Comprehensive evaluation of latest GPM era IMERG and GSMaP precipitation products over mainland China

Following the launch of the Global Precipitation Measurement (GPM) Core Observatory, the latest GPM era high-resolution satellite precipitation products (SPPs), namely Integrated Multi-satellite Retrievals for GPM (IMERG) and Global Satellite Mapping of Precipitation (GSMaP), have been released. A critical evaluation of these newly released SPPs is very important for both end users and data developers. This study aims to evaluate the performance of the IMERG and GSMaP products over mainland China from 1 January 2015 to 31 December 2017, using daily precipitation data from 778 meteorological stations. The SPPs are validated adopting three widely used statistical metrics (CC, RMSE and RB). The assessment also considers the precipitation detection capability (POD, FAR and CSI), probability density function (PDF) of precipitation intensities, and elevation factor. Results show that GSMaP_G performs better in daily, monthly and seasonal scales than other SPPs. In terms of CC and RMSE, all SPPs share similar spatial patterns, with high values in southeastern China and low values in northwestern China. In addition, all SPPs underestimate the precipitation amount in southern China, but overestimate that in northern China, especially in northwestern China. Obviously, the precipitation detection ability of all products is good in southern China, except the Sichuan Basin. All SPPs can better reproduce the PDF in terms of precipitation intensity. Compared with other SPPs, the performance of GSMaP_G is better at all elevations. However, significant elevation-dependent errors exist in the GSMaP_N. Overall, the gauge-corrected GSMaP_G and IMERG_F perform better than other SPPs in reducing various errors, especially the former. This study will be valuable for both algorithm developers and users, as well as for associated products, and it will also better support disaster risk reduction and hydrological applications, especially in areas with a sparse rain gauge network.