Assimilation of evapotranspiration maps derived from Modis satellite data into the Palmer distributed soil moisture model

Soil moisture is considered as an important hydrological parameter which has a significant role in agricultural drought monitoring. Palmer model is a lumped hydrological model and has been applied widely for drought monitoring. Recently its distributed version has been developed for regional drought monitoring in Iran, but correction of evapotranspiration methods which definitely have a high effect on soil moisture residual have been less regarded. In this study, potential evapotranspiration derived from MODIS satellite images and AHAS algorithm was used to reduce the uncertainty of the model caused by the input parameters. The soil moisture simulated using both original and developed Palmer distribution model. Results showed that updating Palmer distribution model during simulation increased correlation compared to simulation without any updates as coefficients of determination improved from 0.304 to 0.5348 in June. The results also showed that using potential evapotranspiration maps for dry periods had a positive effect on the model correction, while it had not significantly improvement in wet periods. The highest amount of coefficient of determination (R2) and the lowest amount of absolute error of 0.53 and 0.0595 respectively was observed in dry period (June) which can be recommended as the best time to update the images of simulated potential evapotranspiration.