使用1961—2019年全国2 510个气象站点日值降水数据和2030—2100年第六次耦合模式比较计划的12个耦合模式的SSP2-4.5未来情景的降水模拟数据，基于核密度函数分别计算4个重现期（五年一遇、十年一遇、二十年一遇和五十年一遇）历史和未来情景中3个年暴雨要素（年暴雨日数、年暴雨量和年暴雨强度）的数值，在此基础上评估了中国未来暴雨危险性的变化。得到的主要结论如下：\mathrm{①}从全国来看，在未来年暴雨日数和雨量呈增加趋势，而年暴雨强度在不同重现期下表现不一样。4个重现期下全国的年暴雨日数变化均值分别为0.36、0.57、0.73和0.92 d；年暴雨量变化均值分别为22.30、36.24、46.92和60.12 mm；年暴雨强度变化均值分别为2.43、0.27、-1.95和 -4.86 mm/d。\mathrm{②}从不同气候区来看，年暴雨量和雨日在青藏高原、东部干旱区、东北地区、华北地区和西南地区呈增加趋势，在西部干旱（半干旱）区、华中地区和华南地区呈减少趋势。4个重现期下年暴雨强度在青藏高原均呈增加趋势，而其他地区以减少为主要趋势。\mathrm{③}热点分析结果显示青藏高原东南部和南部是暴雨危险性增加最显著的区域；年暴雨日数和雨量减少的区域集中分布在西南地区的中部、华中和华南地区的中部和南部，年暴雨强度减少的区域集中在东部干旱区的中部、西南地区西部及华北地区北部和南部。

Using the daily precipitation data of 2 510 meteorological stations in China from 1961 to 2019 and the precipitation simulation data of 12 coupling models of CMIP6 from 2030 to 2100 under the SSP2-4.5 future scenario， the values of three annual rainstorm elements （annual rainstorm days， annual rainstorm rainfall， and annual rainstorm intensity） in the historical and future scenarios were calculated based on the kernel density function under four return periods （5，10，20，50 years）. Based on this， the future hazard change of rainstorms in China was assessed. The main conclusions are as follows： \mathrm{①} From a national perspective， the number of rainstorm days and rainfall amount are estimated to increase in the future， and the annual rainstorm intensity is estimated to change differently in different return periods. Under the four return periods， the mean change of annual rainstorm days in China is expected to be 0.36， 0.57， 0.73 and 0.92 days； the mean change of annual rainstorm rainfall is expected to be 22.30， 36.24， 46.92 and 60.12 mm； and the mean change of annual rainstorm intensity is expected to be 2.43， 0.27， -1.95 and -4.86 mm/d. \mathrm{②} From the perspective of different climatic zones， the annual rainstorm rainfall and rainstorm days showed an increasing trend in the Qinghai-Tibet Plateau， Eastern arid zone， Northeast China， North China， and Southwest China， and a decreasing trend in the western arid （semi-arid） zone， Central China， and South China. The annual rainstorm intensity is expected to increase in the Qinghai-Tibet Plateau and decrease in most other regions under the four return periods. \mathrm{③} The hot spot analysis shows that the southeastern and southern Qinghai-Tibet Plateau are expected to be the areas with the most significant increases in rainstorm hazards. The areas with reduced annual rainstorm days and annual rainstorm rainfall are expected to be concentrated in the middle of Southwest China， the middle and south of Central China and South China. Finally， the areas with reduced annual rainstorm intensities are expected to be concentrated in the middle of the eastern arid zone， the west of Southwest China， and the north and south of North China.