Subglacial drainage systems. Depending on whether the glacier bed is rigid or deformable different types of subglacial drainage systems can be developed. In distributed systems drainage takes place over much of the bed (type 1, 2, 5, 6 and 7), whereas in discrete systems drainage is carried by a few channels (type 4).

1. Bulk movement with deforming till;
2. Darcian porewater flow;
3. Pipe flow;
4. Dendritic network (e.g. R-channels and N-channels);
5. Linked cavity system;
6. Braided canal network;
7. Thin film at the ice-rock interface.

(Derived from Benn & Evans, 1998)

Three main areas of precipitation can be distinguished: the southern, western and northern mountain regions with ³ 1 600 mm, the plains of the South- and West-coast with 800 to 1500 mm, and the North and Northeast with less than 800 mm precipitation per year. (According to Schutzbach, 1985, p. 134)

The principle of radio echo-sounding. Radio waves propagate spherically. Therefore, the resolution of the ice radar reflects the ice thickness over an area (~ 10 m²) on the bed.

There are at least two peaks in each signal displayed on the receiver: First, the direct signal from the transmitter (radio wave directly travelling through ice), and secondly, the reflected signal from the bedrock (cf. Appendix 13).

Ice thickness can be calculated from the travel time, i.e. the difference between the time of transmission and reception of the reflected signal. To position the measurement accurately the Magellan GPS is used between transmitter and receiver.

Comprehensive reviews of the principle of radar signal propagation and reflection can be found in Robin et al. (1969), Smith & Evans (1972) and Bogorodsky et al. (1985).

Appendix 13. Example for an ice radar waveform

Idealised glacier showing the variations in accumulation and ablation. The net accumulation ‘wedge’ (Sugden & John, 1976) lies above the equilibrium line, the net ablation ‘wedge’ below; i.e. net ablation increases down-glacier below the equilibrium line while net accumulation tends to increase upglacier above this point.

The wedges reflect the net balance gradient (mm m^-1 a.s.l.) composed of the rate of increase in accumulation plus the rate of decline of ablation with height.

Note: The significance to glacier flow of net accumulation and ablation over a glacier. The higher the rate of increase in the net balance with altitude up a glacier, the faster is the rate of flow required to maintain an equal surface profile. The higher the net balance gradient the thicker the ‘wedges’. (Modified from Sugden & John, 1976; Benn & Evans, 1998).

Side profile showing the theoretical passage of the Hekla-1947-tephra-layer through Gígjökull according to the general passage-way of ice through a glacier (cf. Appendix 15).

Note: The tephra, once deposited all over the glacier (a), becomes incorporated into the glacier exhibiting a tephra ridge on the surface, and travels downward while rotating. Folds are cut on the surface. Towards the snout the tephra will be washed out (b). Presently, the tephra outcrops as a band in the snout region (c)at an angle of almost 76° and will be washed out in the near future.

Characteristic crevasse patterns in a valley glacier. The diagrams on the top show the principle stresses resulting from shear stress and normal stress near the upper margin.

1. Splaying crevasses form where the glacier is subject to compressive flow which causes the glacier to expand laterally. This sideways expansion results in crevasses being bend upstream whilst meeting the ice edges at angles of less than 45°.
2. Chevron crevasses occur in response to the drag of the valley walls which pulls the ice downglacier from the glacier margins towards the centre-line resulting in crevasses aligned at approximately 45° to the valley walls (Nye, 1952).
3. Transverse crevasses form where the glacier is subject to extending flow. The principle stresses lie almost parallel to the main flow line near the centre resulting in crevasses at right angles to the principle flow direction. Where the stress pattern is influenced by the drag of the valley side transverse crevasses curve downstream.

In general, every crevasse undergoes modification by glacier flow, and flow differences between the centre and the margin cause the initial crevasse to rotate. Transverse and splaying crevasses tend to be straightened.

(Modified from Nye, 1952; Sugden & John, 1967; Paterson, 1994)

Longitudinal velocity (my^-1) in a transverse cross section

1. calculated by Nye (1965) for a glacier with no basal sliding in a parabolic channel
2. observed at the Athabasca Glacier, Canada (Raymond, 1971).

Note: Longitudinal flow velocities increase towards the centre of the glacier and decrease towards the glacier margins owing to friction with bedrock. Case (b) reflects probably best the situation at warm-based ice streams while (a) shows the velocity distribution at cold-based ice streams (Sugden & John, 1976).

(Derived from Sugden & John, 1976)