Atmospheric Environment (v.39, #38)

Organochlorine compounds in the marine atmosphere of Singapore by Oliver Wurl; Jeffrey Philip Obbard (7207-7216).
Organochlorine pesticides (OCPs) and polychlorinated biphenyls (PCBs) were quantified in high-volume air samples collected on polyurethane foam (PUF) over the sea surface of Singapore during the summer monsoon period of 2004. A novel and quality assured microwave-assisted extraction (MAE) procedure for PUF was developed as an alternative to conventional Soxhlet extraction for the analysis of organochlorine compounds (OCs). Average concentrations of ΣPCBs, ΣHCHs and ΣDDTs in air samples were 32, 311 and 8 pg m–3, respectively. Overall, a decline in the ocean atmospheric concentration of DDTs was noted relative to data obtained from the early 1990s, but current data are in a similar range for HCH and PCBs. Levels of OCs in the equatorial atmosphere were compared to equivalent data from temperate regions available in the literature. On a global scale, the levels of DDTs and PCBs in the atmosphere of Singapore are low, but moderate for HCHs.
Keywords: Organochlorine pesticides; Polychlorinated biphenyls; Microwave-accelerated extraction; Polyurethane foam; Air pollution;

The atmospheric oxidation of mercury in the Mediterranean marine boundary layer (MBL) has been studied using the Atmospheric Mercury Chemistry over the Sea (AMCOTS) model. The model results have been compared to measured data obtained during an oceanographic research campaign in 2000, with more success than previous modelling attempts. In light of the often high concentrations of ozone present in the Mediterranean boundary layer, seasonal case studies using typical meteorological conditions and average ozone concentrations have been performed to identify the main oxidants of elemental mercury. The sensitivity of the modelled reactive gaseous mercury (RGM) concentrations to the Hg+O3 rate constant has been assessed using the two most recent rate determinations. The results using the higher of the two literature values gives results inconsistent with measured values of RGM when the reaction between Hg and O3 is assumed to give a gas phase product. This does not necessarily indicate that the rate constant is incorrect but possibly that other rate constants in the model are overestimated or indeed that there may be reduction reactions occurring in the atmosphere which have yet to be identified. Alternatively, when the reaction product of Hg and O3 is assumed to be a solid and therefore not contribute to RGM the modelled and measured results are comparable. The deposition rates calculated by the model when compared with calculated and measured sea surface emission fluxes available in the literature indicate that dry deposition flux of RGM is comparable to the sea surface emission flux. The calculated lifetime of Hg0 in the Mediterranean MBL is between one and two weeks.
Keywords: Mercury; Oxidation; RGM; Chemical modelling; Speciation;

The formation of sulfate and nitrate by heterogeneous reactions of gaseous precursors on mineral dust particles was investigated using positive matrix factorization (PMF) of coarse PM10 (particulate diameters from 2.2 to 10 μm) collected at urban (Hanoi) and rural (Lucnam) sites in northern Vietnam. Air samples were analyzed for ionic and elemental components using ion chromatography and proton induced X-ray emission methods. PMF revealed six similar sources/types of coarse PM10 at the two sites, namely soil dust containing nitrate and sulfate, coal fly ash from distant and local sources, soil dust containing organic matter and ammonium sulfate and marine aerosol. Traffic (road) dust was found only at the urban site.From the PMF factor models, the yields of NO 3 - , SO 4 2 - and NH 4 + can be estimated and their possible chemical forms in different particulate types can be suggested. The yields of nitrate and sulfate formation on mineral dust particles increase with the [Ca]/[Si] ratio, which is greater in soil dust than in coal fly ash. Nitrate is bound to Ca-richest soil dust particles. Ammonium was found in dust particles containing soil organic matter, which also hold the largest amount of sulfate. The comparison of urban and rural receptor models provided synergy for the source identification and insights into the properties of mineral dust particles relevant to their interactions with acidic gases in ambient air.
Keywords: PIXE; Ion chromatography; Positive matrix factorization; Sulfate; Nitrate; Alkalinity;

Vertical differences in PM10 and PM2.5 suspended particles were investigated using a building in Osaka, Japan. Samples were collected on the roof of the building (200 m above ground level) and on the ground during 5–9 August and 2–6 December 2002. In addition to determination of sample mass, concentrations have been analyzed for major chemical components including elemental carbon (EC) and organic carbon (OC). Particle mass concentrations of the samples collected at 200 m were lower than those collected on the ground. “Others” species, defined as the difference between the total particle mass and the sum of the masses of the measured species, dominantly accounted for the vertical difference in mass concentrations in summer, whereas EC and OC were the major contributors in winter. Vertical profiles of relative humidity observed simultaneously indicated that relative humidity on the ground was higher than that at 200 m during the summer sampling period. Hence, it is likely that the higher concentrations of “others” species in the samples collected on the ground were probably caused by water having been absorbed by deliquescent components of the particles. Vertical temperature profiles during the winter sampling period suggested that stable meteorological conditions in winter resulted in the accumulation of primary particles, mainly emitted from vehicle exhaust, leading to the high concentrations of EC and OC on the ground.
Keywords: PM10; PM2.5; Elemental composition; Vertical difference; Relative humidity;

Evaluation of two MCM v3.1 alkene mechanisms using indoor environmental chamber data by R.G. Hynes; D.E. Angove; S.M. Saunders; V. Haverd; M. Azzi (7251-7262).
Photo-oxidation experiments on propene-NO x -air and 1-butene-NO x -air mixtures were performed in the CSIRO indoor environmental chamber. These data were used to test the alkene sub-mechanisms from the Master Chemical Mechanism version 3.1 (MCM v3.1). A comparison of measured and modelled propene-NO x data showed that the mechanisms required the inclusion of O(3P)+parent alkene reactions in order to adequately model the results over a wide range of VOC/NO x ratios. Sensitivity studies were performed on propene-NO x and 1-butene-NO x simulations to determine the effect of uncertainties in the chamber parameters and key photolysis rates on the O3, and NO x profiles. The low VOC/NO x propene simulations were sensitive to the formaldehyde (HCHO) photolysis channel that produces HO2 radicals and to the presence of O(3P) reactions. For the higher VOC/NO x propene simulations, the O3, and NO x results were more sensitive to small changes in the initial HONO and NO2 concentrations.
Keywords: Indoor environmental chamber data; Ozone modelling; Photooxidation; Master chemical mechanism; Tropospheric chemistry;

Quantifying small molecules in secondary organic aerosol formed during the photo-oxidation of toluene with hydroxyl radicals by Jacqueline F. Hamilton; Paul J. Webb; Alastair C. Lewis; Montserrat M. Reviejo (7263-7275).
The composition of secondary organic aerosol (SOA) formed during the photo-oxidation of toluene in a large-volume smog chamber was determined using direct thermal desorption coupled to comprehensive gas chromatography–time of flight mass spectrometry (TD–GC×GC–TOF/MS). TD–GC×GC eliminated offline sample preparation and resulted in a single shot inventory of GC-amenable organic species within the aerosol. Seventy-four species were identified and quantified from chromatographic retention behaviour and mass spectral fragmentation. Functional groups resolved included organic acids, aromatics, dicarbonyls, furans, furanones, furandiones and pyranones. Concentrations of individual species were derived either directly from standards or from chemical surrogates. The major small organic molecule components by mass were formed via the peroxy-bicyclic reaction mechanism (e.g., 5-methyl-2(3H)-furanone=2972 ng m−3 or 0.861% of aerosol mass), although only around 10% of the total aerosol mass could be quantified. In addition, a large number of lower concentration species were observed with ambiguous chemical characteristics. Interpretation of their mass spectra suggests these species may be fragments of larger molecules broken down during the thermal desorption and separation stages.
Keywords: Comprehensive two-dimensional gas chromatography; SOA; Aromatic; Toluene; Heterogeneous reactions; Hydroxyl;

Saharan dust outbreak over southeastern Spain as detected by sun photometer by H. Lyamani; F.J. Olmo; L. Alados-Arboledas (7276-7284).
Characterization of mineral aerosol properties in southeastern Spain is important due to its proximity to the African continent that permit studying the Saharan dust just entering European continent before its possible mixing with anthropogenic emissions in Europe. Two Saharan dust events that occurred during 30 August–3 September and on 10 September 1998 were observed at an inland location on the outskirts of Granada in southeastern Spain. Detailed analyses including back trajectories and TOMS aerosol index maps showed transport of Saharan dust from the Sahara desert and North Africa to the study area. The dust episodes were associated with marked increases in aerosol optical depth (δ a) at all wavelengths examined in this study. Thus δ a (500 nm) increased from a value of ∼0.2 corresponding to normal conditions up to 0.6 in the first event and up to 0.55 in the second. At the end of these events, the arrival of air masses from the Atlantic caused a sharp decrease in δ a (500 nm) that reaches values close to those obtained before the event (∼0.2). The Angström exponent α reached a minimum of 0.36 in the first event and 0.37 in the second. The Angström exponent α and aerosol optical depth values during dust events agree well with those obtained under the same kind of events in AERONET and SKYNET sites, as well as with the results obtained in ACE-2 and ACE-Asia experiments. The aerosol size distributions, retrieved from aerosol optical depth using King's method, demonstrated how the large size fraction of aerosol associated with Saharan dust dominated during these events. When Saharan dust was present, the retrieved aerosol size distributions were bimodal with a well-defined mode centred at a radius of 0.6 μm, and showed an evident increase in the large particles mode with radii in the range 0.4–2 μm. The small particle concentration during the two events did not present any marked change, and was similar to those observed on days without Saharan dust. Large particle concentration was higher by a factor of 8 than during normal conditions.
Keywords: Aerosol properties; Sun photometry; Angström parameter; Saharan dust;

Development of a k – ε model for the determination of air exchange rates for street canyons by Xian-Xiang Li; Chun-Ho Liu; Dennis Y.C. Leung (7285-7296).
A two-dimensional renormalized-group (RNG) k – ε turbulence model was developed to calculate the air exchange rate (ACH) of idealized street canyons. This model was validated against other modelling (large-eddy simulation, LES, and k – ε ) and wind tunnel results. It was demonstrated that the current k – ε model was able to capture the complicated flow structures inside street canyons whose outputs agreed reasonably well with other modelling and wind tunnel results available in literature. This k – ε model was then applied to calculate the ACH for street canyons of aspect (building-height-to-street-width) ratios 0.5 , 1.0 , and 2.0 . The values of ACH calculated by the current k – ε model compared favourably with those calculated by LES in which their differences were less than 20 % . Moreover, there was a remarkable saving in computer resources and computation time by using the current k – ε model.
Keywords: Air exchange rate (ACH); Two-dimensional; Renormalized-group (RNG) k – ε turbulence model;

Tracking changes in carbon monoxide budget over Europe between 1995 and 2000 by T. Mészáros; L. Haszpra; A. Gelencsér (7297-7306).
The variation in carbon monoxide (CO) mixing ratio is of primary interest since it affects the atmospheric abundance of several trace gases, like methane, via extensive chemical feedback mechanisms. From the mid-1990s the global annual average mixing ratio of CO varied by about ±0–1% yr−1 while in 1998 a great pulse of CO emission increased the global CO concentration through mid-1999 by 16% over the background level. This surplus was primarily caused by heavy forest fires in several regions of the world. Our study was performed to reveal the potential causes of changes in the European CO budget and mixing ratio between 1995 and 2000 using a simple tropospheric box model. The primary anthropogenic emission from fossil fuel combustion and industrial processes in Europe showed a fairly constant annual decrease of about 4% yr−1 between 1995 and 2000, while the secondary anthropogenic CO source strength was reduced at a rate of about 3% yr−1. This decrease was partly balanced by sporadic pulses of European forest fire emissions. In spite of the reduction in the sources the region remained a net CO producer. The incremental change in the global CO budget caused by European sources was about 47.2±15.2 Tg CO yr−1 on average.The output of the model was validated against available observational data, both in absolute terms and with respect to the seasonal variation of the CO mixing ratio. The model revealed that the downward tendency of CO mixing ratios in Europe was temporarily offset by the contribution of large-scale forest fires in the northern hemisphere through long-range atmospheric transport.
Keywords: Carbon monoxide mixing ratio; Budget calculation; Box model; Forest fires; Anthropogenic emission;

Pedestrian exposure to air pollution along a major road in Central London, UK by S. Kaur; M.J. Nieuwenhuijsen; R.N. Colvile (7307-7320).
Pedestrian exposure to PM2.5, the loss of reflectance (‘blackness’) of the PM2.5 filters, ultrafine particle counts (particle range: 0.02–1 μm) and carbon monoxide (CO) was investigated along a major road running through the DAPPLE study site in Central London, UK. During an intensive 12-day exposure measurement campaign, groups of four volunteers sampled twice in the morning and twice in the afternoon along Marylebone Road. They were randomly designated a walking direction, walking position (kerbside or buildingside) and a side of the major road to walk along. PM2.5 was sampled using high-flow gravimetric personal samplers, ultrafine particle counts were measured using TSI P-TRAKs and Langans were used to measure CO exposure. PM2.5 samples were analysed gravimetrically and reflectance was measured using a smoke stain reflectometer to obtain a measure of ‘black smoke’. In total 603 acceptable samples were obtained—155 PM2.5 and reflectance, 120 ultrafine particle count and 173 CO. The average pedestrian exposure along the road was 37.7 μg/m3, 12.1 m−1×10−5, 80 009 pt/cm3 and 1.3 ppm for PM2.5, loss of reflectance, ultrafine particle counts and CO, respectively. PM2.5 exposure in the morning was significantly higher than in the afternoon, and there was a significant difference in exposure on the different sides of the road. For both reflectance and ultrafine particle counts, the exposure was significantly different both between the two walking positions on the pavement and the two sides of the street canyon. However there was no significant difference in CO exposure based on walking position, walking direction, canyon side or timing. Filter reflectance was significantly but weakly correlated with PM2.5 exposure ( r = 0.3 , N = 155 ), CO exposure ( r = 0.2 , N = 154 ) and ultrafine particle count exposure ( r = 0.7 , N = 108 ). PM2.5 and CO personal exposure measurements were much higher than those recorded at both the local background fixed monitoring station (FMS) and a kerbside FMS, but PM2.5 personal exposure was significantly correlated with the PM2.5 concentrations at the background FMS ( r = 0.6 , N = 155 ).
Keywords: Personal exposure; PM2.5; Carbon monoxide; Ultrafine particle counts; Reflectance;

Low-molecular weight aldehyde compounds in rain, fog, dew water, and in the gas phase were measured at urban and suburban mountain sites, to characterize the chemical composition of aldehydes in liquid droplets and in the gas phase in the ambient atmosphere, and discuss the factors controlling wet removal processes of aldehydes. Higher concentrations of total aldehydes were found in dew water than in rain and fog water due to the small amount of water volume. Both formaldehyde and acetaldehyde were detected as dominant aldehydes in the gas phase. Secondary formation processes are dominant sources for both aldehydes in the suburban site, whereas primary sources are relatively important for the urban atmosphere. In rainwater, by contrast, formaldehyde was the most abundant aldehyde, followed by glyoxal. Glyoxal was detected as the most dominant aldehyde in fog and dew water. Acetaldehyde was not detected as a main component in liquid droplets in spite of its abundance in the gas phase. Water solubility of each aldehyde compound and dilution effect by water are critical factors that control the compositions and concentrations of these aldehydes in ambient liquid droplets.
Keywords: Formaldehyde; Acetaldehyde; Glyoxal; Henry's law constant; Wet deposition;

The current work presents dry deposition fluxes of trace metals to the Southern and Northern Levantine Basin of the Eastern Mediterranean Sea. The dry depositional inputs were calculated taking into account (i) the spatial gradients in metal aerosol concentrations across the Levantine Basin (using concentrations from two location, Erdemli, Turkey and Tel Shikmona, Israel over a 3 year sampling period), (ii) the air mass origins, (iii) air mass temporal influence across the north and south Levantine basins and (iv) fine/coarse elemental aerosol size mass distributions allowing the use of new elemental settling velocities. Two distinct airflow sectors were defined at each site; North (N) and southwest (SW), having different temporal influences at each site (the temporal ratio of SW:N was 1.85 and 0.43 at Tel Shikmona and Erdemli, respectively). Temporal airflow weighted aerosol concentrations were then calculated for each site. The applied settling velocities calculated using size fractioned aerosol samples ( n = 227 ) collected from Erdemli and from the literature, yielded higher (compared to previous work) settling velocities for Cu, Pb, and Zn being 1.1, 0.8 and 0.9 cm s−1, respectively. Total dry deposition fluxes for the crustal elements (Al, Fe and Mn) were consistent with previous studies for the region; however, Cu, Pb, and Zn were much higher, in part, due to the larger adopted settling velocities. Using sequential leach data, the “exchangeable” as well as total elemental dry inputs were presented for both the north and south Levantine basins. Comparison with limited literature wet deposition would suggest that the dry deposition pathway is a more significant input for Cu, Pb, and possibly Al and Fe. The “exchangeable” dry atmospheric inputs of Pb, Cu and Cd to the southern basin were orders of magnitude greater than dissolved riverine inputs. A similar comparison could not be made for the northern basin owing to absence of riverine input data.
Keywords: Eastern Mediterranean; Trace metals; Aelioan transport; Air mass back trajectories;

The present study examined airborne microbe levels inside and outside passenger cars and public buses during two seasons (winter and summer). The prevalence level of individual fungal genera depended on the fungi and agar type, whereas a constant higher prevalence (>90%) was found for the total bacteria and total fungi. The four most prevalent fungal genera were Cladosporium, Penicillium, Aspergillus, and Alternaria in a descending order. The major parameters associated with the airborne microbe measurements included agar type, vehicle type, and seasonal variation. When compared to the malt extract agar (MEA), the dichloran glycerol 18 agar (DG-18) was found to produce better counts for the target fungi. The summer in-vehicle bacterial concentrations were significantly higher for public buses than for the passenger cars, whereas the reverse was true for summer in-vehicle total fungal concentrations. This pattern was also consistent with summer outdoor results. In contrast, the winter in-vehicle and outdoor airborne microbe levels were similar for cars and buses. Meanwhile, summer was generally found to have higher in-vehicle fungal concentrations than winter. The current in-vehicle airborne microbe concentrations were found to be similar to residential indoor values from other reports, such as bacterial values between 10 and 103  CFU m−3 and total fungal aerosol concentrations ranging from 10 to 103  CFU m−3. It was suggested that motor vehicles in which many individuals spend a certain percentage of their day or week for travel are also a potential microenvironment for individual exposure to airborne microbes.
Keywords: Airborne microbe; In-vehicle; Outdoor; Fungi; Season;

A scheme for drag partition over rough surfaces by Yaping Shao; Yan Yang (7351-7361).
We present a drag partition scheme for surfaces of various roughness densities. The total drag is partitioned into three components: a pressure drag, a skin drag due to momentum transfer to the surfaces of roughness elements and a surface drag due to momentum transfer to the ground surface. We introduce an effective frontal area index to characterize the mutual sheltering effect of roughness elements on drag partition and propose a method for the calculation of the drag components. The drag partition scheme is then applied to the calculation of bulk quantities such as roughness length and zero-plane displacement height. The results are compared with wind-tunnel measurements. Roughness length is found to increase with frontal area index λ to a maximum at about λ = 0.2 , and then decreases with λ . This finding is in agreement with observations. The drag partition scheme has applications to studying scalar exchange in boundary layers over rough surfaces, such as urban boundary layers.
Keywords: Drag partition; Effective frontal area index; Roughness length; Rough surface; Turbulence;

Atmospheric corrosion effects of HNO3—method development and results on laboratory-exposed copper by Farid Samie; Johan Tidblad; Vladimir Kucera; Christofer Leygraf (7362-7373).
The effects of the atmospheric pollutant nitric acid (HNO3) on materials compared to other corrosive gases, e.g. sulphur dioxide (SO2), nitrogen dioxide (NO2) or ozone (O3), have so far received little or no attention. The high sticking coefficient of HNO3 makes this gas one of the most difficult gases to work with. A new apparatus has now successfully been developed for studying the atmospheric corrosion effects of HNO3 on materials. HNO3 concentration measurements up to 1080 μg m−3 (420 ppb) were performed by dissolving the gas in water and analysing the nitrate concentration with ion chromatography (IC). Small changes in relative humidity (RH) largely affect the concentration of this pollutant in the exposure chamber and the high sticking coefficient of this gas on copper and quartz glass has been shown. The quartz glass surface, however, became saturated after a certain time of exposure and at 82% RH, the number of monolayers on the surface was estimated to be 10–13. Initial results of copper samples exposed to HNO3 show that at 63% RH and 25 °C, the deposition of HNO3 on copper is slightly lower than on a perfect absorber. The loss of HNO3 during exposure of the samples showed good agreement with the amount of nitrates dissolved from surfaces of the samples after exposure. FT–IR, XRD and IC analyses of copper exposed to HNO3 and mass loss and mass gain analyses confirmed cuprite (Cu2O) and the basic copper nitrate, gerhardtite, as the main corrosion products. Deposition, as well as the corrosion effect, of HNO3 on copper appeared to be greater than that of any of the other above-mentioned pollutants.
Keywords: Nitric acid; Relative humidity; Gerhardtite; Deposition; Material degradation;

Volatile organic compounds in a multi-storey shopping mall in guangzhou, South China by Jianhui Tang; C.Y. Chan; Xinming Wang; L.Y. Chan; Guoying Sheng; Jiamo Fu (7374-7383).
Volatile organic compounds (VOCs) specified in the USEPA TO-14 list were analysed in microenvironments of a multi-storey shopping mall in Guangzhou city, South China. The microenvironments studied include both indoor (department store, supermarket, fast-food court, electronic games room, children's playground, gallery and book store) and outdoor ones (rooftop and ground level entrance). The characteristics and concentration of VOCs varied widely in differing microenvironments. The average concentrations of the total VOCs in the indoor microenvironments ranged from 178.5 to 457.7 μg m−3 with a maximum of 596.8 μg m−3. The fast-food court and a leather products department store had the highest concentrations of benzene, toluene, ethylbenzene, xylenes and chlorinated hydrocarbons. A high level of 1,4-dichlorobenzene was found in all indoor microenvironments with an average of 12.3 μg m−3 and a maximum of 44.3 μg m−3. The ratios of average indoor to outdoor concentrations (I/O ratio) in all indoor microenvironments fell between 1 and 3, except an average of 24.6 and a maximum of 77.8 in the fashion department store for 1,4-dichlorobenzene. Indoor emission sources of monocyclic aromatic hydrocarbons in the shopping mall might include cooking stoves, leather products and building materials. Chlorinated hydrocarbons, however, were possibly connected with their use as cleaning agents or deodorizers.
Keywords: Volatile organic compounds; Indoor air quality; Shopping mall; Indoor microenvironments; Aromatic hydrocarbons; Chlorinated hydrocarbons;

Traffic-related differences in indoor and personal absorption coefficient measurements in Amsterdam, the Netherlands by Janine Wichmann; Nicole A.H. Janssen; Saskia van der Zee; Bert Brunekreef (7384-7392).
Population studies indicate that study participants living near major roads are more prone to chronic respiratory symptoms, lung function decrements and hospital admissions for asthma. The majority of the studies used proxy measures, such as distance to major roads or traffic intensity in the surroundings of the home. Few studies have communicated findings of concurrently performed measurements of outdoor, indoor and personal air pollution in urban streets with high- and low-traffic density. Measuring light absorption or reflectance of particulate matter (PM) collected on filters is an alternative method to determine elemental carbon, a marker for particles produced by incomplete combustion, compared to expensive and destructive analytical methods. This study sets out to test the null hypothesis that there is no difference in personal and indoor filter absorption coefficients for participants living along busy and quiet roads in Amsterdam. In one study we measured personal and indoor absorption coefficients in a sample of adults (50–70 years) and, in another study, the indoor levels in a population of adults (50–70 years) and school children (10–12 years). In the first study, the ratios of personal and indoor absorption coefficients in homes along busy roads compared with homes on quiet streets were significantly higher by 29% for personal measurements ( n = 16 days, p < 0.001 ), and by 19% for indoor measurements ( n = 20 , p < 0.001 ), while in the second study, the ratio for the indoor measurements was higher by 26% ( n = 25 days, p < 0.05 ). Exposure differences between homes along busy compared to homes along quiet streets remained and significant after adjustment for potential indoor sources (such as cooking and use of unvented heating appliances). This study therefore provides tentative support for the use of the type of road as proxy measure for indoor and personal absorption coefficient measurements in epidemiological studies due to the limitations of the study.
Keywords: Personal; Indoor; Exposure; Traffic-related; Absorption coefficients;

Atmospheric ammonia mixing ratios and the main inorganic ions NH4 +, NO3 and SO4 2− of size-resolved particles in the range from 0.05 to 10 μm were measured at an urban site in Münster, Germany. High mixing ratios of ammonia with a median of 5.2 ppb and a maximum of 50 ppb were detected. The mass fraction of submicron particles was much higher during the day than at night. At night, a greater particle mass and an increased presence of particulate nitrate was measured. Recurring patterns of particle distribution were distinguished and their characteristics analysed. In half of the measurements, the accumulation mode was clearly dominating, which is an indication of aged aerosol. In some measurements, higher concentrations of fine particles were found indicating particle formation. In these cases, a smaller particle mass and about four times greater ratios of ammonia versus ammonium concentrations were observed. These data show that ammonia contributes considerably to the formation of secondary particulate material.
Keywords: Atmospheric ammonia; Ammonium; Accumulation mode; Coarse mode; Particle formation; Particle mass; Urban aerosol;

Review of precipitation monitoring studies in India—a search for regional patterns by U.C. Kulshrestha; L. Granat; M. Engardt; H. Rodhe (7403-7419).
This study is a review of rain chemistry measurements reported in India. Data from nearly 100 stations were reviewed with regard to sampling location, sampling method and chemical analyses. Some characteristic differences were found between the concentrations in different environments in line with known distributions of emissions. This was most clearly seen for SO4 2− with median concentration increasing from rural and suburban to urban and industrial environments, with concentrations of HCO3 and H+ varying the opposite way. Ca2+ concentration was higher in suburban and industrial environments compared to rural and urban. Concentration of NH4 + was lower in rural than in the other locations. Sea salt concentration (Na+ and Cl in approximately constant proportion) was higher in urban than in rural locations at the same distance from the sea. The reported data from rural and suburban locations were used to derive large-scale concentration fields over India. With knowledge of emission fields and with the aid of output from a regional transport model it was possible to see some limited spatial resemblance between modelled and observed concentrations in cases of NH4 +, NO3 - and SO4 2−. For soil dust, current models are very crude but here some resemblance could be seen: primarily limited spatial resemblance with Ca2+, pH and HCO3 with generally increasing values towards NW. Four sites, at least two of them located in forested environments, had a mean rainwater pH close to neutral (concentration of H+ and HCO3 about equal). Other sites had on average an excess of HCO3 , becoming very substantial in the direction of the Thar Desert in NW India. With this review we conclude that there is a need to better assure the quality of the data with regard to sampling methods, chemical analyses and spatial representativeness.
Keywords: Precipitation chemistry; Regional patterns; Sulfate; Data quality analysis; Regional transport model;

Distribution of particle-phase hydrocarbons, PAHs and OCPs in Tianjin, China by Shui-Ping Wu; Shu Tao; Zhi-Huan Zhang; Tian Lan; Qian Zuo (7420-7432).
Aliphatic hydrocarbons, polycyclic aromatic hydrocarbons (PAHs) and organochlorine pesticides (OCPs) were determined in the total suspended particles (TSP) collected from 13 different locations in Tianjin, China, where intensive coal burning for domestic heating in winter takes place and a large quantity of pesticides had been produced and applied. Carbon preference index (CPI), carbon number maximum (Cmax) of n-alkane and plant wax index (%wax C n ) indicate that n-alkanes come from both biogenic and petrogenic sources, and biogenic source contributes more n-alkanes in autumn than in winter. Petroleum biomarkers as indicators of petrogenic source such as hopanes and steranes were also detected in both seasons’ samples. The sum of 16 PAH concentrations ( ∑ PAH 16 ) ranged from 69.3 to 2170 ng m−3 in winter and from 7.01 to 40.0 ng m−3 in autumn. Seasonal variations were mainly attributed to the difference in coal combustion emission and meteorological conditions. The results of a source diagnostic analysis suggest that PAHs in TSP mainly come from coal combustion. Seven OCPs (four hexachlorohexanes (HCHs) and three dichlorodipheny-trichloroethane and metabolites (DDTs)) were detected in most samples. Concentrations of the sum of α-, β-, δ- and γ-HCH ( ∑ HCH ) and the sum of p,p′-DDT, p,p′-DDD and p,p′-DDE ( ∑ DDT ) in autumn varied in the ranges of 0.002–0.9 ng m−3 and 0.025–2.21 ng m−3 with the average±standard deviation values of 0.127±0.241 ng m−3 and 0.239±0.546 ng m−3, respectively. In winter, ∑ HCH and ∑ DDT in TSP ranged from 0.071 to 5.35 ng m−3 and from 0.416 to 3.14 ng m−3 with the average±standard deviation values 1.05±1.88 ng m−3 and 0.839±0.713 ng m−3, respectively. Both of the illegal application of technical HCH and DDT and the volatilization from topsoil contributed to the particle-phase contents of HCHs and DDTs in the atmosphere.
Keywords: Aliphatic hydrocarbons; PAHs; HCHs; DDTs; Seasonal variation;

Measurements of CO2 fluxes from the Mexico City urban landscape by Erik Velasco; Shelley Pressley; Eugene Allwine; Hal Westberg; Brian Lamb (7433-7446).
In a densely populated section of Mexico City, an eddy covariance (EC) flux system was deployed on a tall urban tower to obtain direct measurements of CO2 emissions from an urban neighborhood located in a subtropical megacity. The measured fluxes and boundary layer conditions satisfy EC assumptions of stationarity, and cospectral analyses of the turbulence measurements exhibit the required boundary layer patterns for acceptable flux measurements. Results from a field experiment conducted during April 2003 show that the urban surface is a net source of CO2. The CO2 flux measurements showed a clear diurnal pattern, with the highest emissions during the morning (up to 1.60 mg m−2  s−1), and the lowest emissions during nighttime. The measured fluxes were closely correlated to traffic patterns in the area. The mean daily flux was 0.41 mg m−2  s−1, which is similar to that observed in European and US cities.
Keywords: Carbon dioxide fluxes; Eddy covariance; Mexico; Megacity;

New Directions: Are we prepared for a wet MIRAGE? by Darrel Baumgardner; Graciela B. Raga; Guy Brasseur (7447-7448).