Abstract
The vertical diffusion of a passive tracer released from surface and elevated sources in a neutrally stratified boundary layer has been studied by comparing field and laboratory experiments with a non-Gaussian K-theory model that assumes power-law profiles for the mean velocity and vertical eddy diffusivity. Several important differences between model predictions and experimental data were discovered: 1) the model overestimated ground-level concentrations from surface and elevated releases at distances beyond the peak concentration; 2) the model overpredicted vertical mixing near elevated sources, especially in the upward direction; 3) the model-predicted exponent α in the exponential vertical concentration profile for a surface release [C̄(z) exp(−zα)] was smaller than the experimentally measured exponent. Model closure assumptions and experimental shortcomings are discussed in relation to their probable effect on model predictions and experimental measurements.
