准确模拟和预估我国华北地区极端降水对其防灾减灾工作非常重要。本研究利用一套基于CMIP6的全球逐日降尺度预估数据集（NEX-GDDP-CMIP6）与观测数据中国地面气候资料日值数据集（V3.0）相比较，基于空间相关系数和标准差分析，针对降水日数和降水强度两种降水指标，结合三种降水量级（中小雨、大雨、暴雨）进行资料的历史模拟能力评估和未来预估。结果表明：在历史模拟能力方面，综合各降水量级，NEX-GDDP-CMIP6对7-8月平均降水日数模拟的空间相关系数较高，同观测东南多、西北少的空间分布特征一致,但量级上在中小雨段明显高估而在大雨及以上量级明显低估;对7-8月平均降水强度模拟的标准差较低,对量级的模拟较好,但对空间分布模拟能力较差。在此基础上优选十个模式进行预估，发现未来近期（2021-2040年）和中期（2041-2060年）于三个排放情景下（SSP1-2.6，SSP2-4.5和SSP3-7.0）研究区7-8月极端降水日数和强度在空间分布上有向西向北扩展的趋势，在数值上大雨量级呈现出在未来近期有所减少而在未来中期又明显增加的特征。本研究有助于充分认识NEX-GDDP-CMIP6极端降水模拟效果，预估结果可为未来华北地区极端降水状况提供有效参考。

Precise simulation of extreme precipitation events are vital for preventing, alleviating, and relieving disasters. Using models selected via credibility of NEX-GDDP-CMIP6 in representing precipitation frequency and intensity occurring during July and August of 1981-2014, this study assesses the historical simulation and future change of extreme precipitation over north China. As for credibility of historical simulation, the relative spatial distribution of precipitation frequency is in correspondence with observation, wetter in the southeast and dry in the northwest. Models significantly overestimate the frequency of light or moderate rainy days. The simulation of precipitation intensity is relatively accurate numerically, but not spatial distribution. As for future projection, on this basis, ten models were selected for prediction, and it was found that the number and intensity of extreme precipitation days in July and August in the study area would expand westward to northbound under three emission scenarios (SSP1-2.6, SSP2-4.5 and SSP3-7.0) in the near future (2021-2040) and the middle period (2041-2060). In terms of value, the magnitude of heavy rain shows the characteristics of a decrease in the near future and a significant increase in the middle future..This study reveals the effect of NEX-GDDP-CMIP6 when performing extreme precipitation over North China, and projection results provide effective reference for future.