This dissertation explores key physical mechanisms that control upper ocean and sea ice variability in the Southern Ocean. The first portion of this work presents an observational analysis of wintertime upper ocean stability and pycnocline heat availability in the Antarctic sea ice zone. This analysis reveals that the southern Weddell Sea region, which features a weak upper ocean stratification and relatively strong thermocline, is preconditioned for exceptionally high rates of winter ventilation. In other open-ocean regions, such as the northern Ross Sea, the stronger winter stratification greatly limits the efficiency with which heat may be extracted from the pycnocline. The coupling between winter ice growth and upper ocean ventilation is further explored using an idealized 1D sea ice-ocean model. This model is used to simulate winter ice growth in different regions under identical surface forcing. Consistent with the observational analysis, these simulations show that the unique thermohaline structure of the Weddell Sea, specifically that near Maud Rise, facilitates a strong negative feedback to winter sea ice growth. For this region, the entrainment of heat into the mixed layer can maintain a near-constant ice thickness over much of winter. However, these simulations also reveal that this quasi-equilibrium is attained when the pycnocline is thin and supports a large vertical temperature gradient. Further experimentation demonstrates that the surface stress imparted by a powerful storm may upset this balance and lead to substantial ice melt. In simulations initialized with profiles from more strongly stratified regions, such as near the sea ice edge of the major polar gyres, the entrainment of heat into the mixed layer had weak impact on winter ice growth---even during periods of strong wind forcing. Thus, a key takeaway is that the thermodynamic coupling between winter sea ice growth and ocean ventilation has significant regional variability. This regionality must be considered when evaluating the response of the Antarctic ice-ocean system to future changes in ocean stratification and surface forcing. In the final portion of this dissertation, focus is shifted to variations in Southern Ocean sea surface temperature (SST) and sea ice extent (SIE) on seasonal timescales. This work is motivated by the abrupt reversal of Southern Ocean SST and SIE trends that occurred in 2016 and 2017. The first half of this chapter examines the role of surface winds in the initiation of the anomalous sea ice retreat that occurred in late 2016. This is done via a simple regression analysis that quantifies the linear relationship between seasonal SIC anomalies and near-instantaneous local wind anomalies, using observations and reanalysis. With this empirical relationship, we demonstrate that surface wind anomalies can largely explain the SIC anomalies observed in the winter and spring of 2016. In the Weddell Sea, some of this preconditioning was associated with the winter polynyas that appeared that year. These events are linked to strong upwelling in the Weddell Sea and the passage of powerful winter storms. Lastly, we construct an updated seasonal mixed layer heat budget for the Southern Ocean, which is then used to explain the near-record Southern Ocean SSTs that occurred in the summer of 2016--2017. This analysis reveals that the warming maximum was the combined effect of enhanced air-sea heating, reduced northward Ekman transport, and shallower than normal mixed layer depths. From these results, we conclude that the 2016--2017 Southern Ocean SST and SIE anomalies were primarily caused by a serendipitous sequence of anomalous atmospheric and oceanic conditions. These anomalies coincided with an unusual synchronization of tropical and extratropical modes of climate variability.