Steven Herbette, Yves Morel, and Michel Arhan, SHOM/CMO
This study investigates the behavior of a surface eddy encountering an isolated seamount on the beta-plane. We first show that the erosion mechanism observed within f-plane experiments is still valid. Strong erosion and even splitting of the surface vortex can occur provided it gets close enough to the seamount. Indeed, as on the f-plane, deep fluid parcels can be advected across the isobaths, which generates a pair of cyclonic and anticyclonic eddies in the vicinity of the topography. This pair of vortices, located in the bottom layers, can in turn exert a shear on the initial surface eddy, which induces erosion through the loss of filaments. However, the presence of a background potential vorticity gradient (or beta-effect) leads to substantial differences. The beta-effect provides a way to displace the vortex towards the topography so that, even when located far from the seamount initially, a vortex can get close enough to be strongly eroded. Secondary eddies created by advection of fluid parcels across the beta-plane in the vicinity, and below, the initial surface vortex also contribute to its erosion, even without bottom topography. This latter mechanism makes the vortex/seamount encounter a much more complicate process than it is on the idealized f-plane. Erosion is no more a strict function of the vortex strength, seamount characteristics and distance separating the vortex and the seamount. For similar parameters the erosion rate can be very different and is very sensitive to details of the configuration. This "hypersensitivity" is shown to be associated with the presence of additional eddies generated by advection of planetary potential vorticity, in particular a cyclonic eddy in the bottom layer. The latter modifies the vorticity field of the original vortex in the lower layer, which has some impacts on its path and erosion rate. As a result, the erosion of the vortex becomes impossible to predict.