Stochastic Dynamic Programming-Based Approach for Optimal Irrigation Scheduling under Restricted Water Availability Conditions

This study presents an optimization approach for irrigation scheduling across the different growth stages of crops and estimates crop yield functions considering evapotranspiration uncertainty and water scarcity. A stochastic dynamic programming (SDP) model is developed in which the restriction on total water availability is explicitly accounted for by an additional state variable. The model is used to derive optimal 10-day allocation policies and maximize the expected value of annual yield as the objective function. The application of the proposed approach is illustrated in a multi-crop realistic case downstream of Zayandeh-Rud Dam, in the central plateau of Iran. Simulation of the derived optimal policies shows the significance of the proposed methodology compared with a deterministic linear programming-based approach. Copyright (c) 2017 John Wiley & Sons, Ltd.