Crary Ice Rise plays a critical role in determining the ice flow of the south-eastern Ross Ice Shelf. Lying directly in the path of Ice Stream B, back pressure from the ice rise is transmitted hundreds of kilometers up the ice stream from the grounding line. The ice rise's influence is particularly widespread due to the presence of the “ice plain”, a broad region of shallow-sloped ice at the mouth of the ice stream and loosely coupled to the underlying bed material. For this reason, the current behavior of the West Antarctic ice sheet is affected by the evolution of Crary Ice Rise.

Past analyses have resulted in conflicting interpretations of the history of this ice-shelf feature. Reference MacAyeal and ThomasMacAyeal and Thomas (1980) calculated that a measured profile of elevation across the ice rise made during the RIGGS project was lower than the equilibrium profile, thus inferring current growth of the ice rise. In contrast, Reference JezekJezek (1984) used other RIGGS data and showed that the debris traces in the ice shelf down-stream from the ice rise were not parallel to the current flow direction. He hypothesized a mechanism to explain this difference based on a retreating grounding line (or boundary) of the ice rise. More recent data, analyzed by Reference MacAyeal, Bindschadler, Shabtaie, Stephenson and BentleyMacAyeal and others (1987), indicate an average thickening rate of 0.44 m/a in the region of Crary Ice Rise. Finally, aerial photography of the ice rise taken in 1985 has revealed a discrete “ice raft” separating from the main ice rise (Reference Bindschadler, Vornberger, Stephenson, Roberts, Shabtaie and MacAyealBindschadler and others, 1988). Such a separation can be interpreted as part of an episodic disintegration of the ice rise.

To resolve the question of whether Crary Ice Rise is a remnant of an advanced Wisconsin-aged ice sheet when most or all of the Ross Ice Shelf was grounded and which has since experienced extensive retreat, or whether the ice rise is a recently formed feature, a technique used by Reference Lyons, Ragle and TamburiLyons and others (1972) to date the time of initial grounding for ice rises on the Ward Hunt Ice Shelf was employed. The technique compares the measured temperature versus depth profile of the ice rise to temperature profiles calculated by a numerical model. The model predicts the transient temperatures of the ice as the base of the ice cools after grounding. Initially, the basal ice is in contact with sea-water (at a known temperature of about –2°C at the ice/water contact); after grounding, the basal temperature condition shifts to a specified basal heat flux deep within the underlying bedrock. Our model, similar to that of Reference MacAyeal and ThomasMacAyeal and Thomas (1980), ignores horizontal conduction and advection but includes vertical conduction and advection, as well as non-equilibrium vertical strain and variable density.

Two holes were drilled on Crary Ice Rise in December 1987. The first, 370 m deep, was drilled in the vicinity of the shallowest bedrock known to exist under the ice rise from airborne radar-sounding data. A second hole, 480 m deep, was drilled 15 km south-south-west on the highest ridge of the ice rise. Temperatures were measured in each hole as soon as the thermistor cables were in place. In the shallower hole, measurements were made for [Inline 1]. During this time all thermistors cooled, froze-in, and cooled further. The only exceptions were the two thermistors at the base which remained at the pressure-melting temperature throughout the [Inline 2] period. 10 m above the bed, freeze-in occurred after only 4 d. The rates of cooling decreased markedly with depth, indicating warmer ice temperatures with depth. Rates of cooling near the bed were so small that it is difficult to anticipate basal temperatures much colder than the –2°C temperature which corresponds to the basal temperature at the time of grounding. We have calculated that a basal heat flux of 0.165 W/m² would be required to maintain a basal temperature of -2°C. Because this value of heat flux is unreasonably large, we feel that the warm basal ice at this location is strong evidence that Crary Ice Rise is a recent feature: at most a few centuries old.

This result and other recently published findings are consistent with a recent increase in activity of Ice Stream B. These findings include:

1. Current negative net mass balance of Ice Stream Β (Reference Shabtaie, Bentley, Bindschadler and MacAyealShabtaie and others, 1988; Reference Whillans and BindschadlerWhillans and Bindschadler, 1988).

2. Thickening in the region of Crary Ice Rise (Reference MacAyeal, Bindschadler, Shabtaie, Stephenson and BentleyMacAyeal and others, 1987).

3. Recent age of Crary Ice Rise (this work).

4. Deceleration of the Ice Stream Β “ice plain” (Reference Stephenson and BindschadlerStephenson and Bindschadler, 1989).

5. “Raft” separation from Crary Ice Rise (Reference Bindschadler, Vornberger, Stephenson, Roberts, Shabtaie and MacAyealBindschadler and others, 1988).

A possible evolutionary sequence which ties these results together begins with a recent increase in activity of Ice Stream B, causing a strongly negative net mass balance. The magnitude of this negative mass balance (about 43%) does not suggest a brief period of activity (such as a surge where discharge fluxes are one to two orders of magnitude above balance) but rather a modest increase in activity which could be sustained for well over a century. The excess ice discharged by the negative net balance must have entered the Ross Ice Shelf as a wave of thickening. It was this thickening which, we believe, formed Crary Ice Rise. This thickening appears to be still under way and is responsible for the deceleration of the “ice plain”. We expect this process to be accompanied by an advance of the grounding line but cannot report direct measurements. As Crary Ice Rise grows and presents an ever-increasing obstacle to the discharge of ice from Ice Stream B, it is expected that the ice stream will need to adjust to maintain its discharge. This can be by any number of mechanisms involving its own shape (slope, width, and thickness), as well as the shape of Crary Ice Rise. We observe that the separation of the raft has served to make the ice rise more streamlined to the discharge of Ice Stream B. This streamlining may be a direct result of the interaction between an accelerating ice stream and a forming ice rise.