An 800-kyr Record of Global Surface Ocean δ18Osw and Implications for Ice Volume-Temperature Coupling

Thursday, 18 December 2014: 8:45 AM
Jeremy D Shakun, Boston College, Chestnut Hill, MA, United States, David W Lea, UCSB, Santa Barbara, CA, United States, Lorraine E Lisiecki, University of California Santa Barbara, Santa Barbara, CA, United States and Maureen E Raymo, Lamont-Doherty Earth Obs., New York, NY, United States
We use 49 paired sea surface temperature (SST)-planktonic δ18O records to extract the mean δ18O of surface ocean seawater (δ18Osw) over the past 800 kyr, which we interpret to dominantly reflect global ice volume, and compare it to SST variability on the same stratigraphy. This analysis suggests that ice volume and temperature contribute to the marine isotope record in ~60/40 proportions, but they show consistently different patterns over glacial cycles. Global temperature cools early during each cycle while major ice sheet growth occurs later, suggesting that ice volume may have exhibited a threshold response to cooling and also had relatively little feedback on it. Multivariate regression analysis suggests that the rate of ice volume change through time is largely determined by the combined influence of orbital forcing, global temperature, and ice volume itself (r2 = 0.70 at zero-lag for 0-400 ka), with sea level rising faster with stronger insolation and warmer temperatures and when there is more ice available to melt. Indeed, cross-spectral analysis indicates that ice volume exhibits a smaller phase lag and larger gain relative to SST at the 41 and 23 kyr periods than at the 100 kyr period, consistent with additional forcing from insolation at the obliquity and precession time scales. Removing the surface ocean δ18Osw signal from the global benthic δ18O stack produces a reconstruction of deep ocean temperature that bears considerable similarity to the Antarctic ice core temperature record (r2 = 0.80 for 0-400 ka), including cooler interglacials before 400 ka. Overall, we find a close association between global surface temperature, deep ocean temperature, and atmospheric CO2. Additionally, we find that rapid cooling precedes the gradual buildup of large continental ice sheets, which may then be instrumental in terminating the cycle.