Investigating the plausibility of a PMF source apportionment solution derived using a small dataset: A case study from a receptor in a rural site in Apulia - South East Italy.

Results of a methodological study on the use of Positive Matrix Factorization (PMF) with smaller datasets are being reported in this work. This study is based on 29 PM10 and 33 PM2.5 samples from a receptor in a rural setup in Apulia (Southern Italy). Running PMF on the two size fractions separately resulted in the model not functioning correctly. We therefore, augmented the size of the dataset by aggregating the PM10 and PM2.5 data. The 5-factor solution obtained for the aggregated data was fairly rotationally stable, and was further refined by the rotational tools included in USEPA PMF version 5. These refinements include the imposition of constraints on the solution, based on our knowledge of the chemical composition of the aerosol sources affecting the receptor. Additionally, the uncertainties associated with this solution were fully characterised using the improved error estimation techniques in this version of PMF. Five factors in all, were isolated by PMF: ammonium sulfate, marine aerosol, mixed carbonaceous aerosol, crustal/Saharan dust and total traffic. The results obtained by PMF were further tested inter alia, by comparing them to those obtained by two other receptor modelling techniques: Constrained Weighted Non-negative Matrix Factorization (CW - NMF) and Chemical Mass Balance (CMB). The results of these tests suggest that the solution obtained by PMF, is valid, indicating that for this particular airshed PMF managed to extract most of the information about the aerosol sources affecting the receptor - even from a dataset with a limited number of samples.

Carbonaceous PM10 and PM2.5 and secondary organic aerosol in a coastal rural site near Brindisi (Southern Italy).

Organic and elemental carbon were measured both in daily PM10 and PM2.5 and in 6 h range time PM2.5 samples collected from September 2015 to October 2015 in a coastal rural site near Brindisi in the Apulia region (Italy), in order to determine factors affecting the carbonaceous aerosol variations. Carbon content (total carbon TC) represented a considerable fraction for both PM10 and PM2.5. In particular, in PM10 samples, organic carbon (OC) varied from 1.06 to 18.32 µg m-3 with a mean concentration of 5 ± 4 µg m-3 and EC varied from 0.11 to 0.88 µg m-3 with a mean value of 0.41 ± 0.19 µg m-3. In PM2.5 samples, OC varied from 0.54 to 12.91 µg m-3 with a mean concentration of 3.5 ± 2.8 µg m-3 and EC varied from 0.11 to 0.85 µg m-3 with a mean value of 0.35 ± 0.18 µg m-3. The highest values for both parameters were recorded when the air masses were coming from NE Europe and when Saharan Dust events were recognized. The results show that OC and EC exhibited higher concentrations during the night hours, suggesting that stable atmosphere and lower mixing conditions play important roles for the accumulation of air pollutants and promote condensation or adsorption of semivolatile organic compounds. In samples from a Saharan Dust event and in samples with the lowest and the highest OCsec, ATR-FTIR analysis allowed us to identify organic functional groups including the non-acid organic hydroxyl C-OH group (e.g., sugars, anhydrosugars, and polyols), carbonyl C=O group, carboxylic acid COOH group, aromatic and aliphatic unsaturated C=C-H group, aliphatic saturated C-C-H group, and amine NH2 group. Some inorganic ions were also identified: carbonates, sulfate, silicate, and ammonium. The dusty samples are mainly characterized by the presence of carbonate and hydrogen sulfate ions and by kaolinite (absorption at 914 and 1010 cm-1), while in samples with air masses coming from the NE, OC is mainly characterized by aliphatic and aromatic C-H and O-H and N-H groups (absorptions in the range 3500-2700 cm-1) and by the presence of organonitrate, aromatic amide and amine, and carboxylic acids (absorptions at 1630 and 1770-1700 cm-1). Graphical abstract ᅟ.

Volatilization and oxidative artifacts of PM bound PAHs collected at low volume sampling (1): Laboratory and field evaluation.

Laboratory and field studies were carried out to assess the effects of oxidative degradation and volatilization on PM10 bound polycyclic aromatic hydrocarbons (PAHs), collected at low volume condition according to the EU sampling reference method EN12341:2014 (flow rate 2.3 m3 h-1), on 47 mm quartz filters. For the laboratory experiments, pairs of twin samples were collected in field and, after treatments favoring decomposition or/and volatilization of PAHs on one sample, the PAH amount was compared with that of the corresponding untreated sample. Ozone exposure caused a general PAHs decay with more marked effects on benzo [a]pyrene, perylene and benz [a]anthracene; these compounds showed, similarly to benzo [ghi]perylene, correlations between ozone dose and losses. Treatments with zero air exhibited losses due to volatilization even for 5-ring PAHs up to benzo [a]pyrene, whereas a linear dependence was observed between filter PAH load and losses for benzo [a]anthracene, chrysene and benzofluoranthenes. Concentrations on samples collected simultaneously over 48, 24, 12 and 6 h were compared. Results confirmed a lack of temporal auto-consistency in the PAHs sampling methodology here adopted. In particular higher atmospheric PAH concentrations were ascertained on samples constituted by cumulative filters exposed over shorter sampling times. When 24-h and 2 × 12-h samples were compared, comparable losses were evaluated in the hot and cold seasons. This finding shows that, although in summer meteorology conditions favor sampling artifacts, the effectiveness of these phenomena continue in the winter, probably due to the larger amount of PAH available on the sampling filter (total PAHs ∼ 10 vs 0.5 ng m-3).