A Comparative Analysis of Spatial-Temporal Precipitation Accuracy of TRMM and Synoptic Stations

The Tropical Rainfall Measurement Mission (TRMM) provides an important rainfall database for hydrological applications in aquatic basins. However, its accuracy and usage have not been sufficiently studied due to the occurrence of network-scale rainfall and water basins. In this study, the accuracy of the prediction of TRMM 3B42 V7 rainfall values ​​in Iran is investigated. For conducting this study, the daily net precipitation values ​​of TRMM 3B42 V7 from an Iranian basin were extracted with a 0.25-degree spatial resolution over the period 01.01.1998 to 31.12.2015. Precipitation data recorded at synoptic stations were also provided during this period. The accuracy of TRMM precipitation is evaluated using the nearest neighbor spatial resolution function. The findings show that not only in terms of temporal coherence but also in magnitude, there was a significant difference between the predicted TRMM rainfall values ​​and the rainfall recorded by the stations. The bias value (BIAS), mean absolute error of magnitude (MAE) and root mean the square of error (RMSE) of the Caspian, Urmia, and Persian Gulf basins were reported much higher than in other regions. The Precipitation indicators of the probability of detection (POD), Fault Alert Ratio (FAR), Mean Critical Success Index (CSI) indicates lower accuracy for TRMM 3B42 V7 in the prediction of rainfall at the grid-scale and catchments of Iran.