Improving air quality dispersion models using lidar measurements

Atmospheric boundary layer height is one of the most important parameters in atmospheric dispersion modelling because it has a large effect on predicted air quality. Comparisons between Atmospheric Dispersion Modelling System, version 4 (ADMS 4) and lidar data were carried out on boundary layer height data from central London. The comparison showed that the boundary layer height predicted by the ADMS 4 was, on average, lower than lidar for the subset of data taken. ADMS 4 has a very simple surface scheme which is not representative of complex urban environments and the results from this research imply that there is not sufficient surface roughness within the model to produce a large enough boundary layer height. This research aims to create an improved urban surface scheme to better forecast the growth of the daytime urban boundary layer in air quality dispersion models using lidar measurements. Improved performance of ADMS in an urban area was achieved by a number of steps carried out in this study. They were: establishing the weaknesses of ADMS, developing a combined model for atmospheric boundary layer height, sensitivity analysis and validation. The combined model was developed by combining a surface model and the atmospheric boundary layer height model. Measurements of the atmospheric boundary layer height by lidar used backscatter and vertical velocity variance. However, vertical velocity variance was chosen for comparison with the combined model. Statistically, the combined model was better than ADMS. Sensitivity analysis, which used reasonable ranges of meteorological data for the UK, encompassed surface roughness, friction velocity, sensible heat flux and potential temperature gradient above the atmospheric boundary layer and the atmospheric boundary layer height. A good agreement was observed between the combined model and lidar measurement using data collected at a site in Greater Manchester. The overall conclusion was that the combined model improved the performance of ADMS in urban areas