Atmospheric Environment (v.44, #35)

Size and seasonal distributions of airborne bioaerosols in commuting trains by Ya-Fen Wang; Che-Hsu Wang; Kai-Lin Hsu (4331-4338).
Aerobiological studies in commuting trains in northern Taiwan were carried out from August, 2007 until July, 2008. Two six-stage (>7 μm, 4.7∼7 μm, 3.3∼4.7 μm, 2.1∼3.3 μm, 1.1∼2.1 μm, 0.65∼1.1 μm) cascade impactors of 400 orifices were used to collect viable bacteria and fungi, respectively. The levels of carbon monoxide (CO), carbon dioxide (CO2), formaldehyde (HCHO), temperature, and relative humidity in the commuting trains were also recorded during the sampling period. Results show that bacterial concentrations ranged from 25 to 1530 CFU m−3, and averaged 417 CFU m−3. The fungal concentrations ranged from 45 to 1906 CFU m−3, and averaged 413 CFU m−3. Additionally, the highest fractions occurred in the fifth stage (1.1∼2.1 μm) for both bacteria and fungi. The respirable fractions, R b and R f, for bacteria and fungi were 62.8% and 81.4%, respectively, which are higher than those in other studies. Furthermore, the bacterial concentration reached its highest level in autumn, and its lowest level in winter. However, the fungal concentration was highest in spring and lowest in winter. Though the total bacterial or fungal concentration did not exceed the recommendation standard in Taiwan, the relatively high respirable fraction in commuting trains probably implies a higher adverse health risk for sensitive commuters. This study further conducted multiple regression analysis to determine the relationship of various stage fractions of airborne bacteria and fungi with indoor air pollutants (CO and HCHO) and environmental parameters (CO2, temperature, and relative humidity). The correlation coefficients of multiple regression analysis for total bacteria and fungi concentrations with indoor air pollutants and environmental parameters were 0.707 (p  < 0.00376) and 0.612 (p  < 0.00471), respectively. There are currently no formally regulated laws for indoor air quality (IAQ) in Taiwan, and this preliminary study can provide references to the Taiwan government on IAQ management.
Keywords: Bacteria; Fungi; Commuting trains; Number median diameter; Multiple regression analysis;

Atmospheric water-soluble organic nitrogen (WSON) was determined on size-segregated aerosol particles collected during a two years period (2005–2006) in a remote marine location in the Eastern Mediterranean (Finokalia, Crete island). Average concentration of WSON was 5.5 ± 3.9 nmol m−3 and 11.6 ± 14.0 nmol m−3 for coarse (PM1.3-10) and fine (PM1.3) mode respectively, corresponding to 13% of Total Dissolved Nitrogen (TDN) in both modes. Air masses origin and correlation with tracers of natural and anthropogenic sources indicate that combustion process (biomass burning and fossil fuel) and African dust play an important role in regulating levels of WSON in both coarse and fine aerosol fractions. Chemical speciation of organic nitrogen pool was attempted by analyzing 47 fine aerosol samples (PM1) for 17 free amino acids (N-FAA), dimethylamine (DMA) and trimethylamine (TMA). The average concentration of N-FAA was 0.5 ± 0.5 nmol m−3, while the average concentration of DMA was 0.2 ± 0.8 nmol m−3, TMA was below detection limit. The percentage contribution of N-FAA and DMA to WSON was 2.1 ± 2.3% and 0.9 ± 3.4%, respectively.
Keywords: Organic nitrogen; Sources; Speciation; Free amino acids; Amines; Eastern Mediterranean;

Back-extrapolation of estimates of exposure from current land-use regression models by Hong Chen; Mark S. Goldberg; Dan L. Crouse; Richard T. Burnett; Michael Jerrett; Paul J. Villeneuve; Amanda J. Wheeler; France Labrèche; Nancy A. Ross (4346-4354).
Land use regression has been used in epidemiologic studies to estimate long-term exposure to air pollution within cities. The models are often developed toward the end of the study using recent air pollution data. Given that there may be spatially-dependent temporal trends in urban air pollution and that there is interest for epidemiologists in assessing period-specific exposures, especially early-life exposure, methods are required to extrapolate these models back in time. We present herein three new methods to back-extrapolate land use regression models. During three two-week periods in 2005–2006, we monitored nitrogen dioxide (NO2) at about 130 locations in Montreal, Quebec, and then developed a land-use regression (LUR) model. Our three extrapolation methods entailed multiplying the predicted concentrations of NO2 by the ratio of past estimates of concentrations from fixed-site monitors, such that they reflected the change in the spatial structure of NO2 from measurements at fixed-site monitors. The specific methods depended on the availability of land use and traffic-related data, and we back-extrapolated the LUR model to 10 and 20 years into the past. We then applied these estimates to residential information from subjects enrolled in a case–control study of postmenopausal breast cancer that was conducted in 1996.Observed and predicted concentrations of NO2 in Montreal decreased and were correlated in time. The estimated concentrations using the three extrapolation methods had similar distributions, except that one method yielded slightly lower values. The spatial distributions varied slightly between methods. In the analysis of the breast cancer study, the odds ratios were insensitive to the method but varied with time: for a 5 ppb increase in NO2 using the 2006 LUR the odds ratio (OR) was about 1.4 and the ORs in predicted past concentrations of NO2 varied (OR∼1.2 for 1985 and OR∼1.3–1.5 for 1996). Thus, the ORs per unit exposure increased with time as the range and variance of the spatial distributions decreased, and this is due partly to the regression coefficient being approximately inversely proportional to the variance of exposure. Changing spatial variability complicates interpretation and this may have important implications for the management of risk. Further studies are needed to estimate the accuracy of the different methods.
Keywords: Land use regression model; Back extrapolation; Traffic-related air pollution; Historical exposure; Montreal;

Variations in single scattering albedo and Angstrom absorption exponent during different seasons at Delhi, India by Kirti Soni; Sachchidanand Singh; Tarannum Bano; R.S. Tanwar; Shambhu Nath; B.C. Arya (4355-4363).
Simultaneous measurements of aerosol absorption and scattering coefficients for the PM2.5 aerosols particles were done at Delhi during April 2008–March 2009 to estimate the aerosol single scattering albedo (SSA) and the Angstrom absorption exponents at the surface. The annual average SSA at 0.55 μm was found to be 0.70 ± 0.07 with only slight variations during the four seasons, summer (0.63 ± 0.06), monsoon (0.69 ± 0.07), winter (0.74 ± 0.03) and spring (0.72 ± 0.04). However, large variations in average absorption and scattering coefficients were seen during these four seasons. The average absorption coefficients during summer, monsoon, winter and spring were found to be 62.47 ± 21.27, 50.95 ± 43.61, 189.65 ± 85.94 and 90.65 ± 33.06 Mm−1 respectively. The corresponding scattering coefficients were 110.46 ± 36.15, 95.34 ± 49.46, 565.59 ± 274.59 and 236.56 ± 96.25 Mm−1. The Angstrom absorption exponent (ασ(abs)) remained close to unity throughout the year averaging at 1.02 ± 0.08, 1.02 ± 0.10, 1.04 ± 0.11, and 1.03 ± 0.05 during summer, monsoon, winter and spring seasons respectively, strongly indicating that the absorption at Delhi aerosol is mainly due to the abundance of black carbon of fossil fuel origin.
Keywords: Absorption coefficient; Scattering coefficient; Single scattering albedo; Angstrom absorption exponent;

Determining the destructions of both ozone and odd oxygen, Ox, in the nocturnal boundary layer (NBL) is important to evaluate the regional ozone budget and overnight ozone accumulation. This work develops a simple method to determine the dry deposition velocity of ozone and its destruction at a polluted nocturnal boundary layer. The destruction of Ox can also be determined simultaneously. The method is based on O3 and NO2 profiles and their surface measurements. Linkages between the dry deposition velocities of O3 and NO2 and between the dry deposition loss of Ox and its chemical loss are constructed and used. Field measurements are made at an agricultural site to demonstrate the application of the model. The model estimated nocturnal O3 dry deposition velocities from 0.13 to 0.19 cm s−1, very close to those previously obtained for similar land types. Additionally, dry deposition and chemical reactions account for 60 and 40% of the overall nocturnal ozone loss, respectively; ozone dry deposition accounts for 50% of the overall nocturnal loss of Ox, dry deposition of NO2 accounts for another 20%, and chemical reactions account for the remaining 30%. The proposed method enables the use of measurements made in typical ozone field studies to evaluate various nocturnal destructions of O3 and Ox in a polluted environment.
Keywords: Ozone; Odd oxygen; Destruction; Model; Measurement; Dry deposition; Nocturnal boundary layer; Ozone profiles; Nitrate radical; Dinitrogen penoxside; Nitric acid;

Atmospheric Mercury Depletion Events (AMDE) occur in Arctic and Antarctic regions during polar sunrise. During AMDE, reactive gaseous Hg is rapidly formed through in-situ oxidation of gaseous Hg0 by halogens, notably atomic Br and radical BrO. This leads to high Hg deposition fluxes yet an unknown fraction of deposited Hg is reemitted to the atmosphere through subsequent photo-reduction, so that the net deposition flux related to AMDE is not well constrained. Here, Hg and halogens were measured in lichens hanging in tree branches around Hudson Bay where AMDE were reported. Hg concentrations are strongly correlated to halogen elements Br, Cl and I (r 2 of 0.91, 0.76, 0.81) and decrease with distance from Hudson Bay. We interpret this trend as the result of AMDE, supported by a 1D numerical Br and BrO oxidation model for Hg0. Organic carbon normalized Hg contents of down-core lake sediments reported in the literature also show a decreasing trend away from Hudson Bay. Combined observations suggest that at least 50% of Hg deposited during AMDE is reemitted to the atmosphere. Finally, the latitudinal Hg gradient observed in lake sediments suggests that AMDE were active in the Hudson Bay area during the last 90 to 200 years.
Keywords: Mercury; Atmosphere; Arctic; Lichens;

Three methods for quantifying proximity of air sampling sites to spatially resolved emissions of semi-volatile organic contaminants by J.N. Westgate; C. Shunthirasingham; C.E. Oyiliagu; H. von Waldow; F. Wania (4380-4387).
Passive air samplers have made it possible to measure long-term average air concentrations of semi-volatile organic contaminants (SVOCs) at a large number of sampling sites. In order to use the results of such measurement networks in the derivation of empirical measures of long-range transport, a method is required that quantitatively expresses the proximity of air sampling sites to spatially distributed emissions. We propose three increasingly sophisticated tiers for quantifying proximity to emissions. The ‘static’ method assumes that a sampling site is only influenced by emission taking place in the same 1° of latitude by 1° of longitude cell in which it is located. The ‘dispersion’ method additionally accounts for the influence of emissions in neighboring cells by adding the emissions into each cell weighted by the distance between the cell’s center and the center of the cell containing the sampling site. The ‘air-shed’ method quantifies proximity to emissions by combining the emissions in each cell with the probability that air arriving at the sampling site passed through each cell. The probability is calculated for each sampling site by aggregating a large number of air mass back-trajectories. These new proximity gauges were contrasted against the remoteness index RI, which is derived from global atmospheric tracer transport modeling. The four methods were used to quantify the proximity of the sampling sites of the Global Atmospheric Passive Sampling (GAPS) study to global Polycyclic Aromatic Hydrocarbon (PAH) emissions. The proximity gauges produce markedly different results primarily for sites located near steep gradients in population, such as occur in coastal areas or at the feet of mountain ranges. The dispersion method produces quite similar results to the air-shed method using drastically less computational power and input data, but application of the air-shed method may be necessary where winds are strongly directional.
Keywords: Semi-volatile organic contaminant; Passive air sampling; Proximity gauge; Air-shed; Remoteness index;

Indoor exposure from building materials: A field study by Dafni A. Missia; E. Demetriou; N. Michael; E.I. Tolis; J.G. Bartzis (4388-4395).
The present study has been conducted in the frame of BUMA (Prioritization of Building Materials Emissions as indoor pollution sources), a European funded project, aiming at assessing the exposure to emitted compounds in indoor air. Field campaigns in five (5) European cities (Milan, Copenhagen, Dublin, Athens and Nicosia) were carried out. These campaigns covered weekly winter and summer concentration measurements in two (2) public buildings and two (2) private houses in each city. BTEX, terpenes, and carbonyls were measured using passive sampling in two sites inside the building and one outside. VOC emission measurements on selected building material have also been performed using Field and Laboratory Emission Cell (FLEC). The results on indoor concentrations for compounds such as formaldehyde (1.2–62.6 μg m−3), acetaldehyde (0.7–41.6 μg m−3), toluene (0.9–163.5 μg m−3), xylenes (0.2–177.5 μg m−3) and acetone (2.8–308.8 μg m−3) have shown diversity and relatively significant indoor sources depending on the building type, age etc. Indoor concentrations of these substances are varied depending on the building age and type. The percentage of approximately 40% of the indoor air quality levels originated from building materials.
Keywords: Building materials emissions; Volatile Organic Compounds (VOCs); Indoor Air Quality (IAQ); Field and Laboratory Emission Cell (FLEC);

Comparison among filter-based, impactor-based and continuous techniques for measuring atmospheric fine sulfate and nitrate by Wei Nie; Tao Wang; Xiaomei Gao; Ravi Kant Pathak; Xinfeng Wang; Rui Gao; Qingzhu Zhang; Lingxiao Yang; Wenxing Wang (4396-4403).
Filter-based methods for sampling aerosols are subject to great uncertainty if the gas–particle interactions on filter substrates are not properly handled. Sampling artifacts depend on both meteorological conditions and the chemical mix of the atmosphere. Despite numerous of studies on the subject, very few have evaluated filter-based methods in the Asian environments. This paper reports the results of a comparison of the performances of two filter-based samplers, including a Thermo Anderson Chemical Speciation Monitor (RAAS) and a honeycomb denuder filter-pack system, a Micro Orifice Uniform Deposit Impactor (MOUDI) and a real-time ambient ion monitor (AIM, URG9000B) in measuring atmospheric concentrations of PM2.5 sulfate and nitrate. Field studies were conducted at an urban site in Jinan, Shandong province, during the winter of 2007 and at a rural site near Beijing in the summer of 2008. The AIM was first compared with the honeycomb denuder filter-pack system which was considered to have minimal sampling artifacts. After some modifications made to it, the AIM showed good performance for both sulfate and nitrate measurement at the two sites and was then used to evaluate other instruments. For the un-denuded RAAS, the extent of sampling artifacts for nitrate on quartz filters was negligible, while that on Teflon filters was also minimal at high nitrate concentrations (>10 μgm−3); however, loss through evaporation was significant (∼75%) at low nitrate concentrations under hot summer conditions. The MOUDI using aluminum substrates suffered a significant loss of nitrate (50–70%) under summer conditions due to evaporation. Considering that the aluminum substrates are still being widely used to obtain size-resolved aerosol compositions because of their low cost and accurate mass weighed, caution should be taken about the potential significant under determination of semi-volatile components such as ammonium nitrate.
Keywords: Particulate matter; Comparison; Denuded filter-pack system; AIM; Artifacts; Aluminum substrate; Evaporation loss;

Aerosol optical properties of regional background atmosphere in Northeast China by Peng Wang; Huizheng Che; Xiaochun Zhang; Qingli Song; Yaqiang Wang; Zhonghua Zhang; Xin Dai; Dajiang Yu (4404-4412).
Aerosol optical properties from 2005 to 2008 at the Longfengshan regional background station in Northeast China were measured and analyzed. The annual mean of aerosol optical depth (AOD) at 440 nm for the four years was about 0.27 ± 0.25, 0.39 ± 0.37, 0.35 ± 0.34, and 0.38 ± 0.38, respectively, and the corresponding annual mean for the Angstrom exponent between 440 nm and 870 nm was about 1.43 ± 0.48, 1.23 ± 0.37, 1.53 ± 0.47, and 1.55 ± 0.42. The average monthly AOD440nm showed similar seasonal variation with a maximum in spring and a minimum in autumn. The monthly means of AOD at 440, 675, 870 and 1020 nm increase from the January to March with the maxima about 0.77 ± 0.04, 0.65 ± 0.04, 0.58 ± 0.06, 0.57 ± 0.07, respectively and decrease from September to February with the minima about 0.32 ± 0.12, 0.22 ± 0.09, 0.15 ± 0.08, and 0.13 ± 0.07 in January. The monthly mean of Angstrom exponent shows a minimum in March (0.97 ± 0.52) and a maximum in September (1.66 ± 0.29). Both the AOD and Angstrom exponent presents single peak distributions of occurrence frequencies. The Longfenshan data showed high AODs (>1.00) both clustering in the fine mode growth wing and the coarse mode. Two typical cases under dust and haze conditions showed that the AOD under dusty day decreased from 2.20 to 1.20 and the Angstrom exponent increased from 0.10 to 1.00. On the contrast, the AOD under haze day remained relatively stable about 0.90 and the Angstrom exponent was around 1.40. The 3-day backtrajectory analysis at Longfengshan illustrated that the air-masses near ground on the dust day were from Bohai Sea and passed through Liaodong Peninsula and Northeast plain in China. But the air-masses on 500 m AGL were originated from western Mongolia and crossed Gobi deserts, Otindag Sand Land and Horqin Sand Land in Northeast China. The air-masses at Longfengshan near ground 500 m and 1000 m AGL on the haze days were from North China Region and passed through Northeast Heavy Industrial Base in Northeast China.
Keywords: Aerosol optical depth; Angstrom exponent; Longfengshan background station; Backtrajectory analysis; China;

Two years of continuous particle size distribution measurements at three heights over and in a deciduous forest in southern Indiana exhibit evidence of particle nucleation on 46% of classifiable days, and clear nucleation with consistent subsequent growth on 14% of days. Events characterized by high ultra-fine particle production and clear growth are most frequent in spring. These events most frequently follow cold front passages and exhibit a higher frequency of northerly back-trajectories. Local meteorological conditions, as described using the Nucleation Parameter (NP) of and stability indices, indicate a significant association with ultra-fine particle concentrations, and specifically that nucleation was frequently preceded by destabilization of the atmosphere consistent with entrainment of elevated pollution layers. There appears to be little or no dependence of nucleation frequency or intensity on the in situ condensational sink particularly during spring, which is consistent with very high regional emissions of the precursors of ternary nucleation. Vertical gradients of particle size distributions exhibit evidence for nucleation being focused above canopy and accordingly mean ratios of above and below canopy concentrations of ultra-fine particles (6–30 nm) just subsequent to nucleation are 1.8 and 1.7 respectively during leaf-on and leaf-off periods. The difference in vertical concentration ratios during leaf-on and leaf-off are used to infer that of the order of 25% of ultra-fine particles are captured by foliar elements in the canopy during turbulent mixing through the canopy.
Keywords: Particle nucleation; Synoptic meteorology; Back-trajectories; Vertical profiles; Canopy uptake;

An adaptive reduction algorithm for efficient chemical calculations in global atmospheric chemistry models by Mauricio Santillana; Philippe Le Sager; Daniel J. Jacob; Michael P. Brenner (4426-4431).
We present a computationally efficient adaptive method for calculating the time evolution of the concentrations of chemical species in global 3-D models of atmospheric chemistry. Our strategy consists of partitioning the computational domain into fast and slow regions for each chemical species at every time step. In each grid box, we group the fast species and solve for their concentration in a coupled fashion. Concentrations of the slow species are calculated using a simple semi-implicit formula. Separation of species between fast and slow is done on the fly based on their local production and loss rates. This allows for example to exclude short-lived volatile organic compounds (VOCs) and their oxidation products from chemical calculations in the remote troposphere where their concentrations are negligible, letting the simulation determine the exclusion domain and allowing species to drop out individually from the coupled chemical calculation as their production/loss rates decline. We applied our method to a 1-year simulation of global tropospheric ozone-NO x -VOC-aerosol chemistry using the GEOS-Chem model. Results show a 50% improvement in computational performance for the chemical solver, with no significant added error.
Keywords: Atmospheric chemistry; Multi-scale analysis; Time-scale separation; Reduction of chemical kinetics;

Effect of vegetation removal and water table drawdown on the non-methane biogenic volatile organic compound emissions in boreal peatland microcosms by Patrick Faubert; Päivi Tiiva; Åsmund Rinnan; Sanna Räty; Jarmo K. Holopainen; Toini Holopainen; Riikka Rinnan (4432-4439).
Biogenic volatile organic compound (BVOC) emissions are important in the global atmospheric chemistry and their feedbacks to global warming are uncertain. Global warming is expected to trigger vegetation changes and water table drawdown in boreal peatlands, such changes have only been investigated on isoprene emission but never on other BVOCs. We aimed at distinguishing the BVOCs released from vascular plants, mosses and peat in hummocks (dry microsites) and hollows (wet microsites) of boreal peatland microcosms maintained in growth chambers. We also assessed the effect of water table drawdown (−20 cm) on the BVOC emissions in hollow microcosms. BVOC emissions were measured from peat samples underneath the moss surface after the 7-week-long experiment to investigate whether the potential effects of vegetation and water table drawdown were shown. BVOCs were sampled using a conventional chamber method, collected on adsorbent and analyzed with GC–MS. In hummock microcosms, vascular plants increased the monoterpene emissions compared with the treatment where all above-ground vegetation was removed while no effect was detected on the sesquiterpenes, other reactive VOCs (ORVOCs) and other VOCs. Peat layer from underneath the surface with intact vegetation had the highest sesquiterpene emissions. In hollow microcosms, intact vegetation had the highest sesquiterpene emissions. Water table drawdown decreased monoterpene and other VOC emissions. Specific compounds could be closely associated to the natural/lowered water tables. Peat layer from underneath the surface of hollows with intact vegetation had the highest emissions of monoterpenes, sesquiterpenes and ORVOCs whereas water table drawdown decreased those emissions. The results suggest that global warming would change the BVOC emission mixtures from boreal peatlands following changes in vegetation composition and water table drawdown.
Keywords: Monoterpene; Sesquiterpene; Mire; Peat; Climate change; Sphagnum;

The major components of particles emitted during recycling of waste printed circuit boards in a typical e-waste workshop of South China by Xinhui Bi; Bernd R.T. Simoneit; ZhenZhen Wang; Xinming Wang; Guoying Sheng; Jiamo Fu (4440-4445).
Electronic waste from across the world is dismantled and disposed of in China. The low-tech recycling methods have caused severe air pollution. Air particle samples from a typical workshop of South China engaged in recycling waste printed circuit boards have been analyzed with respect to chemical constituents. This is the first report on the chemical composition of particulate matter (PM) emitted in an e-waste recycling workshop of South China. The results show that the composition of PM from this recycling process was totally different from other emission sources. Organic matter comprised 46.7–51.6% of the PM. The major organic constituents were organophosphates consisting mainly of triphenyl phosphate (TPP) and its methyl substituted compounds, methyl esters of hexadecanoic and octadecanoic acids, levoglucosan and bisphenol A. TPP and bisphenol A were present at 1–5 orders of magnitude higher than in other indoor and outdoor environments throughout the world, which implies that they might be used as potential markers for e-waste recycling. The elemental carbon, inorganic elements and ions had a minor contribution to the PM (<5% each). The inorganic elements were dominated by phosphorus and followed by crustal elements and metal elements Pb, Zn, Sn, and lesser Cu, Sb, Mn, Ni, Ba and Cd. The recycling of printed circuit boards was demonstrated as an important contributor of heavy metal contamination, particularly Cd, Pb and Ni, to the local environment. These findings suggest that this recycling method represents a strong source of PM associated with pollutants to the ambient atmosphere of an e-waste recycling locale.
Keywords: Printed circuit board; Triphenyl phosphate; Flame retardants; Molecular marker; Bisphenol A; Electronic waste; Particulate matter; Indoor air; China;

Degradation kinetics of anthracene by ozone on mineral oxides by Jinzhu Ma; Yongchun Liu; Hong He (4446-4453).
To further understand the role of substrates on the heterogeneous reactions of polycyclic aromatic hydrocarbons, the reactions of ozone with anthracene adsorbed on different mineral oxides (SiO2, α-Al2O3 and α-Fe2O3) and on Teflon disc were investigated in dark at 20 °C. No reaction between ozone and anthracene on Teflon disc was observed when the ozone concentration was ∼1.18 × 1014 molecules cm−3. The reactions on mineral oxides exhibited pseudo-first-order kinetics for anthracene loss, and the pseudo-first-order rate constant (k 1,obs) displayed a Langmuir–Hinshelwood dependence on the gas-phase ozone concentration. The adsorption equilibrium constants for ozone (K O3) on SiO2-1, SiO2-2, α-Al2O3 and α-Fe2O3 were (0.81 ± 0.26) × 10−15 cm3, (2.83 ± 1.17) × 10−15 cm3, (2.48 ± 0.77) × 10−15 cm3 and (1.66 ± 0.45) × 10−15 cm3, respectively; and the maximum pseudo-first-order rate constant (k 1,max) on these oxides were (0.385 ± 0.058) s−1, (0.101 ± 0.0138) s−1, (0.0676 ± 0.0086) s−1 and (0.0457 ± 0.004) s−1, respectively. Anthraquinone was identified as the main surface product of anthracene reacted with ozone. Comparison with previous research and the results obtained in this study suggest that the reactivity of anthracene with ozone and the lifetimes of anthracene adsorbed on mineral dust in the atmosphere are determined by the nature of the substrate.
Keywords: Anthracene; Mineral oxides; Ozone; Heterogeneous reactions; Kinetics;

Analysis of 3-year observations of CFC-11, CFC-12 and CFC-113 from a semi-rural site in China by Fang Zhang; Lingxi Zhou; Bo Yao; Martin. K. Vollmer; Brian R. Greally; Peter G. Simmonds; Stefan Reimann; Frode Stordal; Michela Maione; Lin Xu; Xiaochun Zhang (4454-4462).
In-situ measurements of atmospheric chlorofluorocarbons (CFCs) can be used to the assess their global and regional emissions and to check for compliance with phase-out schedules under Montreal protocol and its amendments. The atmospheric mixing ratios of CFC-11 (CCl3F), CFC-12 (CCl2F2) and CFC-113 (CCl2F–CClF2) have been measured by an automated in-situ GC-ECDs system at the regional Chinese Global Atmosphere Watch (GAW) station Shangdianzi (SDZ), from November 2006 to October 2009. The time series for these three principal CFCs showed large episodic events and background conditions occurred for approximately 30% (CFC-11), 52% (CFC-12) and 56% (CFC-113) of the measurements. The mean background mixing ratios for CFC-11, CFC-12 and CFC-113 were 244.8 ppt (parts per trillion, 10−12, molar) 539.6 ppt and 76.8 ppt, respectively, for 2006–2009. The enhanced CFC mixing ratios compared to AGAGE sites such as Trinidad Head (THD), US and Mace Head (MHD), Ireland suggest regional influences even during background conditions at SDZ, which is much closer to highly-populated areas. Between 2006 and 2009 background CFCs exhibited downward trends at rates of −2.0 ppt yr−1 for CFC-11, −2.5 ppt yr−1 for CFC-12 and −0.7 ppt yr−1 for CFC-113. De-trended 3-year average background seasonal cycles displayed small fluctuations with peak-to-trough amplitudes of 1.0 ± 0.02 ppt (0.4%) for background CFC-11, 1.3 ± 2.1 ppt (0.3%) for CFC-12 and 0.2 ± 0.4 ppt (0.3%) for CFC-113. On the other hand, during pollution periods these CFCs showed much larger seasonal cycles of 11.2 ± 10.7 ppt (5%) for CFC-11, 7.5 ± 6.5 ppt (2%) for CFC-12 and 1.0 ± 1.2 ppt (1.2%) for CFC-113, with apparent winter minima and early summer maxima. This enhancement was attributed to prevailing wind directions from urban regions in summer and to enhanced anthropogenic sources during the warm season. In general, horizontal winds from northeast showed negative contribution to atmospheric CFCs loading, whereas South Western advection (urban sector: Beijing) had positive contributions.
Keywords: CFCs; Shangdianzi; GAW; Seasonality; Data analysis;

Process analysis of ozone formation under different weather conditions over the Kanto region of Japan using the MM5/CMAQ modelling system by Mai Khiem; Ryozo Ooka; Hiroshi Hayami; Hiroshi Yoshikado; Hong Huang; Yoichi Kawamoto (4463-4473).
We have assessed the contributions of individual physical and chemical atmospheric processes on ozone formation under different weather conditions during a typical summer month (August 2005) using the MM5/CMAQ modelling system. We found that the ozone episodes in the Kanto region are dominated by three major patterns, of which Patterns I and II are regular summertime pressure patterns with a 26% and 16% frequency of occurrence, respectively. A process analysis at two typical sites in the Kanto region – one located in the central region of Tokyo and the other located in the rural areas of Kanto – indicates that ozone formation is mainly controlled by advection, vertical diffusion, dry deposition, and chemical processes. The ground-level ozone concentrations are enhanced mainly by the vertical mixing of ozone-rich air from aloft, whereas the dry deposition and chemical processes mainly deplete ozone. By investigating the effects of each process under different weather conditions, we found that the significant decrease in ozone removal due to the chemical and advection processes under conditions of high stagnation is the most important cause of the enhanced levels of ozone in the central region of Tokyo. The results of this study can contribute to a better understanding of ozone formation in the Kanto region, and they may be valuable for local policy makers for further development planning.
Keywords: Ozone; MM5; CMAQ; Process analysis; Weather patterns;

Characterisation of PM10 emissions from woodstove combustion of common woods grown in Portugal by Cátia Gonçalves; Célia Alves; Margarita Evtyugina; Fátima Mirante; Casimiro Pio; Alexandre Caseiro; Christoph Schmidl; Heidi Bauer; Fernando Carvalho (4474-4480).
A series of source tests was performed to evaluate the chemical composition of particle emissions from the woodstove combustion of four prevalent Portuguese species of woods: Pinus pinaster (maritime pine), Eucalyptus globulus (eucalyptus), Quercus suber (cork oak) and Acacia longifolia (golden wattle). Analyses included water-soluble ions, metals, radionuclides, organic and elemental carbon (OC and EC), humic-like substances (HULIS), cellulose and approximately l80 organic compounds. Particle (PM10) emission factors from eucalyptus and oak were higher than those from pine and acacia. The carbonaceous matter represented 44–63% of the particulate mass emitted during the combustion process, regardless of species burned. The major organic components of smoke particles, for all the wood species studied, with the exception of the golden wattle (0.07–1.9% w/w), were anhydrosugars (0.2–17% w/w). Conflicting with what was expected, only small amounts of cellulose were found in wood smoke. As for HULIS, average particle mass concentrations ranged from 1.5% to 3.0%. The golden wattle wood smoke presented much higher concentrations of ions and metal species than the emissions from the other wood types. The results of the analysis of radionuclides revealed that the 226Ra was the naturally occurring radionuclide more enriched in PM10. The chromatographically resolved organics included n-alkanes, n-alkenes, PAH, oxygenated PAH, n-alkanals, ketones, n-alkanols, terpenoids, triterpenoids, phenolic compounds, phytosterols, alcohols, n-alkanoic acids, n-di-acids, unsaturated acids and alkyl ester acids.
Keywords: Biomass burning; Woodstove; PM10; Emissions; Organic tracers; GC–MS;

Prediction of hourly O3 concentrations using support vector regression algorithms by E.G. Ortiz-García; S. Salcedo-Sanz; Á.M. Pérez-Bellido; J.A. Portilla-Figueras; L. Prieto (4481-4488).
In this paper we present an application of the Support Vector Regression algorithm (SVMr) to the prediction of hourly ozone values in Madrid urban area. In order to improve the training capacity of SVMrs, we have used a recently proposed approach, based on reductions of the SVMr hyper-parameters search space. Using the modified SVMr, we study different influences which may modify the ozone prediction, such as previous ozone measurements in a given station, measurements in neighbors stations, and the influence of meteorologic variables. We use statistical tests to verify the significance of incorporating different variables into the SVMr. A comparison with the results obtained using a neural network (multi-layer perceptron) is also carried out. This study has been carried out in 5 different stations of the air pollution monitoring network of Madrid, so the conclusions raised are backed by real data. The final result of the work is a robust and powerful software for tropospheric ozone prediction in Madrid. Also, the prediction tool based on SVMr is flexible enough to incorporate any other prediction variable, such as city models, or traffic patters, which may improve the prediction obtained with the SVMr.
Keywords: O3 concentration prediction; Support vector regression algorithms; Air quality;

We investigate the long-range transport potential (LRTP) of five different classes of hypothetical chemical pollutants (volatile, multimedia, semivolatile, particle-associated and hydrophilic) during a low pressure weather event using a novel 2 (x- and z-axis)-Dimensional Multi-Media Meteorological Model (2D4M). The atmosphere (z-axis) is described by three atmospheric layers, where two layers constitute the boundary layer and the third layer the free troposphere. The 2D4M can describe distinct weather events on a regional scale and calculate the LRTP of chemicals as a function of time during these events. Four weather factors are used to model weather events and their influence on the atmospheric transport of chemicals: (1) temperature, (2) wind speed and mixing dynamics of the troposphere, (3) hydroxyl radical concentrations and (4) precipitation. We have modeled the impact of variability in each of these factors on LRTP of pollutants during a front event associated with a low pressure period that interrupts a dominant high pressure system. The physico-chemical properties of the pollutant determine which specific weather factors contribute most to variability in transport potential during the event. Volatile and multimedia chemicals are mainly affected by changing atmospheric mixing conditions, wind speeds and OH radical concentrations, while semivolatile substances are also affected by temperature. Low-vapor-pressure pollutants that are particle-associated, and water-soluble pollutants are most strongly affected by precipitation. Some chemical pollutants are efficiently transported from the boundary layer into the upper troposphere during the modeled low pressure event and are transported by much higher wind speeds than in the boundary layer. Our model experiments show that the transport potential of volatile, multimedia and semivolatile compounds is significantly increased during a front event as a result of efficient tropospheric mixing and fast wind speeds in the upper troposphere, whereas low-volatility and hydrophilic chemicals are largely scavenged from the atmosphere. In future LRTP assessment of chemical contaminants as required by the Stockholm Convention and the convention on long-range transboundary air pollution, it is therefore advised to prioritize volatile, multimedia and semivolatile chemicals that are identified in initial screening.
Keywords: Long-range transport potential; Chemical pollutants; Meteorology; Multimedia model; Screening assessment;

Enhanced ozone over western North America from biomass burning in Eurasia during April 2008 as seen in surface and profile observations by S.J. Oltmans; A.S. Lefohn; J.M. Harris; D.W. Tarasick; A.M. Thompson; H. Wernli; B.J. Johnson; P.C. Novelli; S.A. Montzka; J.D. Ray; L.C. Patrick; C. Sweeney; A. Jefferson; T. Dann; J. Davies; M. Shapiro; B.N. Holben (4497-4509).
During April 2008, as part of the International Polar Year (IPY), a number of ground-based and aircraft campaigns were carried out in the North American Arctic region (e.g., ARCTAS, ARCPAC). The widespread presence during this period of biomass burning effluent, both gaseous and particulate, has been reported. Unusually high ozone readings for this time of year were recorded at surface ozone monitoring sites from northern Alaska to northern California. At Barrow, Alaska, the northernmost point in the United States, the highest April ozone readings recorded at the surface (hourly average values >55 ppbv) in 37 years of observation were measured on April 19, 2008. At Denali National Park in central Alaska, an hourly average of 79 ppbv was recorded during an 8-h period in which the average was over 75 ppbv, exceeding the ozone ambient air quality standard threshold value in the U.S. Elevated ozone (>60 ppbv) persisted almost continuously from April 19–23 at the monitoring site during this event. At a coastal site in northern California (Trinidad Head), hourly ozone readings were >50 ppbv almost continuously for a 35-h period from April 18–20. At several sites in northern California, located to the east of Trinidad Head, numerous occurrences of ozone readings exceeding 60 ppbv were recorded during April 2008. Ozone profiles from an extensive series of balloon soundings showed lower tropospheric features at ∼1–6 km with enhanced ozone during the times of elevated ozone amounts at surface sites in western Canada and the U.S. Based on extensive trajectory calculations, biomass burning in regions of southern Russia was identified as the likely source of the observed ozone enhancements. Ancillary measurements of atmospheric constituents and optical properties (aerosol optical thickness) supported the presence of a burning plume at several locations. At two coastal sites (Trinidad Head and Vancouver Island), profiles of a large suite of gases were measured from airborne flask samples taken during probable encounters with burning plumes. These profiles aided in characterizing the vertical thickness of the plumes, as well as confirming that the plumes reaching the west coast of North America were associated with biomass burning events.
Keywords: Ozone; Biomass burning; Pollution; Arctic; Transport; Trajectories;

Comparison of summer and winter California central valley aerosol distributions from lidar and MODIS measurements by Jasper Lewis; Russell De Young; Richard Ferrare; D. Allen Chu (4510-4520).
Aerosol distributions from two aircraft lidar campaigns conducted in the California Central Valley are compared in order to identify seasonal variations. Aircraft lidar flights were conducted in June 2003 and February 2007. While the ground PM2.5 (particulate matter with diameter ≤ 2.5 μm) concentration was highest in the winter, the aerosol optical depth (AOD) measured from the MODIS and lidar instruments was highest in the summer. A multiyear seasonal comparison shows that PM2.5 in the winter can exceed summer PM2.5 by 68%, while summer AOD from MODIS exceeds winter AOD by 29%. Warmer temperatures and wildfires in the summer produce elevated aerosol layers that are detected by satellite measurements, but not necessarily by surface particulate matter monitors. Temperature inversions, especially during the winter, contribute to higher PM2.5 measurements at the surface. Measurements of the mixing layer height from lidar instruments provide valuable information needed to understand the correlation between satellite measurements of AOD and in situ measurements of PM2.5. Lidar measurements also reflect the ammonium nitrate chemistry observed in the San Joaquin Valley, which may explain the discrepancy between the MODIS AOD and PM2.5 measurements.
Keywords: MODIS; Aerosol; California; San Joaquin Valley; Lidar;

Effect of fireworks on ambient air quality in Malta by Renato Camilleri; Alfred J. Vella (4521-4527).
Religious festivals (festas) in the densely populated Maltese archipelago (Central Mediterranean) are ubiquitous during summer when 86 of them are celebrated between June and October, each involving the burning of fireworks both in ground and aerial displays over a period of 3 days or longer per festival. We assessed the effect of fireworks on the air quality by comparing PM10 and its content of Al, Ba, Cu, Sr and Sb which materials are used in pyrotechnic compositions. PM10 was collected mainly from two sites, one in Malta (an urban background site) and the other in Gozo (a rural site) during July–August 2005 when 59 feasts were celebrated and September–October 2005 when only 11 feasts occurred. For both Malta and Gozo, PM10 and metal concentration levels measured as weekly means were significantly higher during July–August compared to September–October and there exist strong correlations between PM10 and total metal content. Additionally, for Malta dust, Al, Ba, Cu and Sr correlated strongly with each other and also with total concentration of all five metals. The same parameters measured in April 2006 in Malta were at levels similar to those found in the previous October. Ba and Sb in dust from the urban background site in Malta during July–August were at comparable or higher concentration than recently reported values in PM10 from a heavily-trafficked London road and this suggests that these metals are locally not dominated by sources from roadside materials such as break liner wear but more likely by particulate waste from fireworks. Our findings point to the fact that festa firework displays contribute significantly and for a prolonged period every year to airborne dust in Malta where PM10 is an intractable air quality concern. The presence in this dust of elevated levels of Ba and especially Sb, a possible carcinogen, is of concern to health.
Keywords: Fireworks; Air quality; Malta; PM10; Pyrotechnics;