Scaling and Randomness of Ocean Mesoscale Eddies

Abstract:

The next generation of ocean climate models will reach the eddy-permitting regime such that the fidelity of the simulations should be improved. However despite their increased horizontal resolution, these models will not adequately resolve eddies. Parametrization of sub-grid eddies remain therefore necessary. A representation of the mesoscale eddy forcing as function of the straining and shearing of the resolved flow, wind forcing, stratification and model resolution is developed using probability distribution functions (PDFs) diagnosed from high-resolution eddy-resolving runs. The PDFs represent the eddy-eddy and eddy-mean flow interactions and associated Reynolds stresses necessary to mimic the effects of the eddies. The reconstructed PDFs are used as a parametrization of sub-grid eddy effects, leading to improvements in the mean and variability of the coarse-resolution parametrized simulation over the unparametrized version. Uncertainties in the parametrization and parameters are inherent to the formulation of parametrization, via the PDFs, which allows for a quantification of eddy forcing and model error as function of grid-size and external forcing.