Recent advances in computational fluid dynamics have made numerical simulations a promising tool to study the complex dynamics of land-atmosphere interactions. In particular, Large-Eddy Simulation which solves the 3-D unsteady transport equations for momentum and scalars (e.g., temperature, water vapor and pollutants), proves well suited for the task of computationally simulating the Atmospheric Boundary Layer (ABL) with a parsimonius balance of precision and accuracy. LES accomplishes this by using spatial filtering to remove smaller, unresolved (spatial and temporal) scales. The subfilter scales (or subgrid scales) are not explicitly resolved and their effect on the resolved scales is parameterized using a Subgrid-Scale (SGS) model. Results of simulations have been shown to be very sensitive to the SGS model formulation, especially near surfaces where the subgrid scales account for most of the turbulent fluxes. Improving SGS models has been identified as a priority to make LES a more applicable and reliable tool to study the turbulent fluxes of momentum, heat, moisture and pollutants in the ABL.

LES consists of solving the three-dimensional unsteady transport equations for momentum and scalars (e.g. temperature and water vapor) in turbulent flows. The research group has developed an LES code to study turbulent transport in the atmospheric boundary layer. The code has recently been optimized and is fully parrallelized. Computational resources are provided in part by the Minnesota Supercomputing Institute (MSI) and the National Center for Atmospheric Research (NCAR) .

Among other research issues, we use LES to simulate stable boundary layers, to study the influence of turbulence on atmospheric chemistry, to evaluate the impact of surface heterogeneities on the boundary layer structure, to develop and test new subgrid scale parameterizations.