The last deglaciation (~20–9 kyr BP) was marked by rapid reorganizations of the global ocean circulation that accompanied abrupt climate transitions and rising atmospheric CO₂ concentrations. Here, we synthesize transient Earth system model simulations to investigate the spatiotemporal variability of Southern Ocean (SO) dynamics and their interactions with high-latitude climate processes. Our results demonstrate that SO circulation evolved through distinct dynamical regimes during Heinrich Stadial 1, Bølling–Allerød, Younger Dryas, and early Holocene in response to coupled changes in wind stress, Antarctic sea ice, surface buoyancy forcing, and freshwater fluxes. Beyond the dominant influence of wind forcing, Antarctic sea ice reorganized the spatial extent of buoyancy-loss regions and modified meridional salinity gradients, altering overturning circulation and deep-ocean stratification. A pronounced zonal asymmetry characterized the Last Glacial Maximum, with an austral winter double jet that progressively weakened and transitioned into a single mid-latitude jet during the onset of the Holocene, emphasizing the need to move beyond zonal-mean interpretations of SO circulation. Wind stress over the Indian Ocean sector was approximately 36% stronger than over the Pacific sector, corresponding to ~37% greater upwelling and a close spatial correspondence between wind stress maxima and upwelling cores. In contrast, the Pacific sector exhibited a progressive decoupling between wind stress and upwelling during the late deglaciation, indicating stronger modulation by sea ice, buoyancy forcing, and salinity gradients. Moreover, the zero-surface buoyancy flux boundary is closely aligned with the quasi-permanent sea ice boundary, and it also migrated poleward by ~11° from the Last Glacial Maximum to the onset of the Holocene, with a larger displacement in the Pacific sector. These findings highlight the importance of regionally distinct coupled atmosphere–ice–ocean feedbacks, providing new insights into deglacial overturning variability and sensitivity of SO processes to future climate change.