Distributed simulation for regional evapotranspiration and verification by using remote sensing

Regional evapotranspiration (ET) in river basins, one of the most important hydrological processes, was usually unable to be estimated efficiently due to the lack of reliable observed data. Recently, with the development of remote sensing (RS) technology and hydrological model integration, it becomes possible to estimate ET at regional scale. Remote sensing derived ET (1 km-resolution), which was derived from MODIS satellite data by the institute of remote sensing and application (IRSA) of CAS, was applied to validate the monthly ET simulated by using soil and water assessment tool (SWAT) in the Zhangweinan River basin. Agricultural management practices and parameters for crop growth were then calibrated and analyzed. The calibrated and validated results showed that ET was sensitive to both crop parameters and the soil evaporation compensation coefficient (ESCO, its value was suggested to range from 0.8 to 1.0 for the crop land of winter wheat-summer corn). Simulation had been improved greatly by using remote sensing derived ET. The coefficient of determination R2 reached 0.93, and Nash-Sutcliffe efficiency coefficient reached 0.92. Regional ET simulation and then the hydrological modeling were greatly improved by using remote sensing derived ET.