Roosevelt Island in the eastern Ross Sea (Fig.1) is an ideal location for investigating the deglaciation of the Ross Sea Embayment following the last glacial maximum (LGM). Fieldwork by us during 1997-1998 focused on making geophysical measurements in order to understand the glacial history of the island. An AVHRR image (Fig. 1) guided our field work, which included: radio-echo sounding to map ice thickness and internal stratigraphy; global positioning system (GPS) measurements of surface topography; snow pits, and 16-m firn cores to measure spatial patterns of temperature, density and accumulation.
Figure 1. A: Map showing Roosevelt Island and dated locations used to resolve Holocene grounding-line retreat in the Ross Sea adapted from (Conway et al., 1999). The detailed structure of past grounding-line positions is not known; dashed lines show the simplest pattern consistent with the dated evidence. B: AVHRR image (from Scambos et al 1998) shows locations of radar profiles measured in 1997-98 (T-line, A-line, B-line and Z-line). The T-line crosses the summit (79.39o S, 161.46o W).
Results show that high accumulation (0.25m/yr) 12km NE of the crest decreases almost linearly to 0.10m/yr 20km SW of the crest. At the summit, accumulation is 0.21m/yr ice equivalent, and 10-meter firn temperature is -23.4oC. Radar-detected internal stratigraphy across the summit is smoothly varying (Fig. 2A); there is no evidence of disturbed internal layering or diffractors that are indicative of high strain rates or buried crevasses. Of special interest is the distinctive pattern of layering beneath the divide. Matching the observed layer pattern with an ice-flow model (Fig. 2B) indicates: (i) the divide has been thinning ~0.9m/yr for at least 3kyr; (ii) the location of the divide was remarkably stable during that time.
Figure 2. A: Radar-detected internal stratigraphy across the T-line (see Fig.1). View is looking toward the northwest. The bright reflector ~200m below sea level is the bed. The profile crosses the summit where ice thickness is 744m. The distinctive arch in layering beneath the divide can be used to calculate recent changes in ice thickness; B: Comparison of the measured and modeled bump-amplitude profile indicates that Roosevelt Island has been thinning ~0.9m/yr for at least 3,000 yrs.
In collaboration with colleagues from University of Maine, we used the measurements and models to help constrain grounding-line retreat in the Ross Sea Embayment during the Holocene.
Acknowledgements
Now we would like to extract a new ice core from the summit of Roosevelt
Island. We propose to determine the age-depth relationship through measurements of
stable isotopes, occluded methane, and visual stratigraphy. We will also
measure depth profiles of borehole temperature and sonic velocity (a proxy
for ice fabric), and the modern thinning rate near the summit;
Our idea is to use these observations and ice-flow models to determine
the configuration and thickness of ice in the Ross Sea Embayment during
the last glacial maximum, and the timing and pattern of deglaciation.
We (Ed Brook from Oregon State University, Eric Steig, Ed Waddington and
Howard Conway from University of Washington) will submit a proposal to NSF in June 2008 to help fund this work.
International partners (Dorthe Dahl Jensen from Denmark,
and Nancy Bertler from New Zealand) want to collaborate on the project.
We hosted a planning workshop in Seattle in April 2008:
minutes from the meeting are here.