Optimization of Maize Irrigation Strategies in the Middle Reaches Irrigation Area of the Heihe River Using a Differential Evolution Algorithm

Abstract

Optimizing maize irrigation strategies is essential for improving water use efficiency and crop yields in arid regions. However, limited quantitative research exists on these optimizations. This study focuses on the Heihe River Basin in China, aiming to (1) optimize maize irrigation strategies using a differential evolution (DE) algorithm integrated with the AquaCrop model and remote sensing data; (2) compare the DE algorithm’s performance with the traditional Nelder–Mead (fmin) algorithm regarding yield improvement and irrigation water use; and (3) assess the benefits of different irrigation strategies under limited water availability. Covering 22 irrigation management zones in Zhangye City, Gansu Province, the study utilized soil, weather, and crop data from Google Earth Engine to drive the AquaCrop model. Results indicate that the DE algorithm achieved higher simulated maize yields, increasing by 0.5 to 1 t/ha on average compared to the fmin algorithm, albeit with a 30% rise in irrigation water usage. The integration of both the DE and fmin algorithms with the AquaCrop model facilitates the development of tailored irrigation strategies, providing a scientific foundation for sustainable agricultural water management. These findings can guide efficient irrigation management plans in the region and similar arid systems.