Estimation of high-resolution terrestrial evapotranspiration from Landsat data using a simple Taylor skill fusion method

Estimation of high-resolution terrestrial evapotranspiration (ET) from Landsat data is important for many climatic, hydrologic and agricultural applications. It can also bridge the gap between existing coarse-resolution ET products and point-based field measurements. This study presents a simple Taylor skill (STS) fusion method by merging five Landsat Thematic Mapper (TM)-based ET products produced by the individual algorithms and FLUXNET eddy covariance (EC) observations to improve terrestrial ET estimation. The independent validation results show that at the site scale, large differences were found in the five Landsat TM-based ET products among different plant functional types and the merged ET product has the best performance with a decrease in the averaged root-mean-square error (RMSE) by more than 2 W/m2 when compared to the other products. To evaluate the reliability of the STS fusion method on a regional scale, the weights of the STS fusion method using five ET products and all EC ground-measurements collected at 206 EC flux tower sites were used to map the regional ET. An example of regional ET mapping shows that the STS-merged ET can effectively achieve the goal of ET products integration. This study will provide an effective high-resolution ET product for identify agricultural crop water consumption and provide a diagnostic assessment of global land surface models.