Atmospheric Environment (v.45, #31)

PM2.5 source profiles for black and organic carbon emission inventories by Judith C. Chow; John G. Watson; Douglas H. Lowenthal; L.-W. Antony Chen; Nehzat Motallebi (5407-5414).
Emission inventories for black or elemental (BC or EC) and organic (OC) carbon can be derived by multiplying PM2.5 emission estimates by mass fractions of these species in representative source profiles. This study examines the variability of source profiles and its effect on EC emission estimates. An examination of available profiles shows that EC and OC ranged from 6–13% and 35–40% for agricultural burning, 4–33% and 22–68% for residential wood combustion, 6–38% and 24–75% for on-road gasoline vehicles, and 33–74% and 20–47% for on-road heavy-duty diesel vehicles, respectively. Source profiles from the U.S. EPA SPECIATE data base were applied to PM2.5 emissions from the U.S. EPA National Emissions Inventory for 2005. The total estimated EC emissions of 432 Gg yr−1 was apportioned as 42.5% from biomass burning, 35.4% from non-road mobile sources, 15% from on-road mobile sources, 5.4% from fossil fuel (e.g., coal, oil, and natural gas) combustion in stationary sources, 1% from other stationary industrial sources, and 0.5% from fugitive dust. Considering the variability in available source profiles, BC emission estimates for major sources such as open fires and non-road diesels ranged from 42 to 133 (a factor of 3) and 25 to 100 (a factor of 4) Gg yr−1, respectively. The choice of source profiles can be a major source of uncertainty in national and global BC/EC emission inventories.► PM2.5 emission inventories are used for black carbon inventories related to climate. ► BC content depends on the source profiles selected for large carbon emitters. ► BC profiles are variable for major sources and uncertainties are seldom considered.
Keywords: Source profile; Emission inventory; PM2.5; Black carbon; Elemental carbon; Organic carbon; Climate;

Air quality and public health impacts of UK airports. Part I: Emissions by M.E.J. Stettler; S. Eastham; S.R.H. Barrett (5415-5424).
The potential adverse human health and climate impacts of emissions from UK airports have become a significant political issue, yet the emissions, air quality impacts and health impacts attributable to UK airports remain largely unstudied. We produce an inventory of UK airport emissions – including aircraft landing and takeoff (LTO) operations and airside support equipment – with uncertainties quantified. The airports studied account for more than 95% of UK air passengers in 2005. We estimate that in 2005, UK airports emitted 10.2 Gg [−23 to +29%] of NOx, 0.73 Gg [−29 to +32%] of SO2, 11.7 Gg [−42 to +77%] of CO, 1.8 Gg [−59 to +155%] of HC, 2.4 Tg [−13 to +12%] of CO2, and 0.31 Gg [−36 to +45%] of PM2.5. This translates to 2.5 Tg [−12 to +12%] CO2-eq using Global Warming Potentials for a 100-year time horizon. Uncertainty estimates were based on analysis of data from aircraft emissions measurement campaigns and analyses of aircraft operations.The First-Order Approximation (FOA3) – currently the standard approach used to estimate particulate matter emissions from aircraft – is compared to measurements and it is shown that there are discrepancies greater than an order of magnitude for 40% of cases for both organic carbon and black carbon emissions indices. Modified methods to approximate organic carbon emissions, arising from incomplete combustion and lubrication oil, and black carbon are proposed. These alterations lead to factor 8 and a 44% increase in the annual emissions estimates of black and organic carbon particulate matter, respectively, leading to a factor 3.4 increase in total PM2.5 emissions compared to the current FOA3 methodology. Our estimates of emissions are used in Part II to quantify the air quality and health impacts of UK airports, to assess mitigation options, and to estimate the impacts of a potential London airport expansion.► We develop a comprehensive emissions inventory for UK airports. ► Scientific, technical and operational uncertainties are assessed. ► Confidence intervals and contributions to uncertainties in emissions are derived. ► Part II uses this inventory to assess air quality impacts of UK airports.
Keywords: Aviation; Airport; Emissions; Air quality; Particulate matter;

Estimated accuracy of three common trajectory statistical methods by Vitaliy P. Kabashnikov; Anatoli P. Chaikovsky; Tom L. Kucsera; Natalia S. Metelskaya (5425-5430).
Three well-known trajectory statistical methods (TSMs), namely concentration field (CF), concentration weighted trajectory (CWT), and potential source contribution function (PSCF) methods were tested using known sources and artificially generated data sets to determine the ability of TSMs to reproduce spatial distribution of the sources. In the works by other authors, the accuracy of the trajectory statistical methods was estimated for particular species and at specified receptor locations. We have obtained a more general statistical estimation of the accuracy of source reconstruction and have found optimum conditions to reconstruct source distributions of atmospheric trace substances. Only virtual pollutants of the primary type were considered.In real world experiments, TSMs are intended for application to a priori unknown sources. Therefore, the accuracy of TSMs has to be tested with all possible spatial distributions of sources. An ensemble of geographical distributions of virtual sources was generated. Spearman’s rank order correlation coefficient between spatial distributions of the known virtual and the reconstructed sources was taken to be a quantitative measure of the accuracy. Statistical estimates of the mean correlation coefficient and a range of the most probable values of correlation coefficients were obtained. All the TSMs that were considered here showed similar close results.The maximum of the ratio of the mean correlation to the width of the correlation interval containing the most probable correlation values determines the optimum conditions for reconstruction. An optimal geographical domain roughly coincides with the area supplying most of the substance to the receptor. The optimal domain’s size is dependent on the substance decay time. Under optimum reconstruction conditions, the mean correlation coefficients can reach 0.70–0.75. The boundaries of the interval with the most probable correlation values are 0.6–0.9 for the decay time of 240 h and 0.5–0.95 for the decay time of 12 h. The best results of source reconstruction can be expected for the trace substances with a decay time on the order of several days.Although the methods considered in this paper do not guarantee high accuracy they are computationally simple and fast. Using the TSMs in optimum conditions and taking into account the range of uncertainties, one can obtain a first hint on potential source areas.► Accuracy was assessed by comparing reconstructed sources with known ones. ► Typical values of the accuracy and a range of its variability have been obtained. ► Optimal conditions for reconstruction of spatial distribution of the sources have been found.
Keywords: Trajectory statistical method; Back trajectories; Source reconstruction; Atmospheric trace substance;

Comparison of ozone concentrations on a surface elevation gradient with balloon-borne ozonesonde measurements by Molly Brodin; Detlev Helmig; Bryan Johnson; Samuel Oltmans (5431-5439).
Ambient ozone data collected on a ground-level elevation transect in the Colorado Front Range Mountains, spanning an elevation range from 1660 m above sea level (ASL) to 3530 m ASL, were compared to vertical ozonesonde measurements obtained from a National Oceanic and Atmospheric Administration (NOAA) launch site located just a few kilometers from the lowest elevation site on the surface site elevation transect. Ozonesondes were launched approximately weekly throughout the one-year monitoring period from September 1, 2007 to August 31, 2008, and daily from June 27 to July 11, 2008. Ozonesonde profile data were found to display a variable agreement with the data from the surface sites along the elevation transect. The strongest determinants of the ozone profile behavior were found to be the vertical atmospheric potential temperature profile and season. The surface elevation ozone profile generally matched the sonde profile quite well, capturing all of the same major features. A seasonal relationship for surface–ozonesonde concentration differences was found. This difference is smallest in winter and spring, and tends to be slightly positive (i.e. higher ozone at surface sites), with three-month seasonal median values up to 4 ppbv. The difference turns negative (i.e. lower ozone at surface sites) in summer–fall, with seasonal median values becoming as large as −13 ppbv. The seasonal cycle in surface–ozonesonde differences was minimal for the lowest monitoring site at 1660 m ASL. The magnitude of seasonal change increased with elevation to a maximum of 22 ppbv at 3350 m ASL, before then dropping again by more than half at the highest site, which, at 3530 m ASL, is frequently above the mixed layer boundary. Seasonally dependent surface processes, including ozone production in the surface layer and deposition processes are the most likely determinants of seasonal variation in the surface–ozonesonde data relationship.► One year of surface and balloon-borne ozone measurements along a ∼2000 m elevation transect in the Colorado Rocky Mountains. ► There generally was good agreement in the two data sets. ► Deviations between the two data sets show a seasonal and altitude dependence.
Keywords: Ozone; Surface sites; Mountain transect; High elevation; Ozonesonde;

Sources of mercury in precipitation to Underhill, VT by Lynne E. Gratz; Gerald J. Keeler (5440-5449).
Daily event precipitation samples were collected from 1995 to 2006 in Underhill, VT, USA and analyzed for mercury (Hg) and trace elements. Over the 12-year period, annual Hg deposition levels at Underhill did not decline significantly despite regulatory efforts to reduce Hg emissions in the United States. Mercury and trace element wet deposition data were examined using the multivariate receptor model EPA PMF 3.0 to identify the sources contributing to Hg deposition at the Underhill site. Results indicate that coal combustion, a mixture of incineration and non-ferrous metal smelting, and a phosphorus source contributed to the observed Hg wet deposition. The model consistently indicated that coal combustion contributed ∼60% of the total Hg measured in wet deposition and the contribution from this source did not change significantly over time. Mercury and trace element deposition were further analyzed using a hybrid-receptor model, quantitative transport bias analysis (QTBA), to identify the likely source locations contributing to Hg deposition. Results from QTBA indicate that the majority of Hg deposition at Underhill is due to transport from the Midwestern United States, where the density of coal-fired utility boilers in the U.S. is largest.► We examined 12 years of Hg and trace element wet deposition amounts measured at Underhill, VT. ► Using a multivariate receptor model, we attributed ∼60% of the Hg wet deposition to CFUB emissions. ► A hybrid-receptor model largely associated Hg wet deposition with transport from the U.S. Midwest.
Keywords: Mercury; Precipitation; Trace elements; Source apportionment; Receptor modeling;

Elemental concentration and source identification of PM10 and PM2.5 by SR-XRF in Córdoba City, Argentina by María Laura López; Sergio Ceppi; Gustavo G. Palancar; Luis E. Olcese; Germán Tirao; Beatriz M. Toselli (5450-5457).
24-h samplings of PM10 and PM2.5 have been carried out during the period July 2009–April 2010 at an urban and at a semi-urban site of Córdoba City (Argentina). The samples in the PM2.5 fraction weighted in the average 71 ± 21 μg m−3 and 67 ± 18 μg m−3 respectively, whereas the samples of the same sites in the PM10 fraction weighted 107 ± 31 μg m−3 and 101 ± 14 μg m−3. The chemical composition of aerosol particles was determined by synchrotron radiation X-ray fluorescence (SR-XRF). Elemental composition was different in the two fractions: in the finer one the presence of elements with crustal origin is reduced, while the anthropogenic elements, with a relevant environmental and health impact, appear to be increased. An important but unmeasured component is likely constituted by organic and elemental carbon compounds. Multivariate analysis (Positive Matrix Factorization) of the SR-XRF data resolved a number of components (factors) which, on the basis of their chemical compositions, were assigned physical meanings.► Córdoba has dangerous levels of particulate pollution. ► Toxic metal concentrations are heavily influenced by local sources. ► Motor vehicles and road/construction dust are the major contributors to fine particle mass.
Keywords: Córdoba City; PM2.5; PM10; X-ray fluorescence; Synchrotron radiation;

Spatiotemporal variations of nitrous oxide (N2O) emissions from two reservoirs in SW China by Xiao-Long Liu; Cong-Qiang Liu; Si-Liang Li; Fu-Shun Wang; Bao-Li Wang; Zhong-Liang Wang (5458-5468).
Greenhouse gas emissions from hydroelectric dams have recently given rise to controversies about whether hydropower still provides clean energy. China has a large number of dams used for energy supply and irrigation, but few studies have been carried out on aquatic nitrous oxide (N2O) variation and its emissions in Chinese river-reservoir systems. In this study, N2O spatiotemporal variations were investigated monthly in two reservoirs along the Wujiang River, Southwest China, and the emission fluxes of N2O were estimated. N2O production in the reservoirs tended to be dominated by nitrification, according to the correlation between N2O and other parameters. N2O saturation in the surface water of the Wujiangdu reservoir ranged from 214% to 662%, with an average fluctuation of 388%, while in the Hongjiadu reservoir, it ranged from 201% to 484%, with an average fluctuation of 312%. The dissolved N2O in both reservoirs was over-saturated with respect to atmospheric equilibrium levels, suggesting that the reservoirs were net sources of N2O emissions to the atmosphere. The averaged N2O emission flux in the Wujiangdu reservoir was 0.64 μmol m−2 h−1, while it was 0.45 μmol m−2 h−1 in the Hongjiadu reservoir, indicating that these two reservoirs had moderate N2O emission fluxes as compared to other lakes in the world. Downstream water of the dams had quite high levels of N2O saturation, and the estimated annual N2O emissions from hydropower generation were 3.60 × 105 and 2.15 × 105 mol N2O for the Wujiangdu and the Hongjiadu reservoir, respectively. These fluxes were similar to the total N2O emissions from the reservoir surfaces, suggesting that water released from reservoirs would be another important way for N2O to diffuse into the atmosphere. It can be concluded that dam construction significantly changes the water environment, especially in terms of nutrient status and physicochemical conditions, which have obvious influences on the N2O spatiotemporal variations and emissions.► We examined the nutrition status and the impact of anthropogenic N input on N2O production and saturation in reservoirs. ► The nitrous oxide emissions from reservoirs in subtropical region have been reported. ► Reservoir surface represented a continuous source of N2O to the atmosphere yearly. ► N2O emission by discharged water during hydropower generation was significantly large, and hence can’t be ignored.
Keywords: Nitrous oxide; Reservoirs; Spatiotemporal variation; Flux;

Development of a driving cycle for intra-city buses in Chennai, India by K.S. Nesamani; K.P. Subramanian (5469-5476).
In India the emissions rate and fuel consumption of intra-city buses are estimated using the European driving cycles, which don’t represent Indian driving conditions and in-use operation of vehicles. This leads to underestimation or overestimation of emissions and fuel consumption. In this context, this paper offers some insight into the driving characteristics of intra-city buses using a Global Positioning System. The study has revealed that irrespective of road type and time of travel, a higher percentage of time is spent in idle mode. This is primarily due to alighting and boarding of passengers at regular intervals and fixed delays caused by traffic lights. More than 90 percent of trips have an average speed of less than 30 km h−1. This study has also developed an intra-city bus driving cycle for Chennai and compared it with some well-known international driving cycles. It has revealed that Chennai has unique driving characteristics and, therefore, it may not be appropriate to adopt a driving cycle of another country or city.► Road type and number of stops in each trip influences driving characteristics of intra-city buses. ► More than 90 percent of trips have an average speed of less than 30 km h−1. ► Irrespective of road type and time of travel, a higher percentage of time is spent in idle mode. ► A new intra-city bus driving cycle for Chennai was developed using the real-world GPS data. ► To improve fuel economy from intra-city buses, transit agencies should adopt eco-driving.
Keywords: Driving cycle; Driving characteristics; Intra-city buses; Global positioning system; Chennai;

δ13C and δD of volatile organic compounds in an alumina industry stack emission by Christiane Vitzthum von Eckstaedt; Kliti Grice; Marisa Ioppolo-Armanios; Mark Jones (5477-5483).
Compound specific isotope analysis (CSIA) is becoming more widely accepted as a tool for determining the sources of contaminants and monitoring their transport and fate in the environment. However, measuring δD of volatile organic compounds (VOCs) in atmospheric samples is still underexplored. The present study applies thermal desorption–gas chromatography–isotope ratio mass spectrometry (TD–GC–irMS) for the first time to measure stable hydrogen isotope analyses of VOCs in an alumina refinery emission. δ13C data is also collected. A sampling train was designed using TenaxTA as the adsorbent material to gain reliable and reproducible results for CSIA. δ13C values for VOCs (C6–C14) ranged from −22 to −31‰, which is similar to δ13C value range reported for naturally occurring components. The δD values (21 to −137‰) in this study were consistently more enriched in D compared to δD values of VOCs previously reported making the δ values of VOCs in the industrial stack unique. Therefore δD analysis may provide a means for tracking VOCs in atmospheric samples.► VOC in alumina refinery emission analysed by TD–GC–irMS (δ13C and δD). ► δ13C values of VOCs (C6–C14) similar to naturally occurring components. ► Unique δD values of VOCs. Consistently heavier compared to other reported δD values. ► Designed sampling train gains reliable and reproducible results.
Keywords: Compound specific isotope analysis; Stable hydrogen isotope analysis; Stable carbon isotope analysis; Thermal desorption; TenaxTA;

Estimates of aqueous-phase sulfate production from tandem differential mobility analysis by Joshua L. Santarpia; Don R. Collins; Dean A. Hegg; Kathleen C. Kaku; David S. Covert; Haflidi Jonsson; Gintautas Buzorius (5484-5492).
During June and July of 2004, airborne measurements of size-resolved aerosol properties were made over the coastal Pacific near Marina, California. Tandem differential mobility analysis was used to determine the hygroscopic properties of these aerosols and, subsequently, to examine the change in soluble mass after the aerosol had been cloud-processed. Three of the eight cases analyzed during the field campaign exhibited increased soluble mass attributable to cloud-processing. The calculated change in soluble mass after cloud-processing, derived from measurements made below and above cloud, ranged from 1.08 ϒ g m−3 to 1.40 ϒ g m−3. These values are in agreement with those determined using data collected during previous field studies in the same region. Aerosol size distributions measured throughout the study with a Passive Cavity Aerosol Spectrometer Probe were averaged to create a single representative distribution, which was used to examine the impact of the addition of sulfate to a typical aerosol measured during this project. Mass light-scattering efficiencies were calculated for both the initial and cloud-processed size distributions using Mie-Lorenz theory. These calculations show that the increase in mass light-scattering efficiency following passage through a single non-precipitating cloud to be 14%, which is in agreement with findings of previous studies in the same region. This new technique produces results that are consistent with existing methods for measuring in-cloud sulfate production and has the advantage of intrinsic consistency with the hygroscopicity and CCN activity of the aerosol.► TDMA measurements of the hygroscopic properties of aerosol above and below cloud. ► Change in calculated sulfate mass after cloud passage is calculated. ► Calculated sulfate mass is compared to theoretical calculations and past measurement.
Keywords: Aerosol; Sulfate; Cloud-processing; Tandem differential mobility analyzer;

A Bayesian multiple change-point model is proposed to analyse violations of air quality standards by pollutants such as nitrogen oxides (NO2 and NO) and carbon monoxide (CO). Threshold exceedance occurrences are modelled by a step rate Poisson process fitted after short-range correlations in the exceedance data are removed via declusterisation. The change-points are identified, and the rate function is estimated, using a reversible jump MCMC algorithm adapted from . This technique is applied to the daily concentration data collected in Leeds, UK (1993–2009). Results are validated by running the MCMC estimator on the posterior-replicated data. Findings are discussed in the context of the past environmental actions and events. The proposed methodology may be useful for the air quality management by providing quantitative means to measure the efficacy of pollution control programmes.► Reversible jump Markov chain Monte Carlo (MCMC) method due to is adapted to threshold violations of air pollutants. ► Analysis is based on a simple Poisson (step rate) model with no prior restrictions. ► Improved preprocessing is used to tackle missing data, seasonality and correlations. ► Results are assessed and validated via both classical and Bayesian statistical tools. ► MCMC performance is cross-validated using simulation techniques.
Keywords: Air quality action plan; Urban air pollution; Threshold exceedances; Non-homogeneous Poisson process; Change-points; Reversible jump MCMC;

This paper analyzes PM10 concentration data collected at 10 stations for the period of 2005–2009 by Air Quality Section of Istanbul Municipality. The analysis has been focused on spatial and temporal variations of the pollutants and their possible sources in the urban zones of the city. PM10 concentrations in Istanbul show significant variations across the city, with current PM10 levels at several traffic hot points and industrial zones which exceed EC air quality limit. The general temporal pattern is characterized by high concentrations in winter and low in summer. The number of exceedences allowed by the EU was surpassed at all monitoring sites during the analyzed years which reflect the serious pollution problem in the city.► PM10 concentration data collected at 10 stations for the period of 2005–2009 in Istanbul. ► It has been focused on spatial and temporal variations of the pollutants. ► The number of exceedences allowed by the EU was surpassed at all monitoring sites.
Keywords: High PM10 concentrations; Hierarchical clustering; Wind; Precipitation; Istanbul;

Atmospheric CO2 concentrations and δ13C over western and eastern Mediterranean basins during summer 2007 by Antonio Longinelli; Leonardo Langone; Carlo Ori; Enricomaria Selmo (5514-5522).
During summer 2007, two survey cruises covered large sections of both western and eastern Mediterranean basins. Atmospheric CO2 concentrations were measured with a Siemens Ultramat 5E analyser on both outward and return voyages, and 48 discrete air samples were collected with evacuated Pyrex flasks. The δ13C values of the CO2 from the flask air samples were measured in the laboratory following well-established procedures. CO2 concentrations in the western basin were characterised by marked variations, ranging from about 381 to about 394 ppmv. Peaks of up to 398 and 415 ppmv were occasionally recorded, in particular close to the Strait of Gibraltar. The δ13C values were very homogeneous, with a mean value of −9.70 ± 0.22‰, in contrast with the large δ13C fluctuations expected in this basin. The eastern basin, on average, showed far lower CO2 concentrations than those recorded in the western one. The overall range was between about 375 and 398 ppmv, but most values fell between 375 and 388 ppmv. The δ13C values of the discrete air samples were also very homogeneous, with a mean value of −9.37 ± 0.16‰, less negative than that of the western basin. The CO2 concentrations were slightly lower than those measured in December 2007 in the eastern Mediterranean in contrast with the hypothesis of heavy summer pollution in the Mediterranean basins. No quantitative relationship was found between CO2 concentrations and δ13C values. This discrepancy may be the effect of the large biospheric sink of CO2 on land and of the plant isotopic discrimination against 13C.► Atmospheric CO2 concentrations were measured in both the Mediterranean basins. ► CO2 concentrations in the western basin were higher than those in the eastern one. ► δ13C(CO2) from the western basin are more negative than those from the eastern one. ► No quantitative relationship is apparent between CO2 concentrations and δ13C values. ► The biospheric sink of CO2 on land can be responsible for this discrepancy.
Keywords: Atmospheric CO2 concentration; δ13C of atmospheric CO2; Mediterranean; Western and eastern basins; Summer 2007;

Evaluating AURA/OMI ozone profiles using ozonesonde data and EPA surface measurements for August 2006 by Lihua Wang; M.J. Newchurch; Arastoo Biazar; Xiong Liu; Shi Kuang; Maudood Khan; Kelly Chance (5523-5530).
We evaluate the Ozone Monitoring Instrument (OMI) ozone profile retrieval against ozonesonde data and the Environmental Protection Agency (EPA) surface measurements for August 2006. Comparison of individual OMI ozone profile with ozonesonde indicates that OMI ozone profile can explain the general vertical variation of ozone but is limited in observing the boundary layer ozone, due to weak sensitivity to boundary layer ozone and thick lowest layer (∼2.5 km). We made pair-wise comparison between OMI and ozonesondes on 24 OMI vertical layers, as well as the 39 sigma-P vertical layers of the Community Multi-scale Air Quality (CMAQ) modeling system, respectively. OMI shows reasonable agreement with ozonesonde in the lower- to mid-troposphere. In the upper troposphere, while the bias increases, the normalized bias does not show much variation and remains below 10%. Comparison with EPA’s surface-monitoring data indicates that OMI observations at the lowest layer (surface to 2.5 km altitude) represent the mean values. While OMI underestimates elevated ozone concentrations, it explains the larger-scale spatial variation seen in the surface monitors.► We evaluate OMI ozone profile retrieval against ozonesonde and EPA surface data. ► OMI shows reasonable agreement with ozonesonde in the lower- to mid-troposphere. ► In the upper troposphere, the normalized bias of OMI remains below 10%. ► OMI underestimates elevated ozone concentrations observed by EPA surface monitors. ► OMI explains the larger-scale spatial variation seen in the surface monitors.
Keywords: OMI; IONS06; CMAQ; EPA; Ozone evaluation;

“A method to estimate the contribution of unidentified VOCs to OH reactivity” by Shungo Kato; Tadashi Sato; Yoshizumi Kajii (5531-5539).
GC-FID analysis of polluted urban air samples detects a considerable number of unidentified peaks of volatile organic compounds (VOCs). In this study, a new method is proposed to estimate the contribution of the OH reactivity of such unidentified VOCs using a relative rate method. An air sample was analyzed by two methods: the conventional GC-FID method and a method where the sample was first exposed to the OH radicals generated in the photolysis of water vapor using a pen ray mercury lamp and then analyzed by the same conventional method. The difference in the peak areas of each species with and without OH exposure ideally depends on the OH reactivity of the species of interest. This difference in a standard gas sample that included 56 VOCs was observed through test measurements, and the validity of the system using the proposed technique was confirmed. An ambient air sample obtained in suburban Tokyo was then examined. From the peak area change of the identified species, the contribution of the unidentified VOCs to the OH reactivity was found to be about 22% of that of the identified VOCs. In a separate experiment, simultaneous measurements of the total OH reactivity, various trace gases, and the proposed technique to estimate the contribution of unidentified VOCs were conducted at a Tokyo suburb. The contribution of the unidentified VOCs to the total OH reactivity in suburban air was up to about 6% of the total OH reactivity, and the difference between the observed and estimated total OH reactivity was up to 20%.► We estimate the reaction rate constant with OH of the unidentified VOCs in GC-FID. ► Contribution to OH reactivity of the unidentified VOCs is also estimated. ► Contribution of the unidentified VOCs was 22% of the total VOC for the ambient air.
Keywords: OH lifetime; GC-FID; Relative rate analysis; Urban atmosphere; Hydrocarbon;

The effects of indoor particle exposure on blood pressure and heart rate among young adults: An air filtration-based intervention study by Lian-Yu Lin; Hua-Wei Chen; Te-Li Su; Gui-Bing Hong; Li-Chu Huang; Kai-Jen Chuang (5540-5544).
This study aims to evaluate whether air filtration can modify the effect of indoor particles on blood pressure (BP) and heart rate (HR) in a young, healthy population. We recruited 60 students to participate in a study of multiple, prolonged exposures to either particle-filtered or non-filtered indoor air. We made four home visits in which we took continuous 48-hour measurements of systolic BP (SBP), diastolic BP (DBP), and HR in each participant. Particulate matter less than 2.5 μm in diameter (PM2.5) and total volatile organic compounds (VOCs) were measured at each participant’s home. We used mixed-effects models to associate BP and HR with indoor particles and total VOCs, which were averaged over 1-hour to 8-hour periods prior to physiological measurements. We found that the mean values for indoor PM2.5 exposures at 1-hour to 4-hour were associated with an elevation in SBP, DBP and HR. The effects of indoor PM2.5 on BP and HR were greatest during the visits without air filtration. During visits with air filtration, participants showed no significant change in BP and HR in response to indoor PM2.5 exposure. We concluded that air filtration can reduce indoor PM2.5 concentrations and modify the effect of PM2.5 on BP and HR in a healthy, young population.► We make four home visits to measure blood pressure and heart rate in 60 participants. ► Fine particle and total volatile organic compounds are measured during all visits. ► An filter is added to participants’ air conditioners at the last two visits. ► Indoor particle exposure is associated with elevated blood pressure and heart rate. ► Air filtration can reduce indoor particle level and modify the cardiovascular effect.
Keywords: Indoor air; Particulate matter; Cardiovascular effect; Epidemiology;

Forest fires in a changing climate and their impacts on air quality by A. Carvalho; A. Monteiro; M. Flannigan; S. Solman; A.I. Miranda; C. Borrego (5545-5553).
In a future climate scenario forest fire activity over Portugal will substantially increase and consequently area burned and forest fire emissions to the atmosphere are also expected to increase. This study investigated the impact of future forest fire emissions on air quality over Portugal under the IPCC SRES A2 scenario. Reference and future climate change scenarios were simulated using the MM5/CHIMERE air quality modelling system, which was applied over Europe and over Portugal, using nesting capabilities. The initial and boundary conditions were provided by the HadAM3P model simulations for the reference and the future climate. The forest fire emissions were estimated using a methodology, which included the selection of emission factors for each pollutant, burning efficiency, fuel loads and the predicted area burned. These emissions were added to the simulation grid using specific parameterizations for their vertical distribution. Modelling results for Portugal pointed out that future forest fire activity will increase the O3 concentrations of almost 23 μg m−3 by 2100 but a decrease of approximately 6 μg m−3 is detected close to the main forest fire locations. Future forest fire emissions will also impact the PM10 concentrations over Portugal with increases reaching 20 μg m−3 along the Northern coastal region in July. The highest increases are estimated over the north and centre of Portugal where the area burned projections in future climate are higher.► We investigate the impact of forest fires on air quality over Portugal under climatic scenario. ► we simulated reference and future climate change scenarios using the MM5/CHIMERE modelling system. ► The forest fire emissions were estimated using a standard methodology. ► Modelling results pointed out that future forest fire emissions will impact O3 and PM10. ► The highest increases are estimated over the north and centre of Portugal.
Keywords: Climate change; Future fire activity; Forest fire emissions; Air quality modelling; Ozone; Particulate matter;

Organic/elemental carbon and woodsmoke tracer concentrations following a community wide woodstove changeout program by Tony J. Ward; Christopher P. Palmer; Megan Bergauff; R.K.M. Jayanty; Curtis W. Noonan (5554-5560).
As part of the Libby, Montana community-wide woodstove changeout program conducted between 2005 and 2008, nearly 1200 older model woodstoves were replaced with cleaner burning devices such as EPA-certified stoves. The reduction in overall ambient wintertime PM2.5 mass has been documented. However, much less is known about the changes in the chemical (specifically the organic) makeup of the ambient particles as the emission sources changed within the community with the implementation of the cleaner-burning home heating devices.Overall, PM2.5 mass levels decreased by 25%, while Total Carbon (TC) showed a 26% reduction when comparing the pre- and post-changeout concentrations. As the TC fraction is heavily influenced by Organic Carbon (OC) in Libby, trends were similar. Differential effects were observed for OC fractions according to analytical temperatures. Concentrations of Elemental Carbon (EC) were fairly consistent throughout the changeout, and did not show a dramatic reduction similar to the measured OC. Results from the polar organic compound program showed that the majority of the secondary aerosols were reduced, as were the majority of the woodsmoke markers such as levoglucosan and several methoxyphenols. However, the concentrations of resin acids actually increased following the woodstove changeout, suggesting that the chemistry of the particles have been altered as a result of this community-wide intervention.► TC showed a 26% reduction when comparing the pre- and post-changeout concentrations. ► Differential effects were observed for OC fractions according to analytical temperatures. ► The majority of the secondary aerosols were reduced, as were the majority of the woodsmoke markers such as levoglucosan. ► The chemistry of the particles have been altered as a result of this community-wide intervention.
Keywords: Organic fraction; PM2.5; Woodsmoke; Biomass burning; Polar organics; Libby; Montana;

The role of meteorology on the background air quality at the Athens International Airport by C.G. Helmis; G. Sgouros; H. Flocas; K. Schäfer; C. Jahn; M. Hoffmann; Ch. Heyder; R. Kurtenbach; A. Niedojadlo; P. Wiesen; M. O’Connor; E. Anamaterou (5561-5571).
An experimental campaign has been conducted in the frame of the European Commission funded network of excellence ECATS, at Athens International Airport (AIA), Greece, from 13 to 25 September 2007. In order to quantify the background air quality and study the influence of meteorological conditions and airport emissions to the broader region, the analysis was extended to the greater Messogia Plain of Attica, Greece. Mixing height was estimated taking into account measurements from remote sensing and surface based single point instrumentation. According to the analysis, the air quality over Messogia Plain is mainly controlled by the intensity and the direction of the background flow. Under low background wind conditions, the development of local flows (sea and land breeze cells) over the greater area preserves high concentrations of air pollutants which are mainly attributed to the airport emissions, the local activities and traffic. When the background flow is strong, the diurnal cycle of all concentrations was significantly reduced by more than 50%, due to the advection and the subsequent mixing of the lower atmosphere. In order to identify and quantify the various processes linked with meteorology and air quality, the Hilbert–Huang transform was selected to be applied on the air-quality time-series data. The calculated Hilbert spectra of the main pollutants showed that meteorology plays a prescriptive role on the evolution of air pollutants, determining the influence of the local scale characteristics at each station.► Air quality is mainly affected by the intensity/direction of the background flow. ► Under local flows air quality preserves high concentrations of air pollutants. ► Under weak synoptic flow local station characteristics dominate the energy spectra. ► Under strong winds station characteristics don’t influence the pollutants’ spectra.
Keywords: Airport air quality; Air pollutants; Atmospheric boundary layer; SODAR–RASS; Hilbert–Huang transform; Mixing height;

By means of computational fluid dynamic (CFD) simulations, this paper investigated how rural winds transport relatively clean air into high-rise urban areas to dilute airborne pollutants. Two ventilation concepts for indoor environments were applied: the age of air to quantify the time taken by rural young air to reach a given place after it enters an urban area and the air exchange efficiency to evaluate the efficiency of rural winds flushing the entire urban canopy layer. Several square building arrays with street aspect ratios (building height/street width, H/W) ranging from 1.5 to 5.3 and building area densities of 0.25 (medium) or 0.4 (compact) were considered as the approaching wind is parallel to the main street whose length varies from street scales to neighborhood scales (330–510 m to 1.03–1.65 km in full-scale).Results show that considerable young rural air enters windward entries but a major fraction of air is vertically driven out as flowing deeper into such high-rise building arrays. So air exchange efficiencies are less than 50% in street-scale arrays, and smaller in longer (neighborhood-scale) or narrower arrays. For the neighborhood-scale medium arrays, considering the power-law velocity profile in the upstream free flow, a taller array gains a larger inflow rate across its windward entry and experiences younger air and greater air exchange efficiency than a lower one. If all buildings are theoretically open-based in a neighborhood-scale compact array, air becomes much younger everywhere and the air exchange efficiency doubles. In arrays of buildings with different heights, the secondary streets in front of taller buildings get younger air due to the downward flows within them. Although further investigations are still required before providing a practical framework, this paper is one of the first attempts to find ways in improving the ventilation performance in high-rise cities like Hong Kong.► Age of air and air exchange efficiency are good to study how rural winds ventilate high-rise building arrays. ► Air exchange efficiencies are less than 50% in street-scale (∼300–500 m) arrays and smaller in longer or narrower arrays. ► Taller arrays may get younger air and greater air exchange efficiency than lower ones if street lengths are limited. ► In arrays of buildings with different heights, air in the secondary streets in front of taller buildings is younger. ► If open bases of buildings are introduced, air becomes younger everywhere and air exchange efficiency doubles.
Keywords: Age of air; Air exchange efficiency; Numerical simulation; High-rise building array; Homogeneous emission method; Building area density;

Six-day PM10 air quality forecasts for the Netherlands with the chemistry transport model Lotos-Euros by Martijn de Ruyter de Wildt; Henk Eskes; Astrid Manders; Ferd Sauter; Martijn Schaap; Daan Swart; Peter van Velthoven (5586-5594).
In this work we study the ability of the chemistry transport model Lotos-Euros to forecast, with a range of six days, PM10 concentrations and exceedances thereof. For both rural and non-rural locations in The Netherlands and based on one year of data, model performance does not deteriorate up to a lead time of three days. Thereafter the PM10 forecast is increasingly affected by uncertainty in the meteorological forecast. However, up to a lead time of 6 days the forecast still has skill, beats persistence and complies with several performance criteria. The correlation between forecast and observations is between 0.66 and 0.70 for the first half of the forecast and remains above 0.54 until the end of the forecast range. Exceedances of the PM10 concentration over thresholds are also forecasted with reasonable skill up to a forecast range of three days, after which a gradual deterioration sets in. The stability of the forecast displays the same behaviour. Up to a lead time of three days, the forecast remains reasonably stable with more than 80% of forecasted exceedances still present in all later shorter-term forecasts for the same date. Because exceedances can be forecasted with considerable skill a number of days in advance, the forecast can be used for applications that require a range of a few days, such as outdoor activities and the scheduling and implementation of short-term emission reduction measures.► An operational system for forecasting air quality with a range of 6 days was set up. ► The quality of PM10 forecasts was evaluated using different metrics and benchmarks. ► Generated PM10 forecasts do not deteriorate up to a lead time of 3 days. ► These forecasts still display skill up to a lead time of 6 days.
Keywords: Air quality; Forecast; pm10; Lotos-Euros;

Trans-boundary transport of ozone from the Eastern Mediterranean Coast by David Asaf; Mordechai Peleg; Jihad Alsawair; Aiman Soleiman; Valeri Matveev; Eran Tas; Allan Gertler; Menachem Luria (5595-5601).
As part of an ongoing USAID MERC research program investigating the transport of pollutants in the Eastern Mediterranean region, a joint coordinated campaign involving scientists from Jordan and Israel took place for a three-week period in June 2009. The study was aimed at examining previous modeling results that indicated the presence of elevated ozone levels occurring in Northern Jordan, due to emissions originating in coastal Northern Israel that were transported distances of more than 100 km. In the present investigation, ozone and other pollutants were monitored at five sites in Israel and two in Jordan. The sites were located along the prevailing wind direction that tends to move air-masses eastward from the Mediterranean coast, over the Israel Valley toward the Jordan Valley, and then into Northern Jordan. Analysis of the measurements showed that air masses arriving in Jordan passed over the Israeli metropolitan coastal region, entrapping large amounts of pollutant precursors before transporting them towards Jordan under the predominately westerly wind flows causing peak maximum ozone levels of around 100 ppbv. The present findings confirmed the previous modeling results that indicated higher NOx background levels for the inland stations, peak O3 values at later hours depending on wind speed and distance from the Mediterranean coast and maximum O3 levels over northern Jordan.► Inland areas of Israel and Jordan are strongly affected by elevated O3 levels. ► Emissions from Israel contribute significantly to photochemical pollution in Jordan. ► Maximum O3 levels observed over northern Jordan validated previous model results.
Keywords: Trans-boundary; Ozone; Air-pollution; Transport;

Determination of the equivalent emission parameters of wood-based furniture by applying C-history method by Yuan Yao; Jianyin Xiong; Weiwei Liu; Jinhan Mo; Yinping Zhang (5602-5611).
VOC pollution in indoor air has become a serious problem in urban areas of China. Building materials and furniture used indoors are the main sources of such pollution. How to determine the VOC emission characteristics of furniture rapidly and conveniently is an important but difficult problem. In this paper, we extend the C-history method for determining the VOC emission characteristic parameters (i.e., the initial emittable VOC concentration, C 0, the partition coefficient, K, and the diffusion coefficient, D) of homogeneous building materials developed by to measure the parameters of VOC emissions from wood-based furniture in a 30 m3 full-scale chamber. The conditions under which there are three equivalent parameters, C 0,equ, K equ and D equ for multi-layer board and wood-based furniture is analyzed and presented. A series of tests were performed to determine the three parameters of some commonly used wood-based furniture in China. A run of independent ventilated chamber experiments were carried out to validate the method. The results show that: (1) The agreement of the formaldehyde and six VOCs concentrations between the ventilated experiment and the simulated results based on the determined parameters is very good; (2) The method could be used in testing multi-layer board and wood-based furniture. The testing time is usually less than 3 days as compared with the furniture testing method typically performed under ventilated chamber conditions and take 5,7,14 or 28 days. The possibilities and limitations of applying the method to the existing furniture labeling schemes are discussed.► Conditions under which there are equivalent parameters for wood-based furniture. ► Applying C-history method in wood-based furniture testing. ► Independent validation method was used during the test. ► Formaldehyde and other six VOCs were selected as targets. ► Reducing the typical testing time to within 3 days.
Keywords: Volatile organic compounds (VOC); Indoor air quality (IAQ); Wood-based furniture; Emission; C-history method;

Characterization of macromolecular organic matter in atmospheric dust from Guangzhou, China by Jinping Zhao; Ping’an Peng; Jianzhong Song; Shexia Ma; Guoying Sheng; Jiamo Fu (5612-5620).
The chemical compositions of organic macromolecules in dust are very complex and have not yet been investigated in detail. In this paper, we study the organic macromolecules in 12 dust samples collected from an urban and suburban area of Guangzhou, China. Organic macromolecules in the dust were firstly separated into humic acids (HA), kerogen and black carbon (KB) and black carbon (BC) fractions by chemical treatment with NaOH, HCl/HF and K2Cr2O7/H2SO4, respectively. The isolated fractions were subsequently characterized using elementary analysis (EA), organic petrographic examination (OPE), scanning electron microscopy (SEM), and solid state 13C cross-polarization and magic angle spinning nuclear magnetic resonance spectroscopy (13C-CP/MAS NMR) etc. The results showed that the major organic macromolecules were KB, which accounted for 64.8%–95.8% of total organic carbon (TOC) in the dusts. Isolated HAs were characterized by high H/C, N/C and O/C atomic ratios, and high carboxyl, methoxyl and aliphatic carbon contents in the overall carbon structure; meanwhile, kerogens (K) showed lower H/C, N/C and O/C atomic ratios than those of HAs and had chemical compositions enriched in hydroxyl carbon; further, the optical features of K were similar to those of “vitrinite”. BC is formed from the combustion of biomass and fossil fuels, so that aromatic and aliphatic carbons are the most important components in its carbon skeleton. Under the microscope, BC appeared to belong to the petrographic groups, such as “semifusinte” and “fusinite”, and showed a distinct texture when observed by SEM. Compared with the HAs and K isolated from soils and sediments, HAs in dust were relatively lower in aromatic carbon and K in dust was always “type III”. These results are strongly consistent with the dust samples originating from the photochemical degradation of volatile organic compounds, in contrast to the geological HAs and K which are sourced from the biodegradation of lignin and algae or from bacterial activities in water and sediment.► It is the first time using the isolation and characterization approaches applied to dust, and get HA, KB and BC. ► The main source of HA and K in dust is derived from photochemical reactions of VOCs. ► The main source of BC in dust is formed from the combustion of biomass and fossil fuels. ► The results we obtained strongly suggest HA and K in the dust originating from the photochemical degradation of VOCs. ► HA in dust has a relatively lower aromatic ratio in its structure, and K in dust is always type III.
Keywords: Falling dust; Chemical compositions; Organic macromolecules; Elemental analysis; 13C-CP/MAS NMR;

Diurnal and phenological variations of O3 and CO2 fluxes of rice canopy exposed to different O3 concentrations by Lei Tong; Xiaoke Wang; Chunmei Geng; Wei Wang; Fei Lu; Wenzhi Song; Hongjie Liu; Baohui Yin; Lihua Sui; Qiong Wang (5621-5631).
A dynamic chamber system was designed to measure simultaneously the diurnal and phenological canopy ozone (O3) and carbon dioxide (CO2) fluxes in the paddy field under different O3 concentrations (0, 40, 80 and 120 nmol mol−1). On the diurnal timescale, a decreasing trend of canopy O3 flux was observed from morning to evening and the O3 flux increased with increasing O3 concentration, while canopy CO2 flux generally followed the track of photosynthetic active radiation, with higher values at noon except at the end of the growing season when rice was senescent. The constant CO2 flux among different O3 treatments in this experiment suggested that the photosynthesis of the rice canopy was not affected by short-duration (ca. 10 min) O3 exposure of elevated concentration. The daily mean O3 and CO2 fluxes increased with rice growth until the dough stage and the late jointing stage, respectively, then decreased with rice aging. The peak values of O3 flux appeared later than those of CO2 flux because the latter was closely synchronized with the leaf area index of the rice canopy. Diurnal mean canopy O3 flux varied from 18.7 to 43.3 nmol m−2 s−1, and nocturnal mean canopy O3 flux varied from 2.7 to 17.8 nmol m−2 s−1 and from 7.0 to 25.4 nmol m−2 s−1 for the 40 and 80 nmol mol−1 O3 treatments, respectively. The considerable amount of nocturnal O3 flux indicated a significant contribution of non-stomatal factors to canopy O3 uptake. The adjusted Jarvis multiplicative models were used and well parameterized to fit the measured O3 and CO2 fluxes of our rice cultivar from environmental variables. Although more validation work is needed, the present results suggest that the models can be considered as a tool for canopy flux predictions in the paddy field.► O3 flux of rice canopy correlates with O3 concentration positively. ► Short-duration O3 exposure has little influence on CO2 flux of rice canopy. ► Phenological variation of leaf area dominates that of O3 flux. ► Non-stomatal O3 removal is remarkable for rice canopy. ► Adjusted Jarvis models were well parameterized to fit the measured fluxes of rice canopy.
Keywords: Chamber; Ozone; Carbon dioxide; Flux; Oryza sativa; Jarvis model;

Spatial variability of CH4 and N2O fluxes in alpine ecosystems on the Qinghai–Tibetan Plateau by Tomomichi Kato; Mitsuru Hirota; Yanhong Tang; Eitaro Wada (5632-5639).
The intra- and inter-site spatial variability of methane (CH4) and nitrous oxide (N2O) fluxes were investigated in three alpine ecosystems at Haibei station on the Qinghai–Tibetan Plateau (QTP) in summer 2005 and in sixteen alpine ecosystems with various vegetations across the QTP in summer 2006, respectively. The magnitude of average CH4 emissions from wetlands was at least 100 times larger than average CH4 uptake by grassland, suggesting that the entire QTP is likely to be a source of methane in summertime because of a significant fractional area of wetlands (∼ 2.2%) on the plateau.Intra-site investigation, with sixteen chambers, revealed a significant negative relationship of CH4 emissions with the C/N ratio of aboveground biomass and soil pH in the alpine wetland when all chambers were considered. Moreover, soil oxidation–reduction potential (ORP) had a remarkably strong influence on CH4 emissions for nine chambers above the water level, resulting in a negative exponential relationship. For N2O flux in alpine meadows, a negative relationship with both soil pH and livestock dung biomass was observed. Aboveground plant biomass and soil pH were important variables overall in wetlands.Inter-site investigation found positive and negative relationships between CH4 flux and soil biomass to 5 cm depth in nine grassland and seven wetland sites, respectively. N2O flux showed a moderately strong negative exponential relationship with the C/N ratio of surface soil in the grassland sites. In the wetlands, soil pH was negatively correlated with N2O flux, perhaps due to both reduced N2O release from suppressed nitrification and denitrification.► Spatial variability of CH4 and N2O fluxes were examined in alpine ecosystems. ► Both fluxes changed along soil pH variation among sixteen chambers in wetland. ► Spatial variations were not related by any factors in meadow and shrubland. ► Averaged CH4 flux varied along soil biomass in inter-site comparisons. ► N2O flux related to soil C/N ratio among grasslands and to soil pH among wetlands.
Keywords: Spatial heterogeneity; Methanogenesis; Methanotroph; Nitrifier; Denitrifier;

Impact of ammonium nitrate chemistry on the AOT in Cabauw, the Netherlands by J.W.B. Derksen; G.-J.H. Roelofs; R. Otjes; G. de Leeuw; T. Röckmann (5640-5646).
A 1D column model representing detailed microphysics and chemistry associated with sulfate, ammonia and nitric acid is applied to study local processes which influence aerosol optical thickness (AOT). We focus on Cabauw, the Netherlands, on May 8th 2008 during the Intensive Measurement campaign At Cabauw Tower (IMPACT). The 8th of May was characterized by a strong diurnal cycle of RH and aerosol concentration in the boundary layer. Furthermore, the lower troposphere was characterized by high concentrations of aerosol ammonium and nitrate as well as ammonia and nitric acid. Production of ammonium nitrate in the aerosol contributes to the particle hygroscopicity. This leads to additional condensational growth, which, in turn, enables further uptake of ammonium nitrate. It is found to predominantly affect particles smaller than 0.1 μm, so that the simulated wet aerosol size distribution is considerably narrower compared to a situation without ammonium nitrate. Our results show that the daily evolution of aerosol ammonium nitrate influences the variability of AOT more than the daily variability of RH and aerosol concentration. On this particular day, aerosol ammonium nitrate effectively enhances AOT by a factor of 2–4, depending on the time of the day. A relatively large discrepancy between modeled and observed AOT in the afternoon of May 8th is probably associated with secondary organic aerosol production in the afternoon which is not considered in our study.► We model microphysics and chemistry to study local processes which influence AOT. ► We focus on Cabauw, the Netherlands, on May 8th 2008 during IMPACT. ► Variability in AOT is dominated by variability in ammonium nitrate. ► Variability in RH and aerosol concentration is only of secondary importance. ► The wet aerosol size distribution is narrower with than without ammonium nitrate.
Keywords: Ammonium; Nitrate; AOT; Aerosol;

Concentrations of sulphur dioxide (SO2) have been decreasing in the Netherlands since the beginning of the 1980s, as a result of national and international emission control measures. Since 2007, concentrations observed at the Rotterdam port and industrial areas have shown a large decrease that is in line with recent emission control measures. The average annual SO2 concentration in 2010 was about 50% below 2000–2006 levels. This drop in concentration level corresponds with recent decreases in emissions of SO2 from Dutch refineries and international sea shipping, on top of the gradual decreases in emissions from sources outside the Netherlands. The reduction in the emissions from refineries was initiated by a ceiling on the total amount of emissions from this sector, effective since 2010. Emission reductions from sea shipping result from two types of regulations to reduce the sulphur content in marine fuel, by 2010; regulation by the International Maritime Organization for sea ships on the North Sea, and by the EU directive for ships at berth in ports.
Keywords: Rotterdam; Refineries; Shipping; Sulphur;

New Directions: What do we need to know about brominated flame retardants in indoor dust? by Stuart Harrad; Mohamed Abou-Elwafa Abdallah (5652-5653).