A Multi-Agent System Based Model 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 involves distributed, local decision processes, as well as mechanisms that coordinate individual decisions and manage basin-level issues. The Yellow River Conservancy Commission (YRCC) has implemented a water allocation management plan, namely Unified Water Flow Regulation (UWFR), since 1998, as a guideline for water allocation among the eight provinces located in the basin. This regulation works as a centralized control mechanism that enforces the upper limit of water withdrawals (water permits) for users along the river. It is expected that if all the provinces from upstream to downstream follow the allocation plan, then the flow cutoff events that occurred every year during 1972–1998 can be avoided. In the context of the regulation plan, this study attempts to understand the behaviors of the water users in the basin, explores the socioeconomic and environmental consequences of the regulation, and tests options to improve water allocation management in the basin. To address these objectives, a multi-agent system (MAS) based model is developed for the YRB. The MAS model depicts agents' autonomy in water use decisions and interactions among the agents with consideration of both human and ecological water demands. Several modeling scenarios show possible improvements of the current regulation regarding economic, social (equity) and environmental objectives.