Atmospheric Environment (v.39, #20)

Large-eddy simulations of wind flow and pollutant dispersion in a street canyon by Ellen S.P. So; Andy T.Y. Chan; Anton Y.T. Wong (3573-3582).
The wind flow and pollutant dispersion phenomena in urban streets of different aspect ratios (h/w) and relative canyon height ratios (h 2/h 1) are studied using large-eddy simulations (LES). The concerned large eddies are computed by the filtered Navier–Stokes equations in LES and the unresolved small eddies are modelled using Smagorinsky subgrid scale model. The domain is discretised into uneven staggered grids using marker and cell (MAC) method. The objective of this work is to demonstrate the various flow regimes and their threshold values in urban street canyon using LES for various canyon geometries and Reynolds numbers. All cases are investigated with Reynolds number 400 primarily to obtain information of the three regimes of canyon flow and the Reynolds number is then increased incrementally to 2000 to study the consequent flow fields and pollutant dilution patterns. In low Reynolds numbers, results presented agree with the generally obtained threshold values for different flow regimes. Increase in Reynolds number has smeared the flow regimes boundary. The ease of pollutant dispersion is mainly promoted by better mixings inside the canyon, formation of unstable circulations and higher Reynolds numbers. All results show that the flow regime and pollution pattern demarcations depend on both the varying canyon geometry and Reynolds number.
Keywords: Urban street canyon; Pollutant dispersion; Turbulent flow; Large-eddy simulations; Canyon aspect ratio;

Trace metals in bulk precipitation and throughfall in a suburban area of Japan by H. Hou; T. Takamatsu; M.K. Koshikawa; M. Hosomi (3583-3595).
Throughfall and bulk precipitation samples were collected monthly for 1.5 years over bare land and under canopies of Japanese cedar (Cryptomeria japonica), Japanese red pine (Pinus densiflora), Japanese cypress (Chamaecyparis obtusa), and bamboo-leafed oak (Quercus myrsinaefolia) in a suburban area of Japan. Samples were analyzed for dissolved Al, Mn, Fe, Cu, Zn, Ag, In, Sn, Sb and Bi by ICP-AES and ICP-MS. The metal concentrations were higher in throughfall, especially that of C. japonica, than bulk precipitation. Enrichment ratios (ERs: ratios of metal concentrations in throughfall to those in bulk precipitation) ranged from 2.5 (Zn) to 5.3 (Ag) (3.9 on average), and ERs for slightly soluble metals were generally higher than those for easily soluble metals. Concentrations of Mn, Fe, Cu, and Zn accounted for 99% of the total concentration of heavy metals in rainwater, whereas those of rare metals such as Ag, In, Sn, and Bi totaled <0.23%. Average concentrations of rare metals were 0.002 and 0.010 μg l−1 for Ag, 0.001 and 0.005 μg l−1 for In, 0.062 and 0.21 μg l−1 for Sn, and 0.006 and 0.023 μg l−1 for Bi in bulk precipitation and throughfall, respectively. The metal concentrations in rainwater were negatively correlated to the volume of rainwater, indicating that washout is the main mechanism that incorporates metals into rainwater. From the enrichment factors, that is, (X/Al)rain/(X/Al)crust, metals other than Fe were shown to be more enriched in rainwater than in the Earth's crust, including those present as a result of leaching from soil dust (Mn) and from anthropogenic sources (Cu, Zn, Ag, In, Sn, Sb, and Bi).
Keywords: Silver; Tin; Antimony; Rainwater; Enrichment factor; Tsukuba;

Mobile platform measurements of ultrafine particles and associated pollutant concentrations on freeways and residential streets in Los Angeles by Dane Westerdahl; Scott Fruin; Todd Sax; Philip M. Fine; Constantinos Sioutas (3597-3610).
Recent health studies have reported that ultrafine particles (UFP) (<0.1 μm in diameter) may be responsible for some of the adverse health effects broadly attributed to particulate matter. In urban areas, UFP are produced by combustion sources, such as vehicle exhaust, and by secondary formation in the atmosphere. While UFP can be monitored, few studies have explored the impact of local primary sources in urban areas (including mobile sources on freeways) on the temporal and spatial distribution of UFP. This paper describes the integration of multiple monitoring technologies on a mobile platform designed to characterize UFP and associated pollutants, and the application of this platform in a study of UFP number concentrations and size distributions in Los Angeles. Monitoring technologies included two condensation particle counters (TSI Model 3007 and TSI 3022A) and scanning mobility particle sizers for UFP. Real-time measurements made of NO x (by chemiluminesence), black carbon (BC) (by light absorption), particulate matter-phase PAH (by UV ionization), and particle length (by diffusional charging) showed high correlations with UFP numbers, ( r 2 = 0.78 for NO, 0.76 for BC, 0.69 for PAH, and 0.88 for particle length). Average concentrations of UFP and related pollutants varied strongly by location, road type, and truck traffic volumes, suggesting a relationship between these concentrations and truck traffic density.
Keywords: Particle number concentrations; On-road concentrations; In-vehicle concentrations; Nanoparticles;

Spatial mapping of variables that vary in space and time is a common procedure in many research fields. Very often it is of interest to map the time-average or time-integration of the variable over the whole period of interest. Normally, such a map is produced by spatially interpolating the whole period averages of the observed data. An alternative option is to first spatially interpolate narrow time slice averages of the variable and then sum the resultant maps. This paper discusses the latter option, and the accuracy of the spatio-temporal variable interpolation as a function of the width of the time-averaging window. Theoretically, using a linear and data-value independent operator to interpolate a complete data set (i.e. without missing data), the accuracy is independent of the width of the time-averaging window. However, using a nonlinear or a data-value dependent interpolation operator, and/or in the presence of missing data, the accuracy of the interpolation can vary with the averaging window width. The concept is demonstrated using a set of half-hourly SO 2 concentrations measured at 20 monitoring stations in Haifa Bay area, Israel, during the years 1996–2002. Cross-validated interpolation accuracy measures calculated for this data set vary significantly with the time-averaging window width, showing a clear minimum at daily averaging. The results and their general implications for the interpolation of spatio-temporal variables are discussed.
Keywords: Mapping accuracy; Spatial interpolation; Time-averaging window;

Improved modelling of ship SO2 emissions—a fuel-based approach by Øyvind Endresen; Joachim Bakke; Eirik Sørgård; Tore Flatlandsmo Berglen; Per Holmvang (3621-3628).
Significant variations are apparent between the various reported regional and global ship SO2 emission inventories. Important parameters for SO2 emission modelling are sulphur contents and marine fuel consumption. Since 1993, the global average sulphur content for heavy fuel has shown an overall downward trend, while the bunker sale has increased. We present an improved bottom up approach to estimate marine sulphur emissions from ship transportation, including the geographical distribution. More than 53,000 individual bunker samples are used to establish regionally and globally (volume) weighted average sulphur contents for heavy and distillate marine fuels. We find that the year 2002 sulphur content in heavy fuels varies regionally from 1.90% (South America) to 3.07% (Asia), with a globally weighted average of 2.68% sulphur. The calculated globally weighted average content for heavy fuels is found to be 5% higher than the average (arithmetic mean) sulphur content commonly used. The reason for this is likely that larger bunker stems are mainly of high-viscosity heavy fuel, which tends to have higher sulphur values compared to lower viscosity fuels. The uncertainties in SO2 inventories are significantly reduced using our updated SO2 emission factors (volume-weighted sulphur content). Regional marine bunker sales figures are combined with volume-weighted sulphur contents for each region to give a global SO2 emission estimate in the range of 5.9–7.2 Tg (SO2) for international marine transportation. Also taking into account the domestic sales, the total emissions from all ocean-going transportation is estimated to be 7.0–8.5 Tg (SO2). Our estimate is significantly lower than recent global estimate reported by Corbett and Koehler [2003. Journal of Geophysical Research: Atmospheres 108] (6.49 Tg S or about 13.0 Tg SO2). Endresen et al. [2004. Journal of Geophysical Research 109, D23302] claim that uncertainties in input data for the activity-based method will give too high emission estimates. We also indicate that this higher estimate will almost give doubling of regional emissions, compared to detailed movement-based estimates. The paper presents an alternative approach to estimate present overall SO2 ship emissions with improved accuracy.
Keywords: Ship SO2 emission; Accuracy of SO2 ship emission modelling; Fuel based modelling; Marine bunker inventory;

Short-term human exposure to PM2.5, ultrafine particle counts (particle range: 0.02–1 μm) and carbon monoxide (CO) was investigated at and around a street canyon intersection in Central London, UK. During a four-week field campaign, groups of four volunteers collected samples at three timings (morning, lunch and afternoon), along two different routes (a heavily trafficked route and a backstreet route) via five modes of transport (walking, cycling, bus, car and taxi). PM2.5 was sampled using high-flow gravimetric personal samplers, ultrafine particle counts were measured using TSI P-TRAKs and Langans were used to measure CO exposure. Three hundred and ninety-four samples were collected—197 PM2.5, 86 ultrafine particle count and 111 CO. Arithmetic means of PM2.5 personal exposure were 27.5, 33.5, 34.5, 38.0 and 41.5 μg m−3, ultrafine particle counts were 67 773, 93 968, 101 364, 99 736 and 87 545 pt cm−3 and CO levels were 0.9, 1.1, 0.8, 1.3 and 1.1 ppm for walking, cycling, bus, car and taxi respectively. On the heavily trafficked route, personal exposure was 35.3 μg m−3, 101142 pt cm−3 and 1.3 ppm, and on the backstreet route it was 31.8 μg m−3, 71628 pt cm−3 and 0.6 ppm for PM2.5, ultrafine particle counts and CO, respectively. Personal exposure levels were high during the morning measurements for all three pollutants (34.6 μg m−3, 106 270 pt cm−3 and 1.5 ppm for PM2.5, ultrafine particle counts and CO, respectively).There was a moderately strong correlation between personal exposure of ultrafine particle counts and CO ( r = 0.7 , N = 67 ) but a weaker correlation between PM2.5 and ultrafine particle counts ( r = 0.5 , N = 83 ) and a low correlation between PM2.5 and CO exposure ( r = 0.2 , N = 105 ). The exposure assessment also revealed that the background and kerbside monitoring stations were not representative of the personal exposure of individuals to PM2.5 and CO at and around a street canyon intersection.
Keywords: Personal exposure; Carbon monoxide; PM2.5; Ultrafine particle counts; Transport;

Plume dispersion simulation in low wind conditions in stable and convective boundary layers by Davidson M. Moreira; Tiziano Tirabassi; Jonas C. Carvalho (3643-3650).
The present study proposes a steady-state mathematical model for dispersion of contaminants in low winds that takes into account the along-wind diffusion. The solution of the advection–diffusion equation for these conditions is obtained applying the Laplace transform, considering the planetary boundary layer as a multilayer system. The eddy diffusivities used in the K-diffusion model were derived from the local similarity and Taylor's diffusion theory. The eddy diffusivities are functions of distance from the source and correctly represent the near-source diffusion in weak winds. The performances of the model were evaluated against the field experiments carried out at the Idaho National Engineering Laboratory and during the convective conditions at the Indian Institute of Technology. Furthermore, the study suggests that the inclusion of the longitudinal diffusion, important at short distance diffusion from a continuous point source in low wind conditions, can improve the description of the turbulent transport of atmospheric contaminants.
Keywords: Low winds; Turbulence parameterization; Down-wind diffusion; Eulerian dispersion models; Laplace transform method; Solution of advection–diffusion equation;

Magnetic properties of urban street dust and their relationship with organic matter content in the West Midlands, UK by V.F. Shilton; C.A. Booth; J.P. Smith; P. Giess; D.J. Mitchell; C.D. Williams (3651-3659).
This study demonstrates significant correlations between the organic matter content of urban street dust and certain mineral magnetic properties, which accords with previous work that indicates magnetic parameters offer potential as a proxy for organic content. However, site-specific data demonstrate the relationship can be different for particular roads, even within the same area. This indicates the association may be more complex than previous work proposes and a cautionary note is required. It is recommended that the nature of the relationship between magnetic and organic properties should be fully explored for particular urban environments and individual field settings, before using magnetic measurements as a proxy for organic matter content. Furthermore, whilst soil is believed to significantly contribute to urban street dust, magnetic values in this study are much higher than those previously reported for top-soils and indicate the influence of other sources, such as anthropogenic pollutants. This suggests that using magnetic measurements to discriminate sources of urban particulates has considerable potential for development.
Keywords: Environmental magnetism; Data correlation; Anthropogenic pollutants; Public health; Diesel emissions; Top-soil;

Comparative assessment of regionalisation methods of monitored atmospheric deposition loads by Frido Reinstorf; Maja Binder; Mario Schirmer; Jost Grimm-Strele; Wolfgang Walther (3661-3674).
The objective of this investigation is to assess the suitability of well-known regionalisation methods of data from existing deposition monitoring networks for use in water resources management. For this purpose a comparison of the applicability and accuracy of various regionalisation methods was made. A crucial point is the data demand of the various methods. In this investigation the deterministic and geostatistical methods inverse distance weighting (IDW), ordinary kriging (OK) and external drift kriging (EDK) as well as the chemical transport models METRAS-MUSCAT, EMEP, EDACS and EUTREND have been characterised and evaluated. The methods IDW and OK have been applied to the investigation areas—the German Federal States of Lower Saxony and Saxony. An evaluation of these methods was carried out with a cross-validation procedure. The result was in most cases a higher accuracy for the OK method. The EDK method has been investigated in order to find suitable drift variables from the parameters precipitation amount, altitude and wind direction. With help of a correlation analysis a suitable drift variable could not be found. After the application of OK, verification was carried out by a comparison of the estimated data set with an independently determined data set. The result was a relatively smaller deviation of the estimated data set. The investigation considers data from routine monitoring networks as well as networks for special applications and has been carried out on the basis of monitoring networks of the two states. The investigated database was wet and bulk deposition of the substances NH4 +, SO4 2−, NO3 , Na+, Pb2+, and Cd2+ in Lower Saxony and SO4 2− in Saxony. From this, a consistent database of bulk deposition data was built. From all applied methods OK proved to cope best with the data deficiencies that were found.
Keywords: Deposition loads; Regionalisation; Geostatistics; Monitoring networks; Deposition models;

Size-dependent collection of micrometer-sized particles using nylon mesh by Naomichi Yamamoto; Kazukiyo Kumagai; Minoru Fujii; Derek G. Shendell; Osamu Endo; Yukio Yanagisawa (3675-3685).
Our study explored the size-dependent collection characteristics for micron-sized particles using several kinds of commercially available woven nylon net filters. The particle concentrations with and without the filter were compared to determine the filtration characteristics. The theoretical efficiencies based on a single-fiber theory and a hole model were also computed. Although the theoretical efficiencies were generally consistent with the experimental results, the non-uniformity of air velocity profile within a mesh hole, and a particle's detachment from or bounce off the filters, should be further investigated in future research. Overall, the present study revealed the size-fractionation capability of the nylon wire mesh filters for micron-sized particles from experimental and theoretical points of view. Unlike impactors, the size-fractionation characteristics of the nylon wire mesh filter were determined by particle size, mesh fiber diameter, and a combination of different particle collection mechanisms including impaction, interception, and gravitational settling. Each mechanical process appears interdependently governed in part by the filter dimensions such as filter mesh size (diameter of opening) as well as related variables such as packing density and fiber diameter.
Keywords: Coarse particles; Filtration; Size distribution; Woven net filter; Mesh size;

Assessing genetic structure and diversity of airborne bacterial communities by DNA fingerprinting and 16S rDNA clone library by Pierre-Alain Maron; David P.H. Lejon; Esmeralda Carvalho; Karine Bizet; Philippe Lemanceau; Lionel Ranjard; Christophe Mougel (3687-3695).
The density, genetic structure and diversity of airborne bacterial communities were assessed in the outdoor atmosphere. Two air samples were collected on the same location (north of France) at two dates (March 2003 (sample1) and May 2003 (sample 2)). Molecular culture-independent methods were used to characterise airborne bacterial communities regardless of the cell culturability. The automated-ribosomal intergenic spacer analysis (A-RISA) was performed to characterise the community structure in each sample. For both sampling dates, complex A-RISA patterns were observed suggesting a highly diverse community structure, comparable to those found in soil, water or sediment environments. Furthermore, differences in the genetic structure of airborne bacterial communities were observed between samples 1 and 2 suggesting an important variability in time. A clone library of 16S rDNA directly amplified from air DNA of sample 1 was constructed and sequenced to analyse the community composition and diversity. The Proteobacteria group had the greatest representation (60%), with bacteria belonging to the different subdivisions α - (19%), β -(21%), γ -(12%) and δ -(8%). Firmicute and Actinobacteria were also well represented with 14% and 12%, respectively. Most of the identified bacteria are known to be commonly associated with soil or plant environments suggesting that the atmosphere is mainly colonised transiently by microorganisms from local sources, depending on air fluxes.
Keywords: Bacteria; Air; Diversity; Community structure;

Indoor and outdoor particle size characterization at a family house in Espoo–Finland by Tareq Hussein; Kaarle Hämeri; Maire S.A. Heikkinen; Markku Kulmala (3697-3709).
Aerosol particle number size distributions (3–400 nm) were measured for three weeks both indoors and outdoors in a family house with natural ventilation in Espoo, Finland. We investigated the indoor-to-outdoor relationship of aerosol particles and analyzed the effects of indoor activities on the particle number size distributions. We also estimated the decay rate of indoor aerosol particle number concentrations. As expected, in the absence of indoor sources of aerosol particles the indoor particles originated from outdoors and the number concentrations followed similar patterns as those outdoors. The maximum penetration was found for particles between 100 and 400 nm in diameter. The mean value of the I/O values was 0.36 for ultrafine particles (UFP diameter <100 nm) and 0.60 for particles larger than 100 nm in diameter. Because of the penetration and deposition processes of aerosol particles, the modal structure of indoor particle number size distributions had larger geometric mean diameters and significantly lower number concentrations of UFP than those outdoors. The natural ventilation did not provide a stable and controlled indoor-to-outdoor relationship of aerosol particles and it caused variable I/O values and time-lags (10–45 min). In the presence of indoor activities, the indoor particle number concentrations cannot be directly estimated from the outdoor number concentrations only. The loss rate of UFP in the indoor air ranged from 1 h−1 for 10 nm particles to 0.1 h−1 for 100 nm particles in diameter. The quantitative and qualitative results presented in the current study are building and condition specific. However, the results provide a better understanding of the particle number size distribution characterizations indoors, especially during different indoor activities.
Keywords: IAQ; Deposition; Penetration; Activity patterns; Particle number size distribution;

Interpretation of variations in fine, coarse and black smoke particulate matter concentrations in a northern European city by Mathew R. Heal; Leon R. Hibbs; Raymond M. Agius; Iain J. Beverland (3711-3718).
The PM2.5, PMcoarse and black smoke (BS) particle metrics broadly reflect different source contributions to PM10. The aim of this study was to generate data for PM2.5 at an urban background site in the UK, and to use the daily collocated measurement of PM2.5, PM10 (and hence PMcoarse) and BS to yield insight into source influences on particulate matter for input to developing PM air quality policy. Mean daily PM10, PM2.5 and BS for a year of measurement in Edinburgh were 15.5, 8.5 and 6.6 μg m−3. The PM2.5 data were well-within possible future limit values proposed by the European Commission Clean Air For Europe programme. Daily PM2.5 and PM10 were significantly correlated ( r 2 = 0.75 ) with PM2.5 contributing 54%, on average, to PM10. The daily BS:PM10 and BS:PM2.5 ratios were more variable, and significantly lower in summer than in winter, reflecting the greater contribution of non-black photochemical secondary particles to PM10 in summer. Analysis with respect to wind showed a dominant influence of dispersion on BS and PM2.5 but both dispersion and a wind-driven suspension influence on PMcoarse. The latter was higher than in central England (averaging about one-third of the PMcoarse), and greater for on-shore wind direction, suggesting a sea-salt source for this component in addition to other particle resuspension contributions. Overall, the data showed that excursions in PM10 were driven more by variations in PM2.5 than by PMcoarse or BS. Both PM2.5 and its proportion to PM10 were significantly elevated for air-masses passing over continental Europe and the British Isles, whereas BS varied less with air-mass origin, supporting the conclusion that concentrations of particulate matter, particularly of finer PM, are strongly influenced by regional scale synoptic meteorology (presumed to be predominantly secondary PM), whereas BS is dominated more by local sources. Comparison of BS with a nearby rural site suggested that approximately three-quarters, on average, of the urban BS was local in origin.
Keywords: PM10; PM2.5; Black smoke; Air pollution; Back-trajectory;

Evaluation of the potential source contribution function using the 2002 Quebec forest fire episode by Bilkis Ara Begum; Eugene Kim; Cheol-Heon Jeong; Doh-Won Lee; Philip K. Hopke (3719-3724).
Particle data consisting of PM2.5 (particulate matter <2.5 μm aerodynamic diameter), light scattering coefficient (b scat), organic carbon (OC), elemental carbon (EC), and sulfate were measured for 2-h intervals from 1 July to 3 August 2002 in Philadelphia using semicontinuous monitors. The potential source contribution function (PSCF) combines the particle data with air parcel back trajectories to identify potential source areas and the preferred pathways that give rise to the observed high particle concentrations in Philadelphia. Five day back trajectories arriving 12 times a day at 500 m above ground level were calculated using NOAA HYSPLIT with the vertical mixing model. PSCF analyses were performed based on the average values as criterion values for the measured particle parameters. During this period, there was a major forest fire in central Quebec that penetrated the eastern United States to well south of Philadelphia, Pennsylvania. This event permits the testing of the ability of the PSCF analysis to identify the known source location of this fire, which is observed clearly in the maps for PM2.5, OC and EC. The other maps can also be interpreted in terms of the likely sources of the observed concentrations.
Keywords: Receptor models; Potential source contribution function; PM2.5; Long-range transport; Forest fire; Biomass burning;

The spatial distribution of PAH depositions to peatlands of Eastern Canada by Annekatrin Dreyer; Christian Blodau; Jukka Turunen; Michael Radke (3725-3733).
Studies that document historic and recent polycyclic aromatic hydrocarbons (PAH) deposition rates over large regions are rare. To address this knowledge gap, peat cores of 15 ombrotrophic bogs along a transect across Eastern Canada were dated with 210Pb for the period of about 1850–2000, extracted using accelerated solvent extraction (ASE) and analysed for 14 PAH by GC-MS/MS. Additional parameters, such as sulphur contents, were also determined. In the more populated regions along the St. Lawrence River, deposition rates in individual bogs peaked at 300–1430 μg m−2  yr−1 in the mid-20th century, whereas in the remote regions of northern Québec, northern Ontario, and Nova Scotia deposition rates remained below 150 μg m−2  yr−1. Total deposition rates were moderately but significantly correlated to sulphur deposition rates ( R 2 = 0.54 , P < 0.0001 ), suggesting coal combustion as an important source of PAH depositions. Phenanthrene, benzo[b+k]fluoranthene, fluoranthene and pyrene were the most abundant PAH. In the remote regions phenanthrene dominated (40–60% of total), whereas along the St. Lawrence River benzo[b+k]fluoranthene were most abundant (25–40%) and indeno[1,2,3-c,d]pyrene and benzo[g,h,i]perylene occurred in larger quantities. This pattern only emerged in the 20th century, whereas before, deposition rates and composition were more uniform. The study documents that within Eastern Canada regions of particular PAH deposition patterns exist and suggests that both local and regional emission and transport played a role in the formation of these patterns.
Keywords: Wetlands; Bogs; Organic pollutants; Atmospheric deposition; Deposition history;

The chemical composition of inorganic and carbonaceous materials in PM2.5 in Nanjing, China by Hong Yang; Jian Zhen Yu; Steven Sai Hang Ho; Jinhui Xu; Wai-Shing Wu; Chun Hong Wan; Xiaodong Wang; Xiaorong Wang; Liansheng Wang (3735-3749).
PM2.5 samples were collected at an urban and a suburban site in Nanjing, China in 2001. They were analyzed for inorganic ions, elemental carbon, organic carbon (OC), water-soluble organic carbon (WSOC), and individual WSOC and nonpolar organic species. Sulfate and organic matter were the two most abundant constituents in these samples. Sulfate accounted for an average of 23% (urban site) and 30% (suburban site) of the identified aerosol mass. Organic matter accounted for an average of 37% (urban) and 28% (suburban) of the identified aerosol mass. WSOC was a significant portion of OC, accounting for about one-third of OC at the urban site and 45% of OC at the suburban site. The suburban–urban gradient in the WSOC/OC ratio also reflected that the aerosol OC was more aged at the suburban location. The correlations of WSOC with sulfate and nitrate suggest that the WSOC fraction was dominated by secondary organics. More than 30 individual WSOC species in the compound classes of organic anions, amino acids, aliphatic amines, and carbohydrates were quantified, accounting for approximately 8% of the WSOC on a carbon mass basis. In addition, 46 individual nonpolar organic compounds in the compound classes of n-alkanes, hopanes, and polycyclic aromatic hydrocarbons were quantified using an in-injection port thermal desorption technique. These nonpolar organic species accounted for less than 7% of the OC on a carbon mass basis. The quantification of individual compounds allowed the identification of major aerosol sources through principal component analysis. Coal combustion, vehicular emissions, secondary inorganic and organic aerosols, and road/sea salt were the major contributing sources to the identified PM2.5 aerosol mass.
Keywords: Aerosol chemical composition; Water soluble organic compounds; Chinese aerosols; Coal combustion; Principal component analysis; Aerosol source identification;

Receptor modeling for multiple time resolved species: The Baltimore supersite by David Ogulei; Philip K. Hopke; Liming Zhou; Pentti Paatero; Seung Shik Park; John M. Ondov (3751-3762).
A number of advances have been made toward solving receptor modeling problems using advanced factor analysis methods. Most recently, a factor analysis method has been developed for source apportionment utilizing aerosol compositional data with varying temporal resolution. The data used in that study had time resolutions ranged from 10 min to 1 h. In this work, this expanded model is tested using a data set from the Ponca Street site of the Baltimore supersite with time resolutions ranging from 30 min to 24 h. The nature of this data set implies that traditional eigenvalue-based methods cannot adequately resolve source factors for the atmospheric situation under consideration. Also, valuable temporal information is lost if one averaged or interpolated data in an attempt to produce a data set of the identical time resolution. Each data point has been used in its original time schedule and the source contributions were averaged to correspond to the specific sampling time interval. A weighting coefficient, w24, was incorporated in the modeling equations in order to improve data fitting for the 24-h data in the model. A total of nine sources were resolved: oil-fired power plant (2%), diesel emissions (1%), secondary sulfate (23%), coal-fired power plant (3%), incinerator (9%), steel plant (12%), aged sea salt (1%), secondary nitrate (23%), and spark-ignition emissions (26%). The results showed the very strong influence of the adjacent interstate highways I-95 and I-895 as well as the tunnel toll booths located to the south of the sampling site. Most of the sulfate observed was found to be associated with distant coal-fired power plants situated in the heavily industrialized midwestern parts of the United States. The contribution of the steel plant (<10 miles, 141°SE) to the observed PM concentrations (12%) was also significant.
Keywords: Receptor modeling; Highly time resolved species; Source apportionment; Factor analysis; Multilinear engine (ME); Conditional probability function (CPF);

Quantitative analysis of volatile organic compounds (VOCs) in atmospheric particles by Mustafa Odabasi; Okan Ongan; Eylem Cetin (3763-3770).
There are a number of difficulties associated with the quantitative analysis of volatile organic compounds (VOCs) in atmospheric particles. Therefore, majority of the previous studies on VOCs associated with particles have been qualitative. Air samples were collected in Izmir, Turkey to determine ambient particle and gas phase concentrations of several aromatic, oxygenated and halogenated VOCs. Samples were quantitatively analyzed using thermal desorption–gas chromatography/mass spectrometry. Gas-phase concentrations ranged between 0.02 (bromoform) and 4.65 μg m−3 (toluene) and were similar to those previously measured at the same site. Particle-phase concentrations ranged from 1 (1,3-dichlorobenzene) to 933 pg m−3 (butanol). VOCs were mostly found in gas-phase (99.9±0.25%). However, the particulate VOCs had comparable concentrations to those reported previously for semivolatile organic compounds. The distribution of particle-phase VOCs between fine (d p<2.5 μm) and coarse (2.5 μm<d p<10 μm) fractions was also investigated. It was found that VOCs were mostly associated with fine particles.
Keywords: Volatile organic compounds; Thermal desorption; Gas–particle partitioning; Fine/coarse particles;