Reliability assessment of satellite-based precipitation products for coastal region through comparative analysis across multiple global datasets

The study examines at how accurately multiple reanalysis and satellite-based precipitation datasets represent actual precipitation across Kerala, an Indian coastal state. The IMD gridded rainfall data was used as the reference dataset for comparative analysis. Eight datasets including IMERG (Early Run, Late Run and Final Run), CMORPH CDR, CDR, PDIR, PERSIANN and CCS data products were analyzed for the 21-year period from 2003 to 2023. Four statistical techniques were used to evaluate the performance of these dataset. These techniques include the correlation coefficient, root mean square error (RMSE), mean error (ME) and relative bias (RB). The results show that the IMERG Final Run dataset performs better, especially in regions with heavy rainfall like the Western Ghats. On the other hand, datasets such as PDIR, PERSIANN and CCS tend to underestimate rainfall and show weaker correlation with the IMD gridded data especially in hilly areas. The results illustrates the importance of choosing correct dataset for accurate rainfall analysis which can be used for hydrological and climate studies in complex coastal regions.