由于《大气污染防治行动计划》的实施，2014年~2017年期间中国大气环境整体得到改善，但PM_2.5仍然是中国大气的首要污染物，同时中国大气O₃污染不断加重，使得PM_2.5和O₃复合污染成为中国大气污染的总体特征。本研究利用随机森林回归模型和主成分分析的统计方法，结合气象条件、大气污染监测和污染排放三方面因素，分析了2014年~2017年中国重点区域PM_2.5和O₃复合污染的时空特征及成因。研究结果表明，PM_2.5和O₃污染在华北平原、长江三角洲和珠江三角洲地区此消彼长，而在汾渭平原地区共同消长。结合4个研究区域总体来看，短波辐射强度和比湿、电力部门总排放量、NH₃和有机碳的总排放量分别是影响PM_2.5和O₃污染最重要的气象因素、部门排放因素、污染物排放因素。

The implementation of China’s Air Pollution Prevention and Control Action Plan improves China’s atmospheric environment from 2014 to 2017. However, with ozone pollution growing prominent and PM_2.5 still being China’s primary air pollutant, the compound pollution of PM_2.5 and ozone becomes the unique feature of China's air pollution problem. Taking meteorological conditions, air pollution monitoring and pollutant emissions into consideration, this study applied random forest regression model and principal component analysis to analyzes the temporal and spatial characteristics of PM_2.5 and ozone compound pollution in 4 key areas of China from 2014 to 2017. The results show that the relationship of PM_2.5 and ozone pollution is implementary in the Fenwei Plain and complementary in the North China Plain, the Yangtze River Delta and the Pearl River Delta. In general, shortwave radiation intensity and specific humidity, the emissions of power sector, the emissions of NH₃ and organic carbon are most important respectively in meteorological factors, sector emission factors and pollutant emission factors that affect PM_2.5 and ozone pollution in 4 study areas.