Source apportionment of PM2.5 for 25 Chinese provincial capitals and municipalities using a source-oriented Community Multiscale Air Quality model

Source contributions to fine airborne particulate matter with aerodynamic diameters < 2.5 mu m(PM2.5) during 2013 were determined for 25 Chinese provincial capitals and municipalities using a source-oriented version of the Community Multiscale Air Quality (CMAQ) model. Based on the hierarchical clustering analysis of the observed PM2.5 concentrations, the 25 cities were categorized into nine groups. Generally, annual PM2.5 concentrations were highest in the cities in the north (81-154 mu g m(-3)) and lowest in the cities close to seas in the south and east (27-57 mu g m(-3)). Seasonal PM2.5 observations in the cities were generally higher in winter than in the other seasons. Industrial or residential sources were predicted to be the largest contributor to PM2.5 for all the city groups, with annually fractional contributions of 25.0%-38.6% and 9.6%-27%, respectively. The annual contributions from power plants, agriculture NH3, windblown dust, and secondary organic aerosol (SOA) for the city groups were 8.7%-12.7%, 9.5%-12%, 6.1%-12.5%, and 5.4%-15.5%, respectively. Meanwhile, the annual contributions from transportation, sea salt, and open burning were relatively low (< 8%, < 2%, and < 6%, respectively). Secondary PM2.5 accounted for 47%-63% of total annual PM2.5 concentrations in the cities and contributed to as much as 70% of daily PM2.5 concentrations on PM2.5 pollution days ( daily concentrations > 75 mu g m(-3)). Industrial or residential sources were generally the largest contributor on PM2.5 pollution days for all the city groups in each season, except that open burning, SOA, and windblown dust could be more important on some days, particularly in spring. The results of this study would be helpful to develop measures to reduce annual PM2.5 concentrations and the number of PM2.5 pollution days for different regions of China. (C) 2017 Elsevier B.V. All rights reserved.