Actual Evapotranspiration Estimation Using MODIS and ETM+ Imageries (Case Study: Arak)

In this research, the spatial distribution of evapotranspiration and its relationship with remote sensing in contrast with lysimetric data as control was investigated in Arak, Markazi province in Iran. For estimation of actual evapotranspiration amount in the region based on SEBAL, SSEB and TSEB algorithms, 28 imageries of MODIS and Landsat7 (ETM+) were used for the years of 2000-2004. The multiplicity of MODIS images and its high temporal resolution is the reason of least error for ET estimation. According to the statistical results, the SEBAL model with the lowest RMSE in both TERRA and ETM + sensors (0.97 and 1.38 mm/day) was presented as the superior model in the region. Also, TSEB model showed the weakest results among the proposed models, in both MODIS and ETM + sensors (3.57 And 2.53 mm per day). Comparing the performance of two sensors, the ETM+ satellite images are recommended for ET estimation due to increased spatial resolution and improved resolution of images in the Landsat satellite. In addition, the NDVI vegetation index was at its lowest level at the beginning of the growing period due to germination and vegetation thinness, and it is increased by increasing air temperature and vegetation cover. L factor has a significant effect on SAVI and ET estimation and it is depended on the region vegetation. In this study, the L factor for the studied area was estimated to be 0.6 during the maximum growth period, which had the least amount of error in comparison with other values.