This paper first discusses the aerodynamic effects of trees on local scale flow and pollutant concentrationin idealized street canyon configurations by means of laboratory experiments and Computational FluidDynamics (CFD). These analyses are then used as a reference modelling study for the extension a theneighbourhood scale by investigating a real urban junction of a medium size city in southern Italy.A comparison with previous investigations shows that street-level concentrations crucially depend onthe wind direction and street canyon aspect ratio W/H (with W and H the width and the height ofbuildings, respectively) rather than on tree crown porosity and stand density. It is usually assumed in theliterature that larger concentrations are associated with perpendicular approaching wind. In this study,we demonstrate that while for tree-free street canyons under inclined wind directions the larger theaspect ratio the lower the street-level concentration, in presence of trees the expected reduction ofstreet-level concentration with aspect ratio is less pronounced.Observations made for the idealized street canyons are re-interpreted in real case scenario focusing onthe neighbourhood scale in proximity of a complex urban junction formed by street canyons of similaraspect ratios as those investigated in the laboratory. The aim is to show the combined influence ofbuilding morphology and vegetation on flow and dispersion and to assess the effect of vegetation onlocal concentration levels. To this aim, CFD simulations for two typical winter/spring days show that treescontribute to alter the local flow and act to trap pollutants. This preliminary study indicates that failing toaccount for the presence of vegetation, as typically practiced in most operational dispersion models,would result in non-negligible errors in the predictions.

Chemical mass balance modeling (CMB) was applied to determine the PM10 sources and their contributions.PM10 samples were collected in Lecce (40.338N, 18.108E, a town of South Italy), during two monitoring campaigns performedon July 2005 and February 2006. Nine source profiles and average mass concentration of the following chemicalparameters: elemental carbon (EC), organic carbon (OC), chlorine (Cl-), nitrate (NO3-), sulfate (SO42-), sodium (Na+),ammonium (NH4+), potassium (K+), magnesium (Mg2+), calcium (Ca2+), aluminum (Al), silicon (Si), titanium (Ti), vanadium(V), manganese (Mn), iron (Fe), copper (Cu), lead (Pb), and zinc (Zn) were used to run the CMBmodel. The resultsobtained by application of CMB8.2 are shown. The contributions to PM10 show that dominant contributor was traffic with37% followed by petroleum industry with 19% and field burning with 16%. Minor source contributions were marine aerosol(1%), ammonium sulfate production (4%), ammonium nitrate production (11%), oil-fired power plant (0.1%), gypsumhandling (10%) and crustal (2%).Moreover, the Absolute Principal Component Scores (APCS) model was applied to the PM10 samples collected in order tofind a correlation between the two source profile sets. With APCS model five source profiles were found and a good correlation(correlation coefficient bigger than 0.8) between crustal, marine, industrial profiles of CMB model and the correspondingones of APCS model was found.

Presso il panificio "da Teresa", sito in San Teramo in Colle (Bari), sono stati effettuati campionamenti indoor di PM2,5 (Silent Sequential Air Sampler, FAI Instruments) ed analisi dimensionale mediante un analizzatore ottico di particelle (OPC Monitor multichannel, FAI Instruments) dal 7 al 19 Aprile 2013. Sui campioni di PM2,5 sono state effettuate analisi chimiche per la determinazione di OC (carbonio organico), EC (carbonio elementare), LG (levoglucosano) Cl- (cloruri), NO2- (nitriti), NO3- (nitrati), SO42- (solfati), C2O42- (ossalati), Na+ (sodio), NH4+ (ammonio), K+ (potassio), Mg2+ (magnesio) e Ca2+ (calcio). Il campionamento durante l'arco della giornata ha avuto una durata minima di 4.30 ore e massima di 7 ore in modo da coprire le diverse attività lavorative svolte all'interno del panificio, dotato di due forni: uno a legna ed uno a gas. Le concentrazioni medie massime di OC (36.01 ?g/m3), EC (1.08 ?g/m3) e Levoglucosano (0.76??g/m3) si sono registrate nelle ore 3-9 dei vari giorni di campionamento, ovvero nelle ore di massima attività del panificio. Le particelle più fini, ovvero di dimensioni comprese tra 0.28 e 0.50 ?m, hanno rivelato una maggiore concentrazione durante le prime ore lavorative (3.40-7) ed hanno mostrato picchi più intensi in corrispondenza degli orari di accensione dei forni. Gli autori ringraziano la FAI Instruments per il supporto tecnico prestato ed il titolare del panificio "da Teresa".

Mass concentrations of PM2.5, mineral dust, organic carbon (OC) and elemental carbon (EC), water-soluble organic carbon (WSOC), sea salts and anthropogenic metals have been studied in a city-port of south Italy (Brindisi). This city is characterized by different emission sources (ship, vehicular traffic, biomass burning and industrial emissions) and it is an important port and industrial site of the Adriatic sea. Based on diagnostic ratios of carbonaceous species we assess the presence of biomass burning emissions (BBE), fossil fuel emissions (FFE) and ship emission (SE). Our proposed conversion factors from OC to OM are higher than those reported in the literature for urban site: the reason of this could be due to the existence of aged combustion aerosols during the sampling campaign (WSOC/OC = 0.6 +/- 0.3). (C) 2016 Elsevier B.V. All rights reserved.

Temporary streams are characterised by specific hydrological regimes, which influence ecosystem processes, groundwater and surface water interactions, sediment regime, nutrient delivery, water quality and ecological status. This paper presents a methodology to characterise and classify the regime of a temporary river in Southern Italy based on hydrological indicators (HIs) computed with long-term daily flow records. By using a principal component analysis (PCA), a set of non-redundant indices were identified describing the main characteristics of the hydrological regime in the study area. The indicators identified were the annual maximum 30- and 90-day mean (DH4 and DH5), the number of zero flow days (DL6), flow permanence (MF) and the 6-month seasonal predictability of dry periods (SD6). A methodology was also tested to estimate selected HIs in ungauged river reaches. Watershed characteristics such as catchment area, gauging station elevation, mean watershed slope, mean annual rainfall, land use, soil hydraulic conductivity and available water content were derived for each site. Selected indicators were then linked to the catchment characteristics using a regression analysis. Finally, MF and SD6 were used to classify the river reaches on the basis of their degree of intermittency. The methodology presented in this paper constitutes a useful tool for ecologists and water resource managers in the Water Framework Directive implementation process, which requires a characterisation of the hydrological regime and a 'river type' classification for all water bodies.

During a sampling campaign, carried out during June 2012, inside some traditional households located in four vil- lages (Phakding, Namche, Pangboche and Tukla) of Mt. Everest region in southern part of the central Himalaya (Nepal), particulate matter (PM) depositions and ashes have been collected. Moreover, outdoor PM depositions have also been analyzed. Chemical characterization of PM depositions and ashes for major ions, organic carbon, elemental carbon (EC), metal content and PAHs (Polycyclic Aromatic Hydrocarbons) allowed identifying, as major contributes to indoor PM, the following sources: biomass burning, cooking and chimney ashes. These sources significantly affect outdoor PM depositions: in-house biomass burning is the major source for outdoor EC and K+ as well as biomass burning and cooking activities are the major sources for Polycyclic Aromatic Hydrocarbons.[object Object]

During the month of December 2008 a sampling campaign was performed in Salento agricultural soils whit the aim of an explorative investigation of crustal elements and metals spatial distribution. The results have highlighted that crustal elements concentrations reflect the typical geo - lithological characteristics of the area, while some metal concentrations and distribution reflect the land use.

In this study, the Weather Research and Forecasting model with online coupled chemistry (WRF-Chem) isapplied to simulate an intense Saharan dust outbreak event that took place over the Southern Italy in March 2016.The WRF model is found to reproduce well the synoptic meteorological conditions driving the dust outbreak: an omegalikepressure configuration associated with a weak cyclogenesis in the Iberian Peninsula. At the end of the simulatedperiod the merging of two minima produce a large depression in the Peninsular Italy.The model performances in reproducing the atmospheric desert dust load is evaluated using a multi-platformobservational dataset of aerosol and desert dust properties, including optical properties from satellite and ground-basedsun-photometers, plus in-situ particulate matter mass concentration (PM) data. This comparison allows us to investigatethe model ability in reproducing both the horizontal and the vertical displacement of the dust plume, and its evolutionin time. The preliminary comparison with satellite (MODIS-AQUA) and sunphotometers (AERONET) showed thatthe model is able to reproduce well the horizontal field of the aerosol optical depth (AOD) and its evolution in time.The routinely measurements of ARPA-Puglia revealed the intense dust outbreak with peak PM10 value larger than300 µg/m3 during march 23. On the other side, the model-measurements comparison for PM10 shows a good temporalmatching.The model-to-measurements comparisons allows the evaluation and the tuning of physics-based emission scheme thatis part of the WRF-Chem package release.

Multivariate statistical techniques, including discriminant function analysis (DFA), cluster analysis (CA),principal component analysis (PCA), absolute principal component score (APCS), and radial basis function neural network(RBF-NN), were applied to a data set formed by 905 samples and 15 parameters, including pH, electrical conductivity(EC), total dissolved solids (TDS), dissolved oxygen (O2), chemical oxygen demand (COD), Na+, Ca2+, Mg2+, K+, Cl-,NO3-, SO42-, HCO3, and vital organism at 22°C and 36°C, of groundwater samples collected in the Apulia region (southernItaly). Among all collected samples, only samples showing values for all parameters were used to compose the data set onwhich the multivariate statistical techniques were applied. PCA and APCS allowed us to identify, for each province as wellas the sites diverging from the main cluster, the pollution sources pressuring the sampling sites investigated: they wereidentified as fertilizer applications, the use of unpurified irrigation water, marine water intrusion, and calcareous characteristicsof the soil. We found that the groundwater pollution sources pressuring the sites were similar among the five Apulianprovinces (Foggia, Bari, Brindisi, Lecce, and Taranto). Moreover, for each province, marine water intrusion showedthe highest contribution. The application of DFA to the data set allowed us to obtain good results in discriminating amongfour provinces, with the exception of Taranto. The model also gave good performance results in forecasting. However,RBF-NN provided more accurate results than DFA and confirmed that EC had the greatest relative importance. This isprobably due to the different salinity among the sites (Na+ also showed good discriminant importance). In fact, with PCAand APCS, it was possible to observe that EC, together with Na+, Cl-, and TDS, was the parameter that most often showedhigh loading values, and the scattered samples with these loading values were collected at sites in which marine waterintrusion had been hypothesized. The results obtained by multivariate statistical methods can be useful both in guidingstakeholders and in providing a valid tool to authorities for assessing and managing groundwater resources.

Le particelle di aerosol influenzano il clima modificando sia il bilancio energetico globale attraverso l'assorbimento e la dispersione della radiazione solare (effetti diretti), sia le interazioni con le nuvole e lo sviluppo e il verificarsi di precipitazioni (effetti indiretti). Gli aerosol di origine naturale sono particolarmente importanti perché su scala globale rappresentano circa 70-85 % del totale delle polveri disperse in atmosfera. Le principali componenti dell'aerosol naturale sono lo spray marino, la polvere desertica, i solfati naturali, gli aerosol vulcanici e quelli generati dagli incendi boschivi naturali. Tra gli aerosol naturali il maggior contributo circa il 70% è dato dall'erosione della crosta terrestre, mentre lo spray marino e le polveri di origine vulcanica rappresentano rispettivamente il 15% e l'1,5%.Sebbene le misurazioni in situ e il telerilevamento satellitare e terrestre forniscano informazioni importanti sul carico, la distribuzione e le influenze dell'aerosol, tali misurazioni sono essenzialmente limitate nello spazio e nel tempo e, soprattutto, sono limitate nella loro capacità di distinguere tra le componenti naturali e antropogeniche degli aerosol. D'altra parte le simulazioni numeriche condotte con modelli di trasporto e chimica risentono fortemente delle parametrizzazioni utilizzate. In questo studio, sono state valutate le capacità del modello WRF-Chem nel simulare il trasporto e le caratteristiche spazio-temporali degli aerosol naturali nel bacino del Mediterraneo, considerando alcuni casi particolari. Ovvero, sono stati investigati e simulati separatamente i seguenti tre casi: (i) un episodio sahariano verificatosi nel giugno 2007; (ii) un evento di trasporto di spray marino nell'Italia sud-orientale nel periodo 1-7 luglio 2007 e (iii) l'emissione vulcanica dell'Etna nel periodo 1-7 dicembre 2015.Per ciascun caso è stata effettuata un'analisi di sensitività rispetto ad alcune parametrizzazione interne al modello, in particolare quelle legate agli schemi di emissione. Le simulazioni sono state confrontate con dati al suolo registrati nelle centraline di qualità dell'aria e in alcune campagne sperimentali, e con dati satellitari. I risultati evidenziano delle buone performance del modello nel caso di trasporto sahariano e trasporto di polvere vulcanica. Ulteriori approfondimenti sono invece necessari nel caso dello spray marino.

The PM2.5 and PM10 fractions were simultaneously collected in three sites of the Bari territory (South-East Italy) during winter and summer sampling campaigns.The ionic components and the carbon fractions of the PM samples were determined. The Principal Component Analysis and Absolute Principal Component Scores of the dataset allowed to identify similar PM sources in the two PM fractions: organic and inorganic secondary compounds, local and diffuse combustion processes, calcium carbonate, sea salt. While secondary and combustion compounds were mainly in PM2.5, primary and natural species were prevalent in coarse PM. Seasonal and site-specific differences were investigated.

In the frame of the project EDOC@WORK3.0, Education and Work on Cloud, a monitoring plan has been carried out in the highly industrialized town of Taranto (one of the most polluted sites of Italy) in order to investigate contemporary indoor and outdoor concentrations of NO2 and SO2 by passive sampling devises (Radiello). Simultaneously indoor and outdoor samplings of NO2 and SO2 were performed from 2nd November 2015 to 2nd December 2015 in nine sites scattered in the investigated area at different quotes and distances from the industrial complex. Our findings show substantial differences between the spatial distributions of the two pollutants and support the hypothesis of two different prevalent sources for NO2 and SO2. In particular, we find diffusive sources of NO2 linked mainly to the vehicular traffic and secondarily to industrial sources. In contrast, SO2 was mainly associated to industrial sources present in the area, representing also a proxy of the mixture of air contaminants associated to industrial processes. Our hypothesis is also confirmed by analysis of data measured by ARPA air quality monitoring stations. Comparison between indoor and outdoor concentrations confirms that outdoor pollutants infiltrate to indoor environments, moreover it highlights potential NO2 indoor sources basically linked to cooking activities, representing adverse health effects for population risk categories such as children or cooks. Considering that urban people spend a lot of their time in indoors, attention should be paid both to outdoor pollutant sources and to indoor sources. © 2018 Elsevier B.V.

In the laboratory of a bakery, PM2,5 indoor samplings have been performed every each six hour meanly from 7th to 19th April 2013.For each daily sampling four PM2,5 samples have been collected. In total 40 samples were obtained. On each sample the following PAHs were determined: BaA(228), Bb,kF (252), BaP (252), BgP (276), Ip (276), DbA (278).The mean PAHs concentrations per sampling hourly range have shown, in the first and last sampling range, higher values than the other two hourly ranges. Because of it was not possible to perform samplings in the store of the baker, we have used the Computational Fluid Dynamics model in order to obtain information about dynamics and dispertion of human health hazard pollutants in the store and have an extimation of B(a)P concentration in that area.

Indoor air pollution assessment in work environments remains challenging due to a combination of logistic reasons and availability of costly instrumentation for data acquisition and post-processing. Existing literature focuses on energy production environments, hospitals, and less so on food production spaces. Studies on indoor air quality in bakeries are scarce or even absent. Motivated by this, the present study investigates indoor air quality in a bakery located in Bari province in South Italy, using a combination of approaches including analytical chemistry analyses and compu- tational fluid dynamics to reconstruct the air ventilation in response to air temperature gradients within the working environment. PM2.5 indoor samplings were collected every 6 h from 7 to 19 April 2013 in the proximity of two bakery ovens powered by gas and wood, respectively. For each sampling day, 4 PM2.5 samples were collected: from 3:00 to 9:00 h (first), from 9:00 to 13:30 h (second), from 14:00 to 21:00 h (third), and from 21:00 to 3:00 h (fourth). In total, 40 samples were collected. On each sample, several polycyclic aromatic hydrocarbons (PAHs) were deter- mined such as benzo[a]anthracene (228), benzo[b]fluoranthene (252), benzo[k]fluoranthene (252), benzo[a]pyrene (252), benzo[g,h,i]perylene (276), indeno[1,2,3-cd]pyrene (276), and dibenzo[a,h]anthracene (278), the main com- pounds of 16 priority US Environmental Protection Agency (US-EPA) PAHs in particulate phase. The PAH mean concentrations showed higher values during the first (from 3:00 to 9:00 h) and fourth (from 21:00 to 3:00 h) sampling intervals than the other two with benzo[a]pyrene mean values exceeding the Italian law limit of 1 ng/m3. Taking into account benzo[a]pyrene mean concentration for the first interval and the first plus the second one, which are the hours with the largest working activity, we have estimated that the baker and co-workers are exposed to a cancer risk of 4.3 × 10-7 and 5.8 × 10-7, respectively (these values are lower than US-EPA recommended guideline of 10-6). Our study was complemented by numerical analyses using state-of-the-art computational fluid dynamics to reconstruct at high resolution air movement from the various working places, i.e., the bakery and the selling area which were connected via a door. The numerical simulations were possible given that surface temperature using infrared thermog- raphy as well as air temperature was continuously recorded throughout the sampling acquisition. The use of this approach allowed us to estimate the transport and diffusion of benzo[a]pyrene from one area to the other thus complementing the point sampling information. Computational fluid dynamic simulation results confirm the presence of benzo[a]pyrene in the laboratory as obtained from the measurements and suggests its presence in the sales' area of the bakery with concentrations similar those found in the laboratory.

An interlaboratory comparison was performed to evaluate the analytical methods for quantification of anhydrosugars - levoglucosan, mannosan, galactosan - and biosugars - arabitol, glucose and mannitol - in atmospheric aerosol. The performance of 10 laboratories in Italy currently involved in such analyses was investigated on twenty-six PM (particulate matter)ambient filters, three synthetic PM filters and three aqueous standard solutions

The Weather Research and Forecasting model with online coupled chemistry (WRF-Chem) is applied to simulate a severe Saharan dust outbreak event that took place over Southern Italy in March 2016. Numerical experiments have been performed applying a physics-based dust emission model, with soil properties generated from three different Land Surface Models, namely Noah, RUC and Noah-MP. The model performance in reproducing the severe desert dust outbreak is analysed using an observational dataset of aerosol and desert dust features that includes optical properties from satellite and ground-based sun-photometers, and in-situ particulate matter mass concentration (PM) data. The results reveal that the combination of the dust emission model with the RUC Land Surface Model significantly over-predicts the emitted mineral dust; on the other side, the combination with Noah or Noah-MP Land Surface Model (LSM) gives better results, especially for the daily averaged PM10.

Atmospheric aerosols have potential effects on human health, on the radiation balance, on climate, and on visibility.The understanding of these effects requires detailed knowledge of aerosol composition and size distributionsand of how the different sources contribute to particles of different sizes. In this work, aerosol sampleswere collected using a 10-stage Micro-Orifice Uniform Deposit Impactor (MOUDI). Measurements were takenbetween February and October 2011 in an urban background site near Lecce (Apulia region, southeast of Italy).Samples were analysed to evaluate the concentrations of water-soluble ions (SO4 2-, NO3-, NH4+, Cl-, Na+, K+,Mg2+ and Ca2+) and of water-soluble organic and inorganic carbon. The aerosols were characterised by twomodes, an accumulation mode having a mass median diameter (MMD) of 0.35 ± 0.02 ?m, representing51 ± 4% of the aerosols and a coarse mode (MMD = 4.5 ± 0.4 ?m), representing 49 ± 4% of the aerosols.The data were used to estimate the losses in the impactor by comparison with a low-volume sampler. The averageloss in the MOUDI-collected aerosol was 19 ± 2%, and the largest loss was observed for NO3- (35 ± 10%).Significant losses were observed for Ca2+ (16 ± 5%), SO4 2 - (19 ± 5%) and K+ (10 ± 4%), whereas the lossesfor Na+ and Mg2+ were negligible. Size-segregated source apportionment was performed using Positive MatrixFactorization (PMF), which was applied separately to the coarse (size interval 1-18 ?m) and accumulation (sizeinterval 0.056-1 ?m) modes. The PMF model was able to reasonably reconstruct the concentration in each sizerange.The uncertainties in the source apportionment due to impactor losseswere evaluated. In the accumulationmode, it was not possible to distinguish the traffic contribution from other combustion sources. In the coarsemode, it was not possible to efficiently separate nitrate from the contribution of crustal/resuspension origin.

This paper analyses elemental (EC), organic (OC) and total carbon (TC) concentration in PM2.5 and PM10 samples collected over the last few years within several national and European projects at 37 remote, rural, urban, and traffic sites across the Italian peninsula.The purpose of the study is to obtain a picture of the spatial and seasonal variability of these aerosol species in Italy, and an insight into sources, processes and effects of meteorological conditions.OC and EC showed winter maxima and summer minima at urban and rural locations and an opposite behaviour at remote high altitude sites, where they increase during the warm period due to the rising of the Planetary Boundary Layer (PBL). The seasonal averages of OC are higher during winter compared to summer at the rural sites in the Po Valley (from 1.4 to 3.5 times), opposite to what usually occurs at rural locations, where OC increases during the warm period. This denotes the marked influence of urban areas on the surrounding rural environment in this densely populated region.The different types of sites exhibit marked differences in the average concentrations of carbonaceous aerosol and OC/EC ratio. This ratio is less sensitive to atmospheric processing than OC and EC concentrations, and hence more representative of different source types. Remote locations are characterised by the lowest levels of OC and especially EC, with OC/EC ratios ranging from 13 to 20, while the maximum OC and EC concentrations are observed at road-traffic influenced urban sites, where the OC/EC ratio ranges between 1 and 3. The highest urban impacts of OC and EC relative to remote and rural background sites occur in the Po Valley, especially in the city of Milan, which has the highest concentrations of PM and TC and low values of the OC/EC ratio.

In this paper, the results obtained from multivariate statistical techniques such as PCA (Principal component analysis) and LDA (Linear discriminant analysis) applied to a wide soil data set are presented. The results have been compared with those obtained on a groundwater data set, whose samples were collected together with soil ones, within the project "Improvement of the Regional Agro-meteorological Monitoring Network (2004-2007)". LDA, applied to soil data, has allowed to distinguish the geographical origin of the sample from either one of the two macroaeras: Bari and Foggia provinces vs Brindisi, Lecce e Taranto provinces, with a percentage of correct prediction in cross validation of 87%. In the case of the groundwater data set, the best classification was obtained when the samples were grouped into three macroareas: Foggia province, Bari province and Brindisi, Lecce and Taranto provinces, by reaching a percentage of correct predictions in cross validation of 84%. The obtained information can be very useful in supporting soil and water resource management, such as the reduction of water consumption and the reduction of energy and chemical (nutrients and pesticides) inputs in agriculture.

La deposizione secca è una delle due principali vie di rimozione delle particelle di aerosol dall'atmosfera e trasferimento in altri comparti, diventando il processo dominante in determinate condizioni.Misure di deposizione secca con risoluzione temporale sub-settimanale sono rare. Per questo motivo un set di 7 nuovi raccoglitori di deposizioni secche sono stati installati in diversi siti dell'area del Mediterraneo e delle coste atlantiche (Huelva, Barcelona, Spain; Île-Rousse, France; Gozo Island, Malta; Lecce, Italy; Athens, Greece; Nicosia, Cyprus) per monitorarne la variabilità spaziale e temporale.Le particelle di aerosol si depositano su un filtro (25 mm) adesivo di carbonio, protetto da deposizioni umide mediante uno shelter che, in automatico, si abbassa in caso di piaggia.I filtri di carbonio raccolti vengono analizzati mediante microscopia elettronica (SEM-EDX) per ottenere dimensione, forma e composizione elementare di ogni singola particella. Per ogni campione raccolto vengono analizzate diverse centinaia di migliaia di particelle.

Dry deposition is one of the two major pathways for aerosol particles to be removed from the atmosphere and be transferred into another compartment. Depending on the precipitation pattern and thus, locality, it can dominate then total aerosol flux. Still, measurements of dry deposition with sub-weekly time resolution are rare, as usually the mass collected during this period is too low.For XMed-Dry, a set of seven new dry deposition-only collectors were installed at different locations across the Mediterranean and Atlantic coasts (Huelva, Barcelona, Spain; Île-Rousse, France; Gozo Island, Malta; Lecce, Italy; Athens, Greece; Nicosia, Cyprus) to capture spatial and temporal variability. Sampling was performed on a 3-times-per-week schedule. Particles deposited on a 25 mm pure carbon adhesive protected from wet deposition by a shelter and an active closing mechanism during rain. The carbon adhesive was subject to electron microscopy with energy-dispersive X-ray analysis to obtain size, shape and elemental composition of single particles. Several hundred to several thousand particles were analyzed for each sample.First results show that deposition consists of a highly variable mixture of sea-salt, sulfate, mineral dust, metal oxides and biological material, depending on location, season and meteorological situation. Moreover, different state of ageing of sea-salt was detected by the single particle analysis. Statistical back-trajectory analyses allowed for the discrimination of potential source regions for different compounds. The latter showed that particles of similar chemical composition, but different size can have different origin, e.g., in Gozo large (d>4µm) iron-rich particles (possibly fly-ashes) have probably SE European origin, whereas small iron-rich ones originate from the Saharan desert. For the same receptor location, S-rich sea-salt particles were transported in the boundary layer over Italy and SW Europe, while the more pristine ones were produced locally or coming from the Atlantic Ocean.