MEMO: Description of docketed GIS analyses , fish tissue Hg database and Excel-based risk assessment spreadsheet modeling conducted to support the revised national-scale mercury risk assessment 
December, 2011 
(point of contact: Dr. Zachary Pekar, USEPA, Office of Air Quality Planning and Standards, e-mail: pekar.zachary@epa.gov)
This memo briefly outlines the GIS-related analyses completed in support of the revised national-scale mercury risk assessment and how information generated through those analyses was incorporated into the revised risk assessment spreadsheet. The revised risk assessment spreadsheet is included in the docket with the final rule as "Revised Risk Assessment Model-2016.xlsx". This memo also describes the excel files comprising the fish tissue Hg database used in the risk assessment, which have been docketed as "2011 mercury fish tissue dataset" with the final rule.  This memo is not intended to provide a rigorous overview of the risk assessment approach, which is provided in the "Revised Technical Support Document: National-Scale Assessment of Mercury Risk to Populations with High Consumption of Self-caught Freshwater Fish In Support of the Appropriate and Necessary Finding for Coal- and Oil-Fired Electric Generating Units"  (Revised Mercury Risk TSD). 
A number of GIS-related shapefiles were docketed with Mercury Risk TSD accompanying the proposed rule and were used without additional modification in the revised risk assessment (docket id: EPA-HQ-OAR-2009-0234-3079).  For that reason, only those GIS shapefiles which have been updated as part of the revised risk assessment have been docketed with the final rule, including updated shapefiles of the watershed-level interpolation of mercury deposition for the 2005 and 2016 scenarios. In addition, we have also docketed the revised fish tissue mercury database used in the risk assessment. The updated fish tissue Hg data were incorporated into the analysis as Excel files and used directly in the risk calculation spreadsheet (rather than being incorporated into the GIS analysis).  And for this reason, we have docketed the full fish tissue mercury dataset (see below), but have not docketed any shapefiles related to the fish tissue Hg data, since they were not used in the risk assessment. 
The following procedure was used to generate risk estimates for the revised national-scale mercury risk assessment:
Step 1) Develop an updated (augmented) fish tissue Hg dataset for use in the risk assessment:  The procedure used to derive the 2011 mercury fish tissue dataset is fully documented in the Revised Mercury Risk TSD, section 1.4.2). Figures 1-3 through 1-6 illustrate how the various underlying fish tissue Hg datasets were filtered and integrated to produce the final 2011 mercury fish tissue HUC-level dataset used to derive the 75[th] percentile fish tissue Hg estimates. Each of the underlying datasets identified in these figures used in deriving the final integrated dataset (e.g., USGS States, USGS Others, NRSA) have been docketed as separate Excel files under the title "2011 mercury fish tissue dataset". 
Step 2) Generate HUC-level fish tissue Hg 75[th] percentile values (using the fish tissue Hg dataset described in Step 1) and incorporate those estimates into the "Revised Risk Assessment Model-2016.xlsx" spreadsheet: The procedure used for this is described in detail in the Revised Mercury Risk TSD.  The screening of watersheds to exclude non-air Hg sources was accomplished as part of the process of developing the fish tissue database and will not be discussed further here. The 75[th] percentile fish tissue watershed-level data (i.e., 3,141 watershed-level fish tissue Hg concentrations) are incorporated as a separate sheet in the Revised Risk Assessment Model-2016.xlsx which has been docketed (see <MeHg fish tissue combined> worksheet within "Revised Risk Assessment Model-2016.xlsx").
Step 3) Incorporate the watershed-level CMAQ-derived total and U.S. EGU-related Hg deposition (for the 2005 and 2016 scenarios) into the "Revised Risk Assessment Model-2016.xlsx" spreadsheet:  We have corrected the spatial interpolation of the gridded modeling results to the watersheds, using the CMAQ modeling conducted in support of the Mercury Risk TSD accompanying the proposed rule. These corrected watershed-level Hg deposition concentrations are docketed as shapefiles and have been incorporated into the "Revised Risk Assessment Model-2016.xlsx" spreadsheet as a separate worksheet (presenting watershed-level Hg deposition values - see <2005 Hg dep> and <2016 Hg dep> worksheets).
Step 4) Identify watersheds where each of the female subsistence fish consumer scenarios are active:  We used shapefiles from the March analysis to identify which watersheds potentially have fishing activity by each of the female subsistence fish consuming populations we were considering in the analysis (and consequently, those shapefiles are not redocketed here). However, for this analysis, rather than doing the intersection of watersheds with potential activity by specific subsistence scenarios and those with mercury fish tissue Hg levels in GIS, we implemented this step within the "Revised Risk Assessment Model-2016.xlsx" spreadsheet.  Specifically, we had a sheet identifying which tracts had potential activity for a given female subsistence fish consuming scenario (see <Scenario location by HUC> worksheet) and then we assessed the intersection within the "Revised Risk Assessment Model-2016.xlsx" spreadsheet. For the revised mercury risk assessment, the typical female subsistence fish consumer scenario was evaluated for all watersheds with fish tissue Hg data (i.e., we did not filter out any of these watersheds based on application of census data for this scenario). 
Additional detail on the layout of the Excel risk assessment spreadsheet can be found on the "NOTES" sheet incorporated as part of the spreadsheet. 


