A 3D RANS approach based on a modification of the effective Prandtl number for high resolution mesoscale simulations

The continuous increase of the computational power is making possible mesoscale simulations at sub-kilometer resolutions. This can be beneficial in some situations because it allows a better resolution of the circulations induced by surface hetereogeneitis of the scale of few kilometers like urban areas, small lakes, narrow valleys, etc. However, practice shows that when standard mesoscale models are run at these sub-kilometer resolutions, during convective conditions, some spurious circulations form that have shapes similar to convective rolls or cells, but do not always have the same size, do not necessarily form in the same conditions and are resolution dependents. It can be shown that these structures are the result of an incorrect treatment of the superadiabatic layer that forms in convective conditions (see Ching et al. same session). In this contribution a new approach is proposed, aiming to parameterize the effect of the real convective circulations that form in similar conditions, avoid the formation of the spurious ones, and resolve the circulations induced by the surface hetereogenties. Since the spurious circulations form when the effective Raleigh number (obtained by using the eddy viscosity and turbulent thermal diffusivities produced by the PBL schemes) is greater than a critical value, the approach keeps a sub-critical effective Raleigh by increasing the turbulent thermal diffusivity and, in addition, by using in the horizontal the same eddy viscosities and diffusivities used in the vertical. This is equivalent to ensuring that the diffusion mechanism is always more efficient than the generation of the vertical motions caused by the instability of the superadiabatic layer. This approach is implemented in WRF for the Bougeault and Lacarrere (MWR, 1989) PBL scheme, and its efficiency is tested against LES and compared to the standard approach currently used in mesoscale models, over homogeneous and heterogeneous terrains.