Antarctic ice-sheet sensitivity to obliquity forcing enhanced through ocean connections

Abstract

Antarctic ice sheet (AIS) growth and decay is strongly influenced by astronomical variations, yet it is not known why AIS response to this climate driver varies through time. Here we examine AIS variability from 34 to 5 million years ago through integration of geological records from the Antarctic margin and a novel assessment of sensitivity to changes in Earth's axial tilt (obliquity sensitivity) derived from the oceanic oxygen-isotope proxy for global ice volume. Three phases of AIS development are found: (1) ~34-24 Ma-a largely terrestrial ice sheet with low obliquity sensitivity; (2) 24 to 14 Ma-frequent ephemeral marine ice sheets with amplified obliquity sensitivity; and (3) 14 to 5 Ma-episodes of extensive marine ice sheet advance, persistent sea ice, and a general decrease in obliquity sensitivity. These phases are associated with decreasing atmospheric CO2 and progressively colder mean climate states. Our analysis suggests the AIS is most sensitive to obliquity forcing when it extends into marine environments and sea-ice extent is limited. We infer this is due to obliquity-driven changes in meridional temperature gradient that affect the position and strength of circum-Antarctic easterly flow, which enhances (or reduces) ocean heat transport across the Antarctic continental margin. Insight into causes of Antarctic ice sheet variability-over a range of time scales-is fundamental to our understanding of Earth system response to climate change. Large scale shifts in AIS volume and extent are controlled by changes in atmospheric CO2 1,2 and plate

Midlatitude Temperature Variations in the Oligocene to Early Miocene

“…These ice‐proximal records suggest that the ice sheet was principally extending into the marine realm during this interval and was at some points marine‐grounded. Benthic δ 18 O records have low sensitivity to obliquity forcing during this time interval (Levy et al, ). Subsequently, the warmer average North Atlantic temperatures between 24.5 and 21 Ma correspond with Ross Sea sedimentation dominated by nonglacially deposited sandstones and mudstones, interpreted as period when the ice sheet was predominantly land‐based.…”

“…A large variation in benthic δ 18 O, in the absence of appreciable midlatitude SST change, could reflect appreciable (>3‐fold) polar amplification of SST change in deep‐water formation areas if ice volume were constant, and/or ice volume as a significant component of benthic δ 18 O. This period is marked by low sensitivity of megasplice benthic δ 18 O to obliquity forcing (Figure b), which is interpreted to reflect a state when ice sheets responded to direct insolation forcing rather than regional temperature (Levy et al, ). At the same time, the amplitude of benthic δ 18 O on short eccentricity timescale was also low (Liebrand et al, ).…”

“…(a) Sea surface temperature from Site 1406A and its 16th and 84th confidence intervals. (b) Obliquity sensitivity (Levy et al, ) from obliquity forcing variance of benthic δ 18 O. (c) The blue line shows Antarctic proximal environmental motifs from Levy et al () for the middle Miocene from ANDRILL cores; the black lines show disconformities.…”

“…In the ~1 Myr prior to and following the O/M boundary (from 24.5 to 22 Ma)—an interval encompassing both warmer temperatures and lower average benthic δ 18 O than the preceding Oligocene—the larger amplitude (5°C) variation in midlatitude SST is coupled to a 1‰ range in benthic δ 18 O, with the warmest temperatures coinciding with minima in benthic δ 18 O. The Oligocene Miocene transitionitself is marked by a two‐step decrease in North Atlantic SST, analogous to the two‐step rise in high resolution benthic δ 18 O in orbitally resolved records (Liebrand et al, ; Mawbey & Lear, ), the pacing of which has been recently suggested to be driven by both changes in atmospheric carbon dioxide and obliquity (Greenop et al, ; Levy et al, ). Across the O/M boundary, previous high‐resolution analysis of a tropical Atlantic site between 22.4 to 23.7 Ma showed that intervals of increased δ 18 O sw interpreted as recording ice growth were coupled to intervals of bottom water cooling, inferred from benthic foraminiferal Mg/Ca (Mawbey & Lear, ).…”

Glacial sequence stratigraphy of ANDRILL‐1B core reveals a dynamic subpolar Antarctic Ice Sheet in Ross Sea during the late Miocene

“…Previous studies suggest links with orbital cycles during the Plio–Pleistocene (Naish et al ., ; Patterson et al ., ), whereas early to middle Miocene glacial cycles may be driven by a combination of orbital cycles and complex carbon‐cycle feedbacks (Gasson et al ., ; Levy et al ., ). The dominant late Miocene Antarctic climate driver is thought to be obliquity (Levy et al ., ), however the sedimentary record remains fragmented. This study may provide a few insights: (i) glacial sequences are likely to represent very short time intervals, but some may be linked to obliquity cycles within their respective interpreted polarity reversal zones (Fig.…”