Hermann Engelhardt

Glacier Mechanics

The participation of the Antarctic Ice Sheet in global change may take the form of a rapid collapse, which would have a drastic effect on world-wide sea level. This possibility is presented most clearly by the West Antarctic part of the ice sheet, which is traversed by a number of large ice streams, internal currents moving at speeds ca. 100 times faster than the normal ice- sheet motion. If these ice streams were to enlarge and speed up enough, they could carry a large enough ice flux outward to bring about ice-sheet collapse.

We are concentrating our research on an effort to find the mechanism of the rapid ice streaming flow and the physical variables that control it and couple it to global change. The mechanism generally favored at this time is lubrication of the ice-stream bed by a ~5 m thick layer of glacial till between the base of the ice and bedrock. By drilling through the ice (ca. 1000 m thick) and taking piston cores we have obtained samples of unmistakable till from a layer at least 3 m thick under the ice. The till is unfrozen and water saturated, with a high porosity (ca. 40%), which indicates that it is probably undergoing or has recently undergone substantial deformation. We have measured its shear strength, both in the recovered core material and in situ, under the ice, and find that it is extremely weak, with shear strength in the range 1.5-8 kPa. This strength is much lower than the gravitationally-driven basal shear stress of 20 kPa, so the lubrication of basal motion by till deformation seems an inescapable conclusion. Our measurements show that the ice temperature reaches the pressure melting point at the base, and the basal water pressure is within 10 to 200 kPa of the ice overburden pressure, which are conditions that favor till deformation. The till-lubrication mechanism has been dubbed in some quarters "the new paradigm of glacier motion." It makes the motion of the ice streams mechanically more like that of huge landslides (50 km wide and 500 km long) than the flow of normal glaciers.

---

However, there are complications in this simple picture of a till-lubrication mechanism for ice stream movement. The weakness of the till should be controlled by the basal water pressure, but so far we have not detected any variations of ice stream motion in correlation with marked time- variations in basal water pressure that we do observe. The basal water pressure should be controlled by the water source (basal melting) and a basal water conduit system, but different borehole experiments give seemingly conflicting indications as to the nature of this conduit system. Although theories of the "new paradigm" like to model the till as a viscous fluid deforming uniformly throughout its thickness, the observed mechanical properties correspond to extremely nonlinear viscosity and there are observational indications that the till deformation is concentrated in a thin shear zone at the top, or that some of the motion actually occurs by basal sliding. A till-deformation mechanism with highly nonlinear till rheology can be shown theoretically to be mechanically unstable. Also, the observed weakness of the till makes it incapable of supporting the gravitationally driven, down-slope basal shear stress. These facts point to the necessity for operation of other flow mechanisms in ice stream motion.

The above complications are being addressed by currently ongoing research. In particular we have initiated an effort to find out if the ice stream motion can be controlled by ice deformation in the marginal shear zones instead of by basal shear of weak till: to do this we sample at depth in the marginal shear zone the ice whose deformation is involved, and test it mechanically at the observed marginal shear strain rate so as to measure the marginal shear stress (work of graduate student Miriam Jackson). We have developed and successfully used a new ice coring drill to obtain the needed samples. The marginal ice is found to have a strong preferred orientation of ice crystal c axes, which results from the rapid marginal shear deformation and probably enhances the ice fluidity.

---

Of particular interest to us is the possibility of a relationship between ice streaming and the mechanism of glacier surging, which we have studied in Alaska. Variegated Glacier, near Yakutat, surged in 1982-83, at speeds up to 65 m/day. We found that the surging condition was marked by abnormally high basal water pressure and abnormally low transport rate in the basal water system. On the basis of the observations a new hydromechanical model of the surge mechanism has been developed. Further insight was obtained in 1987 from Columbia Glacier, a large Alaskan tidewater glacier that is in a state of continuous surging at speeds up to 20 m/day. The basal water pressure was continuously high, but measured variations in the pressure did not have a simple relation to glacier motion, as they did in Variegated Glacier during the surge and in "mini-surge" events that occurred before and after the surge. However, Columbia Glacier showed a good correlation between flow velocity fluctuations and the storage of input water (rainwater, meltwater) in or under the glacier. From this evidence the control of rapid glacier motions in continuous and episodic surging is seen to be a matter comparably complicated to that in Antarctic ice streaming. It remains to be seen how closely related these two phenomena are.

---

SRI interferogram of a part of the Rutford Ice Stream, Antarctica.

---

To provide a new means of monitoring the flow velocities and grounding-line positions of ice streams, which are indicators of involvement of the Antarctic ice sheet in global change, we have undertaken a collaboration with Richard Goldstein of JPL in the application of his method of satellite radar interferometry (SRI) to the Rutford Ice Stream, Antarctica. The method uses phase comparison of the radar signal obtained for a pair of SAR images taken a few days apart to plot an interferogram which directly displays relative ground motions that have occurred in the time interval between images. The detection limit is about 1.5 mm for vertical motions and about 4 mm for horizontal motions in the radar beam direction. Comparison of SRI velocities with earlier ground-truth data in the Rutford Ice Stream suggests a secular decrease in velocity of about 2% from 1978-80 to 1992. Ungrounded ice is revealed by large (~2 m) vertical motions due to tidal uplift, and the grounding line can be mapped at a resolution of ca. 0.5 km from the SRI interferogram. The mapped configuration implies grounding line retreat of 1 to several km since 1980. Application of the method to other ice streams and to glaciers is planned.

---

Selected Publications

Kamb, B., 1987. Glacier surge mechanism based on linked cavity configuration of the basal water conduit system, J. Geophys. Res., 92, 9083-9100.

Engelhardt, H., N. Humphrey, B. Kamb, and M. Fahnestock, 1990. Physical conditions at the base of a fast moving Antarctic ice stream, Science, 248, 57-59.

Kamb, B., 1990. Rheological nonlinearity and flow instability in the deforming-bed mechanism of ice-stream motion, J. Geophys. Res., 96, 16585-16595.

Humphrey, N., B. Kamb, M. Fahnestock, and H. Engelhardt, 1993. Characteristics of the bed of lower Columbia Glacier, Alaska, J. Geophys. Res., 98, 837-846.

Kamb, B., M. F. Meier, H. Engelhardt, M. A. Fahnestock, N. Humphrey, et al., 1993. Mechanical and hydrologic basis for the rapid motion of a large tidewater glacier. 2. Interpretation, J. Geophys. Res., 99, 15231-15229.

Goldstein, R. M., H. Engelhardt, B. Kamb, and R. M. Frolich, 1993. Satellite radar interferometry for monitoring ice-sheet motion: application to an Antarctic ice stream, Science, 262, 1525-1530.

Engelhardt, H., B. Kamb, 1993. Vertical temperature profile of Ice Stream B, Antarctic Journal of the US, 28, 63-66.

The tethered stake experiment for measuring sliding speed of Ice Stream B is shown in Fig. 1

Here is our instrument site on the Unicorn (lat. 83 deg, 34.4 min S, long. 138 deg, 8.9 min W).