PROBABILISTIC GOAL DRIVEN WATER QUALITY MANAGEMENT

Water quality management decisions are often based on objectives to achieve strict water quality standards generally defined by numerical values for constituents of interest. However, strict adherence to these quantitative standards in decision making is not possible in many instances. The current study proposes an approach for water quality management that utilizes water quality standards as probabilistic goals. A simulation-optimization model is used to achieve this objective. The simulation model is first calibrated and validated for a specific water quality constituent of interest, and then is used to derive a probability density function (PDF) of that constituent, over a specific simulation period. A separate PDF based on water quality standard is defined and is referred to as goal based PDF. An optimization model is formulated with an objective of minimizing the difference between derived and goal based PDF. To demonstrate this concept a discrete volume method (DVM) based simulation model used for modeling the fate and transport of fecal coliform in streams is adopted and is linked with a genetic algorithm based optimization model. Two 8-digit USGS watersheds that comprise of several 11-digit USGS HUC watersheds in the Big Sandy region of southeastern Kentucky are used to assess the applicability of this management model. Straight pipes and failing septic systems are identified as two main sources of pathogen impairment in this region and are used as major decision variables in the management model. The management model can be used to estimate fecal load reductions due to contributions from point and non-point sources to achieve pre-set probabilistic water quality goals.