The interaction of jets with topography in the Southern Ocean is investigated using 19 years of altimetry data. In particular, the ''jet jumping'' mode of variability, by which two or more jets passing close to the same topographic feature show strongly anticorrelated strengthening and weakening, is studied. Three regional case studies are described-the Southeast Indian Ridge south of Tasmania, the Macquarie Ridge south of New Zealand, and the Pacific-Antarctic Rise-where the jet jumping variability is found to occur. Using principal component analysis, the spatial patterns of variability show a vortex dipole forming on either side of a particular jet. For each regional study, it is found that the variability in strength of these vortices (as measured by the spatially averaged vorticity) is strongly correlated with time series of the principle component that describes the jet jumping variability. The observational analysis is complemented by a suite of idealized numerical experiments using a three-layer quasigeostrophic model with simple topography. The numerical results show similar spatial patterns of variability to those observed in the altimetric data. Internal variability is sufficient to generate jet jumping variability, as there is no time-varying external forcing applied in the model configuration. The simulations are used to investigate the effect of topographic scale and changing bottom friction. The authors find that both have a strong influence on the time scale of the variability, with larger topographic scales and higher bottom friction leading to faster time scales. This study shows that even in regions where the flow is strongly influenced by topography, Southern Ocean jet flow may exhibit low-frequency variability.
internal variability, low frequency variability, model configuration, numerical experiments, observational analysis, Southeast Indian Ridge, three-layer quasigeostrophic models, topographic features