Application of positive matrix factorization in characterization of PM(10) and PM(2.5) emission sources at urban roadside.

The 24-h average coarse (PM(10)) and fine (PM(2.5)) fraction of airborne particulate matter (PM) samples were collected for winter, summer and monsoon seasons during November 2008-April 2009 at an busy roadside in Chennai city, India. Results showed that the 24-h average ambient PM(10) and PM(2.5) concentrations were significantly higher in winter and monsoon seasons than in summer season. The 24-h average PM(10) concentration of weekdays was significantly higher (12-30%) than weekends of winter and monsoon seasons. On weekends, the PM(2.5) concentration was found to slightly higher (4-15%) in monsoon and summer seasons. The chemical composition of PM(10) and PM(2.5) masses showed a high concentration in winter followed by monsoon and summer seasons. The U.S.EPA-PMF (positive matrix factorization) version 3 was applied to identify the source contribution of ambient PM(10) and PM(2.5) concentrations at the study area. Results indicated that marine aerosol (40.4% in PM(10) and 21.5% in PM(2.5)) and secondary PM (22.9% in PM(10) and 42.1% in PM(2.5)) were found to be the major source contributors at the study site followed by the motor vehicles (16% in PM(10) and 6% in PM(2.5)), biomass burning (0.7% in PM(10) and 14% in PM(2.5)), tire and brake wear (4.1% in PM(10) and 5.4% in PM(2.5)), soil (3.4% in PM(10) and 4.3% in PM(2.5)) and other sources (12.7% in PM(10) and 6.8% in PM(2.5)).