Impacts of Different Satellite-Based Precipitation Signature Errors on Hydrological Modeling Performance Across China

Abstract

The quasi-global availability of satellite-based precipitation products (SPPs) holds significant potential for improving hydrological modeling skill. However, limited knowledge exists concerning the impacts of different SPP error type on hydrological modeling skill and their sensitivity across different climate zones. In this study, forcing data sets from 10 SPPs were collected to drive hydrological models during the period 2001–2018 for 366 catchments across China. Here, we analyze the impact of the SPP errors associated with different precipitation intensities (light, moderate, and heavy) and different precipitation signatures (magnitude, variance, and occurrence) on the performance of hydrological simulations, and rank the sensitivities of SPPs errors for four major Köppen-Geiger climate zones. The results show that heavy precipitation in SPPs is generally associated with higher errors than light and moderate precipitation when compared to gauge-based precipitation observations, but hydrological model skill is more sensitive to errors from moderate precipitation than from heavy precipitation. The probability of moderate precipitation detection was identified as the most sensitive metric in determining hydrological model performance, with sensitivities of 0.58, 0.39, 0.59, and 0.47 in the temperate, boreal, arid, and highland climate zones, respectively. The variance error and magnitude error for heavy precipitation from SPPs were also identified as sensitive factors for hydrological modeling in the temperate and arid climate zones, respectively. These findings are crucial for enhancing the understanding of interactions between SPPs uncertainty and hydrological simulations, leading to improved data accuracy of precipitation forcing and the identification of appropriate SPPs for hydrological simulation in China.