Workshop on Cretaceous Climate and Ocean Dynamics

July 14-17, 2002

Florissant, Colorado, USA


What would a Cretaceous glaciation look like?

Author:Karen L Bice
Date Submitted:06/20/2002
Address:Mail Stop 23 Woods Hole
Co-Authors:Pollard, Dave, Penn State University,; Mathieu, Renaud, GEO-RS, Colomiers; Norris, Richard D., Woods Hole Oceanographic Institution,
Affiliation:Woods Hole Oceanographic Institution
Abstract URL:
Keywords:ice glaciation model oxygen isotope sea level eustatic ice-sheet
Abstract:The Cretaceous has traditionally been thought of as an interval during which no continental polar ice-sheets existed, based on quantitative and semi-quantitative evidence for surface ocean and coastal continental temperatures much warmer than the modern in polar regions, as well as a lack of obvious ice-rafted sediments. Although the reconstruction of Cretaceous global sea level change is imperfect and debate over any one feature is common, rapid global sea level changes on the order of 10 to 70 meters are typical of Cretaceous reconstructions. Yet, true eustatic sea level change of 10-70 m on the time scale of tens to hundreds of thousands of years is difficult to explain in the absence of ice volume changes as a mechanism. Now, a growing body of literature exists that suggests that the assumption of a Cretaceous "greenhouse" earth free of continental ice should be re-examined. What is the likely isotopic composition of an ice-sheet formed during intervals of Cretaceous climate conditions and at what size is such an ice-sheet likely to produce a significant shift in seawater isotopic composition?

Using a version of the GENESIS atmospheric general circulation model (GCM) modified to include water isotopes ( and D/H), we have performed two preliminary experiments with mid-Cretaceous boundary conditions. A run with 900 ppm CO2 produces SSTs consistent with the coolest upper ocean temperatures inferred from late Albian-Turonian planktonic oxygen isotope ratios. The GCM run with 4500 ppm CO2 matches maximum inferred upper ocean temperatures. Increasing atmospheric CO2 and decreasing the equator-to-pole surface temperature gradient reduces the gradient in precipitation substantially compared to the modern. Tropical freshwater is 2-3 (SMOW) more depleted than the modern and high latitude winter precipitation is 6 (at 60° lat.) to 24 (at 90° lat.) heavier than modern. Winter precipitation falling over Antarctica has a mean composition of ~ -14 to -20 SMOW (compare to the modern ~ -50).

These preliminary results have at least two important ramifications for interpreting warm Cretaceous climate records:
1) If the mean isotopic composition of ice that formed at high latitudes was no less than -20 , then an ice volume equivalent to 60% of the modern Antarctic ice-sheet would produce only a 0.2 shift in the mean ocean dw value. This means that a substantial volume of ice could form/decay without being detectable in benthic records. In the sequence of events reflected in the record during ice growth, eustatic sea level fall may therefore be the only "given" and is, indeed, generally accepted as having occurred through the Cretaceous.
2) A lower latitudinal gradient in freshwater would result in a lower gradient in upper ocean dw, relative to the modern. Use of the latitudinal correction for dw based on modern GEOSECS data (Zachos et al., 1994) has therefore consistently overestimated (by a few degrees) tropical upper ocean temperatures and dramatically underestimated high latitude temperatures. This seems to exacerbate the problem of Cretaceous polar ice growth.