使用PCA算法与MLR模型实证分解2002—2016年重载货车污染对北京市严重空气污染天气形成的作用路径与贡献度。由方差分析结果得知DHDT、DPM、DNO、DGTKM的方差贡献度均在50%以上，暗示省际公路货物周转需求是引起重载货车数量与行驶里程增加的关键因素，由此引起大量柴油消耗并显著提高北京市NO₂与PM10、PM2.5浓度。系数估计结果显示，重载货车数量、省际货物周转量、NO₂浓度与PM10、PM2.5浓度显著相关，重载货车污染综合变量与CO浓度、柴油消耗量每提高1个标准单位会使北京市重空气污染天数增加1.489个标准单位。因此，在重载货车污染因素中，重载货车数量、省际货物周转量、柴油使用量是北京市重污染天数增加的"元凶"，NO₂、PM10、PM2.5与CO的排放生成是主要污染路径。

PCA algorithm and MLR model are used to figure out the route and degree that heavy duty trucks pollution has contributed to the severe air pollution weather in Beijing from 2002 to 2016.The variance contribution rates of DHDT,DPM,DNO and DGTKM exceed 50% by variance analysis,which indicates that the trans provincial freight turnover demand is the key factor causing the increase of the number and mileage of heavy-duty trucks,and in turn causing a large amount of diesel consumption and significantly increasing the concentration of NO₂ PM10 and PM2.5 in Beijing.The number of heavy-duty trucks,trans provincial cargo turnover and NO₂ concentration are significantly correlated with PM10 and PM2.5 concentration.Each standard unit increase in FAC,CO concentration and diesel consumption of heavy-duty trucks will increase the days of heavy air pollution in Beijing by 1.489 standard units.Therefore,among the pollution factors concerning heavy-duty trucks,the number of heavy-duty trucks,trans-provincial freight turnover and diesel consumption are the main causes for the increase of servere air pollution days in Beijing.The generation and emission of NO₂,PM10,PM2.5 and CO are the main pollution paths.

李云燕(1963-),女,博士,教授,博士生导师,研究方向为环境经济、环境管理与评价、低碳经济、循环经济等,yunyanli@126.com