Summary

 This contribution is cooperation of the Finnish Meteorological Institute (FMI) and the Helsinki Metropolitan Area Council (YTV).

A modelling system was developed for evaluating the traffic volumes, emissions from stationary and vehicular sources, and atmospheric dispersion of pollution in an urban area. The modelling system was refined to allow for the chemical interaction of pollutants, originating from a large number of urban sources. We have investigated the factors leading to air quality episodes and the various possibilities to control these events.

A measurement campaign was conducted in a street canyon in Helsinki during 1997. The hourly concentrations of CO, NO_X, NO₂ and O₃ were measured on the street and roof levels. The experimental results were compared with the results produced by the OSPM model. The overall agreement of measured and predicted concentrations was good for CO and NO_X.

We have also developed a preliminary model in order to evaluate the exposure of population to air pollution. Using a graphical information system, the modelled NO₂ concentrations have been combined with information concerning activities of population, land use and other data. The computed results illustrate the temporal and spatial variation of population exposure to NO₂.

Aim of research

The objective is to improve our ability of establishing source-receptor relationships at the urban scale, both for gaseous pollutants and particulate matter. We have refined atmospheric dispersion models in order to allow for the chemistry of nitrogen oxides and the dispersion of particles. Field measurement campaigns are conducted in order to evaluate and validate dispersion modelling systems. We also aim to model exposure to air pollution, in order to quantify the adverse health effects of air pollution.
 

Activities during the year

We have developed a modelling system for evaluating the traffic volumes, emissions from stationary and vehicular sources, and atmospheric dispersion of pollution in an urban area; containing the UDM-FMI (Karppinen et al., 1998a) and CAR-FMI (Harkonen et al., 1997) models. The modelling system was refined to allow for the chemical interaction of pollutants, originating from a large number of urban sources (MOD 4). The predicted NO_x and NO₂ concentrations were compared with the results of the urban air quality monitoring network of YTV. The argeement of predicted and measured NO₂ concentrations was good at all the stations considered (Karppinen et al., 1997 and 1998 b,c).

We have conducted a measurement campaign in a street canyon in Helsinki during 1997, in cooperation with YTV and National Environmental Research Institute (Denmark). The hourly concentrations of CO, NO_X, NO₂ and O₃ were measured on the street and roof levels. The relevant meteorological parameters (wind direction and speed, temperature and solar radiation) were measured on the roof level. We applied on-site electronic traffic counts. The experimental results were compared with the results produced by the OSPM model (Hertel and Berkowicz, 1989). The overall agreement of measured and predicted concentrations was good for CO and NO_X; however, the model slightly overestimated the measured concentrations of NO₂ (EXP 1).

We have also developed models in order to evaluate the exposure of population to air pollution, particularly for nitrogen oxides. We have utilised emission inventories collected previously, and the above mentioned atmospheric dispersion modelling system. This work was cooperation with YTV and National Public Health Institute (KTL). The computed results illustrate the temporal and spatial variation of population exposure to NO₂ (INT 4).

Principal results

The dispersion models CAR-FMI and UDM-FMI allow for the chemical transformation of nitrogen oxides. However, the plumes originating from various sources also interact chemically with each other and with the background concentrations in the urban area. For instance, it is not uncommon that urban background O₃ concentration vanishes, caused by the depletion of O₃ in the oxidation of NO into NO₂. However, regulatory modelling systems commonly assume for simplicity that the urban background O₃ concentration is equal to the regional O₃ background concentration. We have developed a modelling system, which allows for the chemical interdependence of the NO_x concentrations originating from various sources and the O₃ concentrations (Karppinen et al., 1997 and 1998 b,c) (MOD 4).
 
 

Figure 1. The concentrations of nitrogen oxides (NO₂), carbon monoxide (CO) and thoracic particles (PM₁₀) during a severe air quality episode in December 1995, measured at a station in downtown Helsinki. Source: Helsinki Metropolitan Area Council.
 
 

Work is in progress for extending the urban dispersion modelling system to allow for particle dispersion. We have developed a simple semi-empirical model for evaluating the concentrations of urban fine particles (MOD 4). The modelling system is based on a particle emission model, urban dispersion models and data from the air quality monitoring network of YTV.

Severe air pollution episodes occur in Finnish cities every two to five years; these can last for a few days. The authorities need to be prepared for these events and the public has to be informed on the situation. The most important pollutants in the episode situations are nitrogen dioxide and particles. During an episode these pollutants at the ground level in cities are mainly originated from road traffic. We have investigated the occurrence of episodes nationally (Kukkonen et al., 1998a), meteorological factors leading to episodes (Kukkonen et al., 1998b) and their forecasting, the exhaust emissions from road traffic and the conceivable measures for controlling traffic during an episode (Makela et al., 1997 and 1998). This work was cooperation with Technical Research Centre (VTT).

A very severe episode prevailed in Finland in December 1995. For instance, the measured NO₂ concentrations in Helsinki increased to a substantially high level, even compared internationally, as shown in Figure 1. Figure 2 presents the corresponding predicted spatial concentration distribution of NO₂.
 

Main conclusions

The comparison of measured results in a street canyon in Helsinki with the predictions of the OSPM model provide more evidence for the credibility of the OSPM model. The experimental street canyon database is available also for testing of other street canyon dispersion models.

We have investigated the occurrence of episodes, meteorological factors leading to episodes and conceivable measures for controlling traffic during an episode. For instance, according to dispersion model computations, a feasible NOx emission reduction by 19 % would lower the ambient NO2 concentrations by 0-15 %, in various urban areas (Makela et al., 1998).

A model was developed for evaluating the exposure of population to air pollution, particularly for nitrogen oxides. The results computed for the greater Helsinki area illustrate the temporal and spatial variation of population exposure to NO2.

Aim for the coming year
 

Work is in progress for extending the modelling system to allow for particle dispersion. The main objective for the coming year is to review available literature on aerosol modelling, develop models for aerosol processes in urban environments, extend the existing dispersion modelling system accordingly, and compare particle model predictions with experimental results.
 

Acknowledgements

 Financial support from the Academy of Finland is gratefully acknowledged. The work described here concerning particle dispersion and exposure modelling are cooperation with the EXPAND project within the Finnish Research Programme on Environmental Health (http://www.ktl.fi/sytty) and the EXPOLIS project (http://www.ktl.fi/expolis).
 
 

Figure 2. Predicted spatial concentration distribution of NO₂ in the Helsinki metropolitan area during an air pollution episode, 28 December 1995 at 16.00 (m g/m³).
 

References