Ice-stream flow switching during deglaciation of the southwestern Barents Sea

Abstract

Ice streams dominate the discharge of continental ice sheets. Recent observations and reconstructions have revealed that large-scale reorganizations in their flow trajectory (flow switching) can occur over relatively short time scales. However, the underlying causes of such behavior, and the extent to which they are predictable, are poorly known. This paper documents a major episode of ice-stream flow switching during the late Weichselian deglaciation of the southwestern Barents Sea and explores various hypotheses for its causation. Regional bathymetric data show that two ice streams that had similar, adjoining, topographically constrained source areas had very different trajectories and dynamics on the outer shelf. At the late Weichselian maximum, the Håkjerringdjupet ice stream flowed westward along the cross-shelf trough of Håkjerringdjupet, while the Sørøya Trough ice stream flowed northward into Ingøydjupet, forming a tributary of the Bjørnøyrenna ice stream. Initial retreat of the Håkjerringdjupet ice stream was rapid but with episodic periods of grounding. As it retreated onto the higher, rougher topography of the inner shelf, we infer a reduction in ice velocity and a dramatic decrease in the pace of retreat, as recorded by nested sequences of recessional moraines. Following (and probably in response to) this, we suggest that there was a short-lived surge/readvance of an adjacent lobe onto Fugløybanken. In contrast, the adjacent Sørøya Trough ice stream remained active throughout deglaciation, before retreating rapidly, with no stillstands or readvances. We argue that the different retreat histories of the ice streams were determined by variations in bed topography/bathymetry, which modulated the grounding line response to sea-level variation. Such a mechanism is likely to be an important control on the long-term behavior of marine-based ice streams and outlet glaciers in Antarctica and Greenland and suggests that gathering data on their subglacial topography should be a priority.