Water Resource Management Using Multi-objective Optimization and Rainfall Forecast

This paper proposes an integrated model for water resource management using multi-objective optimization and rainfall forecast. To optimally allocate limited water supply, the objective functions involve maximizing net economic benefit while striving to maintain parity of water shortage in allocated areas. Moreover, because rain water is an important factor especially in monsoon season, rainfall forecast is considered along with water distribution in the model. The auto-regressive integrated moving average model (ARIMA) is used to fit rainfall time series during the year 1974-2003. The amount of forecasted monthly rainfall is useful in predicting real water demand, which is an input to the optimization process. The model is implemented into a decision support system (DSS) as a tool to assist in irrigation decisions in the Rayong province of Thailand, where water must be allocated to three main sectors: agricultural, domestic, and industrial areas.