摘要：In increasing numbers, communities that rely on groundwater for drinking supplies have discovered contamination from agricultural pesticides and herbicides, road salt, underground fuel storage, and septic systems. A variety of short- and long-run remedies are available with highly uncertain outcomes. An appropriate technique for solving a benefit-cost problem of this type is a sequential decision framework using stochastic dynamic programming procedures for solution. The approach is illustrated here by means of an application to the problem of the recent contamination of the groundwater of Whately, Massachusetts by the agricultural fumigant EDB and the pesticide aldicarb.
