GSA Annual Meeting in Seattle, Washington, USA - 2017

Paper No. 325-13
Presentation Time: 4:45 PM


CHRISTIANSON, Knut1, HOLSCHUH, Nicholas D.2, PADEN, John3, SPRICK, Jordan3, PETERS, Leo4, ANANDAKRISHNAN, Sridhar5 and ALLEY, Richard B.6, (1)Earth and Space Sciences, University of Washington, Seattle, WA, (2)Earth and Space Sciences, University of Washington, Johnson Hall Rm-070, Box 351310, University of Washington, Seattle, WA 98195-1310, (3)Electrical Engineering and Computer Science, University of Kansas, Lawrence, KS 66046, (4)Centre for Oceans and Cryosphere, University of Tasmania, Hobart, (5)Dept. of Geosciences and EESI, Pennsylvania State University, 442 Deike Bldg, University Park, PA 16802-2711, (6)Department of Geosciences, The Pennsylvania State Univ, Deike Builiding, University Park, PA 16802,

Deglaciated landscapes, whether subaerial or submarine, are often host to a rich panoply of subglacial landforms, such as drumlims, crags, megascale glacial lineations, grounding-line wedges, deep meltwater channels, and more. These landforms are formed and shaped by interactions between the ice and underlying substrate, and thus have implications for the flow of the overlying ice. Robust interpretations of the relationship between the ice and its substrate based on subglacial landforms that remain after deglaciation have been inhibited by a dearth of high-resolution observations of currently glaciated subglacial landscapes, where ice flow speed is known and where subglacial conditions can be ascertained using geophysical methods. Past direct observations of landforms under currently fast-flowing ice have been limited to a few ice streams, where relatively homogeneous, thick dilatant till layers may favor formation of specific subglacial features, i.e., megascale glacial lineations and grounding-zone wedges. Here we present two detailed gridded subglacial topographies, obtained from ice-penetrating radar measurements, from Thwaites Glacier, West Antarctica, where ice flows over a highly variable bed (in both topography and model-inferred basal shear stress). One grid is located ∼170 km downstream from the ice divide where ice is moving ∼100 m/yr. Here the ice advects over a broad basin and then flows into a subglacial ridge (of several hundred meters amplitude) oriented orthogonally to flow. A deep canyon (~400 m) that cuts through this ridge in roughly the ice-flow direction and relatively soft sediments on the downstream side of the basin (immediately upstream of the canyon) suggest that a large subglacial lake may have formed in this location and drained catastrophically, as has been hypothesized as the formation mechanism for the deep canyons observed on the Amundsen Sea continental shelf. Numerous multiscale glacial lineations are also observed in the subglacial basin. The second grid is located ∼300 km downstream of the ice divide where the ice is moving ∼350 m/yr. A large crag and even more extensive multiscale subglacial lineations are observed in the downstream grid. Our results suggest that multiple subglacial landforms form in close geographic proximity due to heterogeneous basal conditions.