A Rapid in Time

Ice sheets tend to collapse in response to unusually large maxima in insolation forcing that result from the coincidence of high obliquity and alignment of perihelion with Northern Hemisphere summer solstice, consistent with the models hypothesized by Milankovitch¹ and others^{5, 6, 7, 8, 9, 10}. During these forcing maxima, summer insolation is as much as 40 W m⁻² greater at high northern latitudes (Fig. 3b). However, this consistency is not exclusive of all other orbital contributions to deglaciation. For instance, when perihelion aligns with the Northern Hemisphere summer solstice, aphelion occurs during the Southern Hemisphere summer, causing the length of the Southern Hemisphere summer to be longer (Fig. 3b) and, possibly, increasing the escape of CO₂ from the Southern Ocean into the atmosphere^{26, 27, 28, 29}. The climate system is thoroughly interconnected across temporal and spatial scales, and, just as neither obliquity nor precession act in isolation, no one region should be expected to exert exclusive influence upon deglaciation.

A long-standing hypothesis for ice-sheet synchronization invokes sea-level forcing of Antarctic grounding lines driven by fluctuations of NH ice sheets (41, 42), but until now the chronology of the Antarctic ice sheets has been too limited to evaluate this hypothesis, other than for the deglaciation where existing arguments for a 4- to 5-ky lag relative to the start of deglacial sea-level rise (5, 8, 23) would appear to contradict it. Where dating constraints for onset of the [local Last Glacial Maximum] exist, however, they support a sea-level forcing in placing the associated Antarctic margins at their maximum extent when global sea level was approaching or first reached its LGM lowstand (Fig. 3). In particular, we suggest that NH ice-sheet growth that occurred in response to decreases in insolation and Pacific SSTs (9, 30) caused the global mean sea level to fall, allowing Antarctic ice margins to advance across the continental shelf and reach their maximum extent. At the same time, the reduction in NADW formation (Fig. 3) and attendant heat flux would further contribute to advance of Antarctic marine margins.

The subsequent onset of NH deglaciation ~19 ka in response to boreal summer insolation forcing caused an initial rapid global mean sea-level rise of ~5 to 10 m (Fig. 3) (9, 43, 44). Although this sea-level forcing may explain the contemporaneous retreat of Antarctic grounding lines in the Weddell Sea and, perhaps, Amundsen Sea regions, the lack of a response at other dated Antarctic marine margins appears inconsistent with this hypothesis. This spatial variability in response may reflect different geometries of ice shelves, variations in subshelf pinning points, or differences in sedimentary wedges (size and stiffness) that stabilize grounding lines to a rapid sea-level rise of this magnitude (45). In addition, sea-level calculations indicate that gravitational, deformation, and rotational effects associated with the initial melting of NH ice ~19 ka caused enhanced sea-level rise around the Weddell and Amundsen Seas relative to eustatic, whereas it was equal to or less than eustatic around the Ross Sea and Mac Robertson Land margins (Fig. 4) (SOM). This regional enhancement may have been sufficient to overcome the stabilizing effect provided by sedimentary wedges at grounding lines in the Weddell and Amundsen Seas (45), causing early retreat, whereas grounding lines elsewhere remained immune to the lesser sea-level rise.

The decline in the partial pressure of atmospheric carbon dioxide during the [Eocene-Oligocene] climate event was substantial, but absolute CO₂ concentrations depend on the value of ε_f applied. Collectively, CO₂ estimates calculated by using U₃₇^K’ and TEX₈₆ SST estimates and a range of εf values indicate that CO₂ decreased ~40% from 35.5 to 32.5 million years ago (SOM). Application of reasonable εf values (25 to 28‰) indicates that the partial pressure of atmospheric CO₂ fell from 1200 to 1000 ppm to 700 to 600 ppm. Interestingly, the change in CO₂ determined from this study, as well as the boron-isotope methodology (11), is consistent with model estimates for a threshold CO₂ level required for rapid Antarctic glaciation (8, 29).

We conclude that the available evidence supports a fall in CO₂ as a critical condition for global cooling and cryosphere evolution ~34 million years ago. Whether CO₂ acted alone to cause the E-O transition or whether a threshold CO₂ level in combination with favorable orbital configurations (1) ultimately triggered glaciation cannot be determined from our results. However, during the E-O transition both CO₂ decline and enhanced ice albedo account for global temperature changes. Lastly, the long-term permanence of the CO₂ decline (10) and the impermanent inorganic carbon isotope shift (1) implicate the role of silicate weathering rates over the influence of short-term organic-carbon burial rates as the primary cause for long-term change in atmospheric carbon dioxide.