a b s t r a c tThe hypothesis of floating ice shelves covering the Arctic Ocean during glacial periods was developed in the 1970s. In its most extreme form, this theory involved a 1000 m thick continuous ice shelf covering the Arctic Ocean during Quaternary glacial maxima including the Last Glacial Maximum (LGM). While recent observations clearly demonstrate deep ice grounding events in the central Arctic Ocean, the ice shelf hypothesis has been difficult to evaluate due to a lack of information from key areas with severe sea ice conditions. Here we present new data from previously inaccessible, unmapped areas that constrain the spatial extent and timing of marine ice sheets during past glacials. These data include multibeam swath bathymetry and subbottom profiles portraying glaciogenic features on the Chukchi Borderland, southern Lomonosov Ridge north of Greenland, Morris Jesup Rise, and Yermak Plateau. Sediment cores from the mapped areas provide age constraints on the glaciogenic features. Combining these new geophysical and geological data with earlier results suggests that an especially extensive marine ice sheet complex, including an ice shelf, existed in the Amerasian Arctic Ocean during Marine Isotope Stage (MIS) 6. From a conceptual oceanographic model we speculate that the cold halocline of the Polar Surface Water may have extended to deeper water depths during MIS 6 inhibiting the warm Atlantic water from reaching the Amerasian Arctic Ocean and, thus, creating favorable conditions for ice shelf development. The hypothesis of a continuous 1000 m thick ice shelf is rejected because our mapping results show that several areas in the central Arctic Ocean substantially shallower than 1000 m water depth are free from glacial influence on the seafloor.

No abstract

No abstract

X-ray fluorescence (XRF) scanning of sediment cores from the Lomonosov Ridge and the Morris Jesup Rise reveals a distinct pattern of Ca intensity peaks through Marine Isotope Stages (MIS) 1 to 7. Downcore of MIS 7, the Ca signal is more irregular and near the detection limit. Virtually all major peaks in Ca coincide with a high abundance of calcareous microfossils; this is particularly conspicuous in the cores from the central Arctic Ocean. However, the recorded Ca signal is generally caused by a combination of biogenic and detrital carbonate, and in areas influenced by input from the Canadian Arctic, detrital carbonates may effectively mask the foraminiferal carbonates. Despite this, there is a strong correlation between XRF-detected Ca content and foraminiferal abundance. We propose that in the Arctic Ocean north of Greenland a common palaeoceanographic mechanism is controlling Ca-rich ice-rafted debris (IRD) and foraminiferal abundance. Previous studies have shown that glacial periods are characterized by foraminfer-barren sediments. This implies that the Ca-rich IRD intervals with abundant foraminifera were most likely deposited during interglacial periods when glaciers left in the Canadian Arctic Archipelago were still active and delivered a large amount of icebergs. At the same time, conditions were favourable for planktic foraminifera, resulting in a strong covariance between these proxies. Therefore, we suggest that the XRF scanner's capability to efficiently map Ca concentrations in sediment cores makes it possible to systematically examine large numbers of cores from different regions to investigate the palaeoceanographic reasons for the calcareous microfossils’ spatial and temporal variability

No abstract