Source apportionment of urban ambient PM 2.5 in two successive measurement campaigns in Helsinki, Finland

• Published in: Atmospheric environment (1994) Vol. 37; no. 5; pp. 615 - 623
• Main Authors: Vallius, Marko, Lanki, Timo, Tiittanen, Pekka, Koistinen, Kimmo, Ruuskanen, Juhani, Pekkanen, Juha
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 2003
• Subjects: Chemical composition; PCA; Receptor modelling; Trace elements; Urban aerosol; Receptor modelling; Chemical composition; Trace elements; Urban aerosol; PCA
• ISSN: 1352-2310; 1873-2844
• DOI: 10.1016/S1352-2310(02)00925-1

Source apportionment of urban fine particle mass (PM 2.5) was done from data collected during two 6-month measurement periods in 1996–97 and 1998–99 in Helsinki, Finland, using absolute principal component analysis and multivariate linear regression. In addition to PM 2.5 elemental composition data, 24-h average concentrations of ultrafine particles (diameter <0.1 μm) and accumulation mode particles (diameter 0.1–1.0 μm), as well as absorption coefficients of PM 2.5 filters and concentrations of nitrogen oxides (NO x ) and sulphur dioxide (SO 2) were used as input data to PM 2.5 source modelling. Five similar source categories of PM 2.5 were identified separately for both measurement periods: local traffic source characterised by NO x , absorption coefficient and ultrafine particle counts; long-range transboundary air pollution characterised by S, K, Zn, Pb and accumulation mode particle counts; crustal source characterised by Si, Al, Ca, Fe and K; oil combustion characterised by V, Ni and SO 2; and salt source characterised by Na and Cl. Long-range transboundary air pollution was the major contributor to PM 2.5 during both 1996–97 and 1998–99 accounting for 51% and 50%, respectively, of the average PM 2.5. Local traffic accounted for 30% and 23%, oil combustion for 3% and 13%, crustal source for 12% and 5%, and salt for 2% and 7% of the average PM 2.5 during 1996–97 and 1998–99, respectively. Despite differences in atmospheric concentrations and availability of several elements for statistical analyses in 1996–97 and 1998–99, the estimates of PM 2.5 source contributions were both qualitatively and quantitatively comparable for the two measurement periods. Using non-elemental markers proved very useful for both source identification and estimation of source contributions at this measurement site.