Beyond the Last Ice Age: Long-Term Records of Millennial-Scale Variability in the North Atlantic

J. F. McManus, D. W. Oppo, (Woods Hole Oceanographic Institution,Woods Hole, MA 02543;;; J. Cullen (Salem State College, Salem MA, 01970); and D. Hodell (University of Florida, Gainesville, FL 32611)

Deep drilling on abyssal sediment drifts in the North Atlantic Ocean

allows the construction of long, continuous, high resolution climate

records from locations that have been influenced by changes in surface

hydrography, thermohaline circulation, and the size and dynamics of

northern ice sheets. We have examined cores from ODP Sites 980, 983, and 984 for evidence of these changes over the last 0.5 million years. Benthic foraminiferal d18O and ice-rafting indices were used to infer ice sheet conditions. Planktonic foraminiferal d18O (G. bulloides, N. pachyderma sinistral, N. pachyderma dextral) was used to infer sea-surface temperature (SST). Epifaunal benthic foraminiferal d13C (primarily from Cibicidoides wuellerstorfi) was used to infer changes in deep circulation. The benthic d18O at these sites is dominated by variability on 'Milankovitch' time scales (10^4-10^5 years). All other indicators display significant and persistent higher frequency (10^3-10^4) variability. Repeated ice-rafting episodes and SST cycles are modulated by ice-sheet size, resulting in interglaciations that are relatively stable and glaciations that are characterized by catastrophic iceberg discharges and rapid 30-60C SST oscillations. Thermohaline circulation is variable on similar time scales, with less pronounced amplitude modulation in sum over the millennial frequency band. Taken together, the evidence suggests that an interraction between internal and external forcing is responsible for the observed behavior. Orbitally driven changes in ice sheet size appear to be responsible for amplifying the regional climate response, while an internal thermohaline oscillator that is less reponsive overall to ice volume may play a role in pacing the variability on millennial time scales.