Agricultural water production function-driven water resources optimization model for irrigation management under extreme drought- a case study for the upper reaches of Yangtze River, China

Chengdu city as one of the major grain-producing areas in China are located in the upper reaches of Yangtze River. Agricultural irrigation is the primary water use in Chengdu. However, efficient and equitable water allocation remains a critical challenge in agricultural allocation water management, particularly under severe drought events. To precisely optimize water allocation temporally and spatially, an agricultural water production function-driven water resources optimization model was proposed with monthly-water sensitivity indices method and logarithmic-transformed Jensen model approach for simulate water allocation among different crops during growth period in eight different drought scenarios. The model strategically allocates more water to subareas with smaller planting areas, while larger planting areas experience greater yield volatility under water deficit. Water allocation should be prioritized due to critical growth periods for the studied crops. During non-sensitive months, crops water requirement mainly derived from rainfall with less irrigation water. Rice should be prioritized for irrigation to minimize drought losses and ensure staple food security. Under non-drought conditions, the model tended to maximize crop yield. In contrast, during drought events, it adjusted to restricting total water use and maintaining profit optimization and drought loss reduction in balance. • Agricultural water production function-driven water resources optimization model is developed. • Multi-water sources and monthly crop growth stages are employed to address severe drought events. • Smaller crops planting area would have more advantages in reducing drought loss. • The solutions adapt to drought scenarios and balance profits and water use.