Designated as the most harmful for health, PM2.5 aerosol fraction was a subject of our study. It was collected for all four seasons during 2014/15 in the suburban area of Belgrade (Serbia) and analysed for Al, Si, P, S, Cl, K, Ca, Ti, V, Cr, Mn, Fe, Ni, Cu, Zn, Br, As, Ba and Pb elements and for NH4+, NO3- and SO42- ions by particle-induced X-ray emission and ion chromatography techniques, respectively. Obtained concentrations have been treated by a combination of several receptor-oriented models to reveal source contributions to the suburban PM2.5 at different spatial scales. Applied positive matrix factorization analysis indicated five main groups of emission sources: biomass burning (14.5%), traffic (3.9%), regional combustion/secondary sulphates (28.8%), local combustion/secondary nitrates (29.7%) and soil (5.4%). Local heating units had been pointed out as dominant contributors by long-range transport and ground-wind circulation analyses. Air masses circulating over the Balkan Peninsula denoted regional emissions as responsible for the high concentrations of secondary sulphates. Local and long-range transport analyses combined suggested that the BB and the LC/NO3 originated from the wider urban area. Several Saharan dust episodes were detected as well. Presented results might be a basis for the development of air pollution mitigation strategies in the continental Balkan area, considered one of the most polluted and under-investigated European regions.
Keywords: Conditional bivariate probability function; Long-range atmospheric transport; PM2.5; Positive matrix factorization; Source apportionment.