Estimation of regional actual evapotranspiration based on the improved SEBAL model

Abstract

The global food and water crises caused by climate warming are becoming increasingly significant. One of the important aspects of surface water transfer is evapotranspiration. The accurate estimation of evapotranspiration (ET) is of great significance in water resource management. However, accurate estimation of large-scale ET is challenging due to the adequate selection of the hot and cold pixels, the heterogeneity of the subsurface and the instability of the lower atmosphere. Therefore, in this paper, a surface energy balance model (Y-SEBAL) was proposed based on the SEBAL model and improved sensible heat flux calculation, and used to simulate large-scale actual evapotranspiration. Based on the principle of surface energy balance, this model introduced the empirical formula of sensible heat flux and simplifies the selection of limit to address the challenge of calculating the sensible heat flux of the crucial variable in the SEBAL model. The study simulated the spatial and temporal evolution of regional evapotranspiration with the model and verified the model simulation performance. In addition, the sensitivity of the model-driven data was quantified. The results showed that the simulation value of the Y-SEBAL model was highly consistent with the eddy covariance measurement data. In the verification results, the R value was 0.82, the index of agreement was 0.90, and the root mean square error was 0.81 mm·d−1. The performance verification indices were better than the SEBAL model, MOD16 model, and SSEBop model. And the model had the highest sensitivity to wind speed, reaching 0.714. The Y-SEBAL model improved the simulation performance of the original model and simulated the actual evapotranspiration. The study can provide a new solution for improving the retrieval of evapotranspiration by remote sensing at the regional scale.