Progress on Implementation of a Decision-Support System to Assess Critical Loads of Atmospheric S Deposition in the Southeastern US
Paul Hessburg1, Keith Reynolds1*, Timothy Sullivan2, Bill Jackson3, Nick Povak1, Brion Salter1 and Todd McDonnell2
The U.S. Environmental Protection Agency (EPA) and USDA Forest Service (USFS) are developing a decision-support system (DSS) for critical loads (CL) of atmospheric S deposition to protect stream resources against acidification in the southeastern United States. The spatial coverage of the study region includes the Ridge and Valley and Appalachian Plateau ecoregions in Virginia and West Virginia and the Blue Ridge ecoregion in Virginia, West Virginia, North Carolina, and Tennessee. The DSS is an application of the Ecosystem Management Decision Support (EMDS) system, originally developed at the USFS Pacific Northwest Research Station. EMDS is an application framework for knowledge-based decision support for environmental analysis and planning at multiple geographic scales. The system integrates geographic information with logic and decision-modeling technologies to provide a spatial analysis system for data management and environmental risk assessment.
Water chemistry data across the study region were compiled from a variety of EPA, USFS, and other existing water quality databases. The final dataset includes 933 sites. Each site is represented by the most recent spring sample. Estimates of base cation weathering have been developed for 140 of the 933 water chemistry sites using the Model of Acidification of Groundwater in Catchments (MAGIC), and are being extrapolated to the broader region for use in the Steady State Water Chemistry (SSWC) model to estimate CL values throughout the region. Extrapolations of the weathering estimates and the regional distribution of stream water acid neutralizing capacity (ANC) are achieved using a variety of multivariate statistical modeling techniques. General data classes evaluated in the modeling include soil and lithology variables as well as wet and dry atmospheric S deposition, topographic wetness index, surface area ratio, and 36 Ameriflux variables. All data were upslope averaged in a geographic information system (GIS) to develop potential predictor variables above each pour point. Despite data limitations imposed by historic non-random selection of sample sites, initial modeling results for predicting regional distribution of ANC appear promising, and will be presented. Regional modeling of weathering is ongoing.
1 Pacific Northwest Research Station, Wenatchee, WA
1* Pacific Northwest Research Station, Corvallis, OR
2 E&S Environmental Chemistry, Corvallis, OR
3 USDA Forest Service Southern Region, Asheville, NC