Discussion

T. J. Hughes: You claimed that the hardest part of an ice stream is the zone of converging flow at its head, and this hard ice would keep an ice stream surge from propagating inland. The zone of converging flow should be one of the softest parts of the ice stream, because strong lateral convergence deforms the whole bulk of ice, not just the basal and lateral shear zones. So bulk strain-softening should be a maximum, not a minimum, at the heads of ice streams and the surge could easily propagate inland.

On another point, in your surge paper with Weertman, was not the ablation zone at the glacier snout the means whereby you obtained compressive flow in the snout? If so, compressive flow in ice-stream tongues could be provided if the tongues were imbedded in confined and pinned ice shelves, and your surge mechanism might apply.

G. de Q. Robin: I agree that your first point is of major importance, but you misinterpret my remarks. I was discussing the relative hardness of the lateral shear zone and the zone of converging flow. I believe the former is much softer than the latter, and hence cannot propagate inland as a single shear zone. I agree that the converging zone will be softer than the main ice sheet, but as to how much softer, we still need an answer. Certainly the important key to stability of ice streams lies in this zone, which needs further study.

As regards your second point, your argument does not apply, unless ice-stream discharge stops almost completely above the grounding line and then only if the equilibrium line is much higher.

C. Lorius: Your reconstruction of the D-I 0 profile for the elevation of origin of ice is very attractive. However at the 220 m depth there is a difference of about 1.5 km with the elevations from the ice-core 0 values. This difference decreases to zero in a rather rapid way, which may not be contradictory with the existence of surge.

Robin: I accept your point that surface lowering was quite rapid when ice from 220 to 200 m depth in the present core was being laid down, but any difference between us lies in our definition of a surge. 1 regard a surge as involving some dynamic instability leading to excessive surface lowering, followed by gradual recovery as the surface builds up to levels associated with the prevailing climate, rather than a relatively uniform, though considerable, change in surface level from that appropriate to one climatic regime to that associated with another.

w. F. Budd: Since the paper by Reference AllisonAllison (1979) on the Lambert Glacier basin is not being presented I would like to comment that the interior of the Lambert basin appears to be one large region which has a quite definite positive balance. The region of Lambert Glacier which has apparently had substantial ice lowering of the order of 800 m or more seems to be rising at a high rate (c. 0.2 m/a) which, if continued, would reach the previous thickness in the order of 4000 a.

My second point is that there seems to be a problem in matching the large, relatively rapid changes observed in isotope and gas-content profiles, without substantial increases in velocity, which are difficult to explain in terms of steady state or direct reaction to climate change.