Correcting the Cenozoic δ18O deep-sea temperature record

The oxygen isotope signal in benthic foraminifera from deep-sea cores is mainly determined by deep-ocean temperature and land ice volume. Separating the temperature and ice volume signals is a key step in understanding the evolution of Cenozoic climate. Except for the last few million years, fluctuations in land ice volume were determined mainly by the size of the Antarctic ice sheet. Here a simplified Antarctic ice-sheet model and composite deep-sea oxygen isotope records are used to derive a deep-sea temperature history for the Cenozoic. The ice-sheet model is forced with the oxygen-isotope record, which is instantaneously corrected for the calculated ice volume. Therefore, the resulting deep-sea temperature and Antarctic ice volume curves are mutually consistent. The temperature record of Lear et al. [Science 287 (2000) 269], derived from the Mg/Ca ratio in benthic calcite, is also used to force the ice-sheet model. The resulting ice-volume does not agree well with the ice-volume history inferred by Lear et al. from a comparison of Mg/Ca and δ18O records. The most important points emerging from this study are: The Cenozoic δ18O record from benthic forams is a truly mixed signal (ice volume and deep-sea temperature contribute equally to the variance). For the period of 45 to 30 Ma ago, δ18O temperatures are significantly lower than Mg/Ca temperatures. Rapid nonlinear build-up of a continental-scale ice sheet on the Antarctic continent can easily occur for a linear cooling rate.It is unlikely that the method of Lear et al. provides a reliable measure of ice volume. The cooling rate at the beginning of the Pliocene is much larger in the δ18O temperature curves than in the Mg/Ca temperature curve