[Source apportionment of polycyclic aromatic hydrocarbons using two mathematical models for runoff of the Shanghai elevated inner highway, China].

Runoff samples of the Shanghai elevated inner highway were collected and analyzed during the rainy season of 2012 to investigate the pollution characteristics of 16 USEPA priority pollutant polycyclic aromatic hydrocarbons (PAHs). Total PAHs concentrations ranged from 1.585 to 7.523 microg.L-1. The sources of PAHs in the runoff were determined by using 2 source apportionment models, positive matrix factorization (PMF) and principal component analysis with multivariate linear regression (PCA/ MLR). Source profiles and contributions obtained from the two models were comparable. For PMF, PAHs mainly came from vehicle emissions (37. 7% ), followed by petroleum products (21. 9% ) , combustive gas (26. 4%), and other sources (14.0% ). As for PCA/MLR, it was 44. 3% , 28.9% , 18. 3% , 8.5% , respectively. The relation between the modeled total PAH concentrations and the measured concentrations was strong, with the correlation coefficient of 0. 961 for PMF and 0. 997 for PCA/MLR.