Modelling concentration fluctuations in a turbulent boundary layer: sensitivity of a Lagrangian model to dissipation rate estimates

Lagrangian dispersion models require estimates of local production and dissipation rate of turbulent kinetic energy (t.k.e). Both quantities are hardly measurable in atmospheric flows. In laboratory experiments, although the t.k.e. production measurements can be easily achieved, the direct measurement of the dissipation ratio still represents a hard task. Therefore, this quantity is generally estimated indirectly, as a residual of the turbulent kinetic energy budget under a series of simplifying assumptions. In this study we evaluate the sensitivity of a lagrangian model to different estimates of t.k.e. dissipation rate in simulating passive scalar dispersion in a turbulent boundary layer over a rough surface. The estimates of t.k.e. dissipation are obtained by means of hot-wire anemometry velocity measurements in wind tunnel experiments. Two different estimates of epsilon are calculated. These values are used to simulate pollutant dispersion emitted by a linear elevated source with a lagrangian model which integrates a macromixing and a micromixing scheme. Comparison between numerical and experimental results allow us to discuss the performance of the model in predicting mean and fluctuating concentration and to define its sensitivity to the differences in the estimates of t.k.e. dissipation.