Chemical composition and sources of PM1 and PM2.5 in Beijing in autumn

Abstract Beijing, the capital of China, suffers from severe atmospheric aerosol pollution; nevertheless, a comprehensive study of the constituents and sources of PM 1 is still lacking, and the differences between PM 1 and PM 2.5 are still unclear. In this study, an intensive observation was conducted to reveal the pollution characteristics of PM 1 and PM 2.5 in Beijing in autumn. Positive matrix factorization (PMF), backward trajectories and a potential source contribution function (PSCF) model were used to identify the source categories and source areas of PM 1 and PM 2.5 . The results showed that the average concentrations of PM 1 and PM 2.5 reached 78.20 μg/m 3 and 95.47 μg/m 3 during the study period, respectively. PM 1 contributed greatly to PM 2.5 . The PM 1 /PM 2.5 value increased from 73.6% to 90.1% with PM 1 concentration growing from 3 to >150 μg/m 3 . Higher secondary inorganic aerosol (SIA) proportions (31.3%–70.8%) were found in PM 1 . The higher fraction of SIA, OC, EC and typical elements in PM 1 illustrated that anthropogenic components accumulated more in smaller size particles. Three typical weather patterns causing the heavy pollution in autumn were found as follows: (1) Siberian high and uniform high pressure field, (2) cold front and low-voltage system, and (3) uniform low pressure field. A PMF analysis indicated that secondary aerosols and coal combustion, vehicle, industry, biomass burning, and dust were the important sources of PM, accounting for 53.8%, 8.0%, 13.0%, 13.2% and 12.0% of PM 1 , respectively, and for 47.5%, 9.9%, 12.4%, 8.4% and 21.8% of PM 2.5 , respectively. The HYSPLIT and chemical components analysis indicated the potential contribution from biomass burning and fertilization ammonia emissions to PM 1 in autumn. The source areas were similar for PM 1 and PM 1–2.5 under general polluted conditions, but during the heavily polluted periods, the source areas were distributed in farther regions from Beijing for PM 1 than for PM 1–2.5 .