Atmospheric Environment (v.34, #27)

Introduction to Transport and Air Pollution by Peter Sturm (4579-4580).

Simulation of the air quality during a field study for the city of Graz by R.A Almbauer; D Oettl; M Bacher; P.J Sturm (4581-4594).
The research project DATE Graz (Dispersion of Atmospheric Trace Elements taking the city of Graz as an example) aimed at the investigation of mesoscale γ pollution dispersion of a city in complex terrain. The winter episode investigated here was characterised by an anticyclonic fair weather situation. Local wind systems developed together with strong temperature inversions. During such an episode air quality is dominated by emissions from within the city. The city of Graz is situated in the southeast of the Alps in the transition area of mountainous to flat land. The city itself is located in the valley of the river Mur, which forms a basin surrounded by small mountains. In order to understand the emitter–receptor relationship a mesoscale dispersion model was applied for the simulation of air quality during the winter episode. Input for the simulation was the emission inventory and extensive meteorological measurement data. Results of the simulation reflect the distinct patterns of daily variations of air quality level measured. The influence of meteorology, emission patterns and chemical reactions are evident and can be qualitatively and partly quantitatively simulated by the model. A validation attempt was made using air quality data from the monitoring network. The introduction shows the importance of anticyclonic fair weather conditions on pollution dispersion in mountainous regions. Section 2 describes the meteorological situation and the instrumentation during the winter measurement campaign. Section 3 deals with the simulation model, initial and boundary conditions and the emission inventory for the city. In Section 4 the simulation results are compared to measurements. Finally conclusions are drawn.
Keywords: Mesoscale model; Pollution dispersion; Urban air quality; Complex terrain;

An urban transport emission model for the Antwerp area by C Mensink; I De Vlieger; J Nys (4595-4602).
We present a detailed modelling approach which provides hourly emissions of CO, NO x , VOC, PM, SO2 and Pb for individual streets and road segments in the Antwerp area (20 km×20 km). The hourly emissions are computed as a function of road type, vehicle type, fuel type, traffic volume, vehicle age, trip length distribution and the actual ambient temperature. The traffic volumes are derived from an urban traffic flow model for the city of Antwerp, which contains a network with almost 2000 road segments. The traffic flow model has been implemented in a GIS environment. The emission factors used in the model are derived from the COPERT-II methodology. Cold start emissions and evaporation losses are included in the model. Results are shown for the hourly hot- and cold start emissions obtained for CO, NO x and VOC. For these pollutants a partial validation of the model results was carried out by comparing the COPERT-II emission factors with on-the-road traffic emission measurements carried out for gasoline passenger cars with a closed-loop controlled three-way catalyst.
Keywords: Emission modelling; Road transport emissions; Urban air-quality assessment;

Uncertainties of modelling emissions from road transport by J Kühlwein; R Friedrich (4603-4610).
To determine emission data from road transport, complex methods and models are applied. Emission data are characterized by a huge variety of source types as well as a high resolution of the spatial allocation and temporal variation. So far, the uncertainties of such calculated emission data have been largely unknown. As emission data is used to aid policy decisions, the accuracy of the data should be known. So, in the following, the determination of uncertainties of emission data is described. Using the IER emission model for generating regional or national emission data, the uncertainties of model input data and the total errors on different aggregation levels are exemplarily investigated for the pollutants NO x and NMHC in 1994 for the area of West Germany. The results of statistical error analysis carried out for annual emissions on road sections show variation coefficients (68.3% confidence interval) of 15–25%. In addition, systematic errors of common input data sets have been identified especially affecting emissions on motorway sections. The statistical errors of urban emissions with warm engine on town level amount to 35%. Therefore they are considerably higher than the errors outside towns. Error ranges of additional cold start emissions determined so far have been found in the same order. Additional uncertainties of temporally highly resolved (hourly) emission data depend strongly on the daytime, the weekday and the road category. Variation coefficients have been determined in the range between 10 and 70% for light-duty vehicles and between 15 and 100% for heavy-duty vehicles. All total errors determined here have to be regarded as lower limits of the real total errors.
Keywords: Road traffic; Statistical error propagation; Systematic error analysis; Temporally highly resolved emission data; Spatial intersection;

This paper presents speed-dependent hot emission factors for passenger cars equipped with a three-way catalyst. The emission factors were developed using emission data from in-use vehicles based on chassis dynamometer tests conducted over real-world driving cycles. The variability of the emission data and the use of cycles with different characteristics resulted in a low correlation between emission factors and mean travelling speed. Analysis of the data identified differences in vehicle emission performance as the main source of variability. The potential for decreasing emission variability by classifying vehicles according to their engine capacity and mileage was studied by means of an analysis of variance test. Results showed that CO, NO x and HC emissions were dependent on mileage but not on engine size. Emissions of CO2 were found to be strongly dependent on engine size, but insensitive to vehicle mileage.
Keywords: Emission factors; Emission variability; Engine capacity; Mileage; Passenger cars; Pollutants;

Influence of driving cycles on unit emissions from passenger cars by R Joumard; M André; R Vidon; P Tassel; C Pruvost (4621-4628).
Small samples of petrol engine or diesel cars, equipped with or without catalysts, were tested over 36 driving cycles divided into four categories – standard cycles and three sets of cycles more representative of real-world driving conditions. The tests addressed standard gaseous pollutants and fuel consumption and also less frequently measured pollutant such as CH4. In the first part of this paper we examine cold emissions in order to assess the duration of the cold start impact and the representativity of the cold ECE15 cycle. Then unit emissions are compared over the four driving cycle families. As compared to representative cycles, the standardised cycles underestimate hot emissions by almost 50% for petrol engine cars and 30% for diesel vehicles. Conversely, the results obtained for the three representative cycle families are in relatively close agreement with each other – within approximately 10%. However, the cinematic properties of the three families differ. Finally, we demonstrate that weighting all emission data equally, not taking into account the weight of each cycle in overall traffic, introduces significant biases, particularly when plotting emission vs. average speed curves.
Keywords: Vehicle emission; Average speed; Driving cycle; Standard cycle; Cold start; Hot emission;

This paper discusses the use of ‘instantaneous’ high-resolution (1 Hz) emission data for the estimation of passenger car emissions during real-world driving. Extensive measurements of 20 EURO-I gasoline passenger cars have been used to predict emission factors for standard (i.e. legislative) as well as non-standard (i.e. real-world) driving patterns. It is shown that emission level predictions based upon chassis dynamometer tests over standard driving cycles significantly underestimate emission levels during real-world driving. The emission characteristics of modern passenger cars equipped with a three-way catalytic converter are a low, basic emission level on the one hand, and frequent emission ‘peaks’ on the other. For real-world driving, up to one-half of the entire emission can be emitted during these short-lasting peaks. Their frequency depends on various factors, including the level of ‘dynamics’ (speed variation) of the driving pattern. Because of this, the use of average speed as the only parameter to characterize emissions over a specific driving pattern is not sufficient. The instantaneous emissions approach uses an additional parameter representing engine load in order to resolve the differences between driving patterns with comparable average speeds but different levels of ‘dynamics’. The paper includes an investigation of different statistical indicators and discusses methods to further improve the prediction capability of the instantaneous emission approach. The fundamental differences in emission-reduction strategies between different car manufacturers make the task of constructing a model valid for all catalyst passenger cars seemingly impossible, if the model is required to predict both fleet-averaged emission levels and emission factors for driving patterns of short duration for individual vehicles simultaneously.
Keywords: Emission factors; Real-world driving behavior; Instantaneous emissions; Standard driving cycles; Catalyst cars; Modal modeling;

Determination of VOC-components in the exhaust of gasoline and diesel passenger cars by Thomas Schmitz; Dieter Hassel; Franz-Josef Weber (4639-4647).
The composition of VOC emissions from in-use passenger cars with different engine types, i.e. cars with diesel engines, cars with petrol engines equipped with three-way-catalysts, and cars with petrol engines without catalysts was determined. Five cars of each engine type have been measured on a chassis dynamometer under conditions of the US FTP 75 test procedure and the “Autobahn” test developed by TÜV Rheinland. Measurements of C2–C10 hydrocarbons were made with a GC-FID system. In addition, samples on DNPH cartridges were taken and analysed by means of a HPLC-system for the determination of aldehydes and ketones.The influence of cold/warm-conditions on the VOC composition was determined. In the case of cars with diesel engines as well as for the petrol-driven cars without exhaust treatment, the effect caused by the cold start only led to minor changes in the VOC composition. A similar behaviour was observed for these car types at higher speeds. In contrast to the cars without catalysts, the cars with three-way-catalysts showed a great variability of the VOC composition. During the cold start phase the aromatic compounds and the alkenes yielded the main fraction of the VOC. During the warm phase the less reactive alkanes were predominant. With increasing mean velocities the VOC composition changed in favour of the more reactive compounds.
Keywords: Passenger car emissions; C2–C10 hydrocarbons; Aldehydes; Ketones; Ozone forming potential;

Environmental effects of driving behaviour and congestion related to passenger cars by I De Vlieger; D De Keukeleere; J.G Kretzschmar (4649-4655).
Using Vito's on-board measuring system the influence of track, driving behaviour and traffic conditions on fuel consumption and emissions were studied for a small test fleet of passenger cars. City traffic resulted in the highest fuel consumption and emissions. Fuel consumption was about two times higher than for ring roads, which generally gave the lowest values. This was even more pronounced for emissions. Depending on road type and technology, fuel consumption increased with up to 40% for aggressive driving compared to normal driving. Again, this was more pronounced for emissions, with increases up to a factor 8. Driving behaviour had a greater influence on petrol-fuelled than on diesel-fuelled cars.Traffic condition also has a major effect on fuel consumption and emissions. For city driving intense traffic increased fuel consumption by 20–45%. The increase in fuel consumption and emissions during rush hours were the highest on ring roads, with increases between 10 and 200%. In absolute terms, a surplus of up to 5 l fuel per 100 km was measured. More environment-friendly route option requires the use of ring roads and motorways during rush hours instead of short cuts.
Keywords: On-board measurements; Emissions; Fuel consumption; Driving behaviour; Traffic conditions;

On-road optical remote-sensing measurements of gasoline light-duty vehicle (LDV) emissions – CO, HC, NO – were conducted on a freeway ramp in Gothenburg, Sweden, in 1991, 1995 and 1998. Based on almost 30,000 emission measurements, the results show that both catalyst cars and non-catalyst cars emissions deteriorate over time, but also that the emission performance of new TWC-cars has improved significantly in recent years. Furthermore, it was found that fleet age rather than model year determines the rate of emission deterioration for TWC-cars for both CO and NO. The study demonstrates that remote sensing may constitute a powerful tool to evaluate real-world LDV emissions; however, daily field calibration procedures need to be developed in order to assure that the evolution in fleet average emissions can be accurately measured.
Keywords: Light-duty vehicles; Emissions; On-road; Measurements; CO; HC; NO; Optical remote sensing; FEAT;

The past two decades have seen significant advances in mesoscale meteorological modeling research and applications, such as the development of sophisticated and now widely used advanced mesoscale prognostic models, large eddy simulation models, four-dimensional data assimilation, adjoint models, adaptive and targeted observational strategies, and ensemble and probabilistic forecasts. Some of these advances are now being applied to urban air quality modeling and applications. Looking forward, it is anticipated that the high-priority air quality issues for the near-to-intermediate future will likely include: (1) routine operational forecasting of adverse air quality episodes; (2) real-time high-level support to emergency response activities; and (3) quantification of model uncertainty. Special attention is focused here on the quantification of model uncertainty through the use of ensemble simulations. Application to emergency-response dispersion modeling is illustrated using an actual event that involved the accidental release of the toxic chemical oleum. Both surface footprints of mass concentration and the associated probability distributions at individual receptors are seen to provide valuable quantitative indicators of the range of expected concentrations and their associated uncertainty.
Keywords: Air quality modeling; Regional air quality analysis; Emergency response; Uncertainty, ensembles;

Simulation of traffic-induced air pollution on regional to local scales by W Brücher; C Kessler; M.J Kerschgens; A Ebel (4675-4681).
The model system CARLOS (Chemistry and Atmospheric transport in Regional and LOcal Scale) has been developed with the aim to simulate atmospheric transport and air chemistry on local scales taking into account the influence of large-scale transport. The system has been applied to a photochemical smog episode in summer 1997. Multiple nested simulations from the European scale down to a nest containing mainly the city of Wuppertal, Germany were carried out. The emission inventory for Wuppertal is based on microscopic traffic flow simulations. The simulated concentration pattern in Wuppertal is strongly influenced by local properties like terrain and emission structure as well as by specific regional factors such as urban plumes of cities nearby.
Keywords: Nesting; Photochemical smog modelling; Complex terrain; Urban air pollution; Urban plume;

Impact of road traffic emissions on air quality of the Lisbon region by C Borrego; O Tchepel; N Barros; A.I Miranda (4683-4690).
The main purpose of this paper is to present the study of traffic emissions impact on the Lisbon region air quality. Two approaches of emission data generation with high spatial and temporal resolution are presented and compared. Main roads were processed as line sources and hot on-road emissions were calculated based on daily mean traffic and emission factors distinguished for several road classes and vehicle types. Also, the disaggregation of national CORINAIR inventory has been performed on the basis of statistical information of fuel consumption and population density. The comparison of emission data obtained by these two approaches demonstrates a good agreement for total values, but a significant difference for spatial distribution of the data. To ensure completeness of the data, to improve their spatial resolution and also to analyse the impact of the traffic emissions, a combination of the two approaches was applied to generate the emission data used by a photochemical numerical system to simulate the atmospheric circulation and the air pollution pattern in Lisbon under summer meteorological conditions, having different emission scenarios. It was possible to conclude that an air pollution abatement strategy is urgently needed and it should take into account the strong contribution of road traffic emissions to the Lisbon air pollution levels.
Keywords: Traffic emission; Emission inventory; Photochemical modelling; Portugal; Lisbon;

Assessing ozone abatement strategies in terms of their effectiveness on the regional and urban scales by N Moussiopoulos; P Sahm; P.M Tourlou; R Friedrich; D Simpson; M Lutz (4691-4699).
Integrated ozone abatement strategies should take into account that an emission intervention which is effective on the regional scale may not necessarily be effective for a city and its surroundings. In the context of a study performed for the European Commission a methodology was developed for assessing ozone abatement strategies in terms of their effectiveness on both the regional and the urban scale. Towards this aim, the assumptions made for three regional emission reduction scenarios were assumed to be valid also for the emission situation in the urban areas of Athens and Stuttgart and the corresponding emission inventories were compiled. The EMEP MSC-W ozone model was used to calculate the regional scale ozone distribution over a 6 month summer period applying the 1990 meteorology. Local scale transport and chemical transformation processes were analysed with the ozone fine structure (OFIS) model. Both the regional and the local scale simulations were performed for the base case (1990 situation) and the three emission reduction scenarios. The significance of regional scale emission reductions was demonstrated by the aid of a second series of simulations assuming that the emission interventions were implemented only at local scale. The results of the simulations reveal that ozone exposure in conglomerations as the ones considered in this study depends on both urban and regional scale influences. Urban VOC control is found to be effective in reducing ozone primarily on the local or urban scale, whereas urban NO x control may cause an increase of urban peak ozone while contributing to an effective reduction of regional ozone.
Keywords: Pollution abatement; Ozone strategy; Emission reduction scenarios; OFIS model;

Road traffic emissions – predictions of future contributions to regional ozone levels in Europe by S Reis; D Simpson; R Friedrich; J.E Jonson; S Unger; A Obermeier (4701-4710).
As part of the European Commission research project “Assessment of policy instruments for efficient ozone abatement strategies in Europe”, detailed emission projections have been developed for the year 2010 based upon currently adopted measures, and feasible reductions. For road-traffic emissions this projection considers passenger cars, light- and heavy-duty vehicles, mopeds and motorcycles. Here we present model calculations made with the EMEP 3-D Eulerian model to illustrate the relative contribution of each of these road-traffic sectors to ozone concentrations across Europe. The model is run for a six-month period, April–September 1996. The model results clearly suggest that further reduction in road-traffic emissions beyond currently planned measures would be beneficial in reducing ozone over Europe, particularly in the case of heavy-duty vehicles and evaporative emissions. These results do of course depend on the estimated emissions in each sector for the year 2010, and we show that this is a major source of uncertainty in such scenario calculations
Keywords: Tropospheric ozone; Road traffic; Emission projection; NMVOC; NO x ;

The ammonium nitrate particle equivalent of NO x emissions for wintertime conditions in Central California's San Joaquin Valley by William R. Stockwell; John G. Watson; Norman F. Robinson; William Steiner; William W. Sylte (4711-4717).
A new method has been developed to assess the aerosol particle formation reactivity of nitrogen oxide (NO x ) emissions. The method involves using a photochemical box model with gas-phase photochemistry, aerosol production and deposition to calculate the ammonium nitrate particle equivalent of NO x emissions. The yields of ammonium nitrate particles used in the box model were determined from parametric simulations made with an equilibrium model that calculated the fraction of nitric acid that reacts to produce ammonium nitrate from the temperature, relative humidity and ammonium-to-nitrate ratios. For the wintertime conditions of emissions and meteorology in the San Joaquin Valley of central California, approximately 80% of the moles of nitric acid produced was found to be in the particulate nitrate phase and about 33% of the moles of emitted NO x was converted to particulate nitrate. The particle equivalent of NO x emissions was found to be on the order of 0.6 g of ammonium nitrate for each gram of NO x emitted (the mass of NO x calculated as NO2). This estimate is in reasonable agreement with an analysis of field measurements made in central California.
Keywords: PM2.5; Secondary aerosol; Nitrate; San Joaquin Valley; Emissions trading;

Volatile organic compounds (VOCs) were measured with a motor-glider during six days of July and August, 1996, in Switzerland’s Mesolcina valley. Morning and afternoon profiles from the motor-glider reached up to 4000 m above sea level (m asl). Additional measurements were made on the valley floor with the same type of instrument. The concentration differences at various altitudes between morning and afternoon were correlated with traffic emission factors for the individual VOCs. Traffic emissions govern the concentrations of anthropogenic VOCs in this Alpine valley and above the crests. Due to vertical transport of air, VOC concentrations can increase strongly during daytime at high altitudes (∼2000–4000 m asl) over the Alps.
Keywords: VOC; Transport of air pollutants; Vertical mixing; Mountain meteorology; Traffic emission factors;

Vehicular traffic is a common important source of air pollution, traffic accidents, road traffic noise as well as other environmental exposures. The relationship between each of these exposures and their respective impacts are nevertheless most often studied separately. An integrated alternative approach was adopted in the Oslo traffic study to allow people's environmental annoyances to be studied relative to the indicators of air pollution, road traffic noise and residential traffic. These annoyances include annoyance with the smell of exhaust, with dust and grime, feeling insecure in traffic and being annoyed with road traffic noise. A hypothesis was that multiple exposures typical in city areas have combined impacts – that people exposed to both air pollution and road traffic noise will be more annoyed than in the respective single-exposure situations. Three environmental studies in 1987, 1994 and 1996 each year comprising about 1000 respondents after a response rate of 50%, serve as before–after studies of two tunnel projects. Personal interviews were utilised in the before study in 1987 and telephone interviews in 1994 and 1996. Exposure indicators for air pollution as well as road traffic noise and residential traffic levels were produced for each respondent by comprehensive environmental modelling. Exposure–effect logistic regression models for the probability of people being highly annoyed by the smell of exhaust and by road traffic noise, respectively, were estimated. The results indicate that the higher the road traffic noise levels people are exposed to, the more likely they are to be highly annoyed by exhaust smell at a specified air pollution level. The higher air pollution levels people are exposed to the more likely they are to be annoyed by road traffic noise at a specified noise level. Modifying factors were controlled for.
Keywords: Environmental surveys; Road traffic noise; Air pollution; Annoyance; Traffic measures;

Oslo traffic study – part 2: by Jocelyne Clench-Aas; Alena Bartonova; Ronny Klæboe; Marika Kolbenstvedt (4737-4744).
In quantifying the benefits of air pollution reduction measures, it is desirable to compare the size of the benefits with the effects of other individual confounding factors such as smoking or passive smoking. The effect of pollution is rarely very large and in order to quantify it, exposure estimating procedures must be as accurate as possible. Dispersion models, run for hourly time intervals and controlled by measurements, are therefore used to provide estimates for specific receptor points. Results of three consecutive cross-sectional investigations in an area of Oslo characterized by heavy traffic are presented. The study was designed to provide repeated information on the effects of traffic diversion measures on the self-reporting of symptoms of reduced health of 1100 adults living in Oslo. The principal source of air pollution in Oslo is vehicular traffic. The primary pollutants of interest are nitrogen dioxide (NO2) and respirable particles (PM2.5 and PM10). The mean hourly concentration of exposure was estimated at each participant's home by means of a time-dependent finite dispersion model combined with subgrid models to describe the source contribution to the grid concentrations. The study controlled the confounding factors. Using the symptom fatigue, the study illustrates that by controlling the changes in population composition, estimated exposure–effect relationships for health symptoms allow the effect of the studied traffic measures on the population to be evaluated. Since the method is based on individual estimates of exposure to different pollutants, it allows standardizing the exposure to compare effects of different pollutants. The study offers a methodology that is useful in evaluating the benefits of measures by both being able to quantify and compare the effects of different compounds and effects on different population sub-groups.
Keywords: Environmental health impact assessment; Fatigue; Dispersion modeling; Exposure–effect relationship; General health; Noise; Particles and NO2;