Abstract
The key objectives of the present study consisted of: I) the study of the occurrence ofPM2.5, PM10 and PM2.5-associated Polycyclic Aromatic Hydrocarbons (PAH) in the urbanatmosphere of Heraklion Crete, by following a specific protocol implemented in the frameworkof the ESCAPE project (http://www.escapeproject.eu) during 2009-2010; and II) the study ofthe occurrence of toxic organic compounds, such as PAHs, Polychlorinated Biphenyls (PCBs)and Organochlorine Pesticides (OCPs), transported by air masses during the presence orabsence of African dust incursions in the downwind area of Eastern Mediterranean (island ofCrete). To achieve the aforementioned objective, in contrast to previous studies that have beenfocused to a small amount of dust episodes, a meticulous 3-year sampling campaign wasconducted (2012-2015) in the framework of the CHEMISAND project(http://chemisand.chemistry.uoc.gr/).For that purpose, an efficient analytical methodology was developed that allowed thequalitative and q ...
The key objectives of the present study consisted of: I) the study of the occurrence ofPM2.5, PM10 and PM2.5-associated Polycyclic Aromatic Hydrocarbons (PAH) in the urbanatmosphere of Heraklion Crete, by following a specific protocol implemented in the frameworkof the ESCAPE project (http://www.escapeproject.eu) during 2009-2010; and II) the study ofthe occurrence of toxic organic compounds, such as PAHs, Polychlorinated Biphenyls (PCBs)and Organochlorine Pesticides (OCPs), transported by air masses during the presence orabsence of African dust incursions in the downwind area of Eastern Mediterranean (island ofCrete). To achieve the aforementioned objective, in contrast to previous studies that have beenfocused to a small amount of dust episodes, a meticulous 3-year sampling campaign wasconducted (2012-2015) in the framework of the CHEMISAND project(http://chemisand.chemistry.uoc.gr/).For that purpose, an efficient analytical methodology was developed that allowed thequalitative and quantitative determination of a relatively large number of toxic organiccompounds (37 PAH members, 50 PCB congeners and 24 OCP members) by using a highresolution gas chromatographer-mass spectrometer (HRGC-MS) in Electron Ionization (EI)and Electron Capture Negative Ionization (ECNI) mode, respectively. The methodology wascharacterized by high quality assurance and control levels (enhanced repeatability, lowuncertainty, reliable compound separation, high recovery levels).The median adjusted PM2.5 concentration in the urban area of Heraklion was 14.3 and13.2 μg m-3, for the urban traffic (N=20) and background (N=20) sites, respectively, while thecorresponding median adjusted PM10 concentration was 37.7 and 35.7 μg m-3, respectively.We did not observe a statistically significant differentiation in PM2.5 and PM10 levels betweenurban traffic and background sites. We measured lower PM2.5 concentrations than the annualarithmetic mean benchmark level, established by EU (25 μg m-3), which is currently in effectfrom January 1st 2015. On the other hand, relatively higher PM10 concentrations were reportedin a few urban sites, having regard to the corresponding EU PM10 annual limit value (40 μg m-3).We measured the highest daily PM concentrations (PM2.5, PM10, TSP) during intensiveAfrican dust episodes, while relatively lower PM concentrations were measured during dustepisodes mixed with aerosol transport from other areas. Conversely, we measured the lowestdaily PM concentrations during the majority of periods with absence of African dust episodesinto the downwind area (Eastern Mediterranean). We also observed a significant statisticaldifferentiation in PM levels between intensive African dust and mixed episodes and days withabsence of African dust episodes. The aforementioned differentiation was also observedregarding the PM2.5/PM10 ratio (0.32 and 0.54 median values, during African dust and non-dustevents, respectively). A statistically significant positive correlation was observed between PM2.5and PM10 levels measured at the downwind area, demonstrating common emission PMsources, the burden of which is controlled by air masses, transporting African mineral dust.The median ΣPAH concentration measured in PM2.5 urban samples (sum of 10 trafficand background sites) was 0.63 ng m-3, while the corresponding arithmetic mean concentrationwas 0.64 (±0.25) ng m-3. We measured relatively higher ΣPAH concentrations in the urbantraffic, compared to the background sites, without any significant statistical difference betweenthem. Accordingly, the mean seasonal ΣPAH levels did not significantly differ between urbantraffic and background sites, suggesting common PAH sources. The arithmetic mean ΣPAH particulate concentration during pure African dust eventswas 0.86 (±0.98) ng m-3. Among that group of samples, there has been a statisticallyinsignificant variation in ΣPAH levels, relative to the origin of the air masses (72-h backtrajectories) reaching the downwind area. BaP arithmetic mean concentration was 0.05 (±0.06)ng m-3, much lower than the corresponding annual limit value (1.0 ng m-3), established by EU.The arithmetic mean ΣPAH particulate concentration during mixed episodes was 1.10 ng m-3,while the corresponding BaP concentration was 0.07 (±0.07) ng m-3, also lower than the EUlimit value (1.00 ng m-3). During the absence of African dust episodes, the mean ΣPAHparticulate concentration was 1.60 (±2.40) ng m-3. The relative arithmetic (but statisticallyinsignificant) differentiation in ΣPAH levels between African dust episodes (pure and mixed)and non-dust intervals is ascribed to the enhanced influence of air masses originated frommore polluted areas (Northern Europe) which more likely affect the downwind area, comparedto the corresponding air parcels originated from more arid areas, such as the Sahara-Sahelregion.The mean total (gas and particulate phase) ΣPCB concentration measured during thepresence of pure African dust events was 77.1 (±41.1) pg m-3, while the corresponding meantotal ΣPCB concentration during mixed dust episodes was 92.5 (±44.9) pg m-3. We did notobserve a statistically significant variation in ΣPCB levels, relative to the origin of air masses.The mean total (gas and particle) ΣPCB concentration during periods of non-African dustconditions was 81.9 (±38.7) pg m-3. We observed low and statistically insignificant correlationbetween ambient temperature and ΣPCB concentrations, illustrating the lack of seasonality inPCB atmospheric levels in the downwind study area. ΣPCB concentration differences betweendust (pure and mixed) and non-dust conditions were also statistically insignificant. Given thelack of seasonality in PCB (gas & particulate phase) ambient levels in the downwind area, thelow and insignificant correlation between partial vapour pressures of gas phase PCBs andtemperature and the shallow slopes obtained from Clausius-Clapeyron plots, we assume thatPCB atmospheric levels in the Eastern Mediterranean are more likely controlled by acombination of physicochemical processes and long-range transport (LRT) phenomena.Mean total (gas and particulate phase) concentrations of 6DDX (2,4´-DDD, 4,4´-DDD,2,4´-DDT, 2,4´-DDE, 4,4´-DDE and 4,4´-DDT), 3Chlordanes (cis-Chlordane, trans-Chlordaneand oxy-Chlordane), 3Endosulfans (α-Endosulfan, β-Endosulfan and Endosulfan sulfate) and4HCHs (α-HCH, β-HCH, γ-HCH and δ-HCH), measured during periods of pure African dustincursions, were 17.1 (±5.7), 1.1 (±0.3), 3.8 (±2.2) and 3.6 (±1.4) pg m-3, respectively. Thecorresponding mean total (gas and particle) concentrations during mixed African dust eventswere 22.8 (±10.5), 1.6 (±1.0), 5.4 (±4.8) and 4.4 (±2.2) pg m-3, respectively. Most OCPstargeted in this study were found to be predominant in the gas phase, while Endosulfanisomers (β-Endosulfan, Endosulfan sulfate, α-Endosulfan) were mostly distributed in theparticulate phase of the atmosphere. Mean total (gas & particle) concentrations of 6DDX,3Chlordanes, 3Endosulfans and 4HCHs during periods of non-dust conditions were measuredrelatively similar to those detected during pure and mixed dust conditions [32.7 (±13.6), 2.1(±1.3), 7.3 (±4.9) and 11.3 (±7.5) pg m-3, respectively], with statistically insignificantconcentration differences between them. The aforementioned observation illustrates the lackof any kind of OCP emission source in the vicinity of the downwind study area (EasternMediterranean). Finally, long-range transport (LRT) is considered as the main factor controllingthe atmospheric levels of OCPs in this area, owing to the low and statistically insignificant correlations between partial vapor pressures of gas phase OCPs and ambient temperature, aswell as the shallow slope values obtained from Clausius-Clapeyron plots.
show more
