Inferences over the sources and processes affecting polycyclic aromatic hydrocarbons in the atmosphere derived from measured data

Data concerning atmospheric lifetime and relative source contributions of polycyclic aromatic hydrocarbons (PAH) are fragmentary and contradictory. In this study, two datasets of measurements of atmospheric PAH (sum of particulate and gaseous phases), one from a national network, the other from a more local three-site study, were analysed and used to infer processes affecting PAH in the atmosphere, and their sources. PAH congener profiles measured at urban and rural locations were remarkably similar suggesting that atmospheric decay processes are relatively slow. This allows the use of such profiles to elucidate sources. A spatial analysis of two PAH datasets showed a clear influence of industry and road traffic upon local PAH concentrations. When Principal Component Analysis (PCA) was applied to UK national network data, it showed a clear influence of steel industry emissions and of home heating emissions from coal and oil in Northern Ireland. These sites also showed different winter/summer concentration ratios to the main group of sites. In the data from Birmingham (UK), PCA identified separate factors relating to gasoline and diesel vehicles, as well as the influence of wood combustion on “Bonfire night”, and a factor related to home heating emissions which shows up only in the cold season.     

Atmospheric concentrations and air-soil gas exchange of polycyclic aromatic hydrocarbons (PAHs) in remote, rural village and urban areas of Beijing-Tianjin region, North China

Forty passive air samplers were deployed to study the occurrence of gas and particulate phase PAHs in remote, rural village and urban areas of Beijing-Tianjin region, North China for four seasons (spring, summer, fall and winter) from 2007 to 2008. The influence of emissions on the spatial distribution pattern of air PAH concentrations was addressed. In addition, the air-soil gas exchange of PAHs was studied using fugacity calculations. The median gaseous and particulate phase PAH concentrations were 222 ng/m³ and 114 ng/m³, respectively, with a median total PAH concentration of 349 ng/m³. Higher PAH concentrations were measured in winter than in other seasons. Air PAH concentrations measured at the rural villages and urban sites in the northern mountain region were significantly lower than those measured at sites in the southern plain during all seasons. However, there was no significant difference in PAH concentrations between the rural villages and urban sites in the northern and southern areas. This urban-rural PAH distribution pattern was related to the location of PAH emission sources and the population distribution. The location of PAH emission sources explained 56%-77% of the spatial variation in ambient air PAH concentrations. The annual median air-soil gas exchange flux of PAHs was 42.2 ng/m²/day from soil to air. Among the 15 PAHs measured, acenaphthylene (ACY) and acenaphthene (ACE) contributed to more than half of the total exchange flux. Furthermore, the air-soil gas exchange fluxes of PAHs at the urban sites were higher than those at the remote and rural sites. In summer, more gaseous PAHs volatilized from soil to air because of higher temperatures and increased rainfall. However, in winter, more gaseous PAHs deposited from air to soil due to higher PAH emissions and lower temperatures. The soil TOC concentration had no significant influence on the air-soil gas exchange of PAHs.     

Ambient volatile organic compound (VOC) concentrations around a petrochemical complex and a petroleum refinery

Air samples were collected between September 2000 and September 2001 in Izmir, Turkey at three sampling sites located around a petrochemical complex and an oil refinery to measure ambient volatile organic compound (VOC) concentrations. VOC concentrations were 4-20-fold higher than those measured at a suburban site in Izmir, Turkey. Ethylene dichloride, a leaded gasoline additive used in petroleum refining and an intermediate product of the vinyl chloride process in the petrochemical complex, was the most abundant volatile organic compound, followed by ethyl alcohol and acetone. Evaluations based on wind direction clearly indicated that ambient VOC concentrations measured were affected by the refinery and petrochemical complex emissions. VOC concentrations showed seasonal variations at all sampling sites. Concentrations were highest in summer, followed by autumn, probably due to increased evaporation of VOCs from fugitive sources as a result of higher temperatures. VOC concentrations generally increased with temperature and wind speed. Temperature and wind speed together explained 1-60% of the variability in VOC concentrations. The variability in ambient VOC concentrations that could not be explained by temperature and wind speed can be attributed to the effect of other factors (i.e. wind direction, other VOC sources).     

2,3,7,8-TCDD equivalence and mutagenic activity associated with PM10 from three urban locations in New Zealand

Ambient particulate matter (PM(10)) in urban centres varies depending on emission sources, geography, demography, and meteorology. Hence physical (PM(10), wind speed, rainfall, temperature), chemical (polycyclic aromatic hydrocarbons, PAH), and toxicological (Ames Test, H4IIE EROD Assay) analyses were done on daily PM(10) (approximately 1640 m(3)/day) collected from three New Zealand urban sites where winter emissions were predominantly due to domestic home heating. Daily PM(10) levels ranged between 9.7 and 20.8 in summer and between 21.8 and 61.0 microg/m(3) in winter. Daily PAH concentrations were 0.5, 0.45, and 1.5 ng/m(3) in summer and 52.1, 128.9, and 5.8 ng/m(3) in winter at sites Christchurch, Alexandra and Dunedin, respectively. During winter, 74% of PM(10) extracts from all three sites showed significant mutagenicity in the Ames Test (TA 98, -S9), whereas approximately 25% of the daily PM(10) was mutagenic in summer. Benzo[a]pyrene and BaP carcinogenic equivalence concentrations during winter were strongly correlated to both mutagenicity and TCDD-like activity at two sites. Daily levels of TCDD toxicity equivalence concentrations ranged from 0.5 to 3.6 pg TCDD/m(3) air in summer and from 0.3 to 4009 pg TCDD/m(3) air in winter. Chemically and biologically derived TCDD toxicity equivalent concentrations were significantly correlated in all study locations indicating that PAH may represent most of the TCDD-like activity present in the PM(10).     

Polycyclic Aromatic Hydrocarbons (PAHs) in urban topsoils: Concentration and source analysis in Xuzhou,China

Polycyclic aromatic hydrocarbons (PAHs) are ubiquitous organic pollutants in urban environments and are considered as the priority pollutants by the US Environmental Protection Agency (EPA). In this study, surface soils (0–10 cm) samples from 41 sampling sites in Xuzhou (China) were collected and analysed for 16 EPA priority polycyclic aromatic hydrocarbons via gas chromatograph/mass spectrometry (GC-MS). Total PAHs concentrations were in the range of 526 to 8612 ng/g with a mean of 2468 ng/g. Total PAHs concentrations were compared to guideline and results showed that nearly all urban soils in our study can be classified as heavily contaminated (>1000 ng/g). Carcinogenic PAHs were detected in relatively high concentrations and contributed 40.2% of the total PAHs in the urban soils. Multivariate statistical analysis including cluster analysis and factor analysis were performed and grouped the 16 PAHs under 2 clusters, which could indicate that the PAHs had different origins. The values of PAHs isomer indicated that vehicular emission might contribute to the soil PAHs in urban topsoils. The carcinogenic potency of PAH compounds was calculated and was found to be insignificant at the present level of emissions in Xuzhou.     

Factors influencing particle number concentrations, size distributions and modal parameters at a roof-level and roadside site in Leicester, UK

Measurements of urban particle number concentrations and size distributions in the range 5-1000 nm were taken at elevated (roof-level) and roadside sampling sites on Narborough Road in Leicester, UK, along with simultaneous measurements of traffic, NO(x), CO and 1,3-butadiene concentrations and meteorological parameters. A fitting program was used to determine the characteristics of up to five modal groups present in the particle size distributions. All particle modal concentrations peaked during the morning and evening rush hours. Additional events associated with the smallest mode, that were not observed to be connected to primary emissions, were also present suggesting that this mode consisted of newly formed secondary particles. These events included peaks in concentration which coincided with peaks in solar radiation, and lower concentrations of the larger modes. Investigation into the relationships between traffic flow and occupancy indicated three flow regimes; free-flow, unstable and congested. During free-flow conditions, positive linear relationships existed between traffic flow and particle modal number concentrations. However, during unstable and congested periods, this relationship was shown to break-down. Similar trends were observed for concentrations of the gas phase pollutants NO(x), CO and 1,3-butadiene. Strong linear relationships existed between NO(x), CO, 1,3-butadiene concentrations, nucleation and Aitken mode concentrations at both sampling locations, indicating a local traffic related emission source. At the roadside, both nucleation and Aitken mode are best represented by a decreasing exponential function with wind speed, whereas at the roof-level this relationship only occurred for Aitken mode particles. The differing relationships at the two sampling locations are most likely due to a combination of meteorological factors and distance from the local emission source.     

Measurement of atmospheric PCDD/F and PCB distributions in the vicinity area of Waelz plant during different operating stages

A comprehensive sampling program was conducted at four sampling sites in the vicinity area of a Waelz plant prior to and after the retrofitting of this plant for reducing PCDD/F (polychlorinated dibenzo-p-dioxin and polychlorinated dibenzofuran) and polychlorinated biphenyl (PCB) emission to evaluate its effects on atmospheric PCDD/F and PCB concentrations. During the shutdown stage of the Waelz plant investigated, the atmospheric PCDD/F and PCB concentrations measured in the vicinity area (sites A, B and C) range from 24 to 170 fg-I-TEQ/m(3) and 5.0 to 8.2 fg-TEQ(WHO)/m(3), respectively. As the facility restarted to operate, the atmospheric PCDD/F and PCB concentrations measured increased to 570-1460 fg-I-TEQ/m(3) and 23-59 fg-TEQ(WHO)/m(3), respectively. When activated carbon injection (ACI) was adopted in the facility in early 2006, the PCDD/F concentrations measured in stack gas decreased dramatically from 190 to 3.4 ng-I-TEQ/Nm(3), while the PCB concentrations decreased from 4.9 to 0.08 ng-TEQ(WHO)/Nm(3). In the meantime, the atmospheric PCDD/F and PCB concentrations measured at sites A, B and C decreased to 150-340 fg-I-TEQ/m(3) and 13-19 fg-TEQ(WHO)/m(3), respectively. However, relatively high PCDD/F (180-1460 fg-I-TEQ/m(3)) and PCB (9.3-59 fg-TEQ(WHO)/m(3)) concentrations were measured at site D during all sampling stages. Overall, atmospheric sampling results indicate that over 50% of PCDD/Fs distributed in solid phase, while over 90% of PCBs distributed in vapor phase during all sampling stages.     

Polycyclic aromatic hydrocarbons in rainwater

A study on polycyclic aromatic hydrocarbons (PAH) in rainwater was carried out over periods of at least one year at two locations in Belgium with different environments. Maximum amounts and concentrations of PAH were reached during the winter period, mainly due to household heating. This influence was also reflected in the relative proportions of the six PAH measured. For comparable periods the amounts of PAH were almost proportional to the quantity of rain. With the available data, and as most of the rainfall is coming from the same wind direction, it was not feasible to localise possible pollution sources based upon the distribution of PAH over the different windsectors, nor was it possible to identify significant differences for both sampling locations due to their different surroundings, indicating that the rain plays an important role in the transport and distribution of pollutants over a wide area. Between about 100 and 170 μg m^?2 PAH and 2 to 3 g m^?2 of total organic carbon content (TOC) per year were deposited by the rain, corresponding to mean concentrations of about 140 ng l^?1 of PAH and 2.9 mg l^?1 of TOC.     

Comparison of atmospheric pesticide concentrations measured at three sampling sites: local, regional and long-range transport

We present the comparison of atmospheric concentrations of eleven currently used pesticides (HCB, alpha-HCH, gamma-HCH, trifluraline, mecoprop, phosalone, atrazine, carbofuran, carbaryl, diuron and isoproturon) measured in remote (Aubure), rural (Colmar) and urban (Strasbourg) areas of Alsace and Vosges regions (cast France). Pesticides samples were collected simultaneously on two of the three sites during the summer season of 1993 and 1994, using a Hi-Vol sampler with Whatman filter paper and XAD-2 resin. The particle and gas phases were collected separately during 24 h. The relative importance of local emissions and local, regional and long distance transport on the contamination of the atmosphere in the three environments (remote, rural and urban) were investigated. To facilitate the interpretation of the results, the alpha/gamma-HCH ratio was used as a tracer of pesticide emissions.     

Mass balance for polycyclic aromatic hydrocarbons in the urban watershed of Le Havre (France): transport and fate of PAHs from the atmosphere to the outlet

Polycyclic aromatic hydrocarbon (PAH) concentrations have been monitored simultaneously in ambient air, bulk atmospheric deposition and runoff waters during one year in a small urban watershed of the Seine river basin (France). PAH fluxes from the atmosphere to the outlet of the urban watershed have been calculated to establish a mass balance for PAHs. PAH flux in runoff waters was 5.2 kg km(-2) yr(-1) while PAH atmospheric deposition was 0.21 kg km(-2) yr(-1). The comparison between atmospheric input and output by runoff has shown the importance of street deposits that appeared to be the most important source of PAHs for surface waters in urban areas. PAH profiles in the various compartments showed the fate of PAHs in the air-water system: proportion of carcinogenic PAHs was more important in runoff waters (35%) than in bulk atmospheric deposition (22%) and air (6%).     

Fine-scale mercury trends in temperate deciduous tree leaves from Ontario, Canada

This study focused on the value of deciduous leaves as biomonitors of total mercury (THg). Leaf samples were collected from a range of deciduous species from five sampling sites in the province of Ontario, Canada. These included a site in the northwest (the Experimental Lakes Area, ELA), two sites in central Ontario (the town of Dorset and the Centre for Atmospheric Research Experiments, CARE), and two sites in the southeast (Sandbanks Provincial Park, SBPP and the City of Kingston). The sampled species exhibited distinctive species-specific differences with red oaks consistently having lower leaf THg concentrations than all maple species, while black and white ash leaves had the highest concentrations. Spatially, leaves collected across the distance between ELA and SBPP (~ 1500 km apart) had overlapping THg concentrations between 20 and 40 ng/g. Unexpectedly, leaves from urban parks of Kingston had considerably lower THg concentrations (< 25 ng/g) than the other sites, which suggested leaves may not reflect subtle gradients of atmospheric THg found under field conditions. Leaf THg increased with the growing season, with highest THg leaf concentrations found in early autumn just before senescence. Microspatial differences within a single tree, such as position on a branch are important, since higher mercury concentrations were associated with leaves positioned deeper into the canopy relative to outer leaves more exposed to wind turbulence and sunlight. Within any single leaf, THg concentrations were highest in the leaf tissue, and consistently distributed, while the vein and petiole tissue had lower THg concentrations. There was no relationship between THg concentrations and leaf area. Using deciduous tree leaves as regional temporal monitors of bioavailable mercury may be feasible, but careful selection of leaf sampling sites on the tree itself and the timing is of utmost importance for ensuring consistent and high quality biomonitoring data.     

PAH content, toxicity and genotoxicity of coastal marine sediments from the Rovinj area, Northern Adriatic, Croatia

Surface marine sediments collected from 8 sampling sites within the Rovinj coastal area, Northern Adriatic, Croatia, were used for determining priority pollutant polycyclic aromatic hydrocarbons (PAHs) and toxic/genotoxic potential of sediment organic extracts. Total PAH concentrations ranged from 32 microg/kg (protected area) to 13.2 mg/kg dry weight (harbor) and showed clear differences between pristine, urban industrial and harbor areas. PAHs distribution revealed their pyrogenic origin with some biogenic influence in harbor. At all sampling sites sediment extracts showed toxic potential that was consistent with the sediment type. No correlation between toxicity measured by Microtox assay and concentrations of individual or total PAHs was found. Noncytotoxic dose of sediment extracts showed no genotoxic potential in bacterial umu-test. DNA damage is positively related to total PAHs at 4 sampling sites (S-1, S-2, S-3, S-6), but the highest DNA damage was not observed at the site with the highest total sediment PAH content (S-5).     

Trends in alkanes and PAHs in airborne particulate matter from Oporto and Vienna: identification and comparison.

A total of 56 weekly samples from Oporto and 40 from Vienna were collected and analysed, for 23 n-alkanes and 16 polycyclic aromatic hydrocarbons (PAH), by GC-MS after extraction with a toluene/methanol mixture. Total and elemental carbon were in the same range of values for both sampling sites. Although parts of the spectrum of species in both sampling sites were constant over the sampling period there is no evidence for suggesting a universal tracer for alkanes. For both Oporto and Vienna, Principal Component Analysis (PCA) identified two PCs and in both cases the first PC contained only PAHs while the second contained only alkanes. This separation between alkanes and PAHs and the observation of an alternating pattern with higher concentrations of odd carbon numbered from C27 to C30 (natural emissions from biological origin) is believed to result from a separation between anthropogenic and biological contributions associated with the first and second PC, respectively.     

Seasonal variation, sources and gas/particle partitioning of polycyclic aromatic hydrocarbons in Guangzhou, China

Air samples were collected weekly at an urban site and a suburban site in Guangzhou City, China, from April 2005 to March 2006, to measure the concentrations of polycyclic aromatic hydrocarbons (PAHs) in the ambient air and study their seasonal variations, gas/particle partitioning, origins and sources. The concentrations of ∑ 16-PAHs (particle + gas) were 129.9 ± 73.1 ng m⁻³ at the urban site and 120.4 ± 48.5 ng m⁻³ at the suburban site, respectively. It was found that there was no significant difference in PAH concentrations between the urban and suburban sites. Seasonal variations of PAH concentrations at the two sampling sites were similar, with higher levels in the winter that gradually decreased to the lowest levels in the summer. The average concentrations of ∑ 16-PAHs in the winter samples were approximately three times higher than those of the summer samples because in the summer local emissions dominated, and in the winter the contribution from outside sources or transported PAHs is increased. The plot of logK_p versus logP_L⁰ for the data sets of summer and winter season samples had significantly different slopes at both sampling sites. The slopes for the winter samples were steeper than those for the summer samples. It was also observed that gas/particle partitioning of PAHs showed different characteristics depending on air parcel trajectories. Steeper slopes were obtained for an air parcel that traveled across the continent to the sampling site from the northern or northeastern sector, whereas shallower slopes were obtained for air masses that traveled across the sea from the southern or eastern sector. Diagnostic ratio analytical results imply that the origins of PAHs were mainly from petroleum combustion and coal/biomass burning. The anthracene/phenanthrene and benzo[a]anthracene/chrysene ratios in the winter were significantly lower than those in the summer, which indicate that there might be long-range transported PAH input to Guangzhou in the winter.     

Spatial and temporal distribution of polycyclic aromatic hydrocarbons (PAHs) in the atmosphere of Xiamen, China

An intensive sampling program was conducted from October 2008 to September 2009 at the five different environmental sites in Xiamen, Fujian Province, to study the spatial and temporal characteristics of Polycyclic Aromatic Hydrocarbons (PAHs) in the gaseous and particulate phase, respectively. The PAHs concentrations at different sites were quite distinct during four seasons. The average concentrations of PAHs in winter were about 8.4 times higher than those in spring, and the concentrations of background were 0.56 times lower than those of industrial area. In addition, the higher temperature in summer affected the particle/gas partitioning of PAHs and led to the higher concentrations of gaseous PAHs. Diagnostic ratios of PAHs, which were employed to indicate the primary sources of PAHs in Xiamen, showed that the traffic vehicle exhaust was the largest contributor and the primary source for PAHs in Xiamen, especially in urban area; while the stationary combustion processes, such as petrochemical factories and power plants, were mainly responsible for PAHs sources in the industrial areas. The health risk of PAHs in the particulate phase was higher than those of the gaseous phase at the five sampling sites. The average toxic equivalent (BaP_eq) of the benzo[a]pyrene values for PAHs were 0.14, 0.32, 1.38 and 3.59 ng m^− 3 in spring, summer, autumn and winter, respectively. Furthermore, the results of average BaP_eq in all four seasons indicated that the health risks of particulate PAHs were higher than those of the gaseous PAHs at different sampling sites.     

Identifying contemporary and historic sources of soil polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans in an industrial urban setting

A study of soil polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans (PCDD/F) concentrations was undertaken in the vicinity of a municipal solid waste incinerator (MSWI) in Newcastle upon Tyne as a result of concerns raised by local residents about potential contamination from fugitive and stack emissions. The study area was divided into four sectors (north-east (NE), south-east (SE), north-west (NW) and south-west (SW)) around the MSWI, and sampling sites were located up to a distance of 2.25 km. Based on air dispersion modelling, the sampling density was four times greater in the NE (downwind) sector compared to the SW (upwind) direction, and twice as great in the NW and SE sectors. PCDD/F concentrations found in soil samples ranged from 6 to 1911 ng I-TEQ/kg DW with a median of 32 ng I-TEQ/kg DW. There was no evidence of elevated concentrations downwind of the MSWI compared to other directions, nor of any trend in concentration at increasing distance from the MSWI. We concluded, therefore, that the MSWI fugitive and stack emissions were not a major source of PCDD/F contamination. Analysis of PCDD/F homologue profiles showed that samples exhibiting furan-dominated and OCDD-dominated profiles and a profile characteristic of the MSWI ash occurred in distinct clusters. Those samples showing the furan-dominated profile had the largest PCDD/F concentrations measured as I-TEQ, followed by samples with the incinerator profile, the deposition profile, and the OCDD-dominated profile. We identified some contamination hotspots located in the SW and SE sampling sectors (upwind of the MSWI), and potential sources for these hotspots were sought by using historic land use data from maps of the locality dating back to 1856. We concluded that the cluster of very high concentrations of PCDD/F in soils showing the furan homologue profile were most likely to have resulted from the disposal of graphite electrode sludges from brine electrolysis carried out at a chemical works between the 1890s and the 1930s.     

Atmospheric deposition of polycyclic aromatic hydrocarbons (PAHs) in mosses (Hypnum cupressiforme) in Hungary

The atmospheric deposition of polycyclic aromatic hydrocarbons (PAHs) was investigated in Hungary by analyzing a moss (Hypnum cupressiforme) species as a bioindicator. In the autumn of 1997, samples were collected at 29 sites distributed across Hungary. The concentrations of total PAH at these sites were in the range of 0.1567-10.45 x 10(4) microg kg(-1) with a mean value of 1.87 x 10(4) microg kg(-1). More than 99% of the total PAHs atmospheric deposition were low molecular weight PAHs (up to 3 ring compounds). The total PAH values showed no correlation with metal concentrations. However, most of the sites in this region showed a positive linear relationship between PAHs levels and traffic volume (r2 = 0.83; P < 0.001) while no relationship existed between PAH levels and population (r2 = 0.01; P > 0.1). Atmospheric deposition of PAHs at different regions in Hungary may be due to incomplete combustion of fuel. The total concentrations of PAHs were compared to the PAH levels in vegetation samples collected from different regions around the world. The highest PAHs concentrations accumulated were found in Hypnum cupressiforme than other vegetation species. A greater affinity for PAH compounds by Hypnum cupressiforme than other moss species probably caused larger amounts of accumulation. A relationship between accumulations of PAH compounds in Hypnum cupressiforme and octanol-air partition coefficients was obtained and is briefly discussed.     

Characterization of atmospheric concentrations and partitioning of PAHs in the Chicago atmosphere

An intensive sampling program has been undertaken in the absence of precipitation at an urban site, Chicago, to characterize the atmospheric concentration and partitioning of PAHs. Two different sampling programs have been carried out with a large number of samples. Measured ambient concentrations of PAHs were classified as Land and Lake samples based on wind direction and back trajectory calculations. Differences in ambient concentrations of PAHs were observed between Land and Lake samples. The concentrations of PAHs when air originated over the Land were approximately two-four times higher than the concentrations measured when air originated over the Lake. It has been demonstrated that partitioning of PAHs shows a consistent difference between samples taken when wind came from off the land rather than off the water. This was most evident by more shallow slopes for Lake samples compared to the slopes for Land samples, when partition coefficient (K(p)) is plotted on a log-log scale vs. the subcooled liquid vapor pressure (P(L)(0)). Experimentally, determined K(p) values were compared with the results obtained using two different models, one based on absorption into aerosol organic matter and the other adsorption onto soot carbon. Experimental K(p) values generally agreed well with the soot+octanol based model predictions.     

Comparison of MTBE concentrations in groundwater of urban and nonurban areas in Germany

The occurrence of the gasoline oxygenate methyl tert-butyl ether (MTBE) in groundwater samples from known fuel-contaminated sites (n=29 samples), nonurban (n=74) and urban sites (n=67) in Germany was investigated. The analyses revealed detection frequencies of 58% (contaminated sites), 24% (nonurban sites) and 63% (urban sites) at a detection limit of 0.01 microgL(-1). Median (maximum) MTBE concentrations were calculated for nonurban and urban samples as 0.18 microgL(-1) (2.2 microgL(-1)) and 0.06 microgL(-1) (48 microg L(-1)). The data from nonurban samples revealed MTBE detections mainly at public supply wells with higher pumping rates than monitoring wells. MTBE was more frequently detected in urban samples, most probably due to the higher atmospheric input and direct liquid emissions from motorways or gas stations. Higher concentrations above 1.0 microgL(-1) in urban areas were found in wells located at industrial sites, where also a MTBE plume was accidentally detected during the study. The prevalence of MTBE in shallow aquifers was comparable to those in the USA.