Parameterizing Mesoscale Ocean Eddies Over Bottom Topography

Mesoscale ocean eddies play a crucial role in driving large scale ocean circulation (Hallberg and Gnanadesikan 2006) and determining the global ocean stratification (Karsten et al. 2003). Consequently, they may
affect global and localized climate predictions as well as ocean heat and carbon storage capacity (Munday et al. 2014). Accurate simulation of mesoscale ocean eddies in climate models is therefore an important
tool for future climate projections. However, the resolution of climate models is rarely sufficient for eddies to be explicitly simulated and they have therefore been parameterized in climate models.
The prevailing parameterization scheme used in climate models is that of Gent and McWilliams (hereafter referred to as GM90) in which the eddy-induced velocity is parameterized as an additional advection of
tracers via the thickness diffusivity, k (Gent and McWilliams, 1990). GM90 is a hugely successful scheme, having been adopted into most of the CMIP models (Eyring et al. 2016) and leading to improvements in
the accuracy of coarse resolution ocean models (Danabasoglu et al. 1994). One obvious flaw in the GM90 scheme is its inability to respond to variations in ocean floor topography. The problem of improving eddy
parameterization over bottom topography will be the focus of this project.