Source and risk apportionment of selected VOCs and PM2.5 species using partially constrained receptor models with multiple time resolution data

This study was conducted to identify and quantify the sources of selected volatile organic compounds (VOCs) and fine particulate matter (PM2.5) by using a partially constrained source apportionment model suitable for multiple time resolution data. Hourly VOC, 12-h and 24-h PM2.5 speciation data were collected during three seasons in 2013. Eight factors were retrieved from the Positive Matrix Factorization solutions and adding source profile constraints enhanced the interpretability of source profiles. Results showed that the evaporative emission factor was the largest contributor (25%) to VOC mass concentration, while the largest contributor to PM2.5 mass concentration was soil dust/regional transport related factor (26%). In terms of risk prioritization, traffic/industry related factor was the major cause for benzene, ethylbenzene, Cr, and polycyclic aromatic hydrocarbons (29–69%) while petrochemical related factor contributed most to the Ni risk (36%). This indicated that a larger contributor to mass concentration may not correspond to a higher risk. •We applied a partially constrained receptor model to multiple time resolution data.•Hourly VOC, 12-h and 24-h PM2.5 speciation data were combined in the model.•Adding constraints to the model enhanced the interpretability of source profiles.•We applied a risk apportionment approach to obtain the source-specific risk values.•A larger contributor to mass concentration may not correspond to a higher risk. Combining a constrained receptor model and a risk apportionment approach could provide valuable information to design effective control strategies.