Meander dynamics in the Antarctic Circumpolar Current

Meanders formed where the Antarctic Circumpolar Current (ACC) interacts with topography have been identified as dynamical hot-spots, characterised by enhanced eddy energy, momentum transfer, and cross-front exchange. However, few studies have diagnosed the dynamics of ACC standing meanders directly from observations. Therefore, we first use a synoptic, upper-ocean hydrographic survey and satellite altimetry to explore the momentum and vorticity balance of a Subantarctic Front standing meander, downstream of the Southeast Indian Ridge. Second, we validated the processes diagnosed in the first hydrographic survey with a high-resolution model. And finally, we applied the knowledge of the momentum and vorticity balances gained from the Subantarctic Front meander to a second, more recent hydrographic survey of a growing, unstable meander in the Polar Front. This included measurements of full depth temperature, salinity and velocityeanders formed where the Antarctic Circumpolar Current (ACC) interacts with topography have been identified as dynamical hot-spots, characterised by enhanced eddy energy, momentum transfer, and cross-front exchange. However, few studies have diagnosed the dynamics of ACC standing meanders directly from observations. Therefore, we first use a synoptic, upper-ocean hydrographic survey and satellite altimetry to explore the momentum and vorticity balance of a Subantarctic Front standing meander, downstream of the Southeast Indian Ridge. Second, we validated the processes diagnosed in the first hydro[1]graphic survey with a high-resolution model. And finally, we applied the knowledge of the momentum and vorticity balances gained from the Subantarctic Front meander to a second, more recent hydrographic survey of a growing, unstable meander in the Polar Front. This included measurements of full depth temperature, salinity and velocity.
The unique hydrographic survey of a standing meander in the Subantarctic Front revealed along-stream watermass changes linked to the phase of the meander. Using a gravest empirical mode (GEM) approach, we determined the mean cross-stream watermass structure from the observations in the upper 1500 m. Along-stream anomalies of temperature, relative to the GEM mean field, were diagnosed using a Reynolds decomposition to separate anomalies due to heave of isopycnals and watermass property change on isopycnals. The upper ocean (150–600 m) shows cooling entering the trough and warming entering the crest, while warming is observed from trough to crest in the deeper ocean (600–1500 m). Advection of relative vorticity is balanced by vortex stretching, as found in model studies of meandering currents. Meander curvature in this region is sufficiently large that the centripetal term is important and the flow is in gradient wind balance, resulting in ageostrophic horizontal divergence. This divergence drives downwelling of cooler water along isopycnals entering the trough and upwelling of warmer water entering the crest, consistent with the observed evolution of temperature anomalies in the upper ocean. Progressive warming along the meander, observed between 600 and 1500 m, is likely to reflect cyclogenesis in the deep ocean. Vortex stretching couples the upper and lower water column, producing a low pressure at depth between trough and crest and cyclonic flow that carries cold water equatorward in the trough and warm water poleward in the crest. These observational results highlight the importance of the gradient wind balance and cyclogenesis in the dynamics of standing meanders and support earlier model studies of their role in the ACC momentum and vorticity balance.
In a high resolution ocean simulation, we have identified a standing meander at the same location in the Subantarctic Front and at a similar stage of development to investigate the full depth water mass and velocity structure in the meander. In a snapshot of the model’s meander, we find bottom intensified eddies underneath the growing upper ocean meander, as suggested by the hydrographic survey. These bottom eddies are difficult to detect from sea surface dynamic topography. We found that the gradient wind balance applied to the model’s sea surface height is a good approximation for the full model velocities just below the mixed layer in the meander. Horizontal divergence derived from the ageostrophic contributions of the near-surface gradient wind velocity is a good proxy for identifying vertical velocities within the water column and the location of deep cyclones. Patterns of horizontal divergence driven by curvature effects is largest near the surface, but extends throughout the water column. Near the bottom, horizontal divergence is locally intensified near strong topography changes, which indicates another source of divergence, one that is driven by interaction of deep currents with topography.
In contrast to the observations, where temperature anomalies on isopycnals were an order of magnitude larger than anomalies due to isopycnal displacement, the model temperature anomalies due to heave have a similar magnitude to temperature anomalies on isopycnals. The temperature anomalies on isopycnals in the model’s upper ocean show a pattern similar to that seen in the observations: along-stream cooling entering the trough with predominantly downwelling along isopycnals, a warming entering the crest with strong upwelling along isopycnals, and a cooling exiting the crest with predominantly downwelling along isopycnals. At intermediate depths (600-1500 m) the temperature anomalies on isopycnals in the model are similar to the observations. We see a progressive warming from the upstream side of the trough to the downstream side of the crest, which is consistent with rotation of the horizontal velocity vector with depth, and associated cross-stream heat advection that has been documented in previous studies of the Gulf Stream and ACC. The deep ocean (> 1500 m) shows an anomaly pattern that is somewhat opposite to that at intermediate depth around the location of the deep cyclone that lies between the meander trough and crest at the time of the model snapshot. We find slightly warm and salty anomalies upstream of the centre of the deep cyclone and relatively strong, cold and fresh anomalies on the downstream side of the centre of the cyclone. Localised around the cyclone, the meridional gradient in temperature and salinity along isopycnals of the Circumpolar Deep Water changes sign. Instead of the warming from south to north across the front that we see everywhere else, we find cooling from south to north. This localised alteration of the temperature gradient is consistent with cyclonic motion and the temperature anomalies. We speculate that turbulent mixing and flow topography interactions may be responsible for the local alteration of the meridional gradient in properties. A second survey was conducted of a standing meander in the Polar Front between the Southeast Indian Ridge and the Macquarie Ridge. From satellite altimetry and models, this meander is in a location of intense poleward fluxes of heat. It is also less stable and more rapidly evolving than the first meander that was studied in the Subantarctic Front. Temperature variability in the Polar Front meander is dominated by anomalies on isopycnals in the AAIW and UCDW, but a clear along-stream pattern was not detected. Some of the anomalies in the upper water column are attributed to the fragmentation of the SAF and PF downstream of the meander crest. The temperature variability in the LCDW is dominated by heave, which seems to be mostly driven by flow interaction with topography. Surface convergence and divergence in the flow field could not explain the temperature anomalies in the rapidly evolving meander of the Polar Front. In an unstable and rapidly evolving meander, smaller scale dynamics may be important for driving temperature variability in the upper water column.