Rutgers University, United States

During the late Pliocene and Pleistocene pronounced global cooling was accompanied by the growth of large northern hemisphere ice sheets. To explore the climate evolution of the region most impacted by this climate shift, we generated an orbitally-resolved, alkenone-derived, sea surface temperature (SST) record for the past 3.2 Myr at Deep Sea Drilling Project Site 607 (41 degrees N, 33 degrees W) in the North Atlantic Ocean. We compare our new record to existing Mg/Ca derived bottom water temperature (BWT) and stable isotope records from Site 607 and an SST time series from Ocean Drilling Program Site 982 (58 degrees N, 16 degrees W). All records show remarkable similarity in both long-term and orbital scale structure. Comparisons between the different temperature records shows a long-term cooling that is expressed as two distinct cooling events: one during the mid-Pliocene (2 degrees -3 degrees C) and the other during the early Pleistocene (1.5 degrees -2.5 degrees C). On orbital time scales, all records show the classic frequency shift from 41k to 100k cycles across the mid-Pleistocene transition (MPT). All time series are dominated by variability in the 100k band. However, significant 41k power, evidently strongly modulated by obliquity forcing, is also present in each record. During the Pliocene and early Pleistocene, all North Atlantic climate signals were nearly in phase with benthic oxygen isotope variations in the 41k band, suggesting a common northern hemisphere imprint. In contrast, across the MPT, BWT variations in the 100k band developed a significant lead over all other proxies, which were approximately in phase with benthic delta (super 18) O. This difference in phasing between BWT and other proxies suggests that the source of the BWT signal may have shifted from the Northern Hemisphere to the Southern Hemisphere. This change in signal origin is corroborated by the observation that other existing Southern Hemisphere temperature records show similar late Pleistocene phase leads in the 100k band. Overall, our findings suggest a fairly coherent North Atlantic climate response to northern hemisphere glaciation. Our paleoclimate proxies bear a predominantly Northern Hemisphere imprint during the "41k world." However, phase relationships suggest the emergence of a Southern Hemisphere influence on BWT during the transition to the "100k world" of the late Pleistocene.