Chengdu, the capital city of Sichuan Province, is one of the most polluted cities in China. We used single-particle aerosol mass spectrometer to monitor particulate matter pollution in an urban area of Chengdu from December 9, 2015 to January 4, 2016 to determine the characteristics of air pollution during the winter months. The mass concentrations of particulate matter were high during the whole observation period, with mean values for PM2.5 and PM10 of 101 ± 60 and 162 ± 99 μg m-3, respectively. The particles were clustered into nine distinct particle types: dust (3%), potassium-elemental carbon (KEC) (24%), organic carbon (OC) (12%), combined OC and EC (OCEC) (6%), K-organic nitrogen (KCN) (10%), K-nitrate (KNO3) (12%), K-sulfate (KSO4) (18%), K-sulfate and nitrate (KSN) (12%), and metal (3%) particles. Analysis on different types of day showed that: (1) from "excellent" (days with PM2.5 lower than 35 μg m-3) to "light pollution" (PM2.5 between 75 and 115 μg m-3) days, local/regional combustion was the major contributor, whereas the aggravation of pollution from light pollution to "heavy pollution" (PM2.5 higher than 150 μg m-3) days was mainly determined by the combined effect of local/regional combustion and long-distance transport; (2) as the air quality deteriorated, the mixing of sulfate and nitrate in particles increased sharply, especially sulfate; and (3) the relative aerosols acidity increased from excellent to light pollution days, while decreased significantly from light pollution to heavy pollution days. Backward trajectory analysis showed that there were significant differences in PM2.5 concentrations and particle compositions between clusters of trajectories, which affected the level and evolution of PM2.5 pollution in Chengdu. These results give a deeper understanding of PM2.5 pollution in Chengdu and the Sichuan Basin.
Keywords: Chengdu; Mixing state; Pollution evolution; SPAMS; Single particles.