摘要：Abstract Evapotranspiration (ET) plays an important role in water and energy exchange between the atmosphere and the land surface. ET at regional scale is of great importance in hydrology, meteorology, and agriculture. The remotely sensed land surface temperature (LST)‐based two‐source energy balance (TSEB) model is widely used to monitor ET. However, due to the influence of clouds, the quality of remotely sensed LST products is declined and there are also many gaps in the LST products, which make it very difficult to estimate spatio‐temporally continuous ET based on remote sensing data. In this study, we proposed a method to generate hourly continuous ET using Fengyun‐2F geostationary images. First, the hourly continuous real LST are obtained using a clear‐sky LST reconstruction algorithm and radiometric correction algorithm, and then the reconstructed real LST is used to the estimation of latent heat flux using the TSEB model. The results in the Heihe River Basin show that the proposed method can successfully generate hourly continuous ET compared with three eddy covariance sites with different underlying surface. The estimated ET achieves acceptable performance with RMSE = 81.88 W/m2 and R2 = 0.67 at hourly scale and with RMSE = 1.48 mm/d and R2 = 0.65 at daily scale. We believe this can help the community to better understand the water and energy exchange between the atmosphere and the land surface at fine spatio‐temporal scale, especially on the hourly scale.
