Development of conjunctive surface and ground water use model with emphasis on the quality and quantity of water resources

Many real water resources optimization problems involve conflicting objectives. In this study, multiobjective genetic algorithm NSGA-II, has been developed for optimization the conjunctive use of surface water and groundwater resources and optimal management of supply and demand of agricultural water. Here, optimal allocation of land and water resources to the dominant products in Najaf Abad plain, two surrogate models, Artificial Neural Network (ANN) and Genetic Programming (GP), has been linked with NSGA-II. Results according to Mean Squared Error and correlation coefficient values show the efficiency of alternative models for prediction the concentration of Total of Dissolved Solids (TDS) and groundwater level in observation wells. According to the final results of SO model, average drowdown in groundwater level is equal to 0.18 m in optimal conditions, compared to the current(pre-optimal) conditions has been reduced to one third,also average concentration of TDS decreased from 1258 mg/lit to 1229 mg/lit in optimal conditions.