VISIONS OF ICE SHEETS IN THE LATE CRETACEOUS GREENHOUSE WORLD

Backstripped eustatic estimates for the Late Cretaceous (99-65 Ma) of New Jersey show large (>25 m) and rapid (<1 m.y.) changes, requiring a glacioeustatic control. We reconcile records of warm high latitudes during the Late Cretaceous greenhouse world with glacioeustasy by proposing that ice sheets were restricted in area in Antarctica, ephemeral, and paced by Milankovitch forcing. Published modeling evidence indicates that a 5-10 x 106 km3 ice sheet could have developed when atmospheric CO2 fell below a critical threshold. This ice sheet would not have reached the Antarctic coast, hence explaining relative warmth in coastal Antarctica, but it would have significantly influenced sea level by up to ~25 m. A maximum of 25% of deep-sea oxygen isotopic increases of ~1‰ at ca. 71.2, 92–93, and 96 Ma may be attributed to ice (~25 m of eustatic lowering), with ~75% attributed to deep-water cooling of 3–4°C. Unlike the Oligocene and younger icehouse world, these ice sheets probably existed only during short intervals of peak Milankovitch forcing, and the continent was ice-free during much of the greenhouse Late Cretaceous to middle Eocene. Modeling studies predicted glacioeustatic falls from Milankovitch orbital solutions that are remarkably similar to those we obtained from the New Jersey margin. The alternative to invoking Late Cretaceous ice sheets is that global sea-level changes were paced by as yet undefined mechanisms, because none of the other hypothesized mechanisms (temperature effects, storages in lakes, deep-water changes, groundwater, or sea ice) can explain the observed 20-30 m changes in <1 m.y. Thus, because our data require that large, rapid sea-level variations occurred in the Late Cretaceous greenhouse world, we must conclude that either moderate-sized ice sheets paced sea-level changes during this time, or that our understanding of causal mechanisms for global sea-level change is fundamentally flawed.