Drought monitoring utility of satellite-based precipitation products across mainland China

This study mainly evaluated and compared satellite-based quantitative precipitation estimate products (QPEs) for the drought monitoring of mainland China. Two long-term (more than 30 a) satellite-based QPEs, i.e. the Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks-Climate Data Record (PERSIANN-CDR) and the Climate Hazards Group InfraRed Precipitation with Station (CHIRPS), and a short-term (18a) QPE, i.e. the Tropical Rainfall Measuring Mission (TRMM) Multi-Satellite Precipitation Analysis (TMPA) 3B42V7 are considered. Two widely used drought indices, the Standardized Precipitation Index (SPI) and the Palmer Drought Severity Index (PDSI), are chosen to evaluate the drought monitoring utility. The 3B42V7 was only evaluated with PDSI due to the short data records. The results show that all the three QPEs perform satisfactorily in the eastern part of China when using both SPI and PDSI. However, their performances for west China could not be clearly determined due to the sparse gauge networks. 3B42V7 features best performance among the three QPEs in the evaluation using PDSI. To further spatiotemporally evaluate the drought utility of the QPEs, four typical drought-affected regions, i.e. northeast China (NEC), Huang-Huai-Hai plain (3HP), southwest China (SWC), and Loess plateau (LP) were extracted from mainland China for specific case studies. Temporally, all three QPEs are able to detect the typical drought of the four regions with both SPI and PDSI, and 3B42V7 presents the least deviation in PDSI estimate. Spatially, both CHIRPS and 3B42V7 accurately catch the spatial centers and extent of the typical drought events, while PERSIANN-CDR could not match the spatial patterns of drought events well. Generally, the long-term PERSIANN-CDR and CHIRPS perform satisfactorily in drought detection and are suitable for drought utility; however, caution should be applied when studying the spatial variation of drought using PERSIANN-CDR. CHIRPS could also be suitable for near-real-time drought monitoring for its shorter time latency of data release. The short-term 3B42V7 also performs well in many cases, and has thus considerable potential for drought monitoring.