Source apportionment of PM(10) and PM(2.5) using positive matrix factorization and chemical mass balance in Izmir, Turkey

Atmospheric particulate matter (PM) fractions (PM(10) and PM(2.5)) were sampled concurrently between June 2004 and May 2005 at two sites (urban and suburban) in Izmir, Turkey. The elemental composition of PM (Al, Ba, Ca, Cd, Cr, Cu, Fe, K, Mg, Mn, Na, Ni, Pb, Sr, V, and Zn) was determined using inductively coupled plasma-optical emission spectrometer. Elemental compositions of several PM sources were also characterized. Positive matrix factorization (PMF) and chemical mass balance modeling (CMB) were applied to determine the PM sources and their contributions to air concentrations. The major contributors to PM were fossil fuel burning, traffic emissions, mineral industries and marine salt according to the PMF results. However, undetermined parts were more than 40%. On the other hand, the contributions to PM could be determined completely by CMB, and the dominant contributor was traffic with >70% at the two sites. Fossil fuel burning, mineral industries, marine salt and natural gas-fired power plant were the minor contributors.     

Approaching PM(2.5) and PM(2.5-10) source apportionment by mass balance analysis, principal component analysis and particle size distribution

A chemical characterization was carried out for PM(2.5) and PM(2.5-10) samples collected in a suburban area and the concentrations of 12 elements were determined in 8 size segregated fractions using a Berner Impactor. Two main objectives were proposed in this work: 1) to test for closure among chemical and gravimetric measurements of PM(2.5) and PM(2.5-10) and 2) evaluate the performance of Multilinear Regression Analysis (MLRA) and Mass Balance Analysis (MBA) in the determination of source contribution to Particulate Matter (PM) concentrations. The fraction unaccounted for by chemical analysis comprised on average 17% and 34% of gravimetric PM(2.5) and PM(2.5-10), respectively. The lack of closure in PM(2.5) and PM(2.5-10) mass (i.e., constituent concentrations not adding up to gravimetrically measured) could partly result from the presence of water associated with particles and errors in the estimation of unmeasured species. MLRA and MBA showed very similar results for the temporal variation of the source contributions. However, quantitatively important discrepancies could be observed, principally due to the lack of mass closure in PM(2.5) and PM(2.5-10). Both methods indicated that the major PM(2.5) aerosol mass contributors included secondary aerosol and vehicle exhaust. In the coarse fraction, marine and mineral aerosol contributions were predominant.     

Source apportionment of ambient fine particle size distribution using positive matrix factorization in Erfurt, Germany

Particle size distribution data collected between September 1997 and August 2001 in Erfurt, Germany were used to investigate the sources of ambient particulate matter by positive matrix factorization (PMF). A total of 29,313 hourly averaged particle size distribution measurements covering the size range of 0.01 to 3.0 microm were included in the analysis. The particle number concentrations (cm(-3)) for the 9 channels in the ultrafine range, and mass concentrations (ng m(-3)) for the 41 size bins in the accumulation mode and particle up to 3 microm in aerodynamic diameter were used in the PMF. The analysis was performed separately for each season. Additional analyses were performed including calculations of the correlations of factor contributions with gaseous pollutants (O(3), NO, NO(2), CO and SO(2)) and particle composition data (sulfate, organic carbon and elemental carbon), estimating the contributions of each factor to the total number and mass concentration, identifying the directional locations of the sources using the conditional probability function, and examining the diurnal patterns of factor scores. These results were used to assist in the interpretation of the factors. Five factors representing particles from airborne soil, ultrafine particles from local traffic, secondary aerosols from local fuel combustion, particles from remote traffic sources, and secondary aerosols from multiple sources were identified in all seasons.     

Source apportionment of PM10 and PM2.5 in Milan (Italy) using receptor modelling

In this paper a source apportionment of particulate matter pollution in the urban area of Milan (Italy) is given. Results of PM10 and PM2.5 mass and elemental concentrations from a 1-year monitoring campaign are presented. Mean annual and daily PM10 levels are compared with the limits of the EU Air Quality Directive EC/30/1999 and the results show that the limit values established would not be met in the urban area of Milan or the large surrounding area. Moreover, high levels of PM2.5 are registered and this fraction constitutes a high portion of the PM10 mass. In Milan the winter period is characterised by a high degree of air pollution due to a greater contribution of emissions and to adverse meteorological and thermodynamic conditions of the atmosphere. The application of multivariate techniques and receptor modelling (PCFA, APCFA) to the whole data-set led to the identification of the main emitting sources and to the source apportionment of PM10 and PM2.5 in Milan. The most important sources were identified as 'soil dust', 'traffic', 'industry' and 'secondary compounds' for PM10 and as 'soil dust', 'anthropogenic' and 'secondary compounds' for PM2.5, explaining the greatest part of the total variance (91% and 75%, respectively).     

PM10 source apportionment in Milan (Italy) using time-resolved data

In this work Positive Matrix Factorization (PMF) was applied to 4-hour resolved PM10 data collected in Milan (Italy) during summer and winter 2006. PM10 characterisation included elements (Mg-Pb), main inorganic ions (NH₄⁺, NO₃⁻, SO₄²⁻), levoglucosan and its isomers (mannosan and galactosan), and organic and elemental carbon (OC and EC).PMF resolved seven factors that were assigned to construction works, re-suspended dust, secondary sulphate, traffic, industry, secondary nitrate, and wood burning. Multi Linear Regression was applied to obtain the PM10 source apportionment. The 4-hour temporal resolution allowed the estimation of the factor contributions during peculiar episodes, which would have not been detected with the traditional 24-hour sampling strategy.     

Source apportionment of PM2.5 in Beijing using principal component analysis/absolute principal component scores and UNMIX

Source apportionment of fine particulate matter (PM2.5, i.e., particles with an aerodynamic diameter of 2.5 microm or less) in Beijing, China, was determined using two eigenvector models, principal component analysis/absolute principal component scores (PCA/APCS) and UNMIX. The data used in this study were from the chemical analysis of 24-h samples, which were collected at 6-day intervals in January, April, July, and October 2000 in the Beijing metropolitan area. Both models identified five sources of PM2.5: secondary sulfate and secondary nitrate, a mixed source of coal combustion and biomass burning, industrial emission, motor vehicles exhaust, and road dust. On average, the PCA/APCS and UNMIX models resolved 73% and 85% of the PM2.5 mass concentrations, respectively. The results were comparable to previous estimate using the positive matrix factorization (PMF) and chemical mass balance (CMB) receptor models. Secondary products and the emissions from coal combustion and biomass burning dominated PM2.5. Such comparison among various receptor models, which contain different physical constraints, is important for better understanding PM2.5 sources.     

Source apportionment of PM(2.5) and selected hazardous air pollutants in Seattle

The potential benefits of combining the speciated PM(2.5) and VOCs data in source apportionment analysis for identification of additional sources remain unclear. We analyzed the speciated PM(2.5) and VOCs data collected at the Beacon Hill in Seattle, WA between 2000 and 2004 with the Multilinear Engine (ME-2) to quantify source contributions to the mixture of hazardous air pollutants (HAPs). We used the 'missing mass', defined as the concentration of the measured total particle mass minus the sum of all analyzed species, as an additional variable and implemented an auxiliary equation to constrain the sum of all species mass fractions to be 100%. Regardless of whether the above constraint was implemented and/or the additional VOCs data were included with the PM(2.5) data, the models identified that wood burning (24%-31%), secondary sulfate (20%-24%) and secondary nitrate (15%-20%) were the main contributors to PM(2.5). Using only PM(2.5) data, the model distinguished two diesel features with the 100% constraint, but identified only one diesel feature without the constraint. When both PM(2.5) and VOCs data were used, one additional feature was identified as the major contributor (26%) to total VOC mass. Adding VOCs data to the speciated PM(2.5) data in source apportionment modeling resulted in more accurate source contribution estimates for combustion related sources as evidenced by the less 'missing mass' percentage in PM(2.5). Using the source contribution estimates, we evaluated the validity of using black carbon (BC) as a surrogate for diesel exhaust. We found that BC measured with an aethalometer at 370 nm and 880 nm had reasonable correlations with the estimated concentrations of diesel particulate matters (r>0.7), as well as with the estimated concentrations of wood burning particles during the heating seasons (r=0.56-0.66). This indicates that the BC is not a unique tracer for either source. The difference in BC between 370 and 880 nm, however, correlated well exclusively with the estimated wood smoke source (r=0.59) and may be used to separate wood smoke from diesel exhaust. Thus, when multiple BC related sources exist in the same monitoring environment, additional data processing or modeling of the BC measurements is needed before these measurements could be used to represent the diesel exhaust.     

Comparison of methods for converting Dylos particle number concentrations to PM2.5 mass concentrations

The aim of this study was to (a) develop a method for converting particle number concentrations (PNC) obtained by Dylos to PM2.5 mass concentrations, (b) compare this conversion with similar methods available in the literature, and (c) compare Dylos PM2.5 obtained using all available conversion methods with gravimetric samples. Data were collected in multiple residences in three European countries using the Dylos and an Aerodynamic Particle Sizer (APS, TSI) in the Netherlands or an optical particle counter (OPC, GRIMM) in Greece. Two statistical fitted curves were developed based on Dylos PNC and either an APS or an OPC particle mass concentrations (PMC). In addition, at the homes of 16 volunteers (UK and Netherlands), Dylos measurements were collected along with gravimetric samples. The Dylos PNC were transformed to PMC using all the fitted curves obtained during this study (and three found in the literature) and were compared with gravimetric samples. The method developed in the present study using an OPC showed the highest correlation (Pearson (R) = 0.63, Concordance (ρ_c) = 0.61) with gravimetric data. The other methods resulted in an underestimation of PMC compared to gravimetric measurements (R = 0.65‐0.55, ρ_c = 0.51‐0.24). In conclusion, estimation of PM2.5 concentrations using the Dylos is acceptable for indicative purposes.     

Analysis of indoor particle size distributions in an occupied townhouse using positive matrix factorization

From late 1999 to early March 2000, measurements of particle number (particles 0.01-20 microm in aerodynamic diameter) concentrations were made inside of a townhouse occupied by two non-smoking adults and located in Reston, VA (approximately 25 miles northwest of Washington, DC). The particle size measurements were made using an SMPS and an APS as well as a Climet optical scattering instrument. In this study, positive matrix factorization (PMF) was used to study the indoor particle size distributions. The size distributions or profiles obtained were identified by relating the obtained source contributions to the source information provided by the occupants. Nine particle sources were identified, including two sources associated with gas burner use: boiling water and frying tortillas. Boiling water for tea or coffee was found to be associated only with the smallest particles, with a number mode close to the detection limit of the SMPS (i.e., 0.01 microm). Frying tortillas produced particles with a number mode at about 0.09 microm while broiling fish produced particles with a number mode at about 0.05 microm. A citronella candle was often burned during the study period, and this practice was found to produce a 0.2-microm modal number distribution. Other indoor particle sources identified included sweeping/vacuuming (volume mode at 2 microm); use of the electric toaster oven (number mode at 0.03 microm); and pouring of kitty litter (volume mode over 10 microm). Two outdoor sources were also resolved: traffic (number mode at about 0.15 microm) and wood smoke (major number mode at about 0.07 microm). The volume distributions showed presence of coarse particles in most of the resolved indoor sources probably caused by personal cloud emissions as the residents performed the various indoor activities. PRACTICAL IMPLICATIONS: This study has shown that continuous measurements of indoor particle number and volume concentrations together with records of personal activities are useful for indoor source apportionment models. Each of the particle sources identified in this study produces distinct size distributions that may be useful in studying the mortality and morbidity effects of airborne particulate matter because they will have different penetrability and deposition patterns.     

Source apportionment of population representative samples of PM(2.5) in three European cities using structural equation modelling

Apportionment of urban particulate matter (PM) to sources is central for air quality management and efficient reduction of the substantial public health risks associated with fine particles (PM(2.5)). Traffic is an important source combustion particles, but also a significant source of resuspended particles that chemically resemble Earth's crust and that are not affected by development of cleaner motor technologies. A substantial fraction of urban ambient PM originates from long-range transport outside the immediate urban environment including secondary particles formed from gaseous emissions of mainly sulphur, nitrogen oxides and ammonia. Most source apportionment studies are based on small number of fixed monitoring sites and capture well population exposures to regional and long-range transported particles. However, concentrations from local sources are very unevenly distributed and the results from such studies are therefore poorly representative of the actual exposures. The current study uses PM(2.5) data observed at population based random sampled residential locations in Athens, Basle and Helsinki with 17 elemental constituents, selected VOCs (xylenes, trimethylbenzenes, nonane and benzene) and light absorbance (black smoke). The major sources identified across the three cities included crustal, salt, long-range transported inorganic and traffic sources. Traffic was associated separately with source categories with crustal (especially Athens and Helsinki) and long-range transported chemical composition (all cities). Remarkably high fractions of the variability of elemental (R(2)>0.6 except for Ca in Basle 0.38) and chemical concentrations (R(2)>0.5 except benzene in Basle 0.22 and nonane in Athens 0.39) are explained by the source factors of an SEM model. The RAINS model that is currently used as the main tool in developing European air quality management policies seems to capture the local urban fraction (the city delta term) quite well, but underestimates crustal particle levels in the three cities of the current study. Utilizing structural equation modelling parallel with traditional principal component analysis (PCA) provides an objective method to determine the number of factors to be retained in a model and allows for formal hypotheses testing.     

Concurrent assessment of personal,indoor, and outdoor PM2.5 and PM1 levels and source contributions using novel low-cost sensing devices

The intensity, frequency, duration, and contribution of distinct PM_2.5 sources in Asian households have seldom been assessed; these are evaluated in this work with concurrent personal, indoor, and outdoor PM_2.5 and PM₁ monitoring using novel low-cost sensing (LCS) devices, AS-LUNG. GRIMM-comparable observations were acquired by the corrected AS-LUNG readings, with R² up to 0.998. Twenty-six non-smoking healthy adults were recruited in Taiwan in 2018 for 7-day personal, home indoor, and home outdoor PM monitoring. The results showed 5-min PM_2.5 and PM₁ exposures of 11.2 ± 10.9 and 10.5 ± 9.8 µg/m³, respectively. Cooking occurred most frequently; cooking with and without solid fuel contributed to high PM_2.5 increments of 76.5 and 183.8 µg/m³ (1 min), respectively. Incense burning had the highest mean PM_2.5 indoor/outdoor (1.44 ± 1.44) ratios at home and on average the highest 5-min PM_2.5 increments (15.0 µg/m³) to indoor levels, among all single sources. Certain events accounted for 14.0%-39.6% of subjects’ daily exposures. With the high resolution of AS-LUNG data and detailed time-activity diaries, the impacts of sources and ventilations were assessed in detail.     

PM2.5 chemical composition in Hong Kong: urban and regional variations

Chemically speciated PM2.5 measurements were made at roadside, urban, and rural background sites in Hong Kong for 1 year during 2000/2001 to determine the spatial and temporal variations of PM2.5 mass and chemical composition in this highly populated region. Annual average PM2.5 concentrations at the urban and rural sites were 34.1 and 23.7 microg m(-3), respectively, approximately 50-100% higher than the United States' annual average National Ambient Air Quality Standard (NAAQS) of 15 microg m(-3). Daily PM2.5 concentrations exceeded the U.S. 24-h NAAQS of 65 microg m(-3) on 19 days, reaching 131+/-8 microg m(-3) at the roadside site on 02/28/2001. Carbonaceous aerosol is the largest contributor to PM2.5 mass (explaining 52-75% of PM2.5 mass at the two urban sites and 32% at the background site), followed by ammonium sulfate (ranging from 23% to 37% at the two urban sites and 51% at the background site). Ammonium sulfate and crustal concentrations showed more uniform spatial distributions, while the largest urban-rural contrasts found in carbonaceous aerosol (likely due to emissions from on-road gasoline and diesel vehicles). Marine influences accounted for 7% of the mass at the background site (more than twice as much as at the two urban sites). Ternary diagrams are utilized to illustrate the different spatial patterns.     

Comparison of iron accumulation in lakes using sediment core and mass balance calculations

The accumulation of iron (Fe) in several lakes in Ontario, Canada was determined by two independent approaches. First, Fe accumulation was calculated in cores collected from several sites in each lake by integrating Fe concentration profiles with sediment accumulation rates determined from Pb210 dating. These site-specific accumulation rates were corrected for sediment focussing so that whole-lake Fe accumulation values could be derived. Using this approach, recent whole-lake Fe accumulation in eight lakes ranged between approximately 750 and 4000 mg/m2 per year. Second, whole-lake Fe accumulation was estimated from lake mass budgets, which were measured over a maximum of 14 years. Accumulation measured using the mass balances ranged from 10 to 1330 mg/m2 per year. Comparison of the two approaches indicated that retentions calculated from the sediment cores were much greater than those estimated from the mass balances. The most likely explanation for this difference is that, in the two decades since the cores were collected, there has been a substantial decline in Fe retention (in mass units but not percent) in the study lakes, principally as a result of reduced inputs of Fe from the catchments.     

Identifying and quantifying secondhand smoke in source and receptor rooms: logistic regression and chemical mass balance approaches

Identifying and quantifying secondhand tobacco smoke (SHS) that drifts between multiunit homes is critical to assessing exposure. Twenty‐three different gaseous and particulate measurements were taken during controlled emissions from smoked cigarettes and six other common indoor source types in 60 single‐room and 13 two‐room experiments. We used measurements from the 60 single‐room experiments for (i) the fitting of logistic regression models to predict the likelihood of SHS and (ii) the creation of source profiles for chemical mass balance (CMB) analysis to estimate source apportionment. We then applied these regression models and source profiles to the independent data set of 13 two‐room experiments. Several logistic regression models correctly predicted the presence of cigarette smoke more than 80% of the time in both source and receptor rooms, with one model correct in 100% of applicable cases. CMB analysis of the source room provided significant PM_2.5 concentration estimates of all true sources in 9 of 13 experiments and was half‐correct (i.e., included an erroneous source or missed a true source) in the remaining four. In the receptor room, CMB provided significant estimates of all true sources in 9 of 13 experiments and was half‐correct in another two.     

Characterization of chemical species in PM2.5 and PM10 aerosols in suburban and rural sites of central Taiwan.

Aerosol samples for PM2.5 (particulate matter with aerodynamic diameters less than 2.5 microns), PM2.5-10 (particulate matter with aerodynamic diameters between 2.5 and 10 microns) and TSP were collected from June to September 1998 at THU (suburban) and HKIT (rural) sites in central Taiwan. The ratios of PM2.5/PM10 averaged 0.70 for the daytime and 0.63 for the nighttime at THU, respectively. At HKIT, the PM2.5/PM10 ratios averaged 0.56 for the daytime and 0.72 in the nighttime, respectively. These results indicated that the PM2.5 concentrations contribute the majority of the PM10 concentration and PM10 concentrations contribute the majority of the TSP at both sites. The averaged PM2.5 concentrations at THU are higher than those measured at HKIT during the daytime period. However, the average PM2.5-10 concentrations in THU are lower than those measured at HKIT during nighttime. The samples collected were also analyzed by atomic absorption spectrophotometry for the elemental analysis of Ca, Fe, Pb, Zn, Cu, Mn and Cr. Meanwhile ion chromatography was used to analyze for the water-soluble ions: sulphate, nitrate and chloride in the Universal samples. The concentrations of heavy metals in PM10 during daytime were all higher than nighttime at THU. However, the averaged concentrations of metal elements in PM10 during day and night period were distributed irregularly at HKIT. The results indicated that for metal elements collected at HKIT have different emission sources. The concentrations of metal elements during daytime in PM10 at THU were generally higher than HKIT. The phenomena owing to the averaged PM2.5 particle concentrations at THU (suburban) were higher than those measured at HKIT (rural) and PM2.5 occupied the major portions of PM10 for both sites during the day period. For anion species, there are no significant differences between day and night period in PM10 concentrations at both suburban and rural sites.     

Assessing chemical behavior and developing remedial actions using a mass balance model of chemical fate in the Bay of Quinte

A multi-segment model of chemical fate and transport in the Bay of Quinte and a food chain model that simulate average annual conditions, are used to examine the behavior of arsenic (As), pentachlorophenol (PCP) and polychlorinated biphenyls (PCBs) in the Bay, an “Area of Concern” in the Great Lakes. The Bay model was used to predict the status of As and PCP from known loading data, but for PCBs with unknown loadings, the Bay and food chain models were used to “back-calculate” total loadings to the Bay. Chemical behavior depends on the characteristics of the Bay and physical-chemical properties of chemicals. Short water residence times of less than a week to several months result in chemicals being advected, unless subject to other, more rapid processes. In Upper Bay, rapid rates of sediment deposition and resuspension retard losses by advection of persistent chemicals such as PCBs and As, despite As being largely dissolved in the water column. Overall, behavior in Upper Bay is dominated by sediment-water exchange, and in Lower Bay by water exchange with Lake Ontario. Because of PCP's rapid transformation rate in the water column, most chemical is transformed before it reaches the sediments or downstream segments. It is recommended that elevated As inputs from the Moira River must be controlled to reduce in-Bay water and sediment concentrations, and for PCP, industrial discharges must be reduced. If reduced, concentrations in the Bay would respond within 3 months, 1 month and 3 years in the water, and about 6, 3 and 10 years in sediments for As, PCP and PCBs, respectively.     

Source apportionment of human personal exposure to volatile organic compounds in homes, offices and outdoors by chemical mass balance and genetic algorithm receptor models

A number of past studies have shown the prevalence of a considerable amount of volatile organic compounds (VOCs) in workplace, home and outdoor microenvironments. The quantification of an individual's personal exposure to VOCs in each of these microenvironments is an essential task to recognize the health risks. In this paper, such a study of source apportionment of the human exposure to VOCs in homes, offices, and outdoors has been presented. Air samples, analysed for 25 organic compounds and sampled during one week in homes, offices, outdoors and close to persons, at seven locations in the city of Leipzig, have been utilized to recognize the concentration pattern of VOCs using the chemical mass balance (CMB) receptor model. In result, the largest contribution of VOCs to the personal exposure is from homes in the range of 42 to 73%, followed by outdoors, 18 to 34%, and the offices, 2 to 38% with the corresponding concentration ranges of 35 to 80 microg m(- 3), 10 to 45 microg m(- 3) and 1 to 30 microg m(- 3) respectively. The species such as benzene, dodecane, decane, methyl-cyclopentane, triethyltoluene and trichloroethylene dominate outdoors; methyl-cyclohexane, triethyltoluene, nonane, octane, tetraethyltoluene, undecane are highest in the offices; while, from the terpenoid group like 3-carane, limonene, a-pinene, b-pinene and the aromatics toluene and styrene most influence the homes. A genetic algorithm (GA) model has also been applied to carry out the source apportionment. Its results are comparable with that of CMB.     

Concentration and chemical characteristics of PM2.5 in Beijing, China: 2001-2002

Weekly PM2.5 samples were simultaneously collected at a semi-residential (Tsinghua University) and a downtown (Chegongzhuang) site in Beijing from August 2001 through September 2002. The ambient mass concentration and chemical composition of PM2.5 were determined. Analyses including elemental composition, water-soluble ions, and organic and elemental carbon were performed. The annual average concentrations of PM2.5 were 96.5 microg m(-3) and 106.9 microg m(-3) at CGZ and HU site, respectively. More than 80% of the PM2.5 mass concentrations were explained by carbonaceous species, secondary particles, crustal matters and trace elements at the two sites. Carbonaceous species were the most abundant components, constituting about 45% and 48% of the total PM2.5 mass concentrations at CGZ and THU site, respectively. SO4(2-), NO3- and NH4+ were three major ions, accounting for 37%, 23% and 20%, respectively, of the total mass of inorganic water-soluble ions.     

Measurement and source identification of polycyclic aromatic hydrocarbons (PAHs) in the aerosol in Xi'an, China, by using automated column chromatography and applying positive matrix factorization (PMF)

In this study, we measured polycyclic aromatic hydrocarbons (PAHs) in aerosols in Xi'an, China from 2005 to 2007, by using a modified Soxhlet extraction followed by a clean-up procedure using automated column chromatography followed by HPLC/fluorescence detection. The sources of PAHs were apportioned by using the positive matrix factorization (PMF) method. The PM₁₀ concentration in winter (161.1 ± 66.4 μg m^− 3, n = 242) was 1.5 times higher than that in summer (110.9 ± 34.7 μg m^− 3, n = 248). ΣPAH concentrations, which are the sum of the concentrations of all detected PAHs, in winter (344.2 ± 149.7 ng m^− 3, n = 45) was 2.5 times higher than that in summer (136.7 ± 56.7 ng m^− 3, n = 24) in this study. These strong seasonal variations in atmospheric PAH concentration are possibly due to coal combustion for residential heating. According to the source apportionment with PMF method in this study, the major sources of PAHs in Xi'an are categorized as (1) mobile sources such as vehicle exhaust that constantly contribute to PAH pollution, and (2) stationary sources such as coal combustion that have a large contribution to PAH pollution in winter.     

Comparative PM10-PM2.5 source contribution study at rural, urban and industrial sites during PM episodes in Eastern Spain

In this study a set of 340 PM10 and PM2.5 samples collected throughout 16 months at rural, an urban kerbside and an industrial background site (affected by the emissions from the ceramic manufacture and other activities) were interpreted. On the regional scale, the main PM10 sources were mineral dust (mainly Al2O3, Fe, Ti, Sr, CaCO3, Mg, Mn and K), emissions derived from power generation (SO4=, V, Zn and Ni), vehicle exhausts (organic and elemental carbon, NO3- and trace elements) and marine aerosol (Na, Cl and Mg). The latter was not identified in PM2.5. At the industrial site, additional PM10 sources were identified (tile covering in the ceramic production, petrochemical emissions and bio-mass burning from a large orange tree cultivation area). The contribution of each PM source to PM10 and PM2.5 levels experiences significant variations depending on the type of PM episode (Local-urban mainly in autumn-winter, regional mainly in summer, African or Atlantic episode), which are discussed in this study. The results show that it would be very difficult to meet the EU limit values for PM10 established for 2010. The annual mean PM levels are 22.0 microg PM10/m3 at the rural and 49.5 microg PM10/m3 and 33.9 microg PM2.5/m3 at the urban site. The natural contribution in this region, estimated at 6 microg/m3 of natural mineral dust (resulting from the African events and natural resuspension) and 2 microg/m3 of marine aerosol, accounts for 40% of the 2010 EU annual limit value (20 microg PM10/m3). Mineral dust concentrations at the urban and industrial sites are higher than those at the rural site because of the urban road dust and the ceramic-production contributions, respectively. At the urban site, the vehicle exhaust contribution (17 microg/m3) alone is very close to the 2010 EU PM10 limit value. At the rural site, the African dust is the main contributor to PM10 levels during the highest daily mean PM10 events (100th-97th percentile range). At the urban site, the vehicle exhaust product is the main contributor to PM10 and PM2.5 levels during the highest daily mean PM events (100th-85th percentile range). Mineral dust concentrations during African dust events accounts for 20-30 microg/m3 in PM10 and 10-15 microg/m3 in PM2.5. During non-African dust events, mineral dust derived from anthropogenic activities (e.g. urban road dust) is also a significant contributor to PM10, but not to PM2.5.