Decision making using fuzzy sets for optimal water quality management

In the area of management of receiving water quality the focus has shifted away from stringent point source controls and towards a watershed scale approach. The Total Maximum Daily Load (TMDL) procedure is an attempt to control point and nonpoint source pollution by taking into account the assimilative capacity of the receiving water. As part of the TMDL procedure, water quality management models are used to control discharges from sources contributing to the pollution loading to the receiving water body in order to comply with stipulated water quality standards. Information used in the management models is typically sparse and imprecise. Uncertainty in the model prediction and the decision making process should be incorporated in the analysis for determining an optimal pollution reduction strategy.; This study develops a fuzzy sets-based framework to determine an optimal allocation of waste load among contributing point and nonpoint sources to meet dissolved oxygen (DO) standards for a receiving water. The framework consists of a water quality simulation model (FDOM) linked to a multi-objective optimization procedure, (FLOWLAP). The investigation demonstrates that fuzzy sets can effectively represent uncertainties encountered in the modeling and optimization process. Water quality model results are validated with output from the EPA model, QUAL2E. The application of FDOM and FLOWLAP is illustrated in an example involving a hypothetical study area as well as a real case study, Withlacoochee River.; This technique can be used to obtain cost-effective solutions to water quality management problems by incorporating significant sources of uncertainty in the decision-making process. Fuzzy set techniques are combined with a trade-off analysis to address the uncertainty-optimization issue. The framework evaluates uncertainty in the modeling and optimization process to identify a most cost-effective management strategy. Incremental expenditures for acquiring additional data can be weighed against the incremental benefits of reduced pollution control costs thereby considerably reducing information costs.