Assessment of four latest long-term satellite-based precipitation products in capturing the extreme precipitation and streamflow across a humid region of southern China
Qiang Wang,Jun Xia,Dunxian She,Xiang Zhang,Jie Liu,Yin Zhang
DOI: https://doi.org/10.1016/j.atmosres.2021.105554
IF: 5.965
2021-08-01
Atmospheric Research
Abstract:<p>The lack of ground-based observation data usually limits the analysis of extreme precipitation and streamflow, especially for the ungauged or data-limited regions. Satellite-based precipitation products (SPPs) with high temporal and spatial resolutions can provide potential data sources in the regions without adequate precipitation data, and it has received much attention of regional or global hydro-climatic analysis. This study aims to evaluate the accuracy of four widely used long time series SPPs (CHIRPS, TMPA, CMORPH and PERSIANN) in capturing extreme precipitation and streamflow events in a humid region of southern China. It is found that, for extreme precipitation detection, both CMORPH and TMPA can capture most heavy and extreme precipitation, CMORPH works better on the rainfall distribution characteristics, while TMPA shows the best performance in capturing the extreme precipitation events with relatively low RMSE, ME and BIAS, and high R values of extreme precipitation indices. For the streamflow simulation, the combination of different inputs and different models present different model performances. TMPA provides the most accurate hydrological model simulation results, while simulated streamflow forced by CMORPH exhibits considerable underestimation of streamflow. However, the combination of CMORPH inputs and VIC model obtains the best efficiency for detecting extreme streamflow, followed by the combination of TMPA and SWAT model, which indicates the selection of a suitable model and input data is essential to obtain reliable and accurate results for the detection of extreme streamflow. This study is expected to provide a valuable reference for the application and comparison of multiple SPPs and models at watershed scales.</p>
meteorology & atmospheric sciences