明确误差特性(包括系统和随机成分)是改进卫星降水反演算法及发展误差校正技术的基础。以中国自动站融合降水数据集作为参考,在小时尺度上对风云二号卫星反演降水产品(FY-2G与FY-2H)与国际主流的GPM降水产品(IMERG-Final与GSMaP-Gauge)进行系统及随机误差的分解。从空间、时间、雨强、高程四个角度诠释4套降水产品的系统及随机误差的表现。结果表明：(1)FY-2G和FY-2H在中国大陆的系统误差与IMERG-Final与GSMaP-Gauge非常接近,其中以IMERG-Final的系统误差最小。FY-2G和FY-2H仅有红外数据参与降水反演,而IMERG-Final和GSMaP-Gauge的降水反演中不仅有红外数据参与,更包含了精度更高的微波数据。然而FY-2G和FY-2H的系统误差达到了GPM降水产品的水准,这主要是由于FY-2H与FY-2G融合了相对更密集的地面站点数据。(2)4卫星套降水产品中,GSMaP-Gauge在中国大陆的随机误差最低。FY-2G与FY-2H与GPM降水产品的差距主要体现在随机误差上,尤其是IMERG-Final与GSMaP-Gauge在东部季风区和北方干旱区夏季的随机误差远低于FY-2G和FY-2H。(3)IMERG-Final在不同高程分布上的系统误差明显小于其他三种降水数据,而GSMaP-Gauge、FY-2G、FY-2H在不同高程分布上的系统误差则差距不大。4套降水产品在不同高程分布上的随机误差差异较大,其中GSMaP-Gauge的随机误差都保持最低。IMERG-Final则仅在海拔低于3000米时随机误差低于FY-2H与FY-2G,归结于IMERG-Final在青藏高原存在较大的不确定性。总体上,IMERG-Final与GSMaP-Gauge在中国大陆的表现优于FY-2G和FY-2H,风云二号卫星反演降水产品与国际主流的GPM降水产品相比仍存在一定差距。

【Objective】：Characterizing error components (containing systematic and random components) is essential to improve precipitation retrieval algorithms and to develop bias adjustment techniques.【Method】：Benchmarked by the Chinese merged precipitation analysis (CMPA) dataset derived based on the automatic weather stations, errors of satellite precipitation estimates from Fengyun-2-based (FY-2F and FY-2G) and the mainstream GPM-based (IMERG-Final and GSMaP-Gauge) were decomposed into systematic and random components at hourly scale. Comprehensive performances of systematic and random errors of these four products are illustrated from the perspectives of spatial distribution, temporal pattern, rainfall intensity distribution and elevation distribution【Results】：(1) The systematic errors of FY-2G and FY-2H over Chinese mainland are very close to those of IMERG-final and GSMaP-Gauge, among which IMERG-Final has the lowest systematic errors. FY-2G and FY-2H only adopt infrared observations as the data sources in their precipitation retrieval algorithms, while IMERG-Final and GSMaP-Gauge not only adopt infrared data, but also include microwave data with higher accuracy as the data sources. However, the systematic errors of FY-2G and FY-2H reach the level of GPM precipitation products, this is attributed to the relatively dense gauge networks in the satellite-gauge merging procedure of FY-2G and FY-2H. (2) GSMaP-Gauge has the lowest random errors over the Chinese mainland among these four satellite precipitation estimates. The gap between FY-2G and FY-2H with GPM precipitation products is mainly manifested in the random errors. Particularly, the random errors of IMERG-final and GSMaP-Gauge over the eastern monsoon region and the northern arid region are much lower than those of FY-2G and FY-2H in summer. (3) IMERG-final has much lower systematic errors at different elevations than the other three precipitation products, while GSMAP-Gauge, FY-2G, and FY-2H have quite close systematic errors at different elevations. The random errors of these four precipitation products vary greatly at different elevations, and GSMaP-Gauge always has the lowest random errors at different elevations. Besides, IMERG-final has lower systematic errors than FY-2H and FY-2G only below the altitude of 3000 m, which is attributed to the large uncertainty of IMERG-final over the Tibetan Plateau. 【Conclusion】IMERG-Final and GSMaP-Gauge perform better than FY-2G and FY-2H over the Chinese mainland, and FY-2-based precipitation estimates still have much room to improve when compared with the mainstream GPM-based precipitation estimates.