Evapotranspiration is a key component of the global water cycle and significantly impacts water cycling and energy balance. However, existing evapotranspiration data exhibit limitations in their spatial and temporal resolution. To address this issue, this paper employs a new method that combines high-resolution GF-5B VIMI data with the SEBS model. The Dadukou District of Chongqing was selected as the study area. Using GF-5B VIMI imagery data and the ERA5-land climate reanalysis dataset, the daily evapotranspiration of the area was estimated, and its spatial and temporal distribution characteristics, as well as the evapotranspiration from different underlying surfaces, were analyzed. Additionally, the relationship between evapotranspiration and the Normalized Difference Vegetation Index (NDVI) was examined. Experimental results indicate that the estimates from the SEBS model are in good agreement with observations from eddy covariance systems, as evidenced by a determination coefficient (R²) of 0.764 and a root mean square error (RMSE) of 0.348 mm/d. Among the different land use types, water bodies showed the highest daily evapotranspiration rates, while bare lands showed the lowest. The correlation coefficient (R) between daily evapotranspiration and NDVI was 0.908, with a determination coefficient (R²) of 0.824, further validating the model's effectiveness and applicability. This approach significantly enhances the spatiotemporal resolution of evapotranspiration and offers critical scientific insights for the efficient management of water resources.