资源环境科技发展态势分析平台

Evapotranspiration (ET) is an important component of the water and energy cycle. The present study develops a practical approach for generating all-sky ET with the synergistic use of satellite images and meteorological data. In this approach, the ET over clear-sky pixels is estimated from a two-stage land surface temperature (LST)/fractional vegetation cover feature space method where the dry/wet edges are determined from theoretical calculations. For cloudy pixels, the Penman-Monteith equation is used to calculate the ET where no valid remotely sensed LST is available. An evaluation of the method with ET collected at ground-based large aperture scintillometer measurements at the Yucheng Comprehensive Experimental Station (YCES) in China is performed over a growth period from April to October 2010. The results show that the root-mean-square error (RMSE) and bias over clear-sky pixels are 57.3W/m(2) and 18.2W/m(2), respectively, whereas an RMSE of 69.3W/m(2) with a bias of 12.3W/m(2) can be found over cloudy pixels. Moreover, a reasonable overall RMSE of 65.3W/m(2) with a bias of 14.4W/m(2) at the YCES can be obtained under all-sky conditions, indicating a promising prospect for the derivation of all-sky ET using currently available satellite and meteorological data at a regional or global scale in future developments.