Large-eddy simulation (LES) is widely used to replicate atmospheric and other flow behavior for many purposes, such as
1) Prediction of flows in complex topography, 2) Transport and diffusion of pollutants, 3) Convective energy flux estimation.
This research will concentrate on developing a state-of-the-art tool for the LES of atmospheric boundary-layer flow over complex terrain based on Reynolds-averaged Navier-Stokes equations.
Intellectual merit: Simulating flow within a sufficiently large region in Owens Valley, CA with adequate grid resolution will exploit the benefits of the complete subgrid-scale and subfilter-scale turbulence scheme. A new land surface characterization tool to provide better treatment of land surface processes at fine resolution over complex terrain will also be incorporated. Comparison of simulation results with relevant parts of the extensive Terrain-induced Rotor EXperiment (T-REX) dataset will test the realism of the turbulence models and the tiled-land-surface treatment. Specifically, the effects of small-scale variations in surface properties on fluxes and flow dynamics will be assessed. These newly combined modules should provide an improved state-of-the-art tool for atmospheric boundary-layer LES.
Broader impacts: This research will demonstrate that the LES code is suitable for studying the development of atmospheric flow structures, and for sensitivity studies of the effects of grid resolution and land surface treatment on the fluxes and flows of the atmospheric boundary-layer. The results will also be useful for developing parameterizations for numerical simulation at regional to global scales. Improved LES techniques can have wide impacts wherever LES techniques are applied. A Ph.D. student will be trained in the subject area.
Numerical simulation of motions in the atmospheric boundary layer is a mainstay of weather prediction and the study of atmospheric turbulence. Large-eddy simulation [LES] is a staple of weather research and of turbulence modeling. However, in the context of LES, subgrid scales [SGS] arise either implicitly because of the filtering effect of representation of the flow variables on a grid or explicitly due to application of filters. So in LES, the energy producing scales of three-dimensional atmospheric turbulence are explicitly resolved, but the smaller-scale portion of the turbulence spectrum is not resolved, yet it has a significant effect on the results of the simulations. A key aspect of LES is achieving a representation of the SGS such that their impact on the resolved scales is correct. The objective of this grant was to develop an algebraic model for the subgrid-scale motions, which must be represented in the resolved-scale equations in an LES. The outcome of our work was a generalized linear algebraic SGS [GLASS] model that dealt with both the stresses in the fluid and the flux of heat (a scalar) in LES of a dry atmospheric boundary layer. The SGS model components for stresses and flux are fully coupled, so that the scalar heat flux produces a buoyant effect on the stresses and the stresses cause the heat flux to be anisotropic. Our GLASS model is implemented in a FORTRAN subroutine, called MULTISTRESS, in the atmospheric modeling code, the Advanced Regional Prediction System (ARPS). Intellectual Merit: The GLASS model is described in a dissertation, three conference proceedings, and a presentation at the European Geophysical Union 2013 Annual Meeting. Applications were made to a range of atmospheric conditions. Neutral boundary layer, convective boundary layer, and stable boundary layer simulations demonstrate that the GLASS model can (1) perform well in different stability regimes and (2) provide appropriate scalar and momentum flux anisotropies. The GLASS model also overcomes the need to alter model coefficients for different positions in the flow, grid/filter aspect ratios, and atmospheric stabilities, etc. Our simulations include the use of the algebraic model alone and coupled with a reconstruction model. Broader Impacts: There are both educational and scientific impacts. A doctoral student carried out the work. She completed her dissertation in April 2013. In the past, a major community code (WRF) has been modified to include a dynamic reconstruction model; the GLASS model can replace the dynamic portion of that model yielding the combination of models that we have tested. This means that our improved SGS model can be used in any LES code and so is a potential benefit for the entire community.
