U.S. EPA Projects:

EMDS Decision Support Modeling

Decision-support System for Critical Loads of Atmospheric Sulfur Deposition in the Southeastern United States
E&S Environmental Chemistry, Inc., working together with the U.S. Environmental Protection Agency (EPA) and under the direction of the USDA Forest Service (USFS), is assisting the USFS in the development of a decision-support system (DSS) for critical loads (CL) of atmospheric sulfur (S) deposition in the southeastern United States. The spatial coverage of the study region includes the Ridge and Valley and Central Appalachian ecoregions in Virginia and West Virginia and the Blue Ridge Ecoregion in Virginia, West Virginia, North Carolina, and Tennessee. The research team is developing the DSS in this ongoing project as an application of the Ecosystem Management Decision Support (EMDS) system, originally developed at the USFS PNW 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.

E&S compiled water chemistry data across the study region from a variety of EPA and USFS water quality databases. The final data set includes 933 stream sites. Dynamic model estimates of base cation weathering have been developed for 140 of the 933 water chemistry sites, and are being used to develop regional predictions of steady-state CLs, based on tests, currently in progress, of a variety of multivariate 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 to develop potential predictor variables above each pour point in GIS. Despite data limitations imposed by historical non-random selection of sample sites, initial modeling results for predicting regional distribution of CLs appear promising, and will continue to be refined.

The final report can be downloaded here.