The contribution of traffic to PM2,5 levels at a Maltese traffic site

Abstract

The traffic station in Msida (35.90°N, 14.49°E, 2m AMSL), located in the Maltese archipelago (in the Central Mediterranean) forms part of the Maltese air quality network. PM2,5 levels have been monitored at this site since 2008 and have been reported to the European Commission ever since. The receptor is within 10 m of an important traffic node with flows of ~23000 AADT which links Malta's administrative capital (Valletta) to the main Commercial, Entertainment and Tourism hub (Sliema). Additionally the site is surrounded by heavily trafficked areas to both the north and south, including Malta's busiest traffic artery (~41000 AADT) at 2.5 km to the SSE. PM2,5 levels at the receptor are expected to be heavily influenced by traffic. Other contributors to PM2,5 levels are expected to be shipping (due to the presence of a busy shipping lane to the east of the receptor) but also natural sources such as Saharan dust and sea salt. The objective of this study is to isolate and quantify the traffic aerosol source contributions to the PM2,5 levels in Msida. Traffic emissions are Malta's major air quality concern and it is expected that the conclusions of this study will help the policy maker to take the required action. The sampling campaign ran from 01 January 2016 to 31 December 2016, for a total of three hundred and sixty six samples collected on 46.2 mm Teflon filters and every two days on 47 mm quartz fibre filters at a flow rate of 2,3 m3/h. The filters were conditioned, prior to weighing, for 96 h at a relative humidity range >45% and < 50% and at a temperature of 20°C±1°C. A sample of 180 Teflon filters were chosen in such a way that the seasonal weekday and weekend averages were as close as possible to those of the full population of filters. The ratio of weekday : weekend filters was kept at 2.5. The filters were analysed non-destructively for black carbon (BC) using a multi-wavelength absorbance analyse. Subsequently the same filters were analysed for 17 different elements (Al, Si, Cl, K, Ca, Ti, V, Cr, Mn, Fe, Ni, Cu, Zn, Br, Sr, Ba, Pb) by energy dispersive X-ray fluorimetry, and for 5 ions (SO42-, NO3-, NH4+, Na+ and Mg2+) by ion chromatography. Ammonium levels were determined by UV-visible spectroscopy following the conversion to indophenol blue. The maximum absorbance concentration (MAC) was determined from the gradient of the plot of absorption coefficient vs the elemental carbon concentration (determined using the thermal optical transmittance instrument) using 7 QF filters, which sampled during the same sampling period. This was then used to convert the BC readings in equivalent black carbon (eBC). The eBC values were also corrected for the systematic difference between the absorbance of the Teflon and quartz fibre filters at 635nm. This was done by comparing the absorbance of Teflon and quartz fibre filters which were exposed at the receptor on the same day, during a 9 day sampling campaign carried out in August 2017.The carbon measurements were carried out by the Physics Department of the University of Genoa (Italy), while the remaining analysis was carried out by the Atmospheric Contamination Laboratory of the Miguel Hernández University, Elche (Spain). The data with the concentration of the analytes and the respective uncertainties was used as an input into the US EPA positive matrix factorization - PMF model (version 5) in order to identify the sources contributing to PM2,5 levels in Msida. Measured PM2,5 concentrations (in ng/m3) were regressed against resulting PMF factor scores in order to calculate the different sources contributions. In all eight different sources explaining 93.3% of the measured PM2,5 mass, were isolated: Saharan dust (F1), Shipping (F2), Re-suspended dust/Mixed crustal source (F3), a contribution rich in ammonium sulphate (F4), Aged sea salt (F5), Traffic (F6), Fireworks (F7) and Fresh sea salt (F8). The contributions for each source to PM2,5 were as follows 12.8% (1.9 μg/m3) for F1, 8.6% (1.3 μg/m3) for F2, 10.5% (1.6 μg/m3) for F3, 22.0% (3.3 μg/m3) for F4, 11.0% (1.7 μg/m3) for F5, 23.0% (3.5 μg/m3) for F6, 1.8% (0.3 μg/m3) for F7 and 3.7% (0.6 μg/m3) for F8 . 6% of the mass of PM10 was not assigned to any factor. The direct contribution of traffic amounts to 23% of the PM2,5 levels however traffic emissions also affect the aged sea salt component, and the part of the mixed crustal component due to the re-suspension of dust due to traffic.