UC Santa Barbara

The sequence of feedbacks that characterized 100-kyr glacial cycles of the past million years remains uncertain, hampering an understanding of the interconnections between insolation, ice sheets, greenhouse gas forcing, and climate. Critical to addressing this issue is an accurate interpretation of the marine δ18O record, the main template for the Ice Ages. This study uses a global compilation of 49 paired sea surface temperature-planktonic δ18O records to extract the mean δ18O of surface ocean seawater over the past 800 kyr, which we interpret to dominantly reflect global ice volume. The results indicate that global surface temperature, inferred deep ocean temperature, and atmospheric CO2 decrease early during each glacial cycle in close association with one another, whereas major ice sheet growth occurs later in glacial cycles. These relationships suggest that ice volume may have exhibited a threshold response to global cooling, and that global deglaciations do not occur until after the growth of large ice sheets. This phase sequence also suggests that the ice sheets had relatively little feedback on global cooling. Simple modeling shows that the rate of ice volume change through time is largely determined by the combined influence of insolation, temperature, and ice sheet size, with possible implications for the evolution of glacial cycles over the past three million years.