C53A-0276:
Flow of Ice near a Large Melt Channel in the Ross Ice Shelf

Friday, 19 December 2014
Howard Conway1, Richard C A Hindmarsh2, Michelle R Koutnik1, Christopher Stevens3 and Jeremy Paul Winberry4, (1)Univ Washington, Seattle, WA, United States, (2)NERC British Antarctic Survey, Cambridge, CB3, United Kingdom, (3)University of Washington, Seattle, WA, United States, (4)Central Washington University, Ellensburg, WA, United States
Abstract:
There is increasing evidence for the existence of large channels incised beneath ice shelves in West Antarctica. However, the genesis of these channels is still not clear. Measurements from the floating tongue of Pine Island Glacier show that sub-shelf channels formed by localized melting (Stanton et al., 2013); oceanographic models of flow in sub-ice-shelf cavities are capable of producing sub-shelf channels through non-linear feed-back mechanisms (Gladish et al. 2012; Sergienko, 2013). On the other hand, other evidence suggests that sub-shelf channels can initiate at locations where melt water beneath grounded ice crosses the grounding line (Le Brocq et al. 2013).

Here we use ground-based radar to map a channel incised into the base of the Ross Ice Shelf proximal to the grounding zone of Beardmore Glacier, West Antarctica. Results show that in the grounding zone (ice thickness is 1150m), the channel is 100m deep and 0.5 km wide. On the shelf, the channel widens and deepens down stream; 8km down from the grounding line (ice thickness is 600-800 m), the channel is 200m deep and 1 km wide (see Figure); apparently in this case, the channel originates from beneath the grounded ice.

Although there is some indication of radar-detected internal stratigraphy dipping toward the channels indicative of meltin, the pattern of the stratigraphy suggests significant accumulated strain (see Figure). Further, vertical strain-rates calculated from repeat measurements using phase-sensitive radar (pRES), show a more complex pattern over melt-channels than is usually assumed for ice shelves. Specifically, 57 of the 69 pRES measurements exhibit a pattern of compression in the upper 75% of the ice column and extension in the lower 25%. This observation is important because such a pattern affects the ratio of surface- to mean-velocity, which is usually assumed to be unity in shelf-mass balance calculations.

FIGURE CAPTION: Unmigrated, across-channel radar profile 8km downstream from the grounding line. The vertical reflector at km-2 comes from a shot hole drilled for a seismic survey. Here, the shelf is generally ~800m thick, but at km-2 a channel that is 200m deep and 1km wide is incised into the base of the shelf. The off-axis reflectors below 800m indicate that the ice-ocean interface is rough.