Exploring the Potential of Satellite Precipitation Products in Hydrological Application: a Case Study over a Semi-arid Watershed in Northeastern China

Abstract:

The rapid-developed satellite precipitation product (SPP) has unprecedented value in providing spatiotemporal continuous precipitation estimates, which will benefit hydrological application. This study comprehensively evaluated the precipitation accuracy and hydrological simulation performances of the latest versions of various SPPs over a sparsely-gauged watershed in northeastern China, including both gauge-adjusted and gauge-unadjusted products of Tropical Rainfall Measurement Mission Multi-Satellite Precipitation Analysis (TMPA), Global Precipitation Mission (GPM) and Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks (PERISIANN). All SPPs were statistically assessed with ground observations before hydrological implementation, and the monthly bias correction was applied to improve the accuracy of precipitation estimates from SPPs. Then, the Soil and Water Assessment Tool (SWAT) model was used to realize daily and monthly streamflow simulation driven by precipitation data from different sources. Results showed that the monthly bias correction effectively reduced systematic errors of all SPPs, and generally promoted their abilities in depicting streamflow variations at different temporal scales. Specifically, GPM (IMERG) series were superior in precipitation accuracy in both uncorrected and corrected cases, followed by TMPA and PERSIANN series, and such advantages persisted in streamflow simulation in the uncorrected case as well. While after the bias correction, TMPA series gained the greatest improvement and achieved the best performances in hydrological simulation, especially its unadjusted version TMPA_3B42V7RT, which even exhibited comparable applicability against adjusted products. Additionally, despite systematic bias, all the adjusted product-based models outperformed the gauge-based one, indicating considerable potential of satellite products in hydrological application. Whereas, SPPs consistently showed better abilities in wet and normal years but need to be strengthened in dry years. These results will help explore better alternative precipitation sources for local hydrological application, and provide references for regions with similar hydrologic and climatic conditions around the world.