Nonlinear SGS Model

Rotation has a significant influence on large-scale atmospheric and oceanic flows as well as some engineering flows (e.g., turbulence in jet engines), and also, rotating turbulence provides a simple configuration to study characteristic features and turbulent model performance in anisotropic turbulence. Observations and DNS results show that rotating turbulence has many distinctive qualities, such as reduced kinetic energy dissipation, reverse kinetic energy transfer from small scales to large scales, quasi 2D flow at large scales, and cyclone/anti-cyclone asymmetry. The standard eddy-viscosity models yield excessive dissipation in low-Reynolds number isotropic turbulence, boundary turbulence, and many other flows. It is also found that they are excessively dissipative over a wide range of length scales, and this behavior certainly affects the large-scale flow structure in rotating turbulence, such as failure to deliver the cyclone/anti-cyclone asymmetry in favor of cyclones in rotating turbulence.

Further, in the ABL flow, early SGS models have revealed that the mean wind and temperature profiles in the surface layer differ from the Monin-Obukhov similarity forms. Specifically, the non-dimensional vertical gradients of velocity and temperature could be overestimated by more than 20% in the surface layer. Moreover, the real ABL is strongly influenced by temporal variability of buoyancy effects associated with the diurnal cycle of net radiation at the land surface, and also there exist highly non-linear interactions between the complexity of the land surfaces and the ABL turbulence. The main weakness of LES is associated with our limited ability to accurately account for the dynamics that are not explicitly resolved in the simulations (because they occur at scales smaller than the grid size).