OFFSET-BASED DECISION SUPPORT MODELS FOR MITIGATING MERCURY SOURCES IN THE CACHE CREEK WATERSHED, NORTH CENTRAL CALIFORNIA

The U.S. Geological Survey's Western Geographic Science Center, in collaboration with researchers at Stanford University, is developing an adaptive-management approach at the regional watershed scale to assist wastewater-treatment plants in meeting mercury (Hg)-discharge-permit requirements under total-maximum-daily-load (TMDL) guidelines. Offsets, a market-incentive program, allow dischargers facing higher pollution-control costs to fulfill their regulatory obligations by making more cost effective pollutant-reduction decisions. As a result of Hg's serious health threat and the scarcity of remediation funds, offsets may serve as a suitable tool for wastewater-treatment plants to meet discharge permit requirements cost-effectively. An Hg-offset program would be established and succeed only with an active participant. Therefore, the adaptive-management approach described in this report evaluates the offset potential for Hg-related projects by integrating Earth science and economic information to identify a wastewater-treatment plant's interest in becoming an active participant. This approach enhances a TMDL decisionmaker's ability to compare the cost effectiveness of various offset-mitigation choices for meeting discharge-permit requirements.Our study focuses on the use of alternative statistical methods to explicitly state and reduce, where possible, inherent uncertainties in physical, chemical, and biologic processes controlling the fate and transport of Hg in aquatic environments. The combined uncertainty in the relations between these processes and potential mitigation efforts may be better understood and addressed in policy decisionmaking. Analytical results for predicting methylmercury concentrations in water and offset-mitigation costs have been found useful in the economic and environmental analysis of an offset program. An integrated Bayesian-network approach to uncertainty analysis, combining historical data, empirical linear-regression models, conceptual models, and expert judgment, allows easy updating of predictions and inferences when observations of model variables are made. We develop and demonstrate these methods with data from the Cache Creek watershed, a subbasin of the Sacramento River watershed, in north-central California.