Validation of the ADMS dispersion model and assessment of its performance relative to R–91 and ISC using archived LIDAR data. (4th August 2014)

• Record Type: Journal Article
• Authors: Carruthers, D.J.; Edmunds, H.A.; Bennett, M.; Woods, P.T.; Milton, M.J.T.; Robinson, R.; Underwood, B.Y.; Franklin, C.J.; Timmis, R.
• Is Part Of: International journal of environment and pollution. Volume 8:Number 3-6(1997)
• Page Start: 264
• Page End: 278
• Publication Date: 2014-08-04
• Subjects: ADMS -- ISC -- LIDAR data -- R–91 -- atmospheric dispersion models -- air pollution -- environmental pollution -- modelling -- pollutant concentrations -- plume heights -- plume spread; Environmental policy -- Periodicals; Environmental engineering -- Periodicals; Environmental sciences -- Periodicals; Pollution -- Periodicals; 363.73
• Languages: English
• ISSNs: 0957-4352

Abstract

Regulatory authorities have traditionally used Gaussian dispersion models, notably R–91 in the UK and ISC in the USA. These use dispersion coefficients derived from the Pasquill stability category calculated for the prevailing meteorological conditions. Recently, ADMS (Atmospheric Dispersion Modelling System) has been developed; this is one of a new generation of dispersion models in which the structure of the boundary layer and hence dispersion is described as a function of boundary layer height and Monin–Obukhov length. This model has been described in previous workshops. A number of comparisons of ADMS against experimental data of ground–level concentration have already given encouraging results. In this paper we present a validation of ADMS and an assessment of its performance relative to R–91 and ISC against five sets of archived LIDAR data. The LIDAR data have allowed the first fully three–dimensional comparisons between data and model predictions, and include concentrations, plume heights and plume spread. The comparisons are presented statistically by means of standard statistical measures and by x–y plots, scatter diagrams, frequency histograms, and box and whisker plots. In addition the concentration fluctuation model of ADMS is used to estimate the expected scatter of measured data about the predicted ensemble mean concentration.