2007). For air quality and deposition
analysis purposes, the CART technique provides a method for segregating days into categories
that are representative of certain observed meteorological, air quality and deposition conditions. In
this study CART analysis is used to obtain information on the relationships between meteorology
and mercury deposition.
Application of the CART analysis technique requires several data elements to be used as input to
the classification scheme. Of these, one is identified as the "classification" parameter. Days are
segregated according to the value of the classification parameter and each resulting classification
bin corresponds to a specified range of values for this parameter. For this study, the classification
parameter for the CART application was specified to be the REMSAD-simulated daily mercury
deposition. The range for each classification category was determined based on the distribution of
E-9
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
the simulated deposition values for each analysis site; thus, the range in deposition associated with
each classification category is different for each analysis site. The remaining input data for
application of CART are selected to enable the segregation of days with respect to the
classification parameter and are referred to as the "independent" parameters. For this study, these
comprise various meteorological parameters, which were obtained from the MM5 meteorological
files used as inputs for the REMSAD simulations described in this report.
The result of a CART application is typically referred to as a "tree," and the branches lead to the
classification bins. The splits that define the branches are based on the values of certain of the
independent parameters and are chosen (by CART) to provide the best segregation of the data.
CART allows for the possibility that various combinations of meteorological parameters may lead to
similar values of the classification parameter. That is, multiple branches may each lead to
classification bins that represent the same category and, consequently, each category may have
multiple bins associated with it. This feature of CART is especially important for this analysis,
because it accommodates more combinations of meteorological parameters and deposition
amounts, and thus a more detailed reconstruction of the annual deposition.
CART provides information on the independent parameters and the values of these parameters that
are used to assign the days to the classification bins (i.e., branches). The resulting population of each
classification group (or CART classification bin) also provides information on the frequency of
occurrence of the meteorological conditions associated with each classification bin.
For this study, the average deposition for each bin is calculated (based on the REMSAD-simulated
deposition for all days in the bin) and this value was subsequently assigned to other similarly
classified days (from the 10 year analysis period).
The remainder of this section summarizes the application procedures and results for the mercury
deposition CART analysis.
CART APPLICATION PROCEDURES
The dependent parameters used for this study for the application of CART were derived from the
REMSAD-simulated total daily wet and dry mercury deposition at each site. For each day, the
classification variable was assigned a value of 1 to 5 (for wet) or 1 to 4 (for dry), such that each
value corresponded to a different range in daily total mercury deposition. The ranges, which vary
from site to site and for wet deposition vs. dry deposition, were determined based on the
distribution of the data at each site. The ranges represent (approximately) the 70, 90 and 97
percentile values of simulated dry deposition and non-zero wet deposition. For wet deposition, the
additional category is for days with zero wet deposition. These category definitions are typical for
CART applications and emphasize good classification of important high deposition days. These
ranges are presented in Table E-4. The deposition amounts used to define the classification
ranges were extracted from the REMSAD grid cell corresponding to each site/location.
E-10 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
Table E-4a. Range in Daily Mercury Wet Deposition (g km"2)
for Each CART Wet Deposition Classification Category.
Site# MDNSitelD
1 WA03
2 ID03
3 CA75
4 NV99
5 UT97
6 C099
7 SD18
8 MN23
9 IN20
10 TN11
11 OK99
12 LA28
13 FLOS
14 NCOS
15 NJ30
Site Name
Makah National Fish
Hatchery
Craters of the Moon
National Monument
Sequoia National
Park-Giant Forest
Gibb's Ranch
Salt Lake City
Mesa Verde National
Park-Chapin Mesa
Eagle Butte
Camp Ripley
Roush Lake
Great Smoky
Mountains National
Park
Stilwell
Hammond
Chassahowitzka
NWR
Waccamaw
State Park
New Brunswick
CART Classification Category
1 2
0 0 < Hg <
0.076
0 0 < Hg <
0.027
0 0 < Hg <
0.026
0 0 < Hg <
0.075
0 0 < Hg <
0.090
0 0 < Hg <
0.065
0 0 < Hg <
0.072
0 0 < Hg <
0.082
0 0 < Hg <
0.160
0 0 < Hg <
0.070
0 0 < Hg <
0.201
0 0 < Hg <
0.188
0 0 < Hg <
0.160
0 0 < Hg <
0.153
0 0 < Hg <
0.155
3
0.076 < Hg
< 0.184
0.027 < Hg
< 0.138
0.026 < Hg
< 0.081
0.075 < Hg
< 0.250
0.090 < Hg
< 0.212
0.065 < Hg
< 0.149
0.072 < Hg
< 0.140
0.082 < Hg
< 0.236
0.160 0.282
> 0.243
> 0.204
> 0.419
> 0.434
> 0.337
> 0.223
> 0.419
> 0.425
> 0.235
> 0.869
> 0.444
> 0.404
> 0.650
> 0.516
E-11
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
Table E-4b. Range in Daily Mercury Dry Deposition (g km"2)
for Each CART Dry Deposition Classification Category.
Site# MDNSitelD
1 WA03
2 ID03
3 CA75
4 NV99
5 UT97
6 C099
7 SD18
8 MN23
9 IN20
10 TN11
11 OK99
12 LA28
13 FLOS
14 NCOS
15 NJ30
Site Name
Makah National
Fish Hatchery
Craters of the Moon
National Monument
Sequoia National
Park-Giant Forest
Gibb's Ranch
Salt Lake City
Mesa Verde National
Park-Chapin Mesa
Eagle Butte
Camp Ripley
Roush Lake
Great Smoky
Mountains National
Park
Stilwell
Hammond
Chassahowitzka
NWR
Waccamaw State
Park
New Brunswick
CART Classification Category Range (g km-2)
1234
0 < Hg < 0.007
0 0.018
> 0.040
> 0.048
> 0.071
> 0.039
> 0.032
> 0.036
> 0.038
> 0.057
> 0.030
> 0.037
> 0.037
> 0.040
> 0.037
> 0.068
Tables E-5 and E-6 define the surface and upper-air meteorological parameters used in the CART
analysis, respectively. These parameters were extracted from the meteorological files used as
input to REMSAD. The meteorological parameters for each site were extracted for the grid cell in
which the actual meteorological monitoring sites (as paired with the MDN site) are located - in
order to accommodate the use of actual meteorological data in estimating deposition for additional
years.
E-12 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
For the surface parameters, the hourly meteorological values from the REMSAD input were
averaged to provide daily (24-hour average) values. For the upper-air parameters, the input values
for the two times at which sounding data are available (1200 GMT for morning and 0000 GMT
afternoon/evening) were used. Note that the upper-air parameters are for two levels 850 and 700 mb
(which correspond to approximately 1500 and 3000 m above ground level (agl), respectively). For
some high-elevation sites, only the 700 mb parameters were available. Temperature and moisture
parameters for the lower of these two levels were used, as available. In addition, the DT90AM
(difference between the temperature at 900mb and the surface temperature) parameter requires
temperature at an even lower level (900 mb) and this parameter was not available for several high
elevation sites. Temperature at the lowest level available (850 or 700 mb) was used as a surrogate
for this parameter.
Table E-5. Surface Meteorological Parameters used in the Mercury Deposition CART Analysis.
Parameter Name
Description
TMAX Maximum surface temperature (°C)
TMIN Minimum surface temperature (°C)
Q12 Specific humidity at the surface at 1200 1ST (g kg-1)
AVGWS 24-hour average surface wind speed (m s1)
AVGWBIN 24-hour average surface wind direction bin (1=N, 2=E, 3=S, 4=W, 5=Calm)
P12 Surface pressure at 1200 1ST (mb)
TOT_RAIN Total daily rainfall (in)
Table E-6. Upper-Air Meteorological Parameters used in the Mercury Deposition CART Analysis.
Parameter Name
Description
Tp/ei/e/'AVG Average temperature aloft based on the average of the morning (1200 GMT) and
afternoon/evening sounding (0000 GMT) for either 700 or 850 mb on the current day (°C)
Qp/eve/AVG Average specific humidity aloft a based on the average of the morning (1200 GMT) and
afternoon/evening sounding (0000 GMT) for either 700 or 850 mb on the current day (g kg-1)
WSp/eve/AM Wind speed aloft at the time of the morning (1200 GMT) sounding for either 700 or 850 mb on the
current day (m s1)
WSp/eve/PM Wind speed aloft at the time of the afternoon/evening (0000 GMT) sounding for either 700 or 850
mb on the current day (m s1)
YWSp/eve/PM Wind speed aloft at the time of the afternoon/evening (0000 GMT) sounding for either 700 or 850
mb on the prior day (m s1)
WBp/eve/AM Wind direction bin value of 1 through 5, indicating the wind direction [1=N, 2=E, 3=S, 4=W,
5=Calm] at either 700 or 850 mb and the time of the afternoon/evening sounding (1200 GMT) on
the current day.
Wbp/eve/PM Wind direction bin value of 1 through 5, indicating the wind direction [at either 700 or 850 mb
and the time of the afternoon/evening sounding (0000 GMT) on the current day.
YWBp/eve/PM Wind direction bin value of 1 through 5, indicating the wind direction at either 700 or 850 mb and
the time of the afternoon sounding on the prior day (0000 GMT).
plevel = 85 or 70, indicating either 850 or 700 mb
E-13
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
CART ANALYSIS RESULTS
As discussed in the previous section, CART analysis was used to classify or group all days from
the 2001 REMSAD simulation period into bins that were characterized by specified ranges in daily
deposition amount. CART was applied separately for each site/location and for wet and dry
mercury deposition. Given the specified ranges and input variables, CART develops the tree/bin
structure that provides the best segregation and grouping of the data relative to the REMSAD-
simulated deposition amounts. However, some misclassification of days into bins with ranges that
do not correspond to the simulated deposition amounts is expected. For this analysis, the
misclassification is attributable to the inability of the selected meteorological variables to fully
describe the detailed physical and chemical processes that are simulated by REMSAD and result
in the deposition amounts. Nevertheless, classification accuracy is very good and ranges from
about 80 to more than 90 percent, for both wet and dry deposition. This indicates that the
meteorological conditions can be used to describe to a large extent the deposition characteristics
of the simulation. The number of bins and classification accuracy for each CART analysis is
provided in Table E-7.
Table E-7 a. Number of CART Classification Bins and Classification Accuracy
for Mercury Wet Deposition, Using Meteorological Inputs and Simulated Deposition Values
from the REMSAD Base-Case simulation for 2001.
Site#
1
2
MDNSitelD
WA03
ID03
Site Name
Makah National Fish Hatchery
Craters of the Moon National
Number of CART
Classification Bins:
Wet Deposition
31
26
Percent (%) of Days
Correctly Classified by
CART: Wet Deposition
93
91
Monument
3 CA75 Sequoia National Park-Giant Forest
4 NV99 Gibb's Ranch
5 UT97 Salt Lake City
6 C099 Mesa Verde National Park-Chapin
Mesa
7 SD18 Eagle Butte
8 MN23 Camp Ripley
9 IN20 Roush Lake
10 TN11 Great Smoky Mountains National Park
11 OK99 Stilwell
12 LA28 Hammond
13 FLOS Chassahowitzka NWR
14 NCOS Waccamaw State Park
15 NJ30 New Brunswick
24
35
24
28
30
27
24
28
34
30
32
27
30
89
89
90
92
90
92
90
91
90
84
89
85
91
E-14
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
Table E-7b. Number of CART Classification Bins and Classification Accuracy
for Mercury Dry Deposition, Using Meteorological Inputs and Simulated Deposition Values
from the REMSAD Base-Case Simulation for 2001.
Site#
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
MDNSitelD
WA03
ID03
CA75
NV99
UT97
C099
SD18
MN23
IN20
TN11
OK99
LA28
FLOS
NCOS
NJ30
Site Name Number of CART Percent (%) of Days
Classification Bins: Correctly Classified by
Dry Deposition CART: Dry Deposition
Makah National Fish Hatchery
Craters of the Moon National Monument
Sequoia National Park-Giant Forest
Gibb's Ranch
Salt Lake City
Mesa Verde National Park-Chapin Mesa
Eagle Butte
Camp Ripley
Roush Lake
Great Smoky Mountains National Park
Stilwell
Hammond
Chassahowitzka NWR
Waccamaw State Park
New Brunswick
27
33
30
29
27
30
28
27
35
35
25
33
32
34
33
85
88
90
84
88
90
88
86
88
90
89
87
81
84
84
The key classification parameters vary among the areas and for the two different forms of
deposition. For wet deposition, key classification parameters include total daily rainfall, specific
humidity near the surface and aloft, and temperature near the surface and aloft. Wind speed aloft is
also used frequently in the CART trees.
For dry deposition, the key classification parameters include temperature, moisture and wind
speed, both near the surface and aloft. Temperature may be important because it is an indicator of
time of year and consequently of vegetation characteristics which are important in dry deposition.
Wind directions are not used frequently enough in the CART trees to be considered key
parameters to the overall classification, but they are often used near the end of the CART
pathways to distinguish different levels of mercury deposition.
To summarize parameter importance, the individual parameters are grouped into the following
categories: surface temperature (ST), surface moisture (SQ), surface wind speed (SWS), surface
wind direction (SWD), sea level pressure (SLP), total rainfall (RAIN), upper-air temperature (UAT),
upper-air moisture (UAQ), upper-air wind speed (UAWS), and upper-air wind direction (UAWD).
The average importance (averaged over all sites) for each category (on a scale of 0 to 100) is
shown in Figure E-4. Figure E-4a summarizes relative importance for the wet deposition analyses.
Figure E-4b summarizes relative importance for the dry deposition analyses.
E-15 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
Figure E-4a. Relative Importance of the Independent Parameter Categories for the CART Wet
Deposition Analyses.
CART Parameter Importance: Hg Wet Deposition
ST
SQ
SWS
SWD
SLP
RAIN
UAT
UAQ
UAWS
UAWD
I
I
I
!l
I
I
I
I
I
0 20 40 60 80 100
Relative Importance
Figure E-4b. Relative Importance of the Independent Parameter Categories for the CART Dry Deposition
Analyses.
CART Parameter Importance: Hg Dry Deposition
~
SQ |
-
SWS 1 |
SWD
SLP |
RAIN I |
J
UAT |
-
UAQ | |
UAWS |
UAWD | |
0 20 40 60 80 100
Relative Importance
E-16
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
The CART classification trees for each area provided the framework for estimating wet and dry
mercury deposition for days within the ten-year period. This is described in the following section.
Estimation of Mercury Deposition
The methods and results for estimating mercury deposition are discussed in this section.
USE OF THE CART RESULTS TO ESTIMATE MERCURY DEPOSITION
Daily meteorological data for a ten-year period of record extending from 1997 through 2006 were
assembled for each surface and upper-air meteorological monitoring site listed in Table E-3. These
data were processed to represent the CART meteorological input parameters presented in Tables
E-5 and E-6 and discussed earlier in this report.
For each separate CART classification tree (corresponding to a particular site/location and either wet
or dry deposition), the observed meteorological data were used to classify each day for the ten-year
period using the classification parameters provided by CART. In this manner, for each CART tree,
each day was assigned to (or placed in) one of the CART classification bins. The day was then
assigned a deposition value equal to the average of all days within the bin (from the REMSAD
simulation period). The result of this step was that each day for the ten-year period was assigned a
wet and dry deposition amount for each site/location of interest.
In the next step, these values were summed and annual deposition amounts were calculated for
each site (for both wet and dry deposition).
For wet deposition, the tendency for REMSAD to overestimate wet deposition amounts (as presented
in Section 6 of the main report) was also taken into account. It is noted numerous times in the main
report that MDN observed data may be biased low due to missed periods of precipitation, especially
toward the beginning of precipitation events. However, the overestimation by REMSAD is often
greater than what is expected based on the potential underestimation in the observed wet deposition.
Prior to calculating estimated wet deposition, the REMSAD-derived wet deposition was adjusted by
region and by season in accordance with model performance. Since the REMSAD values provide the
basis for estimating deposition for all years, any overestimate of wet deposition by REMSAD would
result in high values for all years.
Specifically, the REMSAD-derived wet deposition estimates were adjusted by the ratio of the mean
observed to mean simulated wet deposition. These ratios were calculated for each season and for
the western, southeastern, and northeastern states. The western region encompasses the
Northwest, West, Southwest, and West North Central climate zones (refer to Figure E-1). The
northeast region encompasses the East North Central, Central, and Northeast climate zones. The
southeast region includes the South and Southeast climate zones. The calculated ratios are as
follows:
West: Wnter = 1.09, Spring = 0.84, Summer = 0.49, Autumn = 0.54
Northeast: Wnter = 0.81, Spring = 0.84, Summer = 0.55, Autumn = 0.61
Southeast: Wnter = 0.71, Spring = 0.49, Summer = 0.0.61, Autumn = 0.53
Estimated wet deposition was recalculated using the adjusted REMSAD wet deposition values. Both
sets of wet deposition estimates are presented in the following sections, and are intended to bound
the uncertainty in the calculations due to model performance.
E-17 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
MERCURY DEPOSITION RESULTS
The CART-based estimates for wet, dry, and total mercury deposition for the selected locations are
presented in Table E-8. The calculated annual mercury wet deposition estimates are presented in
Table E-8a, dry deposition is presented in Table E-8b, and total deposition is given in Table E-8c.
For wet and total deposition, both the adjusted and unadjusted values are presented and represent
lower and upper bounds of the estimate. Due to a lack of meteorological data, values could not be
estimated for NV99, UT97, and SD18 for 1997 and for NV99 for 1998.
When examining the results, the reader should keep in mind that the purpose of this analysis was
to investigate the impact on deposition of year-to-year variability due to meteorological influences.
Other important factors, such as year-to-year changes in U.S. and other global emissions that
would affect changes in deposition from one year to another were beyond the scope of this
analysis.
Table E-8a. Estimated Annual Mercury Wet Deposition (g km"2) for 1997-2006 for Selected MDN Sites.
Site ID
WA03 (Adjusted)
(Unadjusted)
ID03 (Adjusted)
(Unadjusted)
CA75 (Adjusted)
(Unadjusted)
NV99 (Adjusted)
(Unadjusted)
UT97 (Adjusted)
(Unadjusted)
C099 (Adjusted)
(Unadjusted)
SD18 (Adjusted)
(Unadjusted)
MN23 (Adjusted)
(Unadjusted)
IN20 (Adjusted)
(Unadjusted)
TN11 (Adjusted)
(Unadjusted)
OK99 (Adjusted)
(Unadjusted)
LA28 (Adjusted)
(Unadjusted)
FLOS (Adjusted)
(Unadjusted)
NCOS (Adjusted)
(Unadjusted)
NJ30 (Adjusted)
(Unadjusted)
1997
11.6
15.2
2.0
3.5
4.2
5.5
NA
NA
NA
NA
1.1
2.0
NA
NA
4.7
7.6
11.1
15.8
8.7
12.2
14.7
25.3
20.4
34.4
13.7
23.0
11.3
19.3
9.6
14.1
1998
10.3
12.8
1.7
3.0
4.7
5.6
NA
NA
6.2
10.4
1.2
2.2
2.5
4.0
4.6
7.7
10.9
16.3
10.1
14.9
14.3
24.5
23.6
39.3
15.8
26.6
12.1
20.7
9.9
15.0
1999
11.3
13.6
1.4
2.5
4.6
6.1
4.0
8.0
4.5
7.5
1.9
3.5
3.6
5.7
5.2
8.4
11.9
17.2
7.2
10.7
11.0
19.0
18.6
31.0
12.8
21.7
10.8
18.5
10.2
14.7
2000
10.7
12.8
1.3
2.3
4.1
5.3
3.9
7.6
6.0
10.0
2.5
4.5
2.9
4.5
4.9
8.3
12.2
17.3
8.6
12.4
13.7
23.6
19.7
32.8
11.1
18.9
10.7
18.2
11.0
16.9
2001
10.6
13.5
1.4
2.6
4.4
5.6
4.1
8.1
5.6
9.6
2.0
3.5
3.1
5.1
3.9
6.5
12.8
18.3
8.0
11.6
14.0
24.2
17.2
29.2
11.4
19.4
10.4
17.8
10.1
15.2
2002
9.4
11.6
1.4
2.4
3.5
4.4
4.3
8.6
5.1
8.4
1.3
2.3
3.0
4.9
5.0
8.4
11.6
17.1
9.3
13.5
13.4
23.3
18.8
31.7
13.0
22.1
11.5
19.7
10.0
14.3
2003
10.7
13.6
1.7
3.1
3.9
5.0
5.2
10.2
6.3
10.5
1.9
3.5
2.6
4.0
3.4
5.7
11.5
16.8
10.3
15.3
15.6
27.2
18.5
31.3
12.7
21.5
13.7
23.4
11.8
17.5
2004
10.5
13.9
1.5
2.6
4.1
5.2
4.2
8.2
6.0
9.6
1.9
3.4
3.1
5.0
5.4
9.0
12.5
17.9
8.9
13.1
13.9
24.3
18.8
31.9
12.1
20.5
12.3
20.8
12.3
18.8
2005
9.3
12.2
1.4
2.6
3.7
4.6
4.7
9.4
7.5
12.2
1.7
3.0
3.3
5.3
4.3
7.2
12.6
18.2
8.8
12.8
13.2
22.7
23.8
39.9
13.3
22.5
17.8
29.8
9.7
14.0
2006
12.4
14.8
1.3
2.3
4.5
5.7
5.2
10.3
5.5
9.0
2.2
3.9
1.9
3.0
4.7
7.9
11.1
16.1
7.1
10.1
11.1
19.2
18.4
30.8
12.2
20.6
14.6
24.9
10.5
15.7
E-18 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
Table E-8b. Estimated Annual Mercury Dry Deposition (g km"2) for 1997-2006 for Selected MDN Sites.
Site ID
WA03
ID03
CA75
NV99
UT97
C099
SD18
MN23
IN20
TN11
OK99
LA28
FLOS
NCOS
NJ30
1997
1.9
5.7
9.6
NA
NA
4.6
NA
5.2
8.2
4.2
5.5
5.7
7.5
6.4
14.2
Table E-8c. Estimated
Site ID
WA03 (Adjusted)
(Unadjusted)
ID03 (Adjusted)
(Unadjusted)
CA75 (Adjusted)
(Unadjusted)
NV99 (Adjusted)
(Unadjusted)
UT97 (Adjusted)
(Unadjusted)
C099 (Adjusted)
(Unadjusted)
SD18 (Adjusted)
(Unadjusted)
MN23 (Adjusted)
(Unadjusted)
IN20 (Adjusted)
(Unadjusted)
TN11 (Adjusted)
(Unadjusted)
OK99 (Adjusted)
(Unadjusted)
LA28 (Adjusted)
(Unadjusted)
FLOS (Adjusted)
1997
13.5
17.1
7.6
9.2
13.8
15.0
NA
NA
NA
NA
5.7
6.6
NA
NA
9.9
12.8
19.3
24.0
13.0
16.4
20.2
30.9
26.1
40.1
21.1
1998
1.7
5.6
8.7
NA
4.7
5.0
5.1
5.5
8.6
4.1
5.8
6.0
7.2
5.9
14.4
1999
1.7
6.2
10.1
12.1
5.9
5.3
4.6
5.4
8.7
4.3
6.1
6.3
7.3
6.1
14.5
2000
1.7
5.8
9.6
12.4
5.5
5.1
5.1
5.5
8.5
4.4
5.7
6.7
7.5
5.9
14.0
2001
1.6
6.2
10.1
12.2
6.1
5.1
5.1
5.3
8.2
4.3
5.9
6.3
7.0
6.1
13.9
Annual Total Mercury Deposition (g
1998
12.0
14.5
7.4
8.7
13.4
14.3
NA
NA
10.9
15.1
6.2
7.1
7.6
9.1
10.1
13.2
19.5
24.9
14.2
19.0
20.1
30.3
29.6
45.3
22.9
1999
13.0
15.3
7.6
8.7
14.7
16.1
16.1
20.1
10.4
13.4
7.2
8.7
8.2
10.4
10.7
13.8
20.6
25.9
11.5
15.0
17.1
25.1
24.9
37.3
20.2
2000
12.3
14.5
7.1
8.1
13.7
14.9
16.3
20.0
11.5
15.6
7.6
9.6
8.0
9.6
10.4
13.8
20.7
25.8
12.9
16.7
19.4
29.2
26.4
39.5
18.6
2001
12.2
15.1
7.6
8.9
14.5
15.6
16.3
20.3
11.7
15.7
7.1
8.6
8.2
10.2
9.3
11.9
21.0
26.5
12.3
15.9
19.9
30.1
23.5
35.5
18.4
2002
1.9
6.2
9.3
11.1
5.8
5.5
5.2
5.5
8.6
3.9
5.8
5.8
6.9
5.8
14.0
km"2) for
2002
11.3
13.5
7.6
8.7
12.7
13.7
15.4
19.7
11.0
14.2
6.8
7.8
8.2
10.1
10.5
13.8
20.2
25.8
13.2
17.4
19.2
29.1
24.6
37.5
20.0
2003
1.8
5.9
9.6
11.9
6.4
5.1
5.1
5.6
8.8
3.9
6.0
6.0
7.6
5.8
14.0
2004
1.7
5.7
10.0
11.4
6.2
5.3
5.1
5.6
8.7
4.1
6.0
5.8
7.2
6.0
13.6
2005
2.0
5.7
9.6
11.5
5.9
5.2
5.1
5.7
8.9
4.3
6.1
6.2
7.0
6.0
14.5
2006
2.0
5.7
9.7
11.1
6.0
4.9
5.6
5.6
8.3
4.3
5.8
6.3
7.1
6.1
13.8
1997-2006 for Selected MDN Sites.
2003
12.5
15.4
7.6
9.0
13.5
14.6
17.0
22.1
12.7
16.9
7.0
8.6
7.7
9.1
9.0
11.3
20.3
25.6
14.2
19.2
21.6
33.2
24.5
37.2
20.3
2004
12.2
15.6
7.2
8.3
14.0
15.2
15.5
19.5
12.2
15.8
7.2
8.7
8.3
10.1
11.0
14.6
21.2
26.6
13.0
17.2
19.9
30.3
24.5
37.7
19.3
2005
11.3
14.1
7.1
8.3
13.3
14.2
16.3
20.9
13.4
18.1
6.8
8.1
8.4
10.4
10.0
12.9
21.5
27.1
13.1
17.1
19.2
28.7
30.0
46.1
20.3
2006
14.4
16.8
7.0
8.0
14.2
15.4
16.3
21.4
11.5
15.0
7.1
8.9
7.5
8.5
10.3
13.5
19.4
24.4
11.4
14.4
16.9
25.1
24.7
37.1
19.3
E-19 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
Site ID
(Unadjusted)
NCOS (Adjusted)
(Unadjusted)
NJ30 (Adjusted)
(Unadjusted)
1997
30.5
17.8
25.7
23.9
28.3
1998
33.8
18.1
26.6
24.3
29.4
1999
29.0
17.0
24.7
24.7
29.2
2000
26.4
16.6
24.1
25.0
30.9
2001
26.4
16.5
23.9
24.0
29.1
2002
29.0
17.3
25.5
24.0
28.3
2003
29.1
19.6
29.2
25.8
31.5
2004
27.7
18.3
26.8
25.9
32.4
2005
29.5
23.8
35.8
24.2
28.4
2006
27.7
20.7
31.0
24.3
29.6
It is informative to summarize these results in terms of the average value and the standard
deviation over the ten-year period. This information is provided in Table E-9.
Table E-9. Ten-Year Average Estimated Annual Mercury Deposition (g km"2) for 1997-2006
for Selected MDN Sites: Wet, Dry, and Total.
Site ID
WA03 (Adjusted)
(Unadjusted)
ID03 (Adjusted)
(Unadjusted)
CA75 (Adjusted)
(Unadjusted)
NV99 (Adjusted)
(Unadjusted)
UT97 (Adjusted)
(Unadjusted)
C099 (Adjusted)
(Unadjusted)
SD18 (Adjusted)
(Unadjusted)
MN23 (Adjusted)
(Unadjusted)
IN20 (Adjusted)
(Unadjusted)
TN11 (Adjusted)
(Unadjusted)
OK99 (Adjusted)
(Unadjusted)
LA28 (Adjusted)
(Unadjusted)
FLOS (Adjusted)
(Unadjusted)
NCOS (Adjusted)
(Unadjusted)
NJ30 (Adjusted)
(Unadjusted)
Wet
10-Year
Average
(g km^)
10.7
13.4
1.5
2.7
4.2
5.3
4.5
8.8
5.9
9.7
1.8
3.2
2.9
4.6
4.6
7.7
11.8
17.1
8.7
12.7
13.5
23.3
19.8
33.2
12.8
21.7
12.5
21.3
10.5
15.6
Deposition
Standard
Deviation
(g km'2)
0.95
1.10
0.22
0.39
0.41
0.51
0.49
1.03
0.83
1.36
0.46
0.81
0.50
0.85
0.60
0.98
0.70
0.88
1.06
1.66
1.46
2.56
2.23
3.62
1.31
2.16
2.29
3.74
0.91
1.62
Dry
10-Year
Average
(g km'12)
1.8
1.8
5.9
5.9
9.6
9.6
11.7
11.7
5.9
5.9
5.1
5.1
5.1
5.1
5.5
5.5
8.5
8.5
4.2
4.2
5.9
5.9
6.1
6.1
7.2
7.2
6.0
6.0
14.1
14.1
Deposition
Standard
Deviation
(g km'2)
0.15
0.15
0.24
0.24
0.40
0.40
0.51
0.51
0.51
0.51
0.25
0.25
0.24
0.24
0.15
0.15
0.24
0.24
0.16
0.16
0.18
0.18
0.31
0.31
0.24
0.24
0.18
0.18
0.29
0.29
Total
10-Year
Average
(g km-"2)
12.5
15.2
7.4
8.6
13.8
14.9
16.2
20.5
11.7
15.5
6.9
8.3
8.0
9.7
10.1
13.2
20.4
25.7
12.9
16.8
19.4
29.2
25.9
39.3
20.0
28.9
18.6
27.3
24.6
29.7
Deposition
Standard
Deviation
(g km'2)
0.97
1.12
0.26
0.38
0.58
0.73
0.51
0.90
0.94
1.40
0.56
0.89
0.33
0.67
0.59
0.98
0.77
0.98
0.94
1.53
1.42
2.51
2.23
3.60
1.31
2.15
2.27
3.72
0.75
1.43
E-20
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
Overall, year-to-year variability is less for dry deposition than for wet deposition, presumably due to
variations in precipitation that influence wet deposition. There is also considerably less year-to-year
variability in wet deposition (and total deposition) for sites located in the western states (consistent
with less frequent and less variable rainfall).
The CART-based values for 2001 indicate that, based on meteorological variability, this was an
average year for mercury deposition at most sites, with slightly higher than average estimated
deposition for CA75 and OK99, and slightly lower than average deposition for CO99, MN23, TN11,
LA28 FLOS, and NCOS.
The adjusted values are consistently lower than the unadjusted values for wet (and total)
deposition. By accounting for the tendency for REMSAD to overestimate wet deposition these
values represent a lower bound for the estimated mercury deposition.
COMPARISON WITH OBSERVED MERCURY WET DEPOSITION DATA
One of the key motivations for this analysis was the lack of a long-term record of both wet and dry
deposition data for mercury. As indicated, however, in Table E-2, wet deposition data are available
for ten sites for some portion of the analysis period (the number of years with complete data
availability within the 10-year analysis period ranges 0 to 10). For sites with five or more complete
years of data during the analysis period, these data are used as a check on the reasonableness of
the results. The mean values of estimated and observed wet deposition are compared to
determine whether the estimated values are within a reasonable range, compared to the observed
data. The estimated standard deviation in annual mercury wet deposition is also compared with
that based on the observed data. This provides a check of the reasonableness of the
meteorological variability of the CART-based estimates with the overall (meteorological and
emissions based) variability revealed by the observed data (it is expected that the variability
present in the estimated values should be lower than that for the observed data, especially if
emissions changes or fluctuations occurred during the years for which data are available). As
discussed later, this comparison can also be used to infer the role of meteorology versus emissions
in producing the variations in the observed data.
This information is summarized in Table E-10. The adjusted estimates in Table E-10a are followed
by the unadjusted estimates in Table E-10b.
E-21 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
Table E-10a. Summary of Observed and Estimated (Based on Adjusted REMSAD Results) Variability in
Annual Wet Deposition (g km"2) for MDN Monitoring Sites with Five or More Years
of Data During the Analysis Period.
Site ID
C099
MN23
IN20
LA28
FLOS
NCOS
Observed or Number of Analysis
Estimated Years with
Complete Data
Observed
Estimated
Observed
Estimated
Observed
Estimated
Observed
Estimated
Observed
Estimated
Observed
Estimated
5
10
6
8
9
10
Minimum Annual
Value (g km-2)
3.3
1.1
6.7
3.4
8.0
10.9
11.0
17.2
13.1
11.1
9.3
10.4
Maximum Annual
Value (g km-2)
6.0
2.5
12.4
5.4
13.2
12.8
21.3
23.8
21.5
15.8
18.1
17.8
Average
Annual
Deposition (g
km-2)
4.7
1.8
8.8
4.6
10.7
11.8
15.2
19.8
16.0
12.8
12.8
12.5
Standard
Deviation (g
km-2)
1.2
0.5
2.0
0.6
1.9
0.7
3.4
2.2
2.9
1.3
3.0
2.3
Table E-10b. Summary of Observed and Estimated (Based on Unadjusted REMSAD Results) Variability
in Annual Wet Deposition (g km"2) for MDN Monitoring Sites with Five or More Years
of Data During the Analysis Period.
Site ID
C099
MN23
IN20
LA28
FLOS
NCOS
Observed or Number of Analysis
Estimated Years with
Complete Data
Observed
Estimated
Observed
Estimated
Observed
Estimated
Observed
Estimated
Observed
Estimated
Observed
Estimated
5
10
6
8
9
10
Minimum Annual
Value (g km-2)
3.3
2.0
6.7
5.7
8.0
15.8
11.0
29.2
13.1
18.9
9.3
17.8
Maximum Annual
Value (g km-2)
6.0
4.5
12.4
9.0
13.2
18.3
21.3
39.9
21.5
26.6
18.1
29.8
Average
Annual
Deposition (g
km-2)
4.7
3.2
8.8
7.7
10.7
17.1
15.2
33.2
16.0
21.7
12.8
21.3
Standard
Deviation (g
km-2)
1.2
0.8
2.0
1.0
1.9
0.9
3.4
3.6
2.9
2.2
3.0
3.7
Focusing first on the minimum, maximum, and mean values, the results indicate that the CART-
based estimates are reasonably consistent in both magnitude and range with the observed values.
For IN20, LA28, FLOS, and NCOS, the results derived using the adjusted REMSAD values are more
consistent with the observed ranges and averages. The mean estimated wet deposition amounts
for these sites are considerably higher than the mean observed values. This indicates that
REMSAD overestimates wet deposition on types of days that occur with some frequency during the
E-22
August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
analysis period. For CO99 and MN23, the results derived using the unadjusted REMSAD values
are more consistent with the observed ranges and averages.
When compared with the observation-based values, the standard deviation (used here to characterize
year-to-year variability) is less for all sites for the results derived using the adjusted REMSAD values, but
greater for some sites for the results derived using the unadjusted values. The greater variability with the
unadjusted values is consistent with the overall higher deposition amounts. Since it is expected that the
variability present in these results should be lower than that for the observed data, the estimates based
on the adjusted REMSAD values are expected to be more reliable.
Based on the results derived using the adjusted REMSAD values, it is estimated that
meteorological variability accounts for between 30% and 75% of the year-to-year variability in
mercury wet deposition for these six locations for the period 1997-2006.
Summary and Conclusions
In this exercise, a combination of mercury deposition modeling results, observed meteorological data,
and statistical analysis were used to estimate annual wet and dry mercury deposition for a ten-year
period for 15 locations throughout the U.S., and specifically the potential year-to-year variability in
mercury deposition simulation results due to variations in meteorology. This methodology appears to
provide reasonable estimates of annual mercury deposition that account for year-to-year variability in
the meteorological conditions that influence mercury deposition. The estimated deposition amounts
show agreement with annual observed deposition amounts, especially when the tendency for the
REMSAD simulation results for the 2001 simulation period to overestimate wet deposition is taken
into account. Key finding from this analysis include:
Classification and Regression Tree (CART) analysis is able to correctly classify 80 to 90 percent
of days that comprise the 2001 REMSAD simulation period into bins defined by mercury
deposition amount, based on input meteorological parameters.
Key classification parameters for wet deposition include total daily rainfall, specific humidity near
the surface and aloft, and temperature near the surface and aloft.
Key classification parameters for dry deposition include temperature, moisture and wind speed,
both near the surface and aloft.
Overall, estimated year-to-year variability is less for dry deposition than for wet deposition and
less for sites located in the western states.
The CART-based values indicate that, based on meteorological variability, 2001 was an
average year for mercury deposition at most of the selected sites.
The CART-based estimates of mercury deposition are reasonably consistent in both magnitude
and range with available data.
When adjusted to account for REMSAD model performance, the estimated values are
consistently lower than the unadjusted values for wet (and total) deposition. The adjusted
estimates show better agreement with observed annual wet deposition data, but less year-to-
year variability.
Meteorology contributes to observed variability in annual mercury deposition and differences in
meteorological conditions are expected to contribute to year-to-year differences in any
E-23 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
corresponding simulation results. Based on the results derived using the adjusted REMSAD
values for selected sites, it is estimated that meteorological variability accounts for between 30%
and 75% of the year-to-year variability in mercury wet deposition for the period 1997-2006.
References
Breiman, Leo, Jerome Friedman, Richard Olshen, and Charles Stone. Classification and
Regression Trees. Pacific Grove: Wadsworth, 1984.
NCDC, 2007. National Climatic Data Center web site: www.ncdc.noaa.gov/oa/climate/research/.
Karl, T. R., and W. J. Koss. 1984. "Regional and National Monthly, Seasonal, and Annual
Temperature Weighted by Area, 1895-1983." Historical Climatology Series 4-3, National
Climatic Data Center, Asheville, NC, 38 pp.
Myers, T., Y. Wei, B. Hudischewskyj, and S. Douglas. 2003. "Application of the REMSAD Modeling
System to Estimate the Deposition of Mercury in Support of the Wisconsin TMDL Pilot." ICF
International, San Rafael, California (03-050).
Salford Systems. 2007. CART user's guide online at: http://www.salford-
systems.com/doc/CARTtrifold.ddf
Steinberg, Dan, and Phillip Colla. CARTClassification and /regression Trees. San Diego, CA:
Salford Systems, 1997.
E-24 August 2008
-------
Model-Based Analysis and Tracking of Airborne Mercury Emissions to Assist in Watershed Planning
Appendix E: Use of Classification and Regression Tree (CART) Analysis to Examine the
Impact of Year-to-Year Meteorological Variability on Mercury Deposition for Several
Locations throughout the U.S.
This page deliberately left blank.
E-25 August 2008
------- |