<p>Precipitation is the most significant component of the global water and energy cycle associated directly with major Earth system processes- atmospheric circulation, clouds and water vapor and overall regulation of the biogeochemical cycle. Precipitation also has a major contribution in maintaining the socio-economic stability of the world as it is the primary source of freshwater and directly affects the food and water security. Extreme weather events associated with precipitation such as floods, droughts, landslides etc. are likely to intensify under current climate change scenarios which could induce mass migration and human conflicts due to unavailability of food and freshwater resources. Hence, accurate precipitation estimates are crucial in enhancing our understanding of the changing earth system processes and management of water resources through numerical weather predictions and hydrological forecasting. In this study, we evaluated the performance of a satellite-gauge merged rainfall product (GSMaP-IMD, 0.1×0.1) with a gauge based observational data (Indian Meteorological Department (IMD) daily gridded rainfall, 0.25×0.25) and two global satellite-based rainfall products- IMERG Final-run and GSMaP-CPC (standard JAXA product, GSMaP-Gauge) over 4 major river basins of Western India for the southwest monsoon period during 2000-2020.  </p><p>Results indicate that GSMaP-IMD better represents the overall distribution of rainfall over the river basins. The cumulative rainfall distribution over the study area is represented more realistically than other two datasets, especially at higher rainfall intensity (mm/day). GSMaP-IMD has smaller root mean squared error and higher correlation coefficient value than IMERG and GSMaP-CPC during the observation period. The distribution of low and moderate rainfall improved remarkably in case of GSMaP-IMD compared to the other products. Temporally, higher rainfall events are not represented accurately by IMERG and GSMaP-CPC which is improved in GSMaP-IMD. Overall, it is observed that IMERG overestimated the high rainfall events while GSMaP-CPC underestimated it whereas GSMaP-IMD showed improvement in estimating the events over the study area. The probability of detecting true rainfall events is further improved in GSMaP-IMD for all the basins. IMERG shows higher false rainfall bias over regions with high rainfall intensity which is reduced in GSMaP-CPC and further improved in GSMaP-IMD. The total ability of a dataset to capture actual rainfall events (Critical Success Index) is further enhanced for GSMaP-IMD. Finally, IMERG shows a large negative bias in detecting low rainfall events while GSMaP-CPC shows large positive bias in detecting high rainfall spatially. This systematic error is reduced in the GSMaP-IMD rainfall product. The results indicate that the IMERG and GSMaP-CPC have difficulties in detecting low and high rainfall events and further have systematic error which is due to the orographic effects and regional characteristics in southwest monsoon. Overall, satellite-gauge merged rainfall dataset performed better than the satellite-based products over major river basins of Western India. The integration of in-situ rainfall data from gauges and radars in future with satellite products at regional scale is found to improve the bias characteristics in IMERG and GsMAP-CPC which is significant in improving the availability of rainfall dataset for regional hydrological modelling applications, numerical weather predictions and water resource management. </p><p> </p>

No abstract

No abstract