Source identification of atmospheric PCBs in Philadelphia/Camden using positive matrix factorization followed by the potential source contribution function

The concentrations of gas-phase polychlorinated biphenyls (PCBs) in the atmosphere of the Camden, NJ, USA are elevated by as much as 20 times over regional background. These high PCB levels are a concern because they lead to atmospheric deposition loadings of PCBs to the tidal Delaware River that exceed the entire total maximum daily load (TMDL). Two models were applied to the atmospheric PCB concentration data from Camden in an attempt to identify the PCB source types and regions. Positive matrix factorization (PMF) was used to identify the source types. Four factors were identified which are thought to represent sources such as volatilized Aroclors and particle-phase PCBs. The potential source contribution function (PSCF) model was then used to identify the geographic source regions by examining the origination points for air parcels that result in high PCB concentrations at the Camden receptor site. The PSCF model for Sigma PCBs indicates PCB source regions throughout the Philadelphia-Camden metro area, including portions of both Pennsylvania and New Jersey. The PSCF plots for the resolved PMF factors suggest that factors 1-4 show fewer distinct source regions, indicating that their sources are diffuse and/or lie very close to the receptor site. The PSCF plots for factors 2 and 3 reveal very different source regions. Factor 2 primarily arises from the city of Philadelphia, whereas factor 3 originates in southern New Jersey and south of Philadelphia. This study demonstrates the utility of the combined PMF/PSCF approach in identifying atmospheric PCB source types and regions. (c) 2007 Elsevier Ltd. All rights reserved.