Estimation of regional evapotranspiration based on remote sensing: case study in the Heihe River Basin

Accurate estimation of evapotranspiration (ET) has long been an important issue in hydrology. Many experimental observations indicate advection has a great impact on ET in arid and semiarid areas. However, most of the remote sensing models only focus on the vertical energy balance and deviate from reality. A revised Penman equation has been derived to estimate actual ET under normal conditions in order to account for advection. The parameter of the water availability for ET is introduced and the T-s/f (surface temperature and vegetation fraction) space is used to characterize this parameter in the revised formula. The estimates were validated using the observations measured with eddy covariance systems at the Yingke station. In 22 of all 24 days, the difference between the observations and the estimates was smaller than 70 W/m(2). The correlation coefficient is 0.91 and the RMSE is 48.38 W/m(2). This finding reveals that this approach is capable of providing reliable results. In addition, when considering advection, the potential ET can be 83.23 W/m(2) larger than the available energy. This finding indicates that the advection effect needs to be considered in remote sensing models in order to derive more reliable regional ET. (C) 2012 Society of Photo-Optical Instrumentation Engineers (SPIE). [DOI: 10.1117/1.JRS.6.061701