Accuracy evaluation and comparison of GPM and TRMM precipitation product over Mainland China

For evaluating the accuracy of the remotely sensed precipitation products of Tropical Rainfall Measuring Mission 3B42 (TRMM) and latest-generation Global Precipitation Measurement IMERG (GPM), four commonly used error metrics (R, E_RMS, E_MA and E_R) and four category statistical indexes (R_FA, P_OD, B and S_ET) were adopted to quantify the evaluations on GPM, TRMM products by comparing with the gauged daily precipitation data obtained from the 824 meteorological stations within Mainland China. Evaluations were carried out spatially over the Mainland China and the 9 drainage basins at the daily and monthly temporal scales. The results suggested that:① The accuracy of GPM generally outperformed TRMM in Mainland China at daily scale, the R, E_R and S_ET of GPM and TRMM reached to about 0.73, 2.03%, 0.36 and 0.70, 3.75%, 0.33, respectively; ② At daily scale, both GPM and TRMM products exhibited high detection accuracy in the Haihe River basin, the Huaihe River basin, the Yangtze River basin, the Pearl River basin and the Southeast Drainage basins, while the data accuracy in the Songhua and the Liaohe River basin, the Yellow River basin and the Southwest Drainage basins were relatively lower and the detection accuracy in the Inland Drainage basins shown relatively the lowest; ③ As the improved capability of the GPM in detecting light rainfall and solid precipitation significantly benefited the accuracy increase of GPM over Mainland China, GPM performed obviously better than that of TRMM in winter precipitation detections at monthly scale. Overall, as the successor of the TRMM, GPM outperformed TRMM over each drainage basin as well as in whole Mainland China, which implied the great potential of its application in various studies. However, yet rather poor accuracy of the GPM precipitation product in high altitude and arid regions as well as high false alarm ratio for light precipitation over the drainage basins in the Southeast China remained a challenge to improve its IMERG inversion algorithm.