LAUSR

Abstract The Quaternary paleoenvironmental history of the Arctic Ocean remains uncertain, mainly due to the limited chronological constraints, especially beyond the 14C dating limits of accelerator mass spectrometry (AMS). The difficulty in establishing reliable chronostratigraphies is mainly attributed to low sedimentation rates and diagenetic sediment changes, resulting in very poor preservation of microfossils and altered paleomagnetic records. In the absence of independent chronostratigraphic data, the age model of Pleistocene sediments from the Arctic Ocean is mainly based on cyclostratigraphy, which relates lithologic changes to climatic variability on orbital time scales. In this study, we used the Mn/Al record measured from the sediment core ARA03B-41GC retrieved from the Makarov Basin in the western Arctic Ocean. The Mn/Al variation was tuned to the global benthic oxygen isotope stack (LR04) curve under different assumptions for computational correlation. Regardless of assumptions, our computational approach led to similar ages of about 600–1,000 ka for the bottom part of the core. These age models were up to about 200 ka older than those derived from lithostratigraphic approaches. Interestingly, our new age models show that the Ca/Al peak, a proxy for a detrital input from the Laurentide Ice Sheet, first occurred about 150 ka earlier than those previously proposed. Therefore, our results suggest that the glaciers in northern North America developed more extensively at about 810 ka than in earlier glacial periods, and influenced the sedimentary and paleoceanographic environments of the Arctic Ocean much earlier than previously thought. In order to establish a more comprehensive age model, more work is needed to validate our findings with different sediment cores recovered from the western Arctic Ocean.