Stirring in the Deep: The Impact of Ocean Topography on Mesoscale Eddy Dynamics

Abstract

Motion in the ocean occurs on a wide range of spatial scales: from large-scale currents such as the Gulf Stream spanning thousands of kilometers to small-scale turbulence on the order of millimeters. A large portion of this motion happens at the mesoscale, which describes processes on spatial scales of tens to hundreds of kilometers. Mesoscale eddies, swirling ocean vortices, are ubiquitous in the ocean. They can stir and mix water masses with different properties, such as high and low temperature and nutrient concentrations. This stirring influences which parts of the ocean are hotter or colder, and which parts have more or fewer nutrients to sustain marine life. Eddies thus play an important role in the global ocean circulation, marine ecosystems, and the climate system. Given the relevance of mesoscale eddies, it is important to take them into account when investigating Earth's climate. Such investigations are typically done using numerical climate models. A climate model divides the Earth's ocean and atmosphere into a spatial grid and solves the physical equations governing the dynamics for each grid cell. Variations that are happening within individual grid cells cannot be resolved by the model. Ideally, we would thus want the grid cells to be as small as possible, such that even the smallest-scale processes can be described. However, this demands much more computational power, as many more grid cells are then needed to cover the entire Earth. We need to make choices to include the most relevant physical processes while keeping the computational efforts feasible. Unfortunately, many current climate models do not have fine enough grid cells to fully capture the mesoscale, including the stirring caused by mesoscale eddies. To include processes that cannot be explicitly resolved in the model, many models make use of parametrisations. A parametrisation uses the variables that are represented on the grid cell scale to infer how processes occurring on smaller scales should be affecting the grid cells. In the context of mesoscale eddy stirring, a parametrisation can describe the result of the stirring, which is the transport of oceanic properties (e.g. temperature) between grid cells. We quantify the efficiency with which mesoscale eddies transport properties using the eddy diffusivity: a value that represents the 'strength' of the stirring. The outcomes of many climate models thus depend on the value of the eddy diffusivity used in the model. However, there is high uncertainty in what this value should be, and how it varies throughout the ocean and through time. To achieve a better representation of mesoscale eddy stirring in climate models, we need two things. First, we must further our understanding of how mesoscale eddy dynamics are influenced by environmental parameters. Second, we need to improve existing techniques and develop new ones to more accurately measure eddy diffusivities throughout the ocean. The aim of this thesis is to address both of these issues. In particular, the focus of this thesis is on how ocean topography - the shape of the seafloor - impacts mesoscale eddy dynamics.