Momentum balance in the Southern Ocean

Strong, persistent winds over the Southern Ocean drive the Antarctic Circumpolar Current (ACC) on an unblocked eastward path around Antarctica. Observations and reanalyses have shown that Southern Ocean winds have increased over the past 60 years [Thompson, 2002; Swart and Fyfe, 2012]. ACC baroclinic transport appears to remain stable [Böning et al., 2008], however, suggesting that the interior mechanisms that output momentum from the ACC system are counterbalancing changes in the input wind stress. Here we combine four years of Drake Passage observations from the cDrake Experiment with six years of model output from the Southern Ocean State Estimate (SOSE) to describe where and how momentum exits the ACC system.

In SOSE, we find that 95% of the zonal momentum input via wind stress at the surface is balanced by topographic form stress across continents and ocean ridges, while the remaining 5% is balanced via bottom friction and momentum flux divergence at the northern boundary of the Southern Ocean domain at 30^ºS. While the time-mean zonal wind stress field is relatively uniformly distributed over the ACC, the time-mean topographic form stress concentrates at shallow ridges and continents that lie within the ACC latitudes -- primarily Kerguelen Plateau, the Macquarie Ridge region, and South America and the Drake Passage fracture zones. Topographic form stress can be divided into shallow and deep regimes: the shallow regime contributes most of the westward form stress that balances the wind, while the deep regime comprises strong eastward and westward form stresses that largely cancel in the zonal integral.

The time-varying signals for these momentum source and sink terms suggest a rapid vertical communication of zonal momentum from the surface to the seafloor. The SOSE form stress signal, integrated over the ACC latitudes 42ºS to 65ºS, tracks closely with the wind stress signal integrated over the same domain; at zero lag, 88% of the variance in the form stress timeseries can be explained by the wind stress signal. We investigate this rapid mechanism by mapping vertical fluxes of zonal momentum in both SOSE and in the cDrake Experiment observations, where four years of high-resolution, full-depth observations allow us to ground-truth the SOSE estimates in a highly energetic region of the ACC.