Estimation of Actual Evapotranspiration Using Automatic Calibration in PYSEBAL and METRIC Algorithms in Qazvin Plain

Estimation of actual evapotranspiration over large areas with appropriate time intervals is important in the optimal management of water resources. The constant need for evapotranspiration data has led to the development of several methods for estimating it. In recent years, the use of remote sensing method to estimate the rate of evapotranspiration in large areas with the desired spatial and temporal resolution has been proposed. In this study, the automatic calibration efficiency of two common mono-source evapotranspiration algorithms estimated from PYSEBAL and METRIC were compared with the results of a drainage lysimeter planted with grass in the Qazvin plain. In this regard, 15 images TM, 22 images ETM+ and 24 images MODIS without cloud and snow during the years 2000 to 2003 were used, which in total 122 outputs were obtained from both algorithms. The results of this study showed that the METRIC algorithm in all three sensors with RMSE (0.42, 0.42 and 1.05 mm / day) respectively had better performance than the PYSEBAL model. Also, the studies performed from the three sensors showed that the MODIS sensor with standard error value (0.15 mm / day) and correlation coefficient (0.98) compared to the two ETM and TM sensors with correlation coefficient values (0.97 and 0.92), standard error (0.17 and 0.59 mm/day) have been able to produce better results. The executive applications of this study can be used to determine the exact amount of evapotranspiration in irrigated lands for water allocation planning, optimization of crop production, irrigation management and assessment of land use change on water effieiency.