Mesoscale Ocean Large Eddy Simulations Using High-resolution Ocean Models

Brodie Pearson1, Baylor Fox-Kemper1, Scott Bachman2, Frank Bryan3 and David A Bailey3, (1)Brown University, Providence, RI, United States, (2)University of Cambridge, Cambridge, United Kingdom, (3)National Center for Atmospheric Research, Boulder, CO, United States
Abstract:
Inaccurate parameterization of sub-grid eddies can cause excessive damping and spurious diapycnal mixing, especially in high-resolution [O(10km)] ocean models. The Mesoscale Ocean Large Eddy Simulation (MOLES) approach provides a framework for developing resolution- and flow-adaptive parameterizations of eddy effects. Large eddy simulation techniques are commonly used to simulate 3D turbulence, and MOLES is modified to be appropriate for the more two-dimensional nature of mesoscale ocean turbulence.

However, the effect of MOLES in high-resolution ocean models has not been investigated extensively. We will contrast results, and cost, from a suite of idealized simulations of frontal spin-down (MITgcm) and from high-resolution global climate models (0.1o, POP2), under a variety of eddy parameterizations. These include MOLES based upon 2D turbulence theory, MOLES based upon quasi-geostrophic (QG) turbulence theory, and traditional biharmonic schemes.

The idealized simulations show that MOLES (particularly QG) improves the spectral slopes of energy and enstrophy near the grid-scale when compared to more traditional eddy parameterizations, across a range of grid resolutions. In the high-resolution global climate model we compare the effect of different parameterizations on the spectral characteristics of the simulated flow, and on the large-scale transport. Using MOLES in a climate model results in greater energy and variability near the grid scale, and this produces a flow, which, spectrally, is more consistent with an inertial turbulent cascade and observations of eddy behavior.