Improving source identification of Atlanta aerosol using temperature resolved carbon fractions in positive matrix factorization

Daily integrated PM2.5 (particulate matter less than or equal to 2.5 mum in aerodynamic diameter) composition data including eight individual carbon fractions collected at the Jefferson Street monitoring site in Atlanta were analyzed with positive matrix factorization (PMF). Particulate carbon was analyzed using the thermal optical reflectance method that divides carbon into four organic carbon (OC), pyrolized organic carbon (OP), and three elemental carbon (EC) fractions. A total of 529 samples and 28 variables were measured between August 1998 and August 2000. PMF identified 11 sources in this study: sulfate-rich secondary aerosol I (50%), on-road diesel emissions (11%), nitrate-rich secondary aerosol (9%), wood smoke (7%), gasoline vehicle (6%), sulfate-rich secondary aerosol II (6%), metal processing (3%), airborne soil (3%), railroad traffic (3%), cement kiln/carbon-rich (2%), and bus maintenance facility/highway traffic (2%). Differences from previous studies using only the traditional OC and EC data (J. Air Waste Manag. Assoc. 53(2003a)731; Atmos Environ. (2003b)) include four traffic-related combustion sources (gasoline vehicle, on-road diesel, railroad, and bus maintenance facility) containing carbon fractions whose abundances were different between the various sources. This study indicates that the temperature resolved fractional carbon data can be utilized to enhance source apportionment study, especially with respect to the separation of diesel emissions from gasoline vehicle sources. Conditional probability functions using surface wind data and identified source contributions aid the identifications of local point sources. (C) 2004 Elsevier Ltd. All rights reserved.