It is often assumed that the meltwater drainage system beneath the Antarctic Ice Sheet lacks efficient drainage channels that would decrease basal water pressure and reduce glacier sliding. Using an ensemble of subglacial hydrology simulations constrained by radar observations of locations of pooled water, we show that channelized drainage beneath Thwaites Glacier is extensive and substantially lowers friction at the bed of the glacier.
The presence of subglacial channels beneath the Antarctic Ice Sheet enhances sub-ice-shelf submarine melting and increases friction at the bed. Ice-sheet model projections of ice loss from Antarctica should be accounting for the existence of subglacial channels and their enhancement of these processes.
The configuration of meltwater at the base of glaciers is a primary control on friction restraining sliding of a glacier over its bed. It is often assumed that the Antarctic Ice Sheet does not support the development of efficient subglacial channels that would decrease basal water pressure and reduce glacier sliding due to the lack of surface meltwater supply to the base of the ice sheet. We tested this assumption using the MPAS-Albany Land Ice model to run an ensemble of over 130 subglacial hydrology simulations of Thwaites Glacier, Antarctica, across a wide range of physical parameter choices to assess the likelihood of channelization. We evaluated simulations by comparing modeled water thickness to observed radar specularity content, indicative of pooled water at the bed. In the data-compatible simulations, extensive, stable channel networks form; no simulations resembled observed specularity content when channelization was disabled. Our results suggest channelized subglacial drainage increases basal friction, while also amplifying submarine melting of the adjacent ice shelf that impedes glacier flow. The distribution of effective pressure and associated basal friction implied from our modeling differs from parameterizations typically used in large-scale ice sheet models, suggesting the development of more process-based parameterizations may be necessary.