A13C-02 The Sensitivity of Realistic High-Resolution Weather Simulations to Turbulence Parameterizations (Invited)

Recent studies have highlighted that nested Large Eddy Simulations (LES) of the atmospheric boundary layer (ABL) show a relatively weak sensitivity to the subgrid scale (SGS) model, when forced by realistic heterogeneous surface parameters (e.g. topography, surface roughness) and large scale dynamics. In contrast, a much larger sensitivity is found to the parameterizations of turbulence in the parent domains, specifically in the so-called gray zone of turbulence (the range of model resolutions where large turbulent motions are partially resolved).

In this study, we present empirical evidence for both these observations from a recent multiscale large ensemble over Perdigão and investigate the reasons for these behaviors. How do turbulent parameterizations affect the mean wind at different resolutions? Why is the mean wind sensitive to the turbulence parameterizations at gray zone resolutions but much less at LES resolutions? We explore these questions with simple examples and idealized simulations with the Weather Research and Forecasting (WRF) model. Specifically, we implement an inline momentum and heat budget into WRF to diagnose the different contributions to these sensitivities. We find that implementation details of both subgrid-scale (SGS) LES schemes and gray zone parameterizations matter for the calculated sensitivities. We discuss relevant implications of our findings for practitioners interested in running high-resolution weather simulations for operational forecasting and wind energy planning.