Physics-based aerosol source apportionment at the site of Lecce (Italy)

Positive Matrix Factorization was applied to approximately two years (2016–2017) of aerosol physical property measurements conducted in Lecce (southern Italy) to identify emission sources. We used high-temporal resolution data of particle number size distributions (PNSD) spanning 0.01–10 μm diameter, measured using a Scanning Mobility Particle Sizer (SMPS) and an Optical Particle Counter (OPC), single-wavelength absorption measurements (MAAP, converted to equivalent black carbon, eBC) and NO and NO 2 concentrations. Seven sources were identified: nucleation, fresh traffic, aged traffic, a mixed (secondary + biomass burning) factor, biomass burning, sea spray & transported dust, and local & transported dust. Traffic-related sources dominated particle number concentration (PNC), with aged traffic alone accounting for 39% of the total, followed by fresh traffic (26%) and nucleation (15%). The mixed factor, comprising secondary aerosol and biomass burning contributions that could not be clearly resolved, was the primary volume driver (27%), while the sum of the two coarser factors represented 35% of total volume. Comparison with chemistry-based PMF (elemental and ionic composition of PM 10 and PM 2.5 filter samples) was limited to 31 days, restricting seasonal representativeness. Correlations with chemical species provided supportive though not conclusive evidence for the identification of traffic, biomass burning, and sea spray factors. The single eBC input limited the ability to disentangle biomass burning from secondary aerosol and sea spray from Saharan dust. Future studies will incorporate multispectral absorption data and a longer dataset to improve source separation. Despite these limitations, high temporal resolution enabled detection of nucleation episodes and short-term dynamics not captured by offline methods.