Atmospheric Environment (v.36, #30)

A semi-analytical model for mean concentration in a convective boundary layer by Massimo Cassiani; Umberto Giostra (4707-4715).
A model to predict the mean concentration field in convective conditions is proposed. This model is inspired by the probability density function (pdf) models, retaining their assumption of splitting a plume into updraft and a downdraft components. However, the proposed model is more flexible than the pdf models since it includes the effects of turbulence inhomogeneity and it can be adapted to non-stationary conditions. The formulation of the model ensures that in the limit of zero skewness a traditional Gaussian model is obtained. Thus, the model is applicable to conditions ranging from neutral to unstable. The model is solved numerically, but time consumption is negligible. Despite its simplicity, the model shows good performance compared to classical experiments and to more complex and physically consistent numerical models.
Keywords: Turbulent diffusion; Air pollution modelling; Convective boundary layer; Mean concentration model;

Recently, a modified meandering plume model for concentration fluctuations in a convective boundary layer has been formulated (Atmos. Environ. 34 (2000) 3599). This model is based on a hybrid Eulerian–Lagrangian approach and it accounts for the skewed and inhomogeneous turbulence characteristics of the convective flow. Using the same hypotheses, but eliminating the need for Lagrangian particle model, we propose a generalised approach, that only requires the knowledge of mean concentration field. The proposed model is independent from the method used to obtain the mean concentration field. The evaluation of the concentration fluctuation field needs a computational time of only few seconds on a standard PC. Therefore, the model is suitable for practical applications.
Keywords: Air pollution modelling; Mean concentration field; Meandering plume model; Higher order concentration moment modelling;

Development of a reduced speciated VOC degradation mechanism for use in ozone models by Michael E Jenkin; Sandra M Saunders; Richard G Derwent; Michael J Pilling (4725-4734).
A reduced mechanism to describe the formation of ozone from VOC oxidation has been developed, using the master chemical mechanism (MCM v2) as a reference benchmark. The ‘common representative intermediates’ (CRI) mechanism treats the degradation of methane and 120 VOC using ca. 570 reactions of ca. 250 species (i.e. the emitted VOC plus an average of about one additional species per VOC). It thus contains only ca. 5% of the number of reactions and ca. 7% of the number of chemical species in MCM v2, providing a computationally economical alternative. The CRI mechanism contains a series of generic intermediate radicals and products, which mediate the breakdown of larger VOC into smaller fragments (e.g., formaldehyde), the chemistry of which is treated explicitly. A key assumption in the mechanism construction methodology is that the potential for ozone formation from a given VOC is related to the number of reactive (i.e., C–C and C–H) bonds it contains, and it is this quantity which forms the basis of the generic intermediate groupings. Following a small degree of optimisation, the CRI mechanism is shown to generate levels of ozone, OH, peroxy radicals, NO and NO2 which are in excellent agreement with those calculated using MCM v2, in simulations using a photochemical trajectory model applied previously to simulation of episodic ozone formation. The same model is used to calculate photochemical ozone creation potentials for 63 alkanes, alkenes, carbonyls and alcohols using both mechanisms. Those determined with the CRI mechanism show a variation from compound to compound which is remarkably consistent with that calculated with the detailed chemistry in MCM v2. This suggests that the CRI mechanism construction methodology is able to capture both the salient features of the ozone formation process in general, and how this varies from one VOC to another.
Keywords: Speciated VOC oxidation; Tropospheric chemistry; Degradation mechanisms; Ozone modelling;

Metal emissions from road traffic were studied in two heavily trafficked tunnels in Gothenburg, Sweden. Five runs were made in each tunnel, generally extending over several hours, during which air concentrations of metals in tunnel inlet and outlet, traffic flow and composition, and air ventilation were determined. Emission factors for Cu, Zn, Cd, Sb, Ba and Pb in TSP were derived from these data, mainly representing vehicle emissions. For these metals, results from the two tunnels were fairly consistent and the variability of the emission factors was generally within the experimental uncertainty. In contrast, concentrations of TSP, PM10 and a number of other metals differed widely between the two tunnels. This difference was caused by different magnitudes of resuspension, mainly reflecting that the two tunnels were investigated at different seasons and that one of the tunnels was newly cleaned. Aerosol concentrations of larger particles and of several metals are apparently dominated by resuspended matter, rather than by emissions from vehicles. The vehicle-derived metals (Cu, Zn, Cd, Sb, Ba and Pb) mainly derive from wear rather than from combustion; Cu, Ba and Sb are probably dominated by brake wear. There are also indications that heavy duty vehicles are stronger emitters of Ba and Sb, but not of Cu, than light duty vehicles.
Keywords: Emission factors; TSP; Particles; Antimony; Brake linings; Vehicle emissions;

The benzene emission characteristics of six gasoline-driven EURO-2 vehicles, three passenger cars and three light duty vehicles, have been determined by time-resolved chemical ionization mass spectrometry. Aliquots of the exhaust gas were monitored pre- and post-catalyst with two independently operating mass spectrometers. Each vehicle was driven with two different fuels having benzene contents of 1 and 2 vol%. Seven driving cycles—including the European (EDC) and the US (FTP-75) driving cycle—with a total driving time of about 8800 s were studied. Herein, we discuss the average emission characteristics of the entire fleet at transient driving in the velocity range of 0–150 km h−1. The conversion efficiencies of the involved catalytic systems were deduced from the pre- and post-catalyst data. On average, the vehicles showed optimal benzene conversion efficiencies (>95%) in the velocity range of 30–90 km h−1. When driving below 20 or above 100 km h−1 reduced benzene conversion was found (80–82%). No benzene conversion was observed when driving above 130 km h−1. In contrast, the post-catalyst benzene emissions exceeded those of the untreated exhaust gas by 19–49%. Thus on an average, benzene was formed across the catalysts under these conditions. In addition, the influence of the benzene content of the gasoline on the tail-pipe emissions was also studied. The use of the gasoline with 1 vol% benzene instead of 2 vol% induced a 20–30% reduction of the post-catalyst emissions when driving below 50 km h−1. The fuel effect became smaller above 100 km h−1 and was even negative at high engine load (>130 km h−1). Thus under these conditions, when benzene is formed across the catalyst, the amount of the emitted benzene was independent of the benzene level of the fuel.
Keywords: Benzene emission factors; Conversion efficiency; On-line chemical ionization mass spectrometry; Time-resolved exhaust gas analysis; De novo benzene formation;

A wind tunnel experiment was carried out to simulate stack gas diffusion within an unstable atmospheric boundary layer over a coastal region. The wind tunnel floor, 4 m leeward of the entrance of the test section, was heated to 90°C over a length of 6 m in the streamwise direction, and wind tunnel experiments were performed under the flat plate condition with a prototype-to-model length scale ratio of 1200. Three similarity criteria of flow fields in the wind tunnel and in atmosphere, viz., bulk Richardson number, surface Reynolds number and the ratio of the Peclet number to the Richardson number, were considered in the wind tunnel experiment. Tracer gas was released along the coastline at a height of 10 cm, which corresponded to 120 m in height in atmosphere. The obtained wind tunnel experimental results of ground level concentration were compared with 30-min average values of the field experiments, viz., the data from the Tokai 82 field experiment. The maximum ground level concentration and its location were accurately simulated when there was close similarity between the wind tunnel and atmospheric flow conditions. The maximum concentration increased and occurred closer to the source when the level of convection was relatively stronger in atmosphere.
Keywords: Atmospheric diffusion; Unstable stratified flow; Similarity law; Ground level concentration; Field observation;

Statistical analysis of primary and secondary atmospheric formaldehyde by Stephen Friedfeld; Matthew Fraser; Kathy Ensor; Seth Tribble; Dirk Rehle; Darrin Leleux; Frank Tittel (4767-4775).
Regression models coupled with time series data were used to analyze the contribution of primary and secondary sources to formaldehyde (HCHO) concentrations, as determined by statistical analogy to primary (carbon monoxide, CO) and secondary (ozone, O3) compounds measured simultaneously in Houston, TX. Time series analyses substantiated the need for statistical methods of analysis, given the complexity of the data and the rapid fluctuations that occur in atmospheric concentrations. A positive relationship was found for both the auto-correlation function (ACF) and partial auto-correlation function (PACF) of HCHO with either CO or O3. Regression models used to distinguish primary and secondary contributions included a simple linear regression of the three compounds (one lag unit of time, 5 min) on current HCHO concentrations, resulting in a ratio of secondary formation to primary emission of 1.7. A second, more robust model utilized auto-correlated error processes to approximate the true nature of the linear regression; this model also indicates the ratio of secondary to primary contribution at 1.7 as the mean of ten model simulations. From the error processes model, one lag unit of time was most significant for CO predicting HCHO, while simultaneous measurements (lag 0) were most significant for O3 predicting HCHO. Outlying O3 and HCHO concentrations were shown not to affect the results.
Keywords: Formaldehyde; Statistical modeling; Secondary formation; Houston; TX;

An empirical model to predict indoor NO2 concentrations by Milind M. Kulkarni; Rashmi S. Patil (4777-4785).
This paper discusses the development of an empirical model to predict indoor NO2 concentration (C in) from outdoor NO2 concentration (C out) measured just outside the homes. Concentrations were 48-h time integrated averages. Data from a study carried out on measurement of indoor, outdoor and personal exposure to nitrogen dioxide (NO2) in Mumbai were used in the development of the model. Concentrations were measured over 2-day periods both in winter (February 1996) and summer (April 1996) for 43 respondents. The form of the developed model is C ̄ in =k C ̄ out +S/Q, where k, the ventilation factor and S/Q the source term are the model constants. Analysis of variance and regression analysis indicated that type of fuel was the most significant factor influencing indoor concentration and model constants. Measured indoor concentrations were regressed on outdoor concentrations to evaluate model constants for kerosene and liquefied petroleum gas (LPG), which are the most predominantly used cooking fuels in Indian households. Final models, after evaluating the constants suggested that contribution to indoor NO2 concentration due to indoor sources was higher in kerosene using households whereas in the case of LPG using households, the contribution due to outdoor sources was relatively higher. Results of model validation indicated that the predictive power of the models was good.
Keywords: Kerosene; LPG; Outdoor NO2 concentration; Ventilation factor; Source term;

On ozone dry deposition—with emphasis on non-stomatal uptake and wet canopies by Leiming Zhang; Jeffrey R Brook; Robert Vet (4787-4799).
Measurements of O3 fluxes and concentrations over five different sites were used to study O3 dry deposition. It was found that high humidity, dew and rain increase O3 uptake by canopy cuticles. However, the increase by cuticle uptake maybe offset by a decrease in stomatal uptake due to weak solar radiation or stomatal blocking under wet conditions. Thus, during nighttime the overall canopy resistances (R c) for O3 uptake under wet conditions was usually smaller than under dry conditions, while in the daytime, R c for wet canopies could be either larger or smaller compared to dry canopies. This will depend on the relative contributions of the decrease in cuticle resistance and the increase in stomatal resistance. The non-stomatal uptake of O3 was found to be affected by friction velocity, relative humidity, canopy wetness, leaf area index, etc. Parameterizations for non-stomatal resistance for dry and wet canopies were developed based on the five site O3 flux data. These equations were found to provide reasonable predictions of non-stomatal canopy resistance based upon comparisons with the nighttime and daytime measurements.
Keywords: Dry deposition; Non-stomatal uptake; Ozone; Surface resistance; Wet canopies;

Particle deposition in low-speed, high-turbulence flows by Mads Reck; Poul S Larsen; Ulrik Ullum (4801-4809).
The experimental and numerical study considers the concentration of airborne particulate contaminants, such as spores of spoilage fungi, and their deposition on a surface, in a Petri dish, and on a warm box-shaped product placed in a food-processing environment. Field measurements by standard, active and passive samplers provide typical values of airborne concentrations and specific deposition fluxes. Velocity and turbulence data from field studies are used as input in large eddy simulations of the process, and estimates of deposition fluxes are of the same order of magnitude as those deduced from field measurements. Particle deposition is shown to be associated with near-wall coherent structures. Flow reversal, simulated by impulsive start, is shown to give higher deposition rates than steady mean flows.
Keywords: Spoilage fungi; Spores; Food-processing plant; Deposition flux; Large eddy simulation;

Validation and application of obstacle-resolving urban dispersion models by Michael Schatzmann; Bernd Leitl (4811-4821).
The development of micro-scale meteorological models has progressed in recent years. Some of them are already commercially available. With little hesitation, consulting engineers apply them to complex real-world problems. How accurate are the results? Using the example of urban dispersion models, the paper tries to give a critical assessment of the present ‘state of application’.
Keywords: Urban dispersion; Numerical results; Field experiments; Wind tunnel experiments; Validation; Uncertainty;

Chemical and physical characterization of emissions from birch wood combustion in a wood stove by Emma Hedberg; Adam Kristensson; Michael Ohlsson; Christer Johansson; Per-Åke Johansson; Erik Swietlicki; Vaclav Vesely; Ulla Wideqvist; Roger Westerholm (4823-4837).
The purpose of this study was to characterize the emissions of a large number of chemical compounds emitted from birch wood combustion in a wood stove. Birch wood is widely used as fuel in Swedish household appliances. The fuel load was held constant during six experiments.Particles <2.5 μm in diameter were collected and the size distribution of the particles was measured. The results were compared to the size distribution in road traffic emissions. It could be seen that the number distribution differed between the sources. In traffic exhaust, the number of particles maximized at 20 nm, while the number distribution from wood burning ranged from 20 to 300 nm. The ratio K/Ca on particles was found to be significantly different in wood burning compared to road dust, range 30–330 for the former and 0.8±0.15 for the latter. The source profile of common elements emitted from wood burning differed from that found on particles at a street-level site or in long-distance transported particles.The ratio toluene/benzene in this study was found to be in the range 0.2–0.7, which is much lower than the ratio 3.6±0.5 in traffic exhaust emissions.Formaldehyde and acetone were the most abundant compounds among the volatile ketones and aldehydes. The emission factor varied between 180–710 mg/kg wood for formaldehyde and 5–1300 mg/kg wood for acetone. Of the organic acids analyzed (3,4,5)-trimethoxy benzoic acid was the most abundant compound. Of the PAHs reported, fluorene, phenanthrene, anthracene, fluoranthene and pyrene contribute to more than 70% of the mass of PAH. Of the elements analyzed, K and Si were the most abundant elements, having emission factors of 27 and 9 mg/kg wood, respectively.Although fluoranthene has a toxic equivalence factor of 5% of benzo(a)pyrene (B(a)P), it can be seen that the toxic potency of fluoranthene in wood burning emissions is of the same size as B(a)P. This indicates that the relative carcinogenic potency contribution of fluoranthene in wood smoke would be about 40% of B(a)P.
Keywords: Particle size distribution; BTX; PAH; Wood combustion; Aldehydes; Ketones;

Food contamination by C20–C50 mineral paraffins from the atmosphere by Hans-Peter Neukom; Koni Grob; Maurus Biedermann; Anja Noti (4839-4847).
Most foods from plant origin usually contain 1–10 mg/kg (dry weight) of non-resolved isomeric alkanes in the range of the n-alkanes C20–C50 which are assumed to be residues from mineral oil products (in addition to the natural paraffins). In edible vegetable oils, concentrations may exceed 100 mg/kg. Since it was suspected that this contamination was mostly of environmental origin, particulate matter from air was analysed for the same range of paraffins. In a road tunnel, around 5 μg/m3 of such paraffins were found, corresponding to about 3% of the fine dust (PM10). The composition corresponded to that found in the particulate matter from the exhaust of diesel engines, which in turn largely corresponded to engine (lubricating) oil. In Swiss cities, the C20–C50 mineral paraffins in the PM10 from ambient air amounted to 0.1–1.5 μg/m3 (about 1% of the dust) and seemed to primarily originate from incomplete combustion of heating and diesel oil, lubricating oil, and road tar debris. On the countryside, the concentrations were around 0.03 μg/m3 (0.3% of the dust). Soil contained 0.5–10 mg/kg of these paraffins. The similarity of the molecular weight (volatility) distribution suggests that the food contamination with paraffins, is mostly from the air. A substantial proportion probably consists of lubricating oil. If this hypothesis is confirmed, measures should be investigated to reduce this contamination.
Keywords: Mineral paraffins; Particulate matter in air; Edible oils; Lubricating oil; Food contamination;