Lagrangian Stochastic Modeling of Heavy Particle Trajectories in Atmospheric Turbulence

Particle transport in atmospheric boundary layer turbulence is simulated using Lagrangian Stochastic Models (LSM) coupled with a Large Eddy Simulation (LES) model of atmospheric boundary layer flow. The aim of this work is to improve the accuracy of transport modeling of various natural tracers such as snow flakes and pollen and anthropogenic tracers such as reactive and non reactive pollutants. Lagrangian Stochastic Models are based on the assumption that the pair of variables describing position and velocity of a particle (x,u) evolve as a Markovian process. The particle’s trajectories in six-dimensional phase space can be modeled based on concepts similar to those in the modeling of Brownian motion. In this work, the LSM technique, modified by Weil et al. 2004 to be compatible with LES, is extended to heavy particles. In this technique, the particle follows the resolved turbulent motions simulated by the LES, while the motions due to the unresolved scales are included as a stochastic contribution. The effect of the particle’s mass is accounted for by imposing a settling velocity and reducing the Lagrangian autocorrelation time-scale. The model results are validated against experimental data from field measurements of the dispersion of glass beads in the atmospheric boundary layer.