Estimation of Actual Evapotranspiration at Regional - Annual scale using SWAT

The aim of this study was to estimate actual evapotranspiration at regional – annual scale using semidistributed Soil, Water Assessment Tools (SWAT) model in the Neishaboor watershed. For this reason, SWAT calibration and validation was done based on river discharge data from 5 gauging stations, rainfed and irrigated wheat yield data for the period Oct. 2000 to Sep. 2007 and Oct. 2007 to Sep. 2010, respectively. Because of the direct relationship between crop yield and evapotranspiration, calibration of watershed models using crop yield along with river discharge gives more confidence on the partitioning of water between soil storage, actual evapotranspiration, aquifer recharge. The results showed that SWAT provided satisfactory predictions on hydrologic budget and crop yield. Specifically, calibration (R2 = 0.82, NS=0.79) and validation (R2 = 0.71, NS=0.71) periods were quite suitable for the outlet of watershed. It also was able to predict crop yield satisfactorily for irrigated wheat in which R-factor and RMSE values were 0.97 and 0.08 ton ha-1, respectively. The multi-objective calibrated model was then used to estimate and analyze the actual evapotranspiration. Mean ten-year actual evapotranspiration and precipitation was estimated 230 and 270 mm, respectively. The ten-year actual evapotranspiration to precipitation ratio at mountainous part of watershed was 99%, 80% and 77 for 2000-2001 as a dry year, 2001-2002 as a normal year and 2004-2005 as a wet year, respectively. Groundwater is an important source of water supply in the Neishaboor plain. Therefore, estimation of this ratio is not as simple as mountainous part of watershed due to uncertainties in the crop pattern data and their water requirements. It is obvious that this ratio could be more than one in some years especially dry years.