以中国气象局提供的地面降雨观测资料为参考,利用相关系数、均方根误差和相对偏差,降雨误报率、命中率和关键成功指数等6种不同的统计分析指标,从年、季、月、日4种不同的时间尺度对两种高分辨率卫星降雨产品（IMERG和GSMaP）在陕西省的精度进行了对比和评价,并对二者在监测强降雨过程中的表现进行了对比分析。结果表明：（1） 在年尺度上,GSMaP的数据精度高于IMERG。GSMaP和站点观测数据高度相关,而IMERG和站点观测数据中度相关;GSMaP高估了年尺度的降雨,而IMERG低估了年尺度的降雨。（2） 在季节尺度上,两种数据均在夏季的精度最高,总体上,IMERG在季节尺度的精度亦高于GSMaP。（3） 在月尺度上,两种产品与地面观测数据均呈现较高的相关性,而且均存在一定程度的高估,但IMERG比GSMaP更具有相对较高的精度。（4） 在日尺度上,GSMaP的数据精度略高于IMERG。（5） 卫星降雨产品的数据精度与降雨量有关,总体表现为雨量小时高估、雨量大时低估。（6） 卫星降雨产品数据精度呈现了明显的地域差异,GSMaP对陕西省的降雨总体表现为低估,其中对陕北的低估最为明显;IMERG对关中有轻微高估,对陕北和陕南则存在明显低估。（7） 通过对4场次强降雨事件的分析发现,GSMaP对大雨及以上强降雨事件的监测能力比IMERG略强。研究结果可以为该地区的气象水文研究在选择和使用降雨数据资料时提供参考。

The Global Precipitation Measurement (GPM) mission is an international network of satellites that provide next-generation global observations of rain and snow. GPM was initiated by NASA and JAXA as a global successor to the Tropical Rainfall Measurement Mission (TRMM). In the GPM era, IMERG and GSMaP are two main satellite-based precipitation products. Compared with TRMM, GPM has wider coverage and higher spatiotemporal resolution. Understanding the accuracy and reliability of these products is important to further promote the application of new high-resolution satellite precipitation products. Therefore, this study intends to evaluate and compare the accuracy of IMERG and GSMaP in Shaanxi Province, China. In this paper, with surface precipitation observation datasets acquired from the China Meteorological Data Network used as reference, six evaluation indices, namely, correlation coefficient, root mean squared error, relative bias, false alarm ratio, probability of detection, and critical success index, were employed to evaluate and compare their accuracy at different temporal scales. Their performance in monitoring heavy rainfall events was also analyzed. Results show the following: (1) At an annual scale, the accuracy of GSMaP is higher than that of IMERG. GSMaP is highly correlated with the ground observation, while IMERG is moderately correlated with the ground observation. GSMaP overestimates the precipitation at an annual scale, whereas IMERG underestimates the precipitation. (2) At the seasonal scale, both data have the highest accuracy during summer. Overall, IMERG has a higher data accuracy than GSMaP. (3) At a monthly scale, the two products show a high correlation with the ground observation, but a certain degree of overestimation is observed. IMERG has a relatively higher accuracy than GSMaP. (4) At a daily scale, the accuracy of GSMaP is higher than that of IMERG. (5) The accuracy of satellite precipitation products is related to total precipitation volume, which is generally overestimated when the precipitation volume is small and underestimated when the volume is large. (6) The accuracy of satellite precipitation product data shows obvious regional differences. GSMaP underestimates the precipitation in Shaanxi Province as a whole, especially in northern Shaanxi. IMERG slightly overestimates the precipitation in Guanzhong, but significantly underestimates the precipitation in northern and southern Shaanxi. (7) An analysis of four heavy precipitation events shows that the monitoring capability for heavy precipitation events of GSMap is slightly stronger than that of IMERG. This study is expected to provide a reference for the selection and application of precipitation data in meteorological and hydrological studies in this area.