Effects of stability functions in a dynamic model convective boundary layer simulation

Dynamic subgrid models are increasingly being used in simulations of the atmospheric boundary layer. We have implemented several variant forms of dynamic models in the UK Met Office Large Eddy Model (MetLEM), including a state-of-the-art Lagrangian-Averaged-Scale-Dependent (LASD) model. The implementation includes optional use of stability functions in the specification of the eddy viscosity and diffusivity, as well as optional use within the dynamic calculation of the Smagorinsky parameter. This paper reports on the behaviour of the LASD model with different choices for the inclusion and treatment of stability functions in convective boundary layer simulations at different resolutions. Results are compared against a high-resolution Large-Eddy simulation (LES) and against simulations employing the Smagorinsky-Lilly subgrid model. We conclude that the use of stability functions improves the behaviour of the LASD model in the grey zone regime. Moreover, a careful treatment of the stability functions in the calculation of the dynamic parameters, whilst attractive theoretically, is found to be unnecessary in practical terms.