Abstract
Realities of in-stream water quality and its control in Yasu River, Japan are diagnosed by use of an epsilon robust optimization (ε-RO) model with the aid of a Geographic Information System (GIS). The GIS helps delineate catchment boundaries associated with a diagnosed section of the river, find outfalls of wastewater, and estimate both discharge and uncontrollable COD (Chemical Oxygen Demand) load from nonpoint sources. Twelve hydro-environmental scenarios assumed to stochastically occur in a year are generated based on historical data. Calibration of a finite element model for COD and dissolved oxygen transport is conducted to identify some scenario components. The ε-RO model determines the optimal wasteload quotas at point sources (PSs) through allocation of the maximized total COD load, which lead to ameliorating water quality in the river. The diagnostic results show that there is every possibility that the in-stream water quality is upgraded even if total COD loading is increased so long as it is appropriately allocated to PSs.
(Communicated by Toshio SAWADA, M.J.A.)
