Application of diffusion models in the design of monitoring surveys for air quality assessment
The purpose of this investigation was to develop a methodology for designing air pollution monitoring surveys using diffusion models and sampling statistics to provide cost effective surveys which meet specific sampling objectives. A survey was designed for validating highway line source dispersion models and conducted in Nashville, Tennessee. The survey included real-time measurements of ambient concentrations of carbon monoxide, wind speed and direction, traffic volumes, vehicle speeds, and truck mix. Correlation and regression analyses of measured pollution levels compared to those predicted by the models showed that the models tended to overpredict concentrations under parallel wind conditions and underestimate concentrations for perpendicular and oblique wind conditions. Two empirical models were developed which were independent of atmospheric stability conditions. These models show promise as predictive tools, but require testing at other sites.
Thesis75b.M544.pdf
7.63 MB
Adobe PDF
69663bb78147b7ea91656800ee18b1fb