Validation of a Seasonal Satellite Precipitation Correction Model to increase the Accuracy of Water Balance in Arid Regions of Iran

Rainfall always has been considered as one of the most important variables for estimating the balance of watershed. The evaluation and correction of precipitation data as a supplement to the ground data in Iran, which it most regions is located in arid and semi-arid zone, is necessary. In this study, the precipitation data from 6 synoptic stations in 3 climate of Iran (extra arid, arid and semi-arid) were selected as the basis for the period of 20 years (1998-2017). The satellite monthly precipitation data (TMPA-3B43) was corrected with the multiplicative model, and results were evaluated by using R, MBE, MAE, and RSME indices. Based on uncorrected results, the correlation coefficient (R) varied from 0.79 for Garmsar station to 0.95 for Ramhormoz station. In all three climates, satellite data had overestimated values. After applying the multiplicative model, a seasonal parameter has been obtained for satellite data correction. All evaluation indices, especially the MBE, were significantly reduced after the correction. The lowest values for this error in the corresponding pixels of Garmsar, Boshrouyeh, Saveh stations were reduced to 0.1(spring), 0.4(autumn), and 0.7(winter) respectively. According to the results of the correction parameter in all three studied climates, the highest overestimate was observed in the autumn season. Also, this correction coefficient was 0.64, 0.67, and 0.79 for Qazvin(semi-arid), Garmsar(extra-arid), and Saveh(arid) stations, respectively. Based on the obtained results, the calibration model can be used to correct the seasonal data of the remote sensing in dry regions.