Glacial terminations in the oxygen isotope record of deep sea cores: hypothesis of massive Antarctic ice-shelf destruction

Abstract

Glacial Terminations I and II in the deep-sea core foraminiferal oxygen isotope ratios, though not similar in form, are each characterized by a short interval of about 3 ka during which a large negative change occurred in the ratios. After correcting for temperature effects, we estimate that each change was equivalent to a glacial disintegration capable of causing a 50–60 m eustatic sea-level rise. In the Termination I interval, 16-13 ka B.P., the estimated equivalent sea-level rise is much larger than that suggested by the best evidence of deglacial ice loss from the Laurentide Ice Sheet (which is thought to dominate Pleistocene glacial ice volume changes). Moreover, in Termination II, about 125 ka B.P., the estimated equivalent sea-level rise is much larger than that inferred from the stratigraphy of coral reefs and beaches on New Guinea and other locations which suggests quite high sea-levels for 10 ka prior to this termination. To reconcile these severely conflicting lines of evidence, we propose that the rapid negative isotope ratio changes were largely caused by the disintegration of massive, floating Antarctic ice-shelves which masqueraded as land ice in the oceanic isotope ratios, but which contributed no sea-level change. We propose that such shelves were formed when oceanic circulation changes in the North Atlantic under glacial conditions greatly reduced the formation of relatively warmer North Atlantic Deep Water and its injection into circumpolar Antarctic regions, and that the shelves disintegrated rapidly with the resumption of large scale formation of this deep water and its input into the Antarctic. We find that high sea-levels in the 125–135 ka B.P. interval prior to Termination II (which imply very little Northern ice then) are well supported by dated coral reef stratigraphy, although these ages were not accepted by the CLIMAP study of the last interglacial ocean. Minimal Northern ice volumes at that time imply that strong Milankovitch summer insolation is not always required for Northern deglaciation. A promising additional deglacial mechanism is the prolonged viscous depression of the crust beneath the Laurentide Ice Sheet combined with atmospheric—oceanic circulation factors.