عنوان مقاله [English]
Precipitation is one of the important climatic parameters in hydrological models. Therefore, the accurate estimation of its amount and spatial distribution in a watershed is of great importance. In recent years, due to the limited number of raingage stations especially in mountainous areas, the use of satellite precipitation data as an effective tool for predicting regional distribution of rainfall has gained lots of attention by the researchers. In the present study, the accuracy of TRMM 3B43 data, which is one of the TRMM products, was evaluated in 40 raingage stations and 9 synoptic stations in Hormozgan province in a monthly scale. Comparison of satellite data with the observation data was performed for the time period 1998-2012. To assess the agreement between TRMM rainfall data with observation data, the statistical criteria including Spearman correlation coefficients (Rs), Pearson correlation coefficient (Rp), mean absolute error (MAE), root mean square error (RMSE) and mean square error (MSE), as well as probability of detection (POD), false alarm ratio (FAR), true skill statistics (TSS) and critical success index (CSI) were used. Based on the results, the highest value of POD (1) was observed in August and the lowest POD (0.92) was obtained in May. Moreover, the highest FAR (0.91) and the lowest CSI (0.08) were observed in May, and the lowest FAR (0.16) and the highest CSI (0.83) was obtained in January. In addition, the highest Rp (0.64) and Rs (0.76) were seen in December while the lowest Rp and Rs were occurred in April, May and July. These results indicate that the TRMM satellite has the highest accuracy of predicting rainfall in winter and spring while it has the lowest performance in summer. In other words, the TRMM satellite is able to predict rainfall in cold months better than in warm months of the year. The results also showed that TRMM overestimates rainfall in most months of the year, however the results were significantly improved after calibration especially in August and December as seen in spatial distribution maps.