Daily PM10 samples were collected during a one-month sampling campaign from February 13 to March 12, 2008 at eight different sampling sites in Augsburg, Southern Germany. Source apportionment was performed to identify the main sources and related contributions by analysis of organic and inorganic tracers. Nine factors were separated comprising: solid fuel combustion, traffic-related emissions, secondary inorganics, and mixed sources. Spatiotemporal variation of the source contributions was evaluated using the Pearson correlation coefficient (r) and coefficient of divergence (COD). All factors (except hopanes and mixed sources) showed moderate to high (0.6<r>0.8) correlation coefficients between the eight sites and were distributed heterogeneously. Secondary sulfate and secondary nitrate factors were relatively more uniformly distributed (compared to other factors) with lower medians of COD value (0.47 and 0.56, respectively) and higher correlation values (r=0.97 and 0.85, respectively). The maximum daily average contribution for coal & wood combustion factor was observed at the LfU suburban site (4.0 μg m(-3)); wood combustion factor at the LSW residential site (5.1 μg m(-3)) ; diesel & fuel oil consumption factor at the Bifa suburban and BP urban sites (both 2.5 μg m(-3)); road dust & tram factor at the KP traffic site (16.2 μg m(-3)) and the BP urban site (6.6 μg m(-3)); hopanes factor at the BP urban and Bifa suburban sites (both 0.7 μg m(-3)); and de-icing NaCl factor at the KP traffic site (4.8 μg m(-3)). Secondary sulfate and secondary nitrate factors had approximately similar contributions (6.2 μg m(-3) and 4.3 μg m(-3), respectively) at all sites. Mixed sources factor had the highest daily average contribution to PM10 mass at the KP traffic site (7.0 μg m(-3)).
Keywords: PM(10); Particle composition; Particulate organic compounds (POC); Positive matrix factorization (PMF); Spatial variability.
Copyright © 2013 Elsevier Ltd. All rights reserved.