Can Satellite Precipitation Products Comprehend Rainfall Extremes Over Disaster-Sensitive Mountainous Basins? An Exhaustive Inter-comparison and Assessment Over Nepal

The dearth of continuous precipitation time-series presents a formidable hindrance to advancing our understanding of hydrological extremes and their impacts, particularly over multi-hazard-sensitive mountainous landscapes. Despite the availability of remotely-sensed Satellite Precipitation Products (SPPs) as viable alternatives to scarce ground-truth, their assessment has been limited to only local/regional scales. The present study, for the first time, offers a thorough nation-wide evaluation of four state-of-the-art SPPs, namely, CMORPH, CHIRPS v2.0, PERSIANN-CDR, and PDIR-Now, over the entire Nepal. The mean and varying intensities of rainfall from 2000 to 2015 in each SPP are compared with APHRODITE (0.25° × 0.25°) for Nepal. To obtain deeper insights into the performance of SPPs, categorical indices, e.g., POD, FAR, FB, and CCSI, were determined over 28 stations at the basin-scale. Our observations indicate that CHIRPS v2.0 outperforms other SPPs on most occasions, including spotting the intensities and occurrences of extreme rainfall over Nepal, followed by CMORPH. Exhibiting high POD (0.46–0.87) and low FAR (0.02–0.44), CHIRPS v2.0 closely matches observed rainfall, making it the most reliable SPP for Nepal’s diverse landscape. Moreover, the Taylor Diagram plots reveal that CHIRPS v2.0 demonstrates significantly higher correlations (0.50 to 0.70) across most stations, outperforming other SPPs. In contrast, other SPPs showed varying degrees of underestimation or overestimation, with CMORPH emerging as the second-best performing SPP in most analyses. Our study emphasizes the cautious selection of SPP for complex topography in mountainous basins; which if not considered, may result in significant deviations to multi-hazard risks through SPP-rainfall-driven numerical modeling.