Abstract. A mathematical model is presented for the determination of human exposure to ambient air pollution in an urban area; the model is a refined version of a previously developed mathematical model EXPAND (EXposure model for Particulate matter And Nitrogen oxiDes). The model combines predicted concentrations, information on people's activities and location of the population to evaluate the spatial and temporal variation of average exposure of the urban population to ambient air pollution in different microenvironments. The revisions of the modelling system containing the EXPAND model include improvements of the associated urban emission and dispersion modelling system, an improved treatment of the time use of population, and better treatment for the infiltration coefficients from outdoor to indoor air. The revised model version can also be used for estimating intake fractions for various pollutants, source categories and population subgroups. We present numerical results on annual spatial concentration, time activity and population exposures to PM 2.5 in the Helsinki Metropolitan Area and Helsinki for 2008 and 2009, respectively. Approximately 60 % of the total exposure occurred at home, 17 % at work, 4 % in traffic and 19 % in other microenvironments in the Helsinki Metropolitan Area. The population exposure originating from the long-range transported background concentrations was responsible for a major fraction, 86 %, of the total exposure in Helsinki. The largest local contributors were vehicular emissions (12 %) and shipping (2 %).

LUR and dispersion model estimates correlated on average well for NO2 but only moderately for PM10 and PM2.5, with large variability across areas. DM predicted a moderate to large proportion of the measured variation for NO2 but less for PM10 and PM2.5.

Abstract. We present an overview of the modelling of particle number concentrations (PNCs) in five major European cities, namely Helsinki, Oslo, London, Rotterdam, and Athens, in 2008. Novel emission inventories of particle numbers have been compiled both on urban and European scales. We used atmospheric dispersion modelling for PNCs in the five target cities and on a European scale, and evaluated the predicted results against available measured concentrations. In all the target cities, the concentrations of particle numbers (PNs) were mostly influenced by the emissions originating from local vehicular traffic. The influence of shipping and harbours was also significant for Helsinki, Oslo, Rotterdam, and Athens, but not for London. The influence of the aviation emissions in Athens was also notable. The regional background concentrations were clearly lower than the contributions originating from urban sources in Helsinki, Oslo, and Athens. The regional background was also lower than urban contributions in traffic environments in London, but higher or approximately equal to urban contributions in Rotterdam. It was numerically evaluated that the influence of coagulation and dry deposition on the predicted PNCs was substantial for the urban background in Oslo. The predicted and measured annual average PNCs in four cities agreed within approximately ≤ 26 % (measured as fractional biases), except for one traffic station in London. This study indicates that it is feasible to model PNCs in major cities within a reasonable accuracy, although major challenges remain in the evaluation of both the emissions and atmospheric transformation of PNCs.

This study assesses population exposure caused by the emissions of primary fine particulate matter (PM 2.5 ) originated from road traffic and domestic wood combustion in Finland in 2000 and 2020. The evaluations were performed using source-receptor matrices (SRMs) based on the computations using a local and a regional scale atmospheric dispersion model, on two different grid resolutions: 1 and 10 km. Road traffic and domestic wood combustion are nationally the most important emission source categories of primary PM 2.5 ; they were projected to contribute to 42% of the Finnish total emissions in 2020. Although traffic exhaust emissions were projected to decrease considerably in the future, by 91% from 2000 to 2020, non-exhaust emissions were predicted to increase. Traffic emissions were found to cause on the average considerably higher population-weighted concentration (PWC) to primary PM 2.5 , compared with domestic wood combustion emissions. Based on the computation with 1-km resolution SRMs, the exhaust and non-exhaust traffic emissions were projected to cause 5.5% and 62% of the PWC, respectively, of the total combined PWC caused by traffic and domestic combustion in Finland in 2020. Regarding the sub-categories of domestic wood combustion, supplementary wood heating was found to cause relatively high PWC, 22% in 2020. The modeling of traffic emissions and dispersion using the regional scale model on a resolution of 10 km resulted in PWC that is more than an order of magnitude smaller, compared with the corresponding computations using a local scale model on a resolution of 1 km. The general implication of this study is that the PWC values evaluated using integrated assessment models can be sensitive to the methodology, especially these can substantially increase with an increasing spatial resolution.

Abstract. Reliable and self-consistent data on air quality are needed for an extensive period of time for conducting long-term, or even lifetime health impact assessments. We have modelled the urban-scale concentrations of fine particulate matter (PM2.5) in the Helsinki Metropolitan Area for a period of 35 years, from 1980 to 2014. The regional background concentrations were evaluated based on reanalyses of the atmospheric composition on global and European scales, using the SILAM model. The high-resolution urban computations included both the emissions originated from vehicular traffic (separately exhaust and suspension emissions) and those from small-scale combustion, and were conducted using the road network dispersion model CAR-FMI and the multiple-source Gaussian dispersion model UDM-FMI. The modelled concentrations of PM2.5 agreed fairly well with the measured data at a regional background station and at four urban measurement stations, during 1999–2014. The modelled concentration trends were also evaluated for earlier years, until 1988, using proxy analyses. There was no systematic deterioration of the agreement of predictions and data for earlier years (the 1980s and 1990s), compared with the results for more recent years (2000s and early 2010s). The local vehicular emissions were about 5 times higher in the 1980s, compared with the emissions during the latest considered years. The local small-scale combustion emissions increased slightly over time. The highest urban concentrations of PM2.5 occurred in the 1980s; these have since decreased to about to a half of the highest values. In general, regional background was the largest contribution in this area. Vehicular exhaust has been the most important local source, but the relative shares of both small-scale combustion and vehicular non-exhaust emissions have increased in time. The study has provided long-term, high-resolution concentration databases on regional and urban scales that can be used for the assessment of health effects associated with air pollution.