Decentralized Optimization Method for Water Allocation Management in the Yellow River Basin

Abstract

The management of large river basins, such as the Yellow River Basin (YRB) in China, is complicated by distributed, localized decision processes and by mechanisms that coordinate local decisions and manage basin-level issues. Since 1998, the Yellow River Conservancy Commission (YRCC) has launched the Unified Water Flow Regulation (UWFR) as a centralized controlling mechanism that enforces an upper limit on water withdrawals (water use permits) for eight provinces located in the basin. The implementation of UWFR has maintained a prescribed minimum flow in the downstream channel and avoided the flow cutoff events that occurred every year between 1972 and 1998. This study attempts to explore the socioeconomic and environmental consequences of the regulation and test plans to improve water allocation management in the YRB. A decentralized optimization is combined with a multipleagent system (MAS) framework for the YRB, in which water users, reservoirs, and downstream ecological zones are defined as agents. This method iteratively determines water prices for each water use agent in the context of water market. The proposed water market scenario shows possible improvements of UWFR with respect to social, economic, and environmental objectives.