Comparing Two Implementations of a Micromixing Model. Part II: Canopy Flow

Abstract

The sequential particle micromixing model (SPMMM) is used to estimate concentration fluctuations in plumes dispersing into a canopy flow. SPMMM uses the familiar single-particle Lagrangian stochastic (LS) trajectory framework to pre-calculate the required conditional mean concentrations, which are then used by an interaction by exchange with the conditional mean (IECM) micromixing model to predict the higher-order fluctuations of the scalar concentration field. The predictions are compared with experimental wind-tunnel dispersion data for a neutrally stratified canopy flow, and with a previously reported implementation using simultaneous particle trajectories. The two implementations of the LS-IECM model are shown to be largely consistent with one another and are able to simulate dispersion in a canopy flow with fair to good accuracy.

A stochastic Lagrangian micromixing model for the dispersion of reactive scalars in turbulent flows: role of concentration fluctuations and improvements to the conserved scalar theory under non-homogeneous conditions

“…On the other hand, the Lagrangian micromixing models are still poorly investigated, especially in 3D ( [53]. [54], [40]), also due to a general lack of open-field experimental data on concentration fluctuations ( [37], [38], [67], [20]). These models need Eulerian velocity statistics of the main flow in input, as resulting from another code or a diagnostic tool.…”

Concentration Fluctuations from Localized Atmospheric Releases

“…This approach permits the use of simple algorithms amenable of trivial parallelization where each particle is independently advanced (Luhar and Sawford 2005 ; Postma et al. 2011a , b ). This simplified approach does not allow the use of mixing algorithm based on particle interaction nor the inclusion of chemical reactions.…”

History of Lagrangian Stochastic Models for Turbulent Dispersion

“…The unexpected finding was that this model performed no better in regard to simulations of the mean tracer distribution than did models based on a simplification to Gaussian pdf's (namely, the model provided by Thomson [1987] or the alternative closure of Flesch and Wilson). Most subsequent Lagrangian simulations of canopy transport [e.g., Poggi et al, 2006;Cassiani et al, 2007;Postma et al, 2011] have adopted the approximation of Gaussian pdf's. In general, in such acutely inhomogeneous flows, it appears the effects of velocity skewness are secondary, except very close to the source(s).…”