Assessing changes in temperature, ice volume and ice sheet dynamics across the Middle Pleistocene Transition

Earth's climate for the last 2.5 million years has been characterized by the waxing and waning of large continental ice sheets in the Northern Hemisphere. Prior to about a million years ago, the glacial and interglacial cycles marched to the beat of the tilt of Earth's rotational axis (i.e., obliquity), which varied with a cycle of 41,000 years. About a million years ago, the length of glacial-interglacial cycles became longer, averaging about 100,000 years. This change, known as the Middle Pleistocene Transition (MPT), represents an important unsolved problem in paleoclimatology. Why did the frequency of glacial-interglacial cycles suddenly shift in the middle Pleistocene from 41- to 100-kyrs? The size of the ice sheets became larger across the MPT, thereby allowing individual domes of the North American ice sheet to coalesce to form a massive ice sheet. The great volume and thickness of this ice made it less sensitive to changes in boreal summer insolation, thereby lengthening the glacial cycle as the ice sheets survived through insolation maxima and deglaciation began skipping precessional (~21 kyr) and/or obliquity (~41 kyr) cycles. One of the hindrances to furthering our understanding of the MPT is the lack of a quantitative estimate of the change in ice volume that occurred across the MPT. This shortcoming results largely because the benthic oxygen isotope record reflects both temperature and ice volume changes. Using Mg/Ca of infaunal benthic species, we aim to deconvolve the benthic d18O signal and provide a quantitative estimate of changes in the d18O of seawater that is related to global ice volume. As the North American Ice Sheet increased in volume and thickness following the MPT, the dynamical behavior of the ice sheet may have changed. During the last glacial period, Laurentide Ice Sheet dynamics was dramatically expressed by so-called Heinrich layers, which consist of sediment rich in ice rafted debris (IRD) deposited in the North Atlantic stretching from Labrador to Portugal. Heinrich layers represent massive discharges of icebergs from surging of the Laurentide Ice Sheet in the region of Hudson Strait in northern Canada. Although Heinrich events have been studied for the last glacial cycle, little is known about their occurrence in older glaciations of the Pleistocene. At Site U1308 in the central North Atlantic, we found that the first Heinrich layer occurred at 650 ka coinciding with the onset of the 100-kyr cycle. It is not known, however, whether this represented the initiation of surging of the Laurentide Ice Sheet via the Hudson Strait Ice Stream or was it the first time icebergs produced by this process survived the transport to Site U1308. The former interpretation implies a fundamental change in Laurentide ice sheet dynamics before and after the MPT, whereas the latter interpretation indicates a change in atmospheric and/oror surface water conditions that changed the location of iceberg melting. Were Heinrich events limited to the 100-kyr world or did they also occur prior to the MPT? We will address this question by examining Site U1304 located to the north and west of Site U1308. Site U1304 should record Heinrich layers if they were produced prior to the MPT because the IRD belt was probably displaced to higher latitude when ice sheets were smaller and North Atlantic SST was warmer than late Pleistocene glaciations.