Boreal environment research 16: 251–261 © 2011 issn 1239-6095 (print) issn 1797-2469 (online) helsinki 30 august 2011 chemical characterization of Pm in the apulia region: local and long-range transport contributions to particulate matter martino amodio1), eleonora andriani1), lorenzo angiuli2), Giorgio assennato2), Gianluigi de Gennaro1), alessia Di Gilio1), roberto Giua2), miriam intini1), micaela menegotto3), alessandra nocioni4), Jolanda Palmisani1), maria rita Perrone5), claudia marcella Placentino1) and maria tutino1) 1) Chemistry Department, University of Bari, Via Orabona 4, 70126, Bari, Italy 2) Apulia Region Environmental Protection Agency (ARPA Puglia), Corso Trieste 27, IT-70126 Bari, Italy 3) Apulia Region Environmental Protection Agency (ARPA Puglia), Ospedale Testa, Contrada Rondinella, IT-74123 Taranto, Italy 4) Apulia Region Environmental Protection Agency (ARPA Puglia), Via Galanti 16, IT-72100 Brindisi, Italy 5) Physics Department, University of Salento, Via Arnesano, IT-73100 Lecce, Italy Received 12 Nov. 2010, accepted 4 Jan. 2011 (Editor in charge of this article: Veli-Matti Kerminen) amodio, m., andriani, e., angiuli, l., assennato, G., de Gennaro, G., Di Gilio, a., Giua, r., intini m., menegotto m., nocioni, a., Palmisani, J., Perrone, m. r., Placentino, c. m. & tutino, m. 2011: chemical characterization of Pm in the apulia region: local and long-range transport contributions to particulate matter. Boreal Env. Res. 16: 251–261. The chemical composition of PM10 samples collected during a monitoring campaign per- formed at five Apulia Region sites (Torchiarolo, Galatina, Taranto-Tamburi, Palagiano and Montalbano-Ostuni) was investigated. Ionic fraction, metals, carbonaceous compounds and polycyclic aromatic hydrocarbons (PAHs) were determined for all samples. The PM concentrations showed a regional background of PM10 at all the sites except for Taranto and Torchiarolo which are characterized by an important industrial area close to the town and by a local biomass burning source, respectively. A principal component analysis (PCA) applied to the dataset provided four sources relevant for the area under investigation, that were identified as combustion processes, crustal, marine aerosol and secondary particulate matter. The integrated approach used in this work (chemical characterization of the sam- ples, meteorological information and statistical analysis) allowed to evaluate the synergis- tic effect of regional secondary particulate and long-transport contributions to PM10 con- centrations at the sampling sites and to provide useful information in order to characterize the most relevant local emission sources. Introduction emission sources for the area under investiga- tion and how PM concentrations at receptors are Growing interest in particulate matter (PM) pollu- influenced by transport, mixing and transforma- tion is related to its impact on health (Künzli and tion processes (Henry et al. 1984, Chueinta et al. Perez 2009) and its interactions with other pollut- 2000, Kim and Henry 2000, Vallius et al. 2005). ants. Therefore, plans for human health protection The European Union (EU) established con- should take into account the most relevant PM centration limits for particulate matter with aero- 252 Amodio et al • Boreal env. res. Vol. 16 dynamic diameters smaller than 10 µm (PM10), fractions and metals. Meteorological informa- as well as for several other atmospheric pol- tion and remote-sensing data (DREAM, NAAPS lutants. In particular, the 1999/30/EC and the and Hysplit models) were useful to evaluate the 2008/50/EC directives set up 40 µg m–3 and 50 atmospheric dispersion conditions. Finally, the µg m–3 as annual and daily limit values, respec- principal component analysis (PCA) was applied tively; 35 exceedances of this last value are to the dataset in order to obtain information on allowed for a year. It is also possible to discount main sources of PM10. the exceedances due to natural events that add to PM10 background concentrations, especially in southern European regions that are signifi- Material and methods cantly affected by Saharan dust outbreaks. As concerns the Mediterranean areas, different con- Sampling sites tributions to PM10 concentrations were consid- ered relevant: the long-range transport from the The monitoring campaign of PM10 at the Apulia industrial European regions, natural apportion Region sites entitled ‘Taranto-Salento’ was per- from African regions due to Saharan dust out- formed in collaboration with different partners, breaks and local dust suspension, marine aerosol such as the Apulian Agency for Environmen- contribution from the Mediterranean sea, and tal Protection, University of Bari, University of biomass burning particle contributions caused Salento, Institute of Atmospheric Sciences and by forest fires that occur during summer (Papadi- Climate of the Italian National Research Council mas et al. 2008). In this field, the regional back- (CNR-ISAC). Fifteen daily PM10 samples were ground (RB) sites are useful locations to study collected at five sites: Taranto-Tamburi (TA), contributions of atmospheric aerosols due to Palagiano (PA), Torchiarolo (TR), Galatina (GA) long-transport, because the PM concentrations and Masseria Montalbano-Ostuni (MM). Taranto measured at these sites are determined mainly by (TA; 40°28´N, 17°14´E) is the third most popu- the meteorological conditions rather than local lated city of southern Italy. One of the largest emission sources. However, local anthropogenic steel plants in Europe is located there and fur- emissions from vehicular transport, industry and thermore, an important industrial centre with pet- domestic heating significantly affect PM10 levels rochemical and cement plants, military and mer- at urban sites, especially when low dispersion chant harbours and shipyards are located close meteorological conditions occur. Therefore, the to the urban area. The sampling site was placed evaluation of the local contribution to atmos- close to the industrial area and it was identified pheric pollutant concentrations, compared with as Taranto-Taburi. Palagiano (PA; 40°35´0´´N, regional background levels, may help to identify 17°3´0´´), a small town of about 15 000 inhab- high environmental impact areas and to suggest itants in the Taranto Province, was selected as the correct plans for human health protection the suburban site. Torchiarolo (TR; 40°29´0´´N, (Escudero et al. 2006). 18°3´0´´E) is a country with about 5000 inhabit- In order to investigate these different con- ants located on the border between the Brindisi tributions to the Apulia Region PM10 concen- and Lecce provinces. It is characterized by only trations and to evaluate how the long-range agricultural activities, and therefore it was con- transport influence pollutants concentrations at sidered a regional background site. The Galatina the sampling sites, the chemical composition of (GA; 40°10´0´´N, 18°10´0´´E) sampling site, PM10 collected in the Regional area was inves- located in the city centre and characterized by tigated. For this purpose, 15 PM10 daily samples vehicular traffic emissions, was selected as a were collected from February 2009 to March representative of an urban area. Masseria Mon- 2009 at five different regional sites (Torchi- talbano (MM; 40°46´38´´N, 17°28´39´´E), a vil- arolo, Galatina, Taranto-Tamburi, Palagiano and lage of about 4000 inhabitants near Brindisi, was Montalbano-Ostuni). The chemical characteriza- chosen as a regional background site because this tion of the samples was performed to determine area is characterized by only old farmhouses and inorganic components, PAHs, carbonaceous olive trees. Boreal env. res. Vol. 16 • Chemical characterization of PM in the Apulia Region 253 PM sampling X-ray spectra were fitted using the Win Trace software. The excitation X-rays were obtained PM10 daily samples were collected by a dichoto- with an X-ray tube with an Rh anode (Imax = mous low volume sampler SWAM Dual Sampler 1.98 mA, Vmax = 50 kV). The Si (Li) detec- (FAI Instruments s.r.l., Roma, Italy). PM10 sam- tor provided the means for sensing fluoresced ples were collected on quartz fiber filters (What- X-ray photons and converting them to electri- mann, 47 mm diameter) by FAI EN 1234.1 sam- cal signals. A check of the overall accuracy pling heads operating at a flow rate of 2.3 m3 h–1. of our quantitative analysis was performed by The sampling was carried out from 21 February analyzing the SRM NIST 2783 standard (PM2.5 to 7 March 2009 at all five sites and a total of on Nuclepore polycarbonate membrane). Detec- –2 75 PM10 samples were collected. The chemi- tion limits ranged from 0.1 to 60 ng cm on –3 cal characterization of the PM10 daily samples the filter deposit and from 0.38 to18 ng m for was performed in order to determine inorganic 24-hour samples in terms of airborne concentra- – – 2– + components such as ions (Cl , NO3 , SO4 , Na , tion (depending on the element). + + 2+ 2+ NH4 , K , Mg , and Ca ), elements (Si, Fe, Ti, Al, Cu, Zn, As and Mn), polycyclic aromatic hydrocarbons (PAHs) and carbonaceous com- Polycyclic Aromatic Hydrocarbon pounds (OC and EC). The meteorological data analysis were downloaded from the airport of Grottaglie in the town of Taranto. In particular, the wind The extraction of PAHs was realized with a velocity (WV), atmospheric temperature (T), mixture of acetone/hexane by means of a micro- pressure (P) and relative humidity (RH) were wave assisted solvent extraction using Mile- taken into account during