INTEGRATED MANAGEMENT OF RIVER WATER QUALITY UNDER UNCERTAINTY USING OPTIMIZATION MODEL

摘要

A multiobjective optimization model is developed to explore management alternatives on allocation of COD (Chemical Oxygen Demand) load issuing from point sources to outfalls on Yasu River, Japan. Uncertainty inherent in the water quality environment is accounted for with twelve scenarios corresponding to months in a year. Two measures of robustness based on the framework of robust optimization are introduced to control effects of variable and imperfect input data. Employment of Geographic Information System facilitates deriving catchment boundaries associated with the river, finding outlets of wastewater, and estimating uncontrollable discharge and COD load from nonpoint sources. A finite element model for COD and DO transport, contained as equality constraints in the optimization model, is calibrated using observed data in Yasu River. Other constraints of the optimization problem are effluent limitation standards for wastewater and water quality standards in the river. Five alternatives obtained by optimization are shown, which are advantageous over the current COD loading activity in conserving river water quality and/or increasing allowable COD load. The presented method could be applied to other river systems to assist decision-makers in evolving better policies on water quality management.