Frontal Variability in Drake Passage - A Modeling Study

The Antarctic Circumpolar Current (ACC) features three major fronts: the Sub-antarctic Front (SAF), the Polar Front (PF), the Southern ACC Front (SACCF). The locations of these fronts are not stable. The PF can shift away from its historical mean locations on the order of 100 km. The ACC transport in Drake Passage varies over a large range (50 to 60 Sv). Numerical simulations with the Regional Ocean Modeling System are carried out to study the frontal variability under the influence of ACC transport, local wind stress and bottom topography in Drake Passage. Front-embedded numerical experiments are carried out without surface forcing for different ACC transports (from 95 to 155 Sv with an interval of 10 Sv). Large transport shifts the fronts northward while the fronts move southward with small transport. The mean shifting distance of the PF from the historical mean location is minimum with 135 Sv transport. The SAF and the SACC are confined by northern and southern continents, respectively, while the PF is loosely controlled by the topography. Due to impact of the eddies and meanders on the PF at several regions in Drake Passage, the PF may move northward to join the SAF or move southward to combine with the SACCF, especially in the central Scotia Sea. The SAF and PF are more stable with higher transport. The SAF behaves as a narrow, strong frontal jet with large transport while displaying wavy structure with smaller transport. In the model, the relationship between Ertel Potential Vorticity (EPV) and the 2D stream function is examined at different depth. The linear correlation coefficient between the EPV and the stream function is more than 0.9 between 1000 meter and 2500 meter depth while in the upper 500 meters it is less than 0.3, and near the bottom it is around 0.6. The smaller coefficient is caused by the removal of potential vorticity by friction and strong mixing. Similar simulations with 135 Sv transport but with different wind stress applied on the surface are carried out to study the local wind stress effects. Three kinds of surface wind stress are the 6 hourly QSCAT/NCEP blended wind stress, monthly running mean filtered wind stress and zero wind stress. With 6 hourly wind stress, the PF location is more variable than that with the monthly running mean filtered wind stress. The mean PF location changes with different wind stress. This change is different at different locations in the model. The surface elevation to each side of PF changes with the wind forcing. The peak frequencies at which the wind stress is correlated to the surface elevation above the 95% confidence level in the south are the 8 and 30 days with the wind stress change leading the surface elevation change. The peak frequencies to the north of the PF are 8, 15 and 40 days. The positive phase lag at some frequencies might be due to the contamination from the local baroclinic instabilities. The mean 500 m temperature tracked PF location is consistent with mean surface PF location. The surface PF tends to be south of the 500 m PF front. This difference between the surface and 500 m PF locations is modulated by the wind stress and the topography. With stronger wind stress, the difference is reduced. Form drag from large variation of bottom topography shows little change with different wind stress in the model. The form drag in Drake Passage is calculated to be one order of magnitude larger than the local wind stress. The bottom skin stress can be neglected compared to other terms. Form drag is primarily due to remote forcing (the transport variations) instead of the influence of the local wind stress