An inter-comparison of PM2.5 at urban and urban background sites: chemical characterization and source apportionment

A measurement campaign for simultaneous PM2.5 sampling was performed in two different sites in Italy: an urban site in the centre of the town of Lecce (at ground level 40°21'23 N, 18°10'02 E) and an urban background site, located in University Campus on the roof of ISAC-CNR premises at 12m above the ground, about 4 km SW of the town of Lecce, (40°20'09 N, 18°07'29 E). Measurements were performed using 4 low-volume PM2.5 samplers (two at the urban site and two at the urban background site) between 04/03/2013 and 17/07/2013, collecting and analysing in total: 104 PM2.5 48h samples on Teflon filters and 104 PM2.5 samples on quarts fibre filters. Chemical analysis were focused on the detection of: some metals (Cd, Cr, Cu, Zn, Pb, Fe, Pt), Organic Carbon (OC), Elemental Carbon (EC), levoglucosan, and mannosan on quartz fibre filters; soluble ions (Na+, NH4+, K+, Ca2+, Mg2+, Cl-, NO3-, SO42-, and C2O42-) on Teflon filters, allowing a characterization of the PM2.5 mass of about 60% in both sites. Mass closure analysis performed on datasets reveals an aged marine contribution to PM2.5, very similar for both urban and background site, with a chlorine depletion (about 79% of the total Cl) and formation of sodium nitrate. This is in agreement with previous observations on PM10 in the same area (Contini et al, 2010). Both sites are between 10 and 15 km from seaside and Cl-/Na+ ratio was about 0.4, lower than the typical value for sea water (1.81). A correlation between NO3- and Na+ is evident especially at the urban background site but it is limited at the urban site because of the presence of different sources of nitrate (i.e. NOX from traffic and/or home heating). High correlations of SO42- and NH4+ and the large abundance of nss-SO42- (about 95% of SO42- in both sites) indicate a significant fraction of secondary ammonium sulphate and bisulphate and limited or negligible ammonium nitrate at both sites. OC and EC fractions account about respectively 20% and 7% of PM2.5 mass in urban site and 21% and 4% of PM2.5 mass in urban background site. The spatial variability observed in the EC fraction, confirmed by the analysis of the coefficients of divergence (CODs), reflects the strong influence of primary emissions, (mainly road traffic or residential heating), while OC, which does not show this trend, could derive by a larger number of sources and by secondary formation processes. Urban and urban background OC/EC ratios (Fig. 1) are comparable to those reported in scientific literature for Italian traffic and rural site respectively (Sandrini et al, 2014). OC/EC minimum ratio method allowed estimating the Secondary Organic Carbon (SOC) concentration (Fig. 1), finding similar average concentrations at both sites: 9% of PM2.5 mass (46% of OC) in the urban site; 9% of PM2.5 mass (43% of OC) in the urban background site. The spatial homogeneity of SOC reflects the similar behaviour observed for secondary inorganic aerosol (SIA). A Positive Matrix Factorization (PMF) analysis was performed to investigate differences in the sources affecting the two sites. PMF found 6 factors/sources for both sites: crustal-resuspended, marine, secondary sulphate, traffic, biomass burning and nitrate. It is worth to note that, despite the similar PM2.5 concentrations, 14.5 ?g/m3 urban background and 16.0 ?g/m3 urban, PMF reconstructed different contributions for crustal-resuspended, traffic and biomass burning. These could be considered "local" sources. Similar contribution were estimated for marine, secondary sulphate and nitrate that could be considered "regional" sources. An inter-comparison of source apportionment results obtained with PMF3.0 and PMF 5.0 will be presented.