Estimation of Reference Evapotranspiration at Regional Scale using Remote Sensing Data in Shahrekord Plain - Comparing the Results of SEBAL and METRIC Models to Lysimetric Measurements

Conventional methods for estimating evapotranspiration are based on point measurement and suitable for local areas, therefore, cannot be generalized for larger areas or watershed basins. The remote sensing technology is capable of using satellite images and meteorological data to estimate evapotranspiration in a wider area. In this study, estimates of evapotranspiration (ET) by SEBAL and METRIC models based on Landsat 7 ETM+ sensor were compared against ET measured by lysimeter on seven satellites passing time over Shahrekord plain located in Karun basin. The results showed that the lowest indices of NRMSE, MAE and MBE (respectively, 0.317, 1.503 and -0.973 mm per day) and the maximum of d index (0.768) belonged to SEBAL. These indices were 0.420, 2.120, 2.023 and 0.646 for METRIC, respectively. The results showed that the SEBAL was more accurate than METRIC model for estimating ET under Shahrekord plain conditions. As long as the possibility of getting complete hourly meteorological data be provided, or some modifications on METRIC model were done, SEBAL show closer results to reality, and therefore is recommended.