Search

For the simultaneous analysis of short- and long-term effects of air pollution in the Heinz Nixdorf Recall Cohort a sophisticated exposure modelling was performed. The dispersion and chemistry transport model EURAD (European Air Pollution Dispersion) was used for the estimation of hourly concentrations of a number of pollutants for a horizontal grid with a cell size of 1 km² covering the whole study area (three large adjacent cities in a highly urbanized region in Western Germany) for the years 2000–2003 and 2006–2008. For each 1 km² cell we estimated the mean concentration by calculating daily means from the hourly concentrations modelled by the EURAD process. The modelled concentrations showed an overall tendency to decrease from 2001 to 2008 whereas the trend in the single grid cells and study period was inhomogeneous. Participant-related exposure slightly increased from 2001 to 2003 followed by a decrease from 2006 to 2008. The exposure modelling enables a very flexible exposure assessment compared to conventional modelling approaches which either use central monitoring or temporally static spatial contrasts. The modelling allows the calculation of an average exposure concentration for any place and time within the study region and study period with a high spatial and temporal resolution. This is important for the assessment of short-, medium and long-term effects of air pollution on human health in epidemiological studies.

A semi-empirical mathematical model, Urban Street Model (USM), is proposed to efficiently estimate the dispersion of vehicular air pollution in cities. This model describes urban building arrangements by combining building density, building heights and the permeability of building arrangements relative to wind flow. To estimate the level of air pollution in the city of Krasnoyarsk (in Eastern Siberia), the spatial distribution of pollutant concentrations off roadways is calculated using Markov's processes in USM. The USM-predicted numerical results were compared with field measurements and with results obtained from other frequently used models, CALINE-4 and OSPM. USM consistently yielded the best results. OSPM usually overestimated pollutant concentration values. CALINE-4 consistently underestimated these values. For OSPM, the maximum differences were 160% and for CALINE-4 about 400%. Permeability and building density are necessary parameters for accurately modeling urban air pollution and influencing regulatory requirements for building planning.

This paper reports on a preliminary study of the forecast and evaluation of transport-related air pollution dispersion in urban areas. Some modifications of the traditional Gauss dispersion models are provided, and especially a crossroad model is built, which considers the great variation of vehicle emission attributed to different driving patterns at the crossroad. The above models are combined with a self-developed geographic information system (GIS) platform, and a simulative system with graphical interfaces is built. The system aims at visually describing the influences on the urban environment by urban traffic characteristics and therefore gives a reference to the improvement of urban air quality. Due to the introduction of a self-developed GIS platform and a creative crossroad model, the system is more effective, flexible and accurate. Finally, a comparison of the simulated (predicted) and observed hourly concentration is given, which indicates a good simulation.

The model construction is described as well as the preparation of emission and meteorological data for the test period. For this period SO2-concentration statistics were calculated by the model at the positions of four downtown measuring sites. The best agreement between observed and predicted concentrations was found if the stability over the city was always taken as independent of the surrounding rural area and corresponding to an unstable Pasquill-Turner class. Some residual scatter, probably due to measurement and source estimate inaccuracies remains, but the agreement is sufficiently good to justify further study.

A major application of air pollution dispersion models is to aid in the selection of abatement plans. Present technology dictates that air pollution abatement be a result of altering an area's emission characteristics. The need for pollution abatement is most acute during air pollution episodes. This study is an investigation of how air quality in a hypothetical city is affected by the city's emission characteristics during a simulated air pollution episode. Emission characteristics examined include configuration of sources, diurnal emission rate schedules, total emission rates, and the combination of point and area sources. Results are used to gain insight into how models may be used to reduce air pollution emergencies through long term planning and short term abatement schemes. Among the observations are: 1. (1) air quality is improved significantly by selecting configurations whose major sources are too far apart to jointly contaminate a locality. Increasing the mean distance between major sources 75 per cent led to an improvement in air quality as high as 28 per cent; 2. (2) altering diurnal schedules of emission rates to take advantage of diurnal variations in atmospheric dilutive capacity improved air quality by as much as 23 per cent; 3. (3) the set of priorities given to emission rate reductions of different types of sources is an important determinant of air quality improvement attained with a given reduction in the total emission rate. Differences as high as 28 per cent were observed for a given reduction in the total emission rate; 4. (4) pollutant concentration at a receptor location is primarily the result of emissions in contiguous locations; 5. (5) longer time lags between emission rate reductions and air quality improvement during episodes place a premium on controls which can be implemented quickly. The model is verified using St. Louis emission data. Predicted concentrations are within a factor of 1.4 of observed ones.