Development of a Long-Term Sampling Method for Determination of NMHCs in Indoor Air.

Vapor intrusion is detrimental for indoor air quality. One of the most common sources of vapor intrusion is soil contaminated with petroleum hydrocarbons. To evaluate the long-term risk from individual exposure to hydrocarbons it is necessary to measure quantitively and reliably an average concentration level of individual pollutants on a monthly or yearly basis. Temporal variability of vapor intrusion from hydrocarbons poses a significant challenge to determination of average exposure and there is a need for reliable long-term integrative sampling. To this end, an analytical method for determination of 10 selected nonmethane hydrocarbons (NMHCs), including hexane, heptane, octane, decane, benzene, toluene, ethyl-benzene, m,p-xylene, o-xylene, and naphthalene, sampled on active triple-bed tubes filled with Carbograph 2, Carbograph 1, and Carboxen 1003 adsorbents was developed and validated. Extensive laboratory studies proved the absence of breakthrough at 50% HR and ambient temperature for experiments lasting up to 28 days and established a safe sampling time/volume of 20 days/114 L when sampling at a low flow rate of around 4 mL min-1. In addition, the developed method includes detailed uncertainty calculations for determination of concentrations. Finally, the method was tested by measuring NMHC concentrations in indoor air at a former industrial site during a 2-month-long field campaign in Lyon. The results of the field campaign suggest that 4-week integrated concentration measurements can be achieved by using active sampling on triple-bed tubes at 4.5 mL min-1.

Variability and sources of non-methane hydrocarbons at a Mediterranean urban atmosphere: The role of biomass burning and traffic emissions.

Levels and sources of non-Methane Hydrocarbons (NMHCs) were investigated at the urban background Thissio station, close to the historical center of Athens (Greece) from March 2016 to February 2017 (12 months), by means of an automated GC-FID. Alkanes dominated over aromatics and alkenes, with hourly mean levels ranging from detection limit up to 60 µg m-3 for i-pentane and 90 µg m-3 for toluene. Higher levels were recorded in the cold period relative to the warmer one. In addition, NMHCs seasonal diurnal cycles were characterized by a bimodal pattern, following the trend of tracers of anthropogenic sources. The Positive Matrix Factorization (PMF) was used for the allocation of NMHC to their sources. Five factors were identified and quantified, with traffic-related sources being the main one contributing up to 60% to total NMHCs, while biomass burning contributes up to 19%. A supplementary PMF assimilation was applied on a seasonal basis further including α-pinene, C6-C16 alkanes and aromatics. This PMF resulted to a seven-factor solution that allowed the examination of two additional sources, in addition to five already identified, highlighting the main contribution of anthropogenic sources (70%) to α-pinene.

A review of environmental odor quantification and qualification methods: The question of objectivity in sensory analysis.

For several years, various issues have up surged linked to odor nuisances with impacts on health and economic concerns. As awareness grew, recent development in instrumental techniques and sensorial analysis have emerged offering efficient and complementary approaches regarding environmental odor monitoring and control. While chemical analysis faces several obstacles, the sensory approach can help overcome them. Therefore, this latter may be considered as subjective, putting the reliability of the studies at risk. This paper is a review of the most commonly sensory methodology used for quantitative and qualitative environmental assessment of odor intensity (OI), odor concentration (OC), odor nature (ON) and hedonic tone (HT). For each of these odor dimensions, the assessment techniques are presented and compared: panel characteristics are discussed; laboratory and field studies are considered and the objectivity of the results is debated. For odor quantification, the use of a reference scale for OI assessment offers less subjectivity than other techniques but at the expense of ease-of-use. For OC assessment, the use of dynamic olfactometry was shown to be the least biased. For odor qualification, the ON description was less subjective when a reference-based lexicon was used but at the expense of simplicity, cost, and lesser panel-training requirements. Only when assessing HT was subjectivity an accepted feature because it reflects the impacted communities' acceptance of odorous emissions. For all discussed dimensions, field studies were shown to be the least biased due to the absence of air sampling, except for OC, where the dispersion modeling approach also showed great potential. In conclusion, this paper offers the reader a guide for environmental odor sensory analysis with the capacity to choose among different methods depending on the study nature, expectations, and capacities.

Objective odor analysis of incidentally emitted compounds using the Langage des Nez® method: application to the industrial zone of Le Havre.

Environmental odor studies are usually done using two approaches: nuisance impact assessment and source identification. The latter may be done using chemical analysis or sensory analysis. While sensory analyses offer many advantages, they also face the main obstacle: odor nature description still uses conventional methods based on subjective evocations as odor descriptors. This makes the sensory method ineffective especially when the expected outcome is the source identification in the context of an industrial accident. This work wants to fulfill this gap proposing to build an objective database including the odor nature description of selected potentially emitted compounds using a promising approach: the Langage des Nez® (LdN). Using definite odorous compounds as odor referents, this work provides the odor nature description of 44 compounds, reported as potential incidentally released chemical compounds in the industrial zone of Le Havre. The city of Le Havre, France, was chosen as a model due to a history of odorous emissions of industrial origins. A trained panel described the odor of each compound using up to three referents of the LdN referents collection and attributed a score to each referent. A data analysis method was developed based on the frequency of citation of the referents and the attributed scores allowing the categorization of each compound in three types of consensus categories. The data analysis results showed that around 80% of compounds were described with a good consensus, showing the LdN as a well-adapted lexicon. This study does not point to any correlation between the chemical structures of the compounds of interest and their relative referents. When compared to conventional methods, LdN revealed a more objective and precise approach. The proposed experimental method and the results provided in this work offer the first insight for time-efficient approaches to objectively describe environmental odors, especially potentially emitted odors during incidents. This work may be supplemented by abatement and mixture effect investigations for a complete understanding of odor dispersion.

Atmospheric reactivity of biogenic volatile organic compounds in a maritime pine forest during the LANDEX episode 1 field campaign.

Trace gas measurements were performed during the LANDEX (the LANDes EXperiment) Episode 1 field campaign in the summer 2017, in one of the largest European maritime pine forests (> 95% Pinus pinaster) located in southwestern France. Efforts have been focused on obtaining a good speciation of 20 major biogenic volatile organic compounds (BVOCs, including pinenes, carenes, terpinenes, linalool, camphene, etc.). This was made possible by the development of a new and specific chromatographic method. In order to assess the role of BVOCs in the local gas phase chemistry budget, their reactivity with the main atmospheric oxidants (hydroxyl radicals (OH), ozone (O3) and nitrate radicals (NO3)) and the corresponding consumption rates were determined. When considering the OH reactivity with BVOCs, isoprene and linalool accounted for 10-47% of the OH depletion during daytime, and monoterpenes for 50-65%, whereas monoterpenes were the main contributors during the night (70-85%). Sesquiterpenes and monoterpenes were the main contributors to the ozone reactivity, especially ß-caryophyllene (30-70%), with a maximum contribution during nighttime. Nighttime nitrate reactivity was predominantly due to monoterpenes (i.e. 90-95%). Five specific groups have been proposed to classify the 19 BVOCs measured in the forest, according to their reactivity with atmospheric oxidants and their concentrations. The total amount of BVOCs consumed under and above the forest canopy was evaluated for 7 BVOCs (i.e. isoprene, α-pinene, ß-pinene, myrcene, limonene + cis-ocimene and Δ3-carene). The reactivity of atmospheric oxidants and BVOCs at a local level are discussed in order to highlight the compounds (BVOCs, other VOCs), the atmospheric oxidants and the main associated reactive processes observed under the canopy of a maritime pine forest.

Comparison of the chemical composition of aerosols from heated tobacco products, electronic cigarettes and tobacco cigarettes and their toxic impacts on the human bronchial epithelial BEAS-2B cells.

The electronic cigarettes (e-cigs) and more recently the heated tobacco products (HTP) provide alternatives for smokers as they are generally perceived to be less harmful than conventional cigarettes. However, it is crucial to compare the health risks of these different emergent devices, in order to determine which product should be preferred to substitute cigarette. The present study aimed to compare the composition of emissions from HTP, e-cigs and conventional cigarettes, regarding selected harmful or potentially harmful compounds, and their toxic impacts on the human bronchial epithelial BEAS-2B cells. The HTP emitted less polycyclic aromatic hydrocarbons and carbonyls than the conventional cigarette. However, amounts of these compounds in HTP aerosols were still higher than in e-cig vapours. Concordantly, HTP aerosol showed reduced cytotoxicity compared to cigarette smoke but higher than e-cig vapours. HTP and e-cig had the potential to increase oxidative stress and inflammatory response, in a manner similar to that of cigarette smoke, but after more intensive exposures. In addition, increasing e-cig power impacted levels of certain toxic compounds and related oxidative stress. This study provides important data necessary for risk assessment by demonstrating that HTP might be less harmful than tobacco cigarette but considerably more harmful than e-cig.

Investigation on the near-field evolution of industrial plumes from metalworking activities.

In a context where a significant fraction of the population lives near industrial areas, the main objectives of this study are to provide (a) new data on PM2.5 chemical compositions, heavy-metal concentrations and trace gases released by metalworking activities and (b) new information on the near-field evolution (up to about a thousand meters) of such industrial plumes in terms of particle chemical composition and size distribution. For that purpose, a one-month field campaign was performed in an industrial area near the city of Dunkirk (Northern France), combining measurements of atmospheric dynamics and physico-chemical characterization of air masses. Comparisons between several elemental ratios (mainly Mn/Fe), particle size distributions and volatile organic compound (VOC) concentrations at the stacks and at a near-field site suggest that plumes of a ferromanganese alloy plant were quickly mixed with pollutants emitted by other sources (mainly other industries, possibly traffic and sea spray), in particular a neighboring steelworks, before reaching the sampling site. This led to the emergence of secondary particles related to condensation and/or aggregation phenomena inside the plumes. Metalworking emissions were also identified as a source of new particle formation, formed through the emission of gaseous precursors and their fast transformation and condensation, over a timescale of minutes before reaching the near-field site 800 m downwind. Ultrafine particles emitted at the stacks also quickly agglomerated to form larger particles before reaching the near-field site. These results show that, even over short distances, the chemical composition and size distribution of metalworking plumes may evolve rapidly and the characteristics of particles at the boundary of an industrial area (especially in contiguous urban areas) may differ from those emitted directly at the stacks.

Influence of puffing conditions on the carbonyl composition of e-cigarette aerosols.

Owing to their harmful effects on human health, the presence of carbonyl compounds in e-cigarette aerosols raises concerns. To date, the reported concentration levels in e-vapors vary greatly between studies and several factors that markedly influence carbonyl emission during vaping have been highlighted including the heating temperature, the power supply, the device architecture, the filling level of the tank and the main e-liquid constituents. This study investigated the impact of puffing regimen parameters on the carbonyl composition of e-cigarette aerosols with the aim of: (1) better estimating the variability of carbonyl emissions depending on puffing conditions; (2) highlighting puffing profiles that increase the exposure to carbonyls; and (3) estimating to what extent puffing topography could be implied in the variability of carbonyl concentrations reported in the current literature. E-vapors from a single e-liquid were generated from two e-cigarette models with a smoking machine. A total of 7 different puffing regimens were used to individually study the influence of the puff volume, duration and frequency. Carbonyls were collected by DNPH cartridges and analysed by HPLC-UV. E-liquid consumption and e-vapor temperature were also monitored. E-vapor concentrations of formaldehyde, acetaldehyde, acetone, acrolein, propionaldehyde and methylglyoxal were affected, sometimes differently, by the modification of the puffing regimen, as well as by the e-cigarette model. For example, formaldehyde concentration ranged from 20 to 255 ng/puff depending on the puffing conditions. The results of principal component analyses, applied to the concentration data sets for the 6 carbonyls, suggest that the studied parameters interact and highlight some "carbonyl-emitting" combinations of concern (e-cigarette model/puffing regimen). However, the highest concentrations measured in the present study remain far lower than those observed in conventional cigarette mainstream smoke. This study confirms that the chosen puffing regimen contributes a part of the observed variability in the carbonyl levels reported in the scientific literature, hampering comparisons between studies and making interpretation difficult. Thus, harmonized and realistic protocols for the assessment of e-cigarette toxicity by physicochemical or experimental approaches are clearly needed.

Impact of material emissions and sorption of volatile organic compounds on indoor air quality in a low energy building: Field measurements and modeling.

The assessment of VOC emission rates and sorption coefficients was performed for ten surfaces present within a classroom, using field and laboratory emission cells (FLEC) coupled to online and off-line VOC quantification techniques. A total of 21 identified VOCs were emitted by the different surfaces. VOC emission rates measured using PTR-ToF-MS were compared to gas chromatographic measurements. The results showed that the two methods are complementary to one another. Sorption parameters were also successfully measured for a mixture of 14 VOCs within a few hours (<17 hours per surface). A study of the spatial and temporal variability of the measured parameters was also carried out on the two surfaces that presented the most potential for interaction with VOCs, accounting for the largest surface areas within the room. The dataset of emission rates and sorption parameters was used in the INCA-Indoor model to predict indoor air concentrations of VOCs that are compared to experimental values measured in the room. Modeling results showed that sorption processes had a limited effect on indoor concentrations of VOCs for these field campaigns. Modeled daily profiles show good agreement with the experimental observations for VOCs such as toluene (indoor source) and xylenes (outdoor source) but underestimate concentrations of methanol (both indoor and outdoor sources).

Bidirectional Ecosystem-Atmosphere Fluxes of Volatile Organic Compounds Across the Mass Spectrum: How Many Matter?

Terrestrial ecosystems are simultaneously the largest source and a major sink of volatile organic compounds (VOCs) to the global atmosphere, and these two-way fluxes are an important source of uncertainty in current models. Here, we apply high-resolution mass spectrometry (proton transfer reaction-quadrupole interface time-of-flight; PTR-QiTOF) to measure ecosystem-atmosphere VOC fluxes across the entire detected mass range (m/z 0-335) over a mixed temperate forest and use the results to test how well a state-of-science chemical transport model (GEOS-Chem CTM) is able to represent the observed reactive carbon exchange. We show that ambient humidity fluctuations can give rise to spurious VOC fluxes with PTR-based techniques and present a method to screen for such effects. After doing so, 377 of the 636 detected ions exhibited detectable gross fluxes during the study, implying a large number of species with active ecosystem-atmosphere exchange. We introduce the reactivity flux as a measure of how Earth-atmosphere fluxes influence ambient OH reactivity and show that the upward total VOC (∑VOC) carbon and reactivity fluxes are carried by a far smaller number of species than the downward fluxes. The model underpredicts the ∑VOC carbon and reactivity fluxes by 40-60% on average. However, the observed net fluxes are dominated (90% on a carbon basis, 95% on a reactivity basis) by known VOCs explicitly included in the CTM. As a result, the largest CTM uncertainties in simulating VOC carbon and reactivity exchange for this environment are associated with known rather than unrepresented species. This conclusion pertains to the set of species detectable by PTR-TOF techniques, which likely represents the majority in terms of carbon mass and OH reactivity, but not necessarily in terms of aerosol formation potential. In the case of oxygenated VOCs, the model severely underpredicts the gross fluxes and the net exchange. Here, unrepresented VOCs play a larger role, accounting for ~30% of the carbon flux and ~50% of the reactivity flux. The resulting CTM biases, however, are still smaller than those that arise from uncertainties for known and represented compounds.

Quantitative cancer risk assessment and local mortality burden for ambient air pollution in an eastern Mediterranean City.

Health risks posed by ambient air pollutants to the urban Lebanese population have not been well characterized. The aim of this study is to assess cancer risk and mortality burden of non-methane hydrocarbons (NMHCs) and particulates (PM) based on two field-sampling campaigns conducted during summer and winter seasons in Beirut. Seventy NMHCs were analyzed by TD-GC-FID. PM2.5 elemental carbon (EC) components were examined using a Lab OC-EC aerosol Analyzer, and polycyclic aromatic hydrocarbons were analyzed by GC-MS. The US EPA fraction-based approach was used to assess non-cancer hazard and cancer risk for the hydrocarbon mixture, and the UK Committee on Medical Effects of Air Pollutants (COMEAP) guidelines were followed to determine the PM2.5 attributable mortality burden. The average cumulative cancer risk exceeded the US EPA acceptable level (10-6) by 40-fold in the summer and 30-fold in the winter. Benzene was found to be the highest contributor to cancer risk (39-43%), followed by 1,3-butadiene (25-29%), both originating from traffic gasoline evaporation and combustion. The EC attributable average mortality fraction was 7.8-10%, while the average attributable number of deaths (AD) and years of life lost (YLL) were found to be 257-327 and 3086-3923, respectively. Our findings provide a baseline for future air monitoring programs, and for interventions aiming at reducing cancer risk in this population.

Data on comparison between FLEC and CLIMPAQ methods used for fast sorption measurements of VOCs on building materials.

A test emission chamber called CLIMPAQ has been coupled to a chromatography analyzer GC to measure volatile organic compounds (VOC) concentration during a sorption experiments (Fast sorption measurements of VOCs on building materials: Part 2 - Comparison between FLEC and CLIMPAQ methods, (Rizk et al., In press) [1]). The equations used to calculate the mass transfer coefficient and the thickness of the boundary layer developed on the surface of a material are presented. In addition, the experimental profiles obtained using the CLIMPAQ chamber is also presented in the presence and the absence of a building material. Finally, the impact of chamber size on the obtained concentration profile using different chambers is shown using 3 types of chambers having different volumes, 1 m(3), 30 m(3) and a micro chamber of 40 mL.

Improvement in 8h-sampling rate assessment considering meteorological parameters variability for biogas VOC passive measurements in the surroundings of a French landfill.

Passive sampling technology has been extensively used for long-term VOC atmospheric concentrations' monitoring. Its performances regarding the short-term measurements and related to VOC from biogas were evaluated in this work: laboratory scale experiments have been conducted in order to check the suitability of Radiello® diffusive samplers for the assessment of 8 h-VOC levels in highly changeable meteorological conditions; in a second step a short pilot field campaign was implemented in the vicinity of a West-French landfill. First of all, it was assessed that amongst a diversified list of 16 characteristic compounds from biogas, mercaptans, some halogenated, oxygenated compounds and terpenes could not be measured accurately by this passive technique either because they are not captured by the sorbent or they are not quantitatively desorbed in the chosen mediated analytical conditions. Moreover, it has been confirmed that sampling rates (SR) related to isopentane, THF, cyclohexane, toluene, p-xylene and n-decane are influenced by environmental factors: the main influence concerns the wind speed. From 2 m s(-1), when the velocity increases by 1 m s(-1), the SR increases from 12 to 32% depending on the COV (considering a linear dependence between 2 and 7 m s(-1)). Humidity has no effect on SR, and temperature influence is rather limited to less than 3% per degree. A comprehensive uncertainty estimation, including uncertainties linked to meteorological changes, has led to global relative uncertainties comprising between 18% and 54% from one VOC to another: a quite high value comparatively to those obtained without considering meteorological condition influences. To illustrate our results, targeted VOC were quantified in the field, on a single day: concentrations range between LD to 3 µg m(-3): relatively very low concentrations compared to those usually reported by literature.

Chemical Mass Balance model applied to an olfactory annoyance problematic.

The source-receptor model Chemical Mass Balance (CMB) generally used for air pollution studies is applied for the first time to solve odor signature issues. The olfactory annoyance perceived at a receptor site of an industrial area is often the result of a combination of different smells, emitted by several industrial sources. To identify the main responsible for the inconvenience, tools able to determine the contribution of each source directly where the annoyance is perceived are widely developed and deployed. In this work, the contributions of volatile organic compound sources, coming from a waste treatment plant containing three potential sources of olfactory annoyance (waste storage, production of biogas, and compost piles of green wastes) and perceived at a village located downwind, are studied by chemical analyses. The CMB methodology is applied, and the results are finally compared to olfactometric methods, in order to validate whether it is appropriate to use this model for olfactory problems or not.

Determination of 14 amines in air samples using midget impingers sampling followed by analysis with ion chromatography in tandem with mass spectrometry.

An Ion Chromatography-Mass Spectrometry (IC-MS) method was developed for the simultaneous quantification of 14 volatile amines in air. The method includes collection of compounds into two midget impingers in a row filled with 15 ml of ultrapure water. The analytical performances with mass spectrometry detection were compared to those obtained with classical conductivity detection. The use of mass spectrometry detection (in SIM mode) overcomes most of the coelutions encountered with conductivity detection. Although the linearity domain of calibrations is reduced for the MS detection as compared with the CD detection, the detection limits in MS detection are highly lowered allowing the quantification of amines at the levels of µg m(-3) in air with a good accuracy for most compounds (RSD of less than 10%). This method was successfully applied to the analysis of amines released from polyurethane foams. Seven amines were identified and some in high concentrations, like dimethylaminoethanol, NIAX and TEDA.

Development of a methodology examining the behaviours of VOCs source apportionment with micro-meteorology analysis in an urban and industrial area.

During summer 2009, online measurements of 25 Volatile Organic Compounds (VOCs) from C6 to C10 as well as micro-meteorological parameters were simultaneously performed in the industrial city of Dunkerque. With the obtained data set, we developed a methodology to examine how the contributions of different source categories depend on atmospheric turbulences, and the results provided identification of emission modes. Eight factors were resolved by using Positive Matrix Factorization model and three of them were associated with mixed sources. The observed behaviours of contributions with turbulences lead to attribute some factors with sources at ground level, and some other factors with sources in the upper part of surface layer. The impact of vertical turbulence on the pollutant dispersion is also affected by the distance between sources and receptor site.

Development of a sampling method for the simultaneous monitoring of straight-chain alkanes, straight-chain saturated carbonyl compounds and monoterpenes in remote areas.

Studies have shown that biogenic compounds, long chain secondary compounds and long lifetime anthropogenic compounds are involved in the formation of organic aerosols in both polluted areas and remote places. This work aims at developing an active sampling method to monitor these compounds (i.e. 6 straight-chain saturated aldehydes from C6 to C11; 8 straight-chain alkanes from C9 to C16; 6 monoterpenes: α-pinene, ß-pinene, camphene, limonene, α-terpinene, & γ-terpinene; and 5 aromatic compounds: toluene, ethylbenzene, meta-, para- and ortho-xylenes) in remote areas. Samples are collected onto multi-bed sorbent cartridges at 200 mL min(-1) flow rate, using the automatic sampler SyPAC (TERA-Environnement, Crolles, France). No breakthrough was observed for sampling volumes up to 120 L (standard mixture at ambient temperature, with a relative humidity of 75%). As ozone has been shown to alter the samples (losses of 90% of aldehydes and up to 95% of terpenes were observed), the addition of a conditioned manganese dioxide (MnO(2)) scrubber to the system has been validated (full recovery of the affected compounds for a standard mixture at 50% relative humidity--RH). Samples are first thermodesorbed and then analysed by GC/FID/MS. This method allows suitable detection limits (from 2 ppt for camphene to 13 ppt for octanal--36 L sampled), and reproducibility (from 1% for toluene to 22% for heptanal). It has been successfully used to determine the diurnal variation of the target compounds (six 3 h samples a day) during winter and summer measurement campaigns at a remote site in the south of France.

VOC in an urban and industrial harbor on the French North Sea coast during two contrasted meteorological situations.

Two measurement campaigns of volatile organic compounds (VOC) were carried out in the industrial city of Dunkerque, using Radiello passive samplers during winter (16-23 January) and summer (6-13 June) 2007. 174 compounds were identified belonging to six chemical families. Classifying sampling sites with similar chemical profiles by hierarchical ascending classification resulted in 4 groups that reflected the influence of the main industrial and urban sources of pollution. Also, the BTEX (Benzene, Toluene, Ethylbenzene and Xylenes) quantification allowed us to map their levels of concentration. Benzene and toluene (BT) showed high concentrations in Northern Dunkerque reflecting the influence of two industrial plants. Differences among spatial distributions of the BT concentrations over contrasted meteorological conditions were also observed. An atypical ratio of T/B in the summer samples led us to investigate the BTEX origins shedding light on the contribution of pollutants transported across various zones of VOC emissions situated in Europe.

Using a source-receptor approach to characterise VOC behaviour in a French urban area influenced by industrial emissions. Part II: source contribution assessment using the Chemical Mass Balance (CMB) model.

In Part I of this study (Badol C, Locoge N, Leonardis T, Gallo JC. Using a source-receptor approach to characterise VOC behaviour in a French urban area influenced by industrial emissions, Part I: Study area description, data set acquisition and qualitative data analysis of the data set. Sci Total Environ 2007; submitted as companion manuscript.) the study area, acquisition of the one-year data set and qualitative analysis of the data set have been described. In Part II a source profile has been established for each activity present in the study area: 6 profiles (urban heating, solvent use, natural gas leakage, biogenic emissions, gasoline evaporation and vehicle exhaust) have been extracted from literature to characterise urban sources, 7 industrial profiles have been established via canister sampling around industrial plants (hydrocarbon cracking, oil refinery, hydrocarbon storage, lubricant storage, lubricant refinery, surface treatment and metallurgy). The CMB model is briefly described and its implementation is discussed through the selection of source profiles and fitting species. Main results of CMB modellings for the Dunkerque area are presented. (1) The daily evolution of source contributions for the urban wind sector shows that the vehicle exhaust source contribution varies between 40 and 55% and its relative increase at traffic rush hours is hardly perceptible. (2) The relative contribution of vehicle exhaust varies from 55% in winter down to 30% in summer. This decrease is due to the increase of the relative contribution of hydrocarbon storage source reaching up to 20% in summer. (3) The evolution of source contributions with wind directions has confirmed that in urban wind sectors the contribution of vehicle exhaust dominate with around 45-55%. For the other wind sectors that include some industrial plants, the contribution of industrial sources is around 60% and could reach 80% for the sector 280-310 degrees , which corresponds to the most dense industrial area. (4) The pollution in Dunkerque has been globally characterised taking into account the frequency of wind directions and contributions of sources in each wind direction for the whole year. It has been concluded that contribution of industrial sources is below 20% whereas vehicle exhaust contribution is superior to 40%.