Examination of the challenges to one-way nesting of large-eddy simulations using the Weather Research and Forecasting Model

The technique of nesting fine-scale large-eddy simulations (LES) inside more coarsely-resolved bounding simulations is being increasingly employed to capture smaller-scale features of both the underlying surface and the flow above, an approach that can, in principle, lead to improved prediction of quantities within the atmospheric boundary layer (ABL). However, errors can be introduced into the nested domain from both the outer domain solution, and from the transition of flow and model properties near nest interfaces. While grid nesting has been successfully employed to downscale large-scale simulations to the mesoscale, a different set of issues challenges the extension of this strategy to turbulence-resolving resolutions. We examine the prediction of several ABL flow parameters, including wind speed, stress, and turbulence kinetic energy (TKE), in a variety of one-way nested LES using the Weather Research and Forecasting model. We investigate the role of the SFS stress model formulation by examining several different closure approaches, and provide recommendations for best practices using existing approaches, and suggest areas for future research.

This work performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.