Developments in Mesoscale and Turbulence Modeling in Critical Geophysical Flows

Recent developments of idealized theory of perturbed mesoscale flows over changes in surface roughness, temperature, and elevation, together with mesoscale numerical simulations and field and laboratory data, have been applied to atmospheric flows over polar orography (and ice edges), the Himalayas, and urban areas because of their very large sensitivities to climate and environmental changes. (e.g. Hunt et al. QJRMS 2004,Orr et al. J. Atm. Sci.2008; Hunt AMS Meeting 2009)

Recent experimental, numerical and theoretical studies with colleagues at ASU, Europe and Nagoya have shown up some new critical features of turbulence that are important for modeling geophysical flows:

a) There is a sudden relative drop in surface friction (Τ/U²) when convective velocity w* exceeds the mean velocity U. One explanation is because of the interaction between entrainment into the isolated moving plume structures and the surface layer. Is this the point when MO scaling breaks down (Owinoh et al. BLMet 2005)?

b) Turbulent shear layers above or below stably stratified layers induce wave motions with shear stresses (typically about 1/5 of those in the shear layer) (Mahalov et al. Theor. Comp. Fl. Dyn. 2007). These stresses which are often overlooked in models, induce significant mean flows (e.g. double layers at night ) and contribute to intermittency near the critical Richardson number. Ocean flows below the thermocline may also be driven by this mechanism.

c) Models of thin shear layers within turbulent flows (Hunt et al. J. Fl. Mech. 2006) can now be linked to the very high Reynolds number simulations by Prof. Kaneda and colleagues (Ann. Rev 2009), which show no evidence of a Richardson like cascade, but an intermittent structure of shear layers where intense microscale vortices tend to form. A new deterministic, non-cascade, scaling and statistical analysis is proposed for intermittent processes at the smallest scales and in the inertial range, where fast up and down scale transfer processes are in approximate equilibrium. This approach could probably be developed to model small scale mixing and particle processes in non-equilibrium turbulence.