Simulation-based optimization for spatiotemporal allocation of irrigation water in arid region

Abstract Optimal allocation of irrigation water both spatially and temporally is a great challenge in arid and semi-arid regions with limited water available for irrigation. In this study, a simulation-based optimization model for spatiotemporal allocation of irrigation water in arid region was developed, which integrated a distributed farmland water balance model for simulating hydrological processes in farmland and an optimization model to maximize irrigation and precipitation use efficiency (IPUE). A typical arid region, the middle reach of the Heihe River basin in Northwest China with a total farmland area of 2624 km2, was selected as the study area. Field experiment data in 2012 and 2013 were used to calibrate and validate the simulation model. Then, the simulation-based optimization model was used to optimize the spatial and temporal allocation of irrigation water under various scenarios. Results indicate that the model is applicable for reflecting the complexities of water consumption and irrigation water allocation in the study area. The optimal irrigation water depths averaged for different farmland units are 326 mm, 327 mm, 471 mm, and 495 mm for spring wheat, barley, vegetable, and maize, respectively. After optimal allocation of irrigation water, IPUEs can exceed 0.8 for all climate conditions, and IPUE in the wet year was lower than those in other circumstances because of higher precipitation in the wet year. Higher IPUE can be achieved when the gross irrigation water are 1.72, 1.90 and 1.97 billion m3 for the representative wet, normal, and dry years, respectively. The results can be used to guide irrigation water management by providing the information on when and how much to irrigate for each crop in different farmland units.