Improving crop evapotranspiration estimation in the Yellow River Basin: A modified SEBAL model approach and spatiotemporal trend analysis

Crop evapotranspiration (ET) is a vital component of the water cycle and energy balance in agricultural ecosystems. Scientifically-based and accurate estimation of regional crop evapotranspiration plays a crucial role in regional agricultural water management. To improve the accuracy of ET in the Yellow River Basin (YRB), the Surface Energy Balance Algorithm for Land (SEBAL) was modified through structural reconfiguration of the LST-NDVI feature space. Based on the modified SEBAL model, the actual ET of wheat and maize was estimated in the YRB over 2003–2022. Additionally, we identified the key meteorological factors affecting the interannual ET variation. Results showed that the modified SEBAL model yielded robust estimates of ET for grain crops in the YRB. The spatially averaged ET for winter wheat, summer maize, spring wheat and spring maize in the YRB were 585, 380, 591, and 695 mm, respectively. Moreover, the ET of winter wheat and summer maize exhibited a decreasing trend in the Fenwei Irrigation District and all the downstream irrigation districts, and that of spring wheat and spring maize showed a significant decreasing trend in Jingdian and Qingtongxia irrigation districts over the past two decades. Wind speed and precipitation were the primary driver for the ET of wheat and maize in the YRB. Overall, this study established an improved framework for crop ET estimation at the regional-to-basin scale, and these findings provided a scientific reference for agricultural water management in the YRB. • Reconfiguration of the LST-NDVI feature space by nonlinear dry/wet edge equations. • A modified SEBAL model was established to simulate ET in the YRB. • The ET of wheat and maize exhibited a decreasing trend in the YRB.