Late Pliocene-Pleistocene Antarctic climate variability at orbital and suborbital scale: ice sheet ocean and atmospheric interactions

Continental margin drill core and seismic data indicate that between 3.0 and 2.5 Ma, high-latitude climate cooling drove both the West and East Antarctic Ice Sheets towards their present expanded cold polar state. Ice margins developed permanent marine termini with ice shelves. Direct physical sedimentary records of Antarctic Ice Sheet variability (e.g. ice-rafted debris, proximal glacimarine cycles) and more distal ocean records of sea-ice distribution (e.g. diatom palaeoecology), thermohaline circulation (e.g. sortable silt), ocean temperatures (e.g. δ18O), frontal dynamics and surface circulation (e.g. palaeoecological assemblages and sea-surface temperature (SST) reconstructions) all show a strong covariance with the 41 kyr cycle in Earth's obliquity between 3 and 1 Ma. Glacial periods are characterised by northward expansion of seasonal sea ice, SSTs up to 6°C colder than now, equatorward migration of frontal zones by 5–10° latitude, intensification of zonal westerly winds, invigorated surface circulation (e.g. Antarctic Circumpolar Current) and intensified abyssal currents. These processes lead the δ18O ice volume maximum by 3–7 kyr, at which time Antarctic Ice Sheets were fully extended onto the continental shelf. Antarctic ice volume changes were in part controlled by the effect of Northern Hemisphere glacio-eustasy on its marine margin, and this mechanism accounts for much of the orbital variability of the last 2.6 myr. An enigmatic interval of foraminiferal ooze and coccolith-bearing sediments in Weddell Sea and Prydz Bay sediment cores, together with a bioclastic limestone in the Ross Sea at 1 Ma, imply a significant warming and change in ocean chemistry around the periphery of Antarctica. This event, which occurs within the short normal polarity Jaramillo Subchron in Ross Sea cores, is correlated with warm Marine Isotope Stage 31. The anomalous warming implies an increase of 4–6°C in SST, possible incursion of Subantarctic Surface Waters and depression of the lysocline – an event that is apparently unique in the last 3 myr. The proximal geological record implies a second cooling step occurred at the Mid-Pleistocene Climate Transition (ca. 900 ka). Glacial-interglacial cycles, spanning the last 1 myr, recovered in drill cores from the Ross Sea and Prydz Bay are dominated by coarse-grained, ice-proximal diamicts of cold, polar subglacial affinity. Open-ocean pelagites and hemipelagites are rare on the shelf. The atmospheric temperature and greenhouse gas records from Antarctic ice cores show a pronounced increase in amplitude at the 100 ka periodicity, that is coherent and in phase with marine temperature records, but lags ice volume signal. The records of greenhouse gases from the Dome C ice core show very strong congruence with many features of the temperature record, and are consistent with CO2 in particular playing a significant role in temperature amplification. This also suggests that the Southern Ocean plays a leading role in controlling the atmospheric concentration of CO2 on glacial-interglacial time scales. Higher dust flux to Antarctica during glacial maxima compared with warm periods is considered to have radiative affects over Antarctica by providing nutrients (e.g. Fe) to the Southern Ocean promoting higher algal productivity and atmospheric CO2 drawdown. Most of the variation in total grounded ice volume in Antarctic glacial-cycle ice sheet models is due to expansion and contraction of grounding lines across continental shelves, mostly in the West Antarctic Ice Sheet (WAIS) Ross and Weddell sectors, and is equivalent to 15–20 m of sea level. The East Antarctic Ice Sheet interior responds in the opposite sense, contracting slightly at Northern Hemispheric glacial maxima due to lower model snowfall rates. In general millennial-scale cycles occur in high-resolution Antarctic ice cores, particularly in methane and temperature records. However, they appear lower amplitude and the timing may differ from their Northern Hemisphere counterparts (Dansgaard–Oeschger or D-O events). Although uncertainties between the age of the gas (methane) and the ice enclosing it in ice cores makes evaluation of the precise timing of inter-hemispheric, millennial-scale cycles difficult, some D-O cycles appear to be synchronous. The pattern of millennial-scale variability superimposed on G-I climate cycles in Antarctic ice cores occurs both during and prior to the last 100 kyr cycle. Millennial-scale Antarctic warm periods (A1–A7) are well expressed in SST, sortable silt and δ18O records from the southern mid-latitude ocean. One of these events, the Antarctic Cold Reversal, at 14.2–12.4 ka, is associated with an abrupt cooling of 2°C in the Southern Ocean, an expansion of ice shelves and sea ice, modest intensification of winds and intensification in deep abyssal inflow along eastern New Zealand through the Pacific gateway

Details