ADMS-Urban has been used in approximately half of the pilot studies of air pollution recently carried out in the UK. The pilot areas varied in spatial extent and in the nature of the sources, the balance between traffic, industrial and domestic sources, and the pollutants which are important. The studies used local emission inventories, and outputs included time-series of predicted concentrations which could be compared with measured data at monitoring sites, as well as contour concentration plots across the study area.

This paper describes the use of urban emission inventory data and an urban scale dispersion model (ADMS-Urban) to calculate concentrations of NOx and NO2 in London. The dispersion model used in this study is a second generation Gaussian dispersion model which is characterised by the use of boundary layer similarity profiles to parameterise the variation of turbulence with height within the boundary layer. The model has an integral chemistry model which is used to predict concentrations of nitrogen dioxide and ozone. The paper examines the performance of the integrated modelling system to predict concentrations from emission sources in Greater London. Predictions have been compared with observed data at four locations, two locations in Central London and two in East London. Predicted concentrations for a summer and winter period have been calculated and modelled and measured times series data have been compared. Statistical analyses have been carried out to assist in the comparison of model predictions with monitored data. Although no absolute significance can be attached to the numerical values of these measures, taken cumulatively, some conclusions regarding the emissions inventory data and the models’ performance can be made. The applicability of the model of atmospheric chemistry used in this study to calculate the concentrations of nitrogen dioxide is appraised and a comparison has been made with concentrations of nitrogen dioxide predicted from oxides of nitrogen using empirical relationships.

Ammonia is a reactive pollutant emitted primarily by agricultural sources near ground level in the rural environment. The consequence of these factors is that, in addition to the effects of long-range pollutant transport, ammonia has major effects at a local scale, with emission and receptor areas often closely located in the rural landscape. There is a substantial local spatial variability that needs to be considered in effects assessments, while variations in local deposition may affect the amount of ammonia available for impacts further afield. The wide-ranging UK programme ADEPT (Ammonia Distribution and Effects ProjecT) has addressed these issues through a combination of measurement and modelling activities concerning the distribution of emissions, atmospheric transport, deposition and effects assessment. The results are illustrated here by summarizing the findings of a joint experiment at Burrington Moor, Devon, and wider modelling contrasting the variability at a field scale with 5 km resolution estimates for the UK. The fraction of emitted NH3 deposited locally is shown to depend critically on the downwind land-cover, with fluxes being dependent on interactions with the ammonia compensation point. This will restrict deposition back to agricultural land, but may mean that non-conservation woodlands could be of benefit to recapture a significant fraction of emissions. The generalized models demonstrate the high spatial variability of ammonia impacts, with a case study being used to show the consequences at a field scale. In source regions substantial variability occurs at sub-1 km levels and this will have major consequences for the emission reduction targets needed to protect ecosystems.

Under the (UK) Environment Act 1995, each local authority in the UK has a duty to review and assess air quality. Modelling is one of the tools available to Environmental Health Officers to assist in this process. Many of the models available are, however, difficult to use and often do not reflect the current understanding of dispersion within the atmospheric boundary layer. ADMS Urban is an example of a new generation of Windows based systems which aims to bring the latest modelling techniques to the non-specialist user. At the heart of the system is an extensively validated, receptor specific model (ADMS) for calculating concentration from point, line, area and volume sources. This is enhanced by an integrated street canyon model derived from the Danish model OSPM and a simplified model of atmospheric chemistry. ADMS is characterised by the use of boundary layer similarity profiles to parameterise the variation of turbulence with height within the boundary layer and the use of a skewed-Gaussian concentration profile, which can bring material from elevated sources rapidly down to the surface, as observed in the field. The model is integrated with an emissions inventory relational database framework within a GIS system, in this case ESRI's ArcView. The use of a commercially available GIS system significantly enhances the possibilities for display and analysis of model predictions and, if part of a networked system enables air quality data to be shared between different departments within the local authority. ADMS Urban is now being used in cities throughout the UK, including Oxford, Cambridge, Bristol, Birmingham, Cardiff and Newcastle.