UNCERTAINTY OF GROUNDWATER VULNERABILITY ASSESSMENTS FOR AGRICULTURALREGIONS IN HAWAII - REVIEW

There are important challenges associated with assessing potential gro undwater vulnerability hazards that may result from regional scale app lications of agrochemicals. The increasing availability of Geographic Information System (GIS) software to those involved in assisting with landuse decisions has resulted in the widespread production of multico lored risk management maps for many environmentally sensitive issues. Soil-based GIS's have recently been coupled to various solute-leaching models to make near-surface groundwater vulnerability assessments for guidance in pesticide regulation in several states. In general, these assessments rest on soil, climatic, and chemical data that are extrem ely sparse and contain considerable uncertainty. It is also important to acknowledge the uncertainty associated with the transport/fate proc esses that are not accounted for by the modeling approach used to make the assessment. In this paper, we review the results from a series of papers that have focused on characterization of uncertainty in pestic ide mobility estimates, using the attenuation and retardation indices (AF and RF), for the Pearl Harbor Basin on the Hawaiian island of Oahu . Relative to data error uncertainties, we discuss the impacts of: (i) soil, climatic, and chemical data base uncertainties, (ii) reductions in data base uncertainties, (iii) extrapolation of soil data base inf ormation based on soil taxonomy and soil survey, and (iv) importing in formation from outside the region of interest. Relative to model error uncertainties, we compare pesticide leaching estimates from the simpl e AF and RF mobility indices with simulations from the EPA's Pesticide Root Zone Model (PRZM) and field observations. Finally, we outline a Regional Integrated Risk Assessment approach for characterizing region al scale groundwater vulnerability for near-surface nonpoint sources.