CONJUNCTIVE USE OF SURFACE AND GROUND WATER USING FUZZY NEURAL NETWORK AND GENETIC ALGORITHM

Abstract

Semiarid regions with their exceptional weather conditions, low precipitation, and high evapotranspiration pose a great challenge to water resources managers. One possible way to face this challenge is the conjunctive use of both surface water and groundwater resources in these regions. This paper proposes a conjunctive use model which has been implemented in Najafabad plain in central Iran. The model is one of simulation-optimization in which the simulation portion combines the Fuzzy inference system and Neural Networks (FNN) in order to take the climate conditions and the uncertainty in the relevant data into consideration while the optimization portion consists of a multi-objective Genetic Algorithm (GA). The objectives of the optimization model include not only minimizing water shortages in meeting the irrigation demands by the three irrigation systems operating in the region but also minimizing groundwater drawdown in order to control groundwater extraction in the aquifer. These objectives are subject to constraints on the maximum amount of surface and groundwater allocated to the irrigation zones and the maximum capacity of surface irrigation systems and also maximum and minimum allowable cumulative drawdown in the planning horizon. The results of the proposed FNN-GA model demonstrate the importance of the interactions between surface water and groundwater resources considered in a conjunctive use model for the planning and management of water resources in semiarid regions.