Managing water quality in a river basin with uncertainty

Abstract

The effects of both climate change and the geographic location of Taiwan have influenced the perceived variability of river flow and increased uncertainty and complexity in the management of river basins. In this study, a genetic algorithm (GA) optimizer was integrated into a stochastic river basin model to develop a stochastic optimization river basin management model (SORBMM). Firstly, the flow probability density function was determined through statistical analysis of the hydrological data. A Monte Carlo simulation was then conducted to evaluate the effect of flow variability, and a GA was implemented to obtain an optimal river pollution reduction strategy. A true case involving multi-objective management of a river basin under conditions of high spatiotemporal flow variation was tested to demonstrate the feasibility of the SORBMM. The results revealed that a reduction in pollution removal would lead to higher risks for river basin management due to the dilution effect in the river downstream and the objective of lowering pollution removal costs.