Time-dependent source apportionment of submicron organic aerosol for a rural site in an alpine valley using a rolling positive matrix factorisation (PMF) window

Abstract. We collected 1 year of aerosol chemical speciation monitor (ACSM) data in Magadino, a village located the south Swiss Alpine region, one Switzerland's most polluted areas. analysed mass spectra organic (OA) by positive matrix factorisation (PMF) using Source Finder Professional (SoFi Pro) to retrieve origins OA. Therein, we deployed rolling algorithm, which is closer measurement, account for temporal changes source profiles. As first-ever application PMF with multilinear engine (ME-2) analysis on yearlong dataset that was from rural site, resolved two primary OA factors (traffic-related hydrocarbon-like (HOA) and biomass burning (BBOA)), mass-to-charge ratio (m/z) 58-related (58-OA) factor, less oxidised oxygenated (LO-OOA) more (MO-OOA) factor. HOA showed stable contributions total through whole ranging 8.1 % 10.1 %, while contribution BBOA an apparent seasonal variation range 8.3 %–27.4 (highest during winter, lowest summer) yearly average 17.1 %. OOA (sum LO-OOA MO-OOA) contributed 71.6 mass, varying 62.5 (in winter) 78 spring summer). The 58-OA factor mainly contained nitrogen-related variables appeared be pronounced only after filament switched. However, since this insignificant (2.1 %), did not attempt interpolate its potential work. uncertainties (?) modelled (i.e. rotational uncertainty statistical variability sources) varied ±4 maximum ±40 (LO-OOA). Considering (showing influences had significant suggest reducing controlling biomass-burning-related residential heating as mitigation strategy better air quality lower PM levels region or similar locations. In Appendix A, conduct head-to-head comparison between conventional mechanism. find slightly improved results terms concentrations, correlations external tracers, profiles constrained POA factors. show smaller scaled residuals enhanced corresponding inorganic salts compared those solutions, likely because can capture variations oxidation processes components. Specifically, time-dependent MO-OOA well explain viabilities main ions factors, m/z 44 (CO2+) 43 (mostly C2H3O+). Therefore, provides realistic apportionment (SA) solution sources. also good agreement offline Aerodyne spectrometer (AMS) SA filter samples, except winter. latter discrepancy online measurement fast sources, contrast 24 h samples. This study demonstrates strengths mechanism, comprehensive criterion list ACSM users obtain reproducible results, role model analyses such worldwide available data.