Estimation of Precipitation Using Satellite-based Surface Soil Moisture (SSM) in Semi-Arid and Humid Climates of Iran

One of the new methods for estimation of rainfall is SM2Rain algorithm which calculates rainfall using soil moisture variations and inverse solution of soil water balance equation. This research addressed the efficiency of SM2Rain algorithm for rainfall estimation over the semi-arid (Khorasan-Razavi) and humid (Mazandaran) climate regions of Iran using ASCAT surface soil moisture dataset during 2006-2013. Findings indicate that the basin-averaged value of correlation coefficient (CC) between the estimated and observed datasets for Khorasan-Razavi and Mazandaran areas is 0.70 and 0.62, respectively. Results in the south and south-west regions of Khorasan-Razavi showed that the SM2Rain algorithm with the CC value of 0.84 and RMSE value of 3.9 mm/day (basin-averaged) performs very well, while in the north parts of the province with the CC value of 0.54 and RMSE value of 7.7 mm/day, the performance of this algorithm is relatively low. Also, the performance of SM2Rain algorithm in most parts of the Mazandaran province, especially in east and central parts, is acceptable and the basin-averaged values of CC and RMSE are 0.72 and 3.9 mm/day, respectively. The results also showed that by adding evapotranspiration term to SM2Rain algorithm, the efficiency of modified algorithm in estimation of rainfall increases about 10-18% in both regions. Furthermore, by using the modified SM2Rain algorithm over the Khorasan-Razavi, the basin-averaged value of relative bias (RBias) decreases from -21.9% to 9.3% and in Mazandaran region, the RBias decreases from -36.9 to 7.9%. The findings of this research indicate that the estimated rainfall with the SM2Rain algorithm can be considered as an alternative or supplementary dataset for ground-based observations, especially in ungauged catchments or data-limited areas.