Large eddy simulations and Lagrangian stochastic modelling
Üllar Rannik, Sofia Mei-Kuei Tu, Annika Nordbo, Leena Järvi, Timo Vesala
Large eddy simulations (LES) are a popular fluid dynamics tool for simulating turbulent flow. LES resolves explicitly large eddies and parameterizes small eddies using a subgrid-scale model. The parallelized LES model (PALM, Raasch and Schisröter, 2001) is designed for simulating atmospheric and oceanic flows on massively parallel computer architectures. The latest version of PALM provides a high level of realism to the turbulent atmospheric flow which allows the study of exchange processes between atmosphere and surface (Figure 1). PALM can be used in concurrence with Lagrangian Stochastic (LS, see below) modelling: PALM can resolve a desired 3-dimensional flow field which can be transferred into a LS model that calculates the composition, distribution and trajectory of particles, which have been released during the simulation. The spatial distribution of particles is used for calculating the scalar concentration and flux, in addition to their source areas (footprints). In our group, this modelling setup is used for simulating flow over an urban area and over an idealized chess-board like pattern of grass and forest. In both cases, the footprints are calculated based on released particles.
Lagrangian Stochastic modelling (LS) solves the paths of particles in a given turbulent flow. LS models are used for dispersion simulations within a canopy and the atmospheric boundary layer. The model is capable to account for turbulence statistics and dispersion inside canopy, driven by the canopy leaf area profiles and density, and skewed turbulence in the convective boundary layer (Rotach et al., 1996). The research areas of interest are dispersion and deposition of gases and atmospheric aerosols, including particles of biogenic origin, and also evaluation of the source statistics contributing to micrometeorological flux and concentration measurements, the footprint functions.
Figure 1: LES modelling of turbulent kinetic energy (TKE) of flow that hits a forest edge. x is the along-wind distance and the forest is marked with a green square.