Scale-aware, energetically consistent techniques for eddy-permitting ocean modelling - Scott Bachman (NCAR)

Scale-aware, energetically consistent techniques for eddy-permitting ocean modelling
Scott Bachman, National Center for Atmospheric Research (USA)


CMIP-scale ocean models are progressing ever further into a regime where the largest mesoscale eddies are permitted by the model grid, which introduces new challenges in the development of turbulence closures. If these challenges are handled inappropriately, the consequences can be profound: for example, weak large-scale flows and unphysical gridscale features can emerge, which leaves the integrity of the entire model solution in doubt. In terms of model diagnostics, these issues are manifested in kinetic energy spectra that are depressed at small wavenumbers and exhibit ”ramps” or ”ski jumps” at large wavenumbers. In this talk I will discuss recent progress in handling these issues by employing turbulence

closures which are energetically consistent, scale- and flow-aware. These concepts are key to eddy-permitting modeling, as it allows closures to naturally turn on and off depending on the grid resolution and properties of the resolved flow, while remaining consistent with the large-scale ocean energy cycle and quasigeostrophic turbulent cascades. I will present two closures in particular which are modifications of the well-known Gent and McWilliams (GM) scheme: ”QG Leith”, which adjusts the GM coefficient based on the rate of enstrophy dissipation in the forward enstrophy cascade, and ”GM+E”, which re-injects energy extracted by GM into the resolved barotropic flow. Individually, these closures provide elegant solutions to rectify the kinetic energy at small (QG Leith) and large (GM+E) scales, and it will be shown that together they significantly improve the kinetic energy spectra, probability distributions of flow structures, and restratification rate within an idealized baroclinic channel. I will also discuss results from employing QG Leith in an eddy-permitting global POP simulation, and prospects for including it and GM+E in upcoming experiments using MOM6.