Southern Ocean Deep-Convection as a Driver of Centennial-to-Millennial-Scale Climate Variability at Southern High Latitudes

Abstract:

Antarctic Isotope Maxima (AIM) are centennial-to-millennial scale warming events observed in Antarctic ice core records from the last glacial period and deglaciation. Mounting evidence links AIM events to parallel variations in atmospheric CO₂, Southern Ocean (SO) sea surface temperatures and Antarctic Bottom Water production. According to the prevailing view, AIM events are forced from the North Atlantic by melt-water discharge from ice sheets suppressing the production of North Atlantic Deep Water and associated northward heat transport in the Atlantic. However observations and model studies increasingly suggest that melt-water fluxes have the wrong timing to be invoked as such a trigger. Here, drawing on results form the Kiel Climate Model, we present an alternative hypothesis in which AIM events are forced via internal oscillations in SO deep-convection. The quasi-periodic timescale of deep-convection events is set by heat (buoyancy) accumulation at SO intermediate depths and stochastic variability in sea ice conditions and freshening at the surface. Massive heat release from the SO convective zone drives Antarctic and large-scale southern hemisphere warming via a two-stage process involving changes in the location of Southern Ocean fronts, in the strength and intensity of the Westerlies and in meridional ocean and atmospheric heat flux anomalies. The potential for AIM events to be driven by internal Southern Ocean processes and the identification of time-lags internal to the southern high latitudes challenges conventional views on the North Atlantic as the pacemaker of millennial-scale climate variability.