为研究中国第一代全球大气再分析降水产品（CRA）在四川地区的适用性。利用四川地区155个国家气象站1979-2021年降水观测数据为参考，以欧洲中期天气预报中心再分析产品ERA5和美国国家环境预报中心再分析产品CFSR为对比评估对象，计算平均误差、均方根误差、标准差、相关系数，对三套再分析资料降水的空间分布、时间变化、误差分析等进行分析，评估CRA在四川的适用性。研究表明：（1）关于年降水空间分布特征，三套再分析资料均能反映出川西高原降水偏少、四川盆地偏多的现状，但是仅CRA能较好反映降水极值区的位置和量级；（2）关于四季降水空间分布特征，CRA四季降水分布几乎和观测一致，仅夏季降水比观测偏少，ERA5和CFSR降水分布及量级大多和观测不符，二者降水量大多比观测偏大；（3）关于降水时间变化特征，三套再分析资料均能反映出降水的季节性差异，但是仅CRA降水能较好反映月和季节降水的年际变化，CRA比EAR5和CFSR能更好反映四川降水的线性变化特征；（4）关于日降水均方根误差的空间分布，CRA的均方根误差最小，ERA5和CFSR在不同区域均方根误差大小各不同，从四季降水的相关系数、标准差、均方根误差，CRA在四季的相关系数均最大、在四季的标准差和均方根误差均最小。综合来看，在反映四川降水特征方面，CRA 优于 ERA5、ERA5优于CFSR，CRA可以代替ERA5和CFSR使用。

In order to study the applicability of China΄s first generation global atmospheric reanalysis（CRA）precipitation dataset in Sichuan region. The applicability of CRA has been evaluated with the precipitation observations from 155 national meteorological stations in Sichuan region and further compared with the ERA5 and CF‐ SR datasets in terms of spatial distribution of precipitation，temporal variations，error analysis. The mean error （ME），root mean square error（RMSE），Standard deviation（STD），Correlation coefficient（R）of the three re‐ analysis datasets are computed to compare the accuracy and evaluate the application. The comparative results suggest that：（1）In the simulation of the spatial distribution of annual precipitation，the distribution of less precipitation in the western Sichuan Plateau and more precipitation in the Sichuan Basin can be simulated by the three reanalysis datasets，but the location and magnitude of the extreme precipitation area can only be better simulated by CRA；（2）In the simulation of the spatial distribution of four seasons precipitation，CRA is almost consistent with the observations，and the summer precipitation of CRA is less than the observations，the location and magnitude of ERA5 and CFSR is mostly inconsistent with the observations，and the precipitation of ERA5 and CFSR is larger than the observations；（3）In the characteristics of temporal changes in precipitation，the seasonal variation of precipitation can be simulated by the three reanalysis datasets，but annual variability of the monthly and seasonal precipitation can be only be simulated by CRA. Compared with ERA5 and CFSR，the linear variation characteristics of Sichuan precipitation can be better reflected by CRA；（4）In the spatial distribution of daily precipitation RMSE，the RMSE of CRA is the smallest，the RMSE of ERA5 and CFSR in different regions are different. In terms of correlation coefficient，standard deviation，and root mean square error for precipitation in four seasons，CRA has the largest correlation coefficient in all seasons，and the smallest standard deviation，root mean square error. On the whole，CRA is better than ERA5 in the simulation of precipitation in Sichuan region， ERA5 is better than CFSR. And CRA can be used to replace ERA5 and CFSR.