Modifying the triangle method and MOD16 algorithm for improving evapotranspiration estimation using MODIS data

A new approach for estimating evapotranspiration (ET) driven by MODIS satellite data is proposed based on the integration of the triangle method and MOD16 algorithm. To improve the ET estimation, several pre-processing methods, including destriping for MODIS albedo data, denoising for MODIS EVI data, and topographic correction for MODIS land surface temperature (T_s) data are applied, and several ways determining the cold (wet) edge of EVI~ T_s space are compared. The new approach is applied to the Yihe River watershed, and the estimates with the new approach are compared with the estimates with the triangle method and MOD16 algorithm as well as pan evaporation observations, which shows the reasonability and applicability of the new approach.