Emplacement and fluctuations of the Antarctic Ice Sheet across the Eocene-Oligocene transition

Abstract:

A ~1.5‰ positive increase in deep-sea oxygen isotopic values marks the establishment of a continental-scale Antarctic Ice Sheet (AIS) at the Oi-1 event (~33.6 Ma) across the Eocene-Oligocene transition (EOT). Quantifying the magnitude and timing of EOT AIS volume variations by means of δ18O records is hampered by the fact that the latter reflect a coupled signal of temperature and ice-sheet volume. Furthermore, bathymetric variations based on marine geologic sections are affected by large uncertainties and, most importantly, reflect the local response of relative sea level (rsl) to ice volume fluctuations rather than the global eustatic signal. AIS proximal and Northern Hemisphere (NH) shallow marine settings show, in fact, an opposite trend of rsl change across the EOT. In particular, low-latitude NH sequences record a 70 ± 20m rsl fall, suggesting that AIS volume could have expanded to either near modern dimensions or as much as 25% larger than present day. Conversely, sedimentary facies from shelf areas in the proximity of the AIS margins show that local rsl rose up to ~150 m during the EOT glaciation. In this work we reconcile near- and far-field rsl sites by solving the gravitationally self-consistent Sea Level Equation that accounts for Glacial Isostatic Adjustment (GIA). Our GIA simulations show that the deviations from eustasy significantly increase towards Antarctica. Accordingly, the cyclo-chronological record of sedimentary cycles retrieved from Cape Roberts Project Drillcore CRP-3 indicates a deepening across the EOT. CRP-3 record also shows that full glacial conditions only occurred at the Oi-1a event (32.0 Ma). The strong gradients of rsl change around AIS result in heterogeneous bathymetric variations that might affect the circumpolar ocean flow. In fact, our novel GIA-forced ocean current model predicts along-meridian shifts of current frontal patterns of the order of several degrees. Also, differences in current velocity are locally more than 100%, and the zonal transport decreases in mean and variability. Our modeling results therefore suggest that GIA-induced ocean flow variations alone could have significant impacts on local nutrient variability as well as on erosion and sedimentation rates and ocean heat transport.