Estimating Rice Actual Evapotranspiration Using METRIC Algorithm in a part of the North of Iran

With respect to that water resources shortage in recent years in Iran has become a major challenge and the major water losses are in the agriculture section and due to evapotranspiration, sustainable management of water resources requires identification and correct estimation of evapotranspiration. Among the different methods to estimate evapotranspiration, the remote sensing technique is an appropriate option for evapotranspiration estimating because of appropriate coverage of place and time in regional scale and using this technique, it is possible to appropriately estimate the regional evapotranspiration without requiring the historical information about soil, crop, and type of farm management. Among the various algorithms based on remote sensing, in this research, relatively new METRIC algorithm has been used for estimating rice evapotranspiration in the Rasht city.
To estimating the rice evapotranspiration in the Rasht city, meteorological data of agricultural weather station of Rasht including variables of air temperature, relative humidity, sunshine hours, rainfall and wind speed in the 3-hour time scale, MODIS Satellite images and Lysimeter data were used for 8 non-clouds days of the August, 2014. Using the mentioned data and images and by adopting the energy balance based algorithms of METRIC, The rice crop evapotranspiration was estimated and for validation the results of the METRIC algorithm, the Lysimeter measured data at the Rasht Rice Research Institute in the summer of 2014 were used.
The results of this study showed that by an implementation of the METRIC algorithm and by using MODIS satellite images, it is possible to estimate the actual rice daily evapotranspiration with a reasonable accuracy. Comparing between the estimated amounts of rice evapotranspiration by the METRIC algorithm and the measured Lysimeter values showed that the amounts of three applied error indices including RMSE, MAE, and MBE were respectively 1.21, 1.14 and -1.17 mm/day. These extents represent an acceptable accuracy of the METRIC algorithm for estimating rice evapotranspiration, but in general, this algorithm tends to underestimate the amount of rice evapotranspiration so that the mean of evapotranspiration at the 8 mentioned days using Lysimetery data and the METRIC algorithm were 9.2 and 8 mm, respectively.
In general, use of remote sensing technique and METRIC algorithm as a new method for estimating evapotranspiration have an appropriate and acceptable accuracy and the source of errors between calculated and measured values of evapotranspiration and underestimating of METRIC algorithm as well can be attributed to some factors such as low spatial resolution of MODIS images and selecting procedure of hot and cold pixels.