Investigating Twelve Mainstream Global Precipitation Datasets: Which One Performs Better on the Tibetan Plateau?
Yi Lyu,Bin Yong,Fan Huang,Weiqing Qi,Fuqiang Tian,Guoqing Wang,Jianyun Zhang
DOI: https://doi.org/10.1016/j.jhydrol.2024.130947
IF: 6.4
2024-01-01
Journal of Hydrology
Abstract:Acquiring precipitation information is a critical yet challenging task for the Tibetan Plateau (TP) with sparse gauges and high altitudes. Presently, there is a multitude of operational precipitation products merged from different data sources. A major concern within the TP science community is which one of these available precipitation datasets performs best on the TP. To address this concern, we selected and compared twelve widely used global precipitation datasets, including six satellite-based precipitation estimates (GSMaP-Gauge, IMERG-Final, TMPA-v7, CMORPH-BLD, PERSIANN-CDR, CHIRPS), five reanalysis data (ERA5, JRA-55, MERRA2, GLDAS, NCEP2), and one multi-source fusion product (MSWEP). Benchmarked against both automatic weather stations and independent field observations, we comprehensively investigate the spatiotemporal accuracy and regional applicability of these global precipitation products on the TP. The evaluation results indicate that almost all the datasets have a consistent tendency to overestimate lower precipitation rates and underestimate higher ones. However, four out of the twelve studied datasets (i.e., GSMaP-Gauge, IMERG-Final, ERA5, and MSWEP) significantly outperform others. Specifically, the two GPM-based satellite precipitation products, GSMaP-Gauge and IMERG-Final, can better reflect the difference between different rain intensities and perform better at moderate rain rates. In contrast, the reanalysis-based ERA5 exhibits a more clustered data series, with relatively higher accuracy at low precipitation rates. Moreover, ERA5 demonstrates a stable superiority in the mountainous areas characterized by complex local terrains. Furthermore, we find that the applicability of different precipitation datasets has an obvious region-related feature. Thus, a regional distribution map was made to clearly outline the optimal choice of current mainstream global precipitation datasets over different climate zones within the TP. This study can offer valuable insights for TP researchers, especially aiding them to select the appropriate precipitation data for their specific studies.