Evidence for Barotropic Wave Radiation from the Gulf Stream

Abstract

Highly energetic velocity fluctuations associated with topographic Rossby waves are frequently observed over the continental slope and rise off the United States and Canadian east coast. It has been suggested that the energy source for these waves could be eastward-propagating Gulf Stream meanders, which can couple to the westward-propagating Rossby waves if the meander shape is time dependent. In this study, a historical archive of all available current meter data from the western North Atlantic has been examined for evidence of energy radiation from the Gulf Stream via barotropic/topographic Rossby waves. Maps of abyssal eddy kinetic energy (EKE) and Reynolds stress were constructed for four frequency bands (including motions with periods between 256 and 7.8 days) to examine distributions of these quantities over the largest possible geographical area. Maximum eddy kinetic energy is observed along the mean axis of the Gulf Stream at low frequencies (50?250 days), but this maximum shifts north and west with increasing frequency. The westward shift may be due to the fact that the phase speed of Gulf Stream meanders decreases with increasing downstream distance. The northward shift of maximum EKE to a position over the continental slope and rise is discussed in terms of the refraction and convergence of barotropic Rossby wave energy rays emanating from the region of the Gulf Stream. The Reynolds stress maps show strong evidence of onshore energy radiation north of the Gulf Stream over a large geographical area and at all frequencies considered. The velocity components are found to be statistically coherent and 18O° out of phase at many locations when viewed in a coordinate system aligned with the local ambient potential vorticity gradient. Energy radiation away from the Stream to the south, although expected, is not apparent in the observations, perhaps due to the dominance of other eddy-generating mechanisms there such as baroclinic instability of the westward recirculation. Large-scale features of the maps compare poorly with similar maps generated from a stochastic wave radiation model, and it is suggested that such models need to include more realistic forcing and basin geometry before detailed model-data comparisons can be made.