The spatial representativeness of air quality monitoring stations is a crucial parameter when the observed concentration levels are used in an air quality assessment. Spatial representativeness defines to what extent the monitoring data is meaningful and useful in a spatial context. Within this paper a generic and robust methodology is presented for the assessment of the spatial representativeness of air pollution monitoring sites. The methodology relies on a statistical approach that links annual averaged concentration levels with land use characteristics. The methodology is demonstrated for the monitoring sites in the Belgian telemetric air quality network and then applied to define a set of zones with a given confidence level. Within such a zone the concentrations deviate to a maximum percentage from the measured values at the monitoring sites. Furthermore, the relevance of spatial representativeness for model validation is addressed and the technique is illustrated for the validation of the results of the regional air quality model BelEUROS. In general, the overall improvement of the model validation by taking into account spatial representativeness can be quantified as in the order of 20%.

Despite the scarce effect of speed limit reduction measures on total mass PM10 and PM2.5 concentrations, it is shown that the effect on the probably more hazardous component elemental carbon (EC) is more important which means that, from the viewpoint of health benefit, speed reductions during smog episodes may well be justified. Especially in the very dense highway network in Flanders with a 60% share of diesel cars (the highest in Europe) a speed limit reduction from 120 to 90km/h during winter smog episodes on selected sections of Flemish highways leads to a significant decrease of the EC concentrations near those highways. Key findings are that the decrease in EC depends on the distance from the highways. In the direct vicinity of the highways, a decrease compared to the base-line scenario where no speed limit changes were implemented of up to 30% of the EC concentrations is modeled. Within a distance of 1500m of the concerned highway sections there is an average decrease in EC of 0.18μgm−3 affecting about 1 million people living in these areas. When the speed limit reduction measure is in force, the EC exposure of about 300,000 people decreases by at least 5% and 7500 people experience a decrease of 15% or more.

Real-time assessment of the ambient air quality has gained an increased interest in recent years. To give support to this evolution, the statistical air pollution interpolation model RIO is developed. Due to the very low computational cost, this interpolation model is an efficient tool for an environment agency when performing real-time air quality assessment. Beside this, a reliable interpolation model can be used to produce analysed maps of historical data records as well. Such maps are essential for correctly checking compliance with population exposure limit values as foreseen by the new EU Air Quality Directive. RIO is an interpolation model that can be classified as a detrended Kriging model. In a first step, the local character of the air pollution sampling values is removed in a detrending procedure. Subsequently, the site-independent data is interpolated by an Ordinary Kriging scheme. Finally, in a re-trending step, a local bias is added to the Kriging interpolation results. As spatially resolved driving force in the detrending process, a land use indicator is developed based on the CORINE land cover data set. The indicator is optimized independently for the three pollutants O3, NO2 and PM10. As a result, the RIO model is able to account for the local character of the air pollution phenomenon at locations where no monitoring stations are available. Through a cross-validation procedure the superiority of the RIO model over standard interpolation techniques, such as the Ordinary Kriging is demonstrated. Air quality maps are presented for the three pollutants mentioned and compared to maps based on standard interpolation techniques.

A methodology is presented for the assessment of the spatial representtativeness of air pollution monitoring data. The methodology relies on a statistical approach that links air quality expectation values with land use characteristics. The relevance of this issue for model validation is addressed and the technique is illustrated for the validation of BelEUROS model results.

Real-time assessment of the ambient air quality has gained an increased interest in recent years. To give support to this evolution, the statistical air pollution interpolation model RIO is developed. Due to the very low computational cost this interpolation model is an efficient tool for an environment agency when performing real-time air quality assessments. Beside this, a reliable interpolation model can be used to produce analysed maps of historical data records as well. RIO is an interpolation model that can be classified as a detrended Kriging model. In a first step the local character of the air pollution sampling values is removed in a detrending procedure. Subsequently, the site-independent data is interpolated by an Ordinary Kriging scheme. Finally, in a retrending step a local bias is added to the Kriging interpolation results. As spatially resolved driving force in the detrending process, a land use indicator is developed based on the CORINE land cover data set. The indicator is optimized independently for the three pollutants O3, NO2 and PM10. As a result, the RIO model is able to account for the local character of the air pollution phenomenon at locations where no monitoring stations are available. Through a cross-validation procedure the superiority of the RIO model over standard interpolation techniques, such as the Ordinary Kriging is demonstrated. Air quality maps are presented for the three pollutants mentioned and compared to maps based on standard interpolation techniques.

Aerosol Optical Depth (AOD) data from the MERIS sensor is used in an operational air quality forecast model for Belgium. The spatial pattern of AOD is combined with PM10 in-situ measurements or forecast values at point locations. The RIO interpolation tool is used as a data fusion process to produces maps of PM 10 for the Belgian territory.

Seasonal changes in aerosol compositions over Belgium and Europe are simulated with an extended version of the EUROS model. EUROS is capable of modelling mass and chemical composition of aerosols in two size fractions (PM2.5 and PM10-2.2). The chemical composition is expressed in terms of seven components: ammonium, nitrate, sulphate, primary inorganic compounds, elementary carbon, primary organic compounds and Secondary Organic Compounds (SOA). A comparison of modelled and measured aerosol concentrations showed that modelled concentrations are generally consistent with observed concentrations. The chemical composition of the aerosol showed a strong dependence on the season. High aerosol concentrations during the summer were mainly due to high concentrations of the secondary components nitrate, ammonium, sulphate and SOA in the size fraction PM2.5. In contrast, during autumn and winter, increased PM-concentrations were mainly due to higher concentrations of primary components, especially in the size fraction PM10-2.5.

The Eulerian Chemistry-Transport Model BelEUROS was used to calculate the concentrations of airborne PM10 and PM2.5 over Europe. Both primary as well as secondary particulate matter in the respirable size-range was taken into account. Especially PM2.5 aerosols are often formed in the atmosphere from gaseous precursor compounds. Comprehensive computer codes for the calculation of gas phase chemical reactions and thermodynamic equilibria between compounds in the gas-phase and the particulate phase had been implemented into the BelEUROS-model. Calculated concentrations of PM10 and PM2.5 are compared to observations, including both the spatial and daily, temporal distribution of particulate matter in Belgium for certain monitoring locations and periods. The concentrations of the secondary compounds ammonium, nitrate and sulfate have also been compared to observed values. BelEUROS was found to reproduce the observed concentrations rather well. The model was applied to assess the contribution of emissions derived from the sector agriculture in Flanders, the northern part of Belgium, to PM10- and PM2.5-concentrations. The results demonstrate the importance of ammonia emissions in the formation of secondary particulate matter. Hence, future European emission abatement policy should consider more the role of ammonia in the formation of secondary particles.

The European Operational Smog (EUROS) integrated air quality modelling system has been extended to model fine particulate matter (PM). From an extended literature study, the Caltech Atmospheric Chemistry Mechanism and the Model of Aerosol Dynamics, Reaction, Ionisation and Dissolution were selected and recently coupled to EUROS. Currently, modelling of mass and chemical composition of aerosols in two size fractions (PM2.5 and PM10–2.5) is possible. The chemical composition is expressed in terms of seven components: ammonium, nitrate, sulphate, elementary carbon, primary inorganic compounds, primary organic compounds and secondary organic compounds. Calculated PM10 concentrations and chemical composition are presented for two summer months of the year 2003 (1 July to 31 August).

An interpolation model, called RIO, is described. The interpolation scheme is based on the Kriging technique. RIO produces ozone estimates on a 5x5 km grid. Database is of ambient ozone concentrations that are systematically sampled by the three Belgian Regions at more than 30 sites.

Due to scientific interest on the one hand and political and regulatory obligations on the other hand the monitoring of ozone in the troposphere is an important issue. To this end, in Belgium as in many other countries, a fixed network of monitoring stations is operated. In order to estimate the ozone concentrations over the whole territory, a model is needed to spatially complement the sparse measurements. This paper describes the development of an interpolation scheme which is aimed at fast operational use. The model uses the population density as auxiliary data to remove a spatial trend due to titration by nitric oxide. The residuals are interpolated by kriging. As a benchmark the inverse distance weighting interpolation method is used with and without the detrending. The proposed model systematically improves the interpolation and makes a significant difference when estimating human exposure to ozone. It is generic in design, easy to implement and flexible to changes in the monitoring network.

Over the past years, the health impact of airborne particulate matter (PM) has become a very topical subject. In the environmental sciences a lot of research effort goes towards the understanding of the PM phenomenon and the ability to forecast ambient PM concentrations. In this paper, we describe the development of a neural network tool to forecast the daily average PM10 concentrations in Belgium one day ahead. This research is based upon measurements from ten monitoring sites during the period 1997–2001 and upon ECMWF simulations of meteorological parameters. The most important input variable found was the boundary layer height. A model based on this parameter currently operational online serves to monitor the daily average threshold of 100μgm−3. By extending the model with other input parameters we were able to increase the performance only slightly. This brings us to the conclusion that day-to-day fluctuations of PM10 concentrations in Belgian urban areas are to a large extent driven by meteorological conditions and to a lesser extend by changes in anthropogenic sources.

In this paper we summarize a survey that was made to determine which variables are most relevant as input data for short-term PM10 forecasting based on a neural network model. Since the health impact of airborne particulate matter is becoming a topic of increasing interest, this study was performed as a first step towards the design of an operational system that can inform the media when the PM10 concentration is expected to exceed a given level of concern. The research is based on ambient PM10 measurements from different monitoring sites in Belgium during the period 1997-2001. Besides these data, we used ECMWF-forecasts of meteorological parameters. This parameter set includes standard (ground-level) meteorological variables but also several variables related to turbulence in the boundary layer. To quantify the value of these parameters in a PM10 forecasting model, each of them is used as an input variable of an artificial neural network (feed-forward multi-layer perceptron). After the training, the networks are compared by cross-validation on an independent dataset. The main conclusion of this work is that for Belgium the most relevant meteorological parameter is the boundary layer height (corresponding to a critical bulk Richardson number of 0.5). When examining some emission related variables, we found that the impact of these parameters on the forecast was limited.

In this paper we summarize a survey that was made to determine which variables are most relevant as input data for short-term PM10 forecasting based on a neural network model. Since the health impact of airborne particulate matter is becoming a topic of increasing interest, this study was performed as a first step towards the design of an operational system that can inform the media when the PM10 concentration is expected to exceed a given level of concern. The research is based on ambient PM10 measurements from different monitoring sites in Belgium during the period 1997-2001. Besides these data, we used ECMWFforecasts of meteorological parameters. This parameter set includes standard (ground-level) meteorological variables but also several variables related to turbulence in the boundary layer. To quantify the value of these parameters in a PM10 forecasting model, each of them is used as an input variable of an artificial neural network (feed-forward multi-layer perceptron). After the training, the networks are compared by cross-validation on an independent dataset. The main conclusion of this work is that for Belgium the most relevant meteorological parameter is the boundary layer height (corresponding to a critical bulk Richardson number of 0.5). When examining some emission related variables, we found that the impact of these parameters on the forecast was limited.

Every summer, ground level ozone concentrations rise in Belgium and cause episodes of photochemical summer smog. This phenomenon is the cause of well recognised public health distress, especially for people suffering from respiratory diseases. To warn groups of sensitive people against forthcoming smog episodes, VITO (the Flemish Institute for Technological Research) and VMM (the Flemish Environmental Agency) have joined forces to create an ozone pollution forecasting model, called SMOGSTOP (Statistical Model Of Groundlevel Short Term Ozone Pollution). SMOGSTOP is used by Belgian government agencies to issue ozone reports and warnings in the media (TV, radio, internet, etc.). Ozone pollution levels, as well as concentrations of other air pollutants, are monitored in Belgium by the telemetric air quality measuring networks of the three Belgian regions: Flanders, Wallonia and Brussels. The major meteorological variables, such as the windvector, temperature, pressure, humidity and precipitation, are also monitored by the same networks. The historical time series of those variables, generated by the networks, are the source of input data for SMOGSTOP. Photochemical ozone pollution is the result of complex non-linear interactions between atmospheric pollutants and meteorology. As a consequence, it is extremely difficult to determine relationships between source emissions (ozone precursors) and ambient pollutant concentrations (ozone). To deal with this complexity, SMOGSTOP was constructed as an empirical model, applying a VITO tailor-made methodology called stratified pattern matching, to link meteorological and precursor information into ozone forecasts. In this paper, an overview of the process of ozone formation in Belgium is given, followed by the definition of the explanatory variables which will be used in the model. Then the methodology behind the model is reported and finally the results of the forecasting efforts in the period of 1.5.1995 till 31.8.1995 are presented. Copyright (C) 2000 John Wiley & Sons, Ltd.

Every summer, ground-level ozone concentrations rise in Belgium and cause episodes of photochemical summer smog. This phenomenon is the cause of well recognized public health distress, especially for people suffering from respiratory diseases. In order to warn groups of sensitive people of forthcoming smog episodes, VITO (the Flemish Institute for Technological Research) and VMM (the Flemish Environmental Agency) have joined forces to create an ozone pollution forecasting tool called SMOGSTOP (Statistical Model Of Ground-level Short Term Ozone Pollution). SMOGSTOP is used by Belgian government agencies to issue ozone reports and warnings in the media (TV, radio, internet, etc). In this paper, an overview of the process of ozone formation in Belgium is given, followed by the definition of the explanatory variables which will be used in the approach. Then the methodology behind the approach is reported and finally the results of the forecasting efforts in the period from 1 May to 31 August 1995 are presented. The 1995 summer was selected because of a relatively large number of ozone days (29) compared to other summers in Belgium.

A systematic study has shown that under normal flight conditions and with a forward facing sample air intake, the measurements of atmospheric pollutants (ozone, NOx, SO2, dust) on board helicopters are not subject to contamination by the turbine exhaust.

A simple mathematical relationship has been established between the ambient SO2 concentration and some meteorological parameters (wind speed and direction, temperature). The application of this relationship permits the quantification of the influence of the buildings heating and the long-range transport on the SO2 levels measured.

A simple mathematical relationship has been established between the ambient SO2 concentration and some meteorological parameters (wind speed and direction, temperature). The application of this relationship permits the quantification of the influence of the buildings heating and the long-range transport on the SO2 levels measured.