Abstract:

Evapotranspiration (ET) is a critical component of the land surface energy balance system and hydrologic processes. Analysis of spatiotemporal variations and influencing factors of ET is of great importance to evaluate the growing environment for crops and to effectively use water resources in Anhui Province. The traditional methods are based on point measurement, while remote sensing provides extensive surface information. The development of remote sensing has promoted the study of regional ET. SEBAL model is based on Surface Energy Balance Algorithm for Land and its physical meaning is clear. This model was developed to show the spatial variability of surface evapotranspiration. SEBAL model was applicable to large areas in conjunction with Moderateresolution Imaging Spectroradiometer (MODIS) data products (i.e., MOD11A1, MOD11A2, MOD13A2, MOD43B3, and MCD43B3). According to the shortcomings of the traditional method of calculating ET, based on SEBAL model, the daily regional evapotranspiration of Anhui Province was estimated with 1 km spatial resolution by using MODIS products and meteorological data (including temperature and wind speed) collected in meteorological stations distributed across the study area. Firstly, the estimation results of March 26th and August 5th, 2010  were taken as examples in the analysis of daily ET characteristics of different land covers in the study area by using the spatial analysis module of GIS. Secondly, based on meteorological data of March 26 th, 2010 and August 5 th, including daily temperature, relative humidity, sunshine hours and wind speed of 80 meteorological stations in Anhui province, crop reference evapotranspiration (ET0) was calculated by using the PenmanMonteith (PM) formula, which was nominated by FAO in 1998 and Kriging method. Because of lacking observed data from the lysimeter, the results of PM were compared with the estimation results based on SEBAL model in this research. The comparison of the evapotranspiration estimated with MODIS products and field observation showed that the former results were lower than the latter results on the whole, and demonstrated that there existed a certain correlation between the two results, the average relative error was different at different land surfaces. The average relative error was below 20%. It was concluded that evapotranspiration of water body was high, which was 36 mm. It was followed by paddy field, which was 20 mm. Evapotranspiration of dry lands was the lowest, which was only about 165 mm. The ET computation method based on Remote Sensing proves that this model has strong practicality in Anhui Province, and it will show great potential in this field with better optimizing the model parameters