Basal ice deformation

Examination of deformation processes at the base of glaciers is beset by numerous problems, which include limited accessibility, structural complexity and spatial variability in physical properties and temporal variability of deformation processes. Access to subglacial locations is a significant problem because deformation of basal ice and subglacial sediment takes place at the ice-

substrate interface beneath a substantial thickness of ice, which makes direct observation of deformation processes very difficult. Even where access is possible there are considerable uncertainties associated with measurements that are made remotely. If boreholes are used it is often not clear where instruments are located or observations made with respect to the glacier bed. In the case of excavations made to access to the bed they are usually limited to glacier margins where ice is relatively thin and the presence of unsupported walls may result in substantial disturbance of subglacial deformation processes by strain-relief creep. In addition, the flow behaviour of basal ice is likely to vary in time and space in response to both flow perturbations as well as variations in the composition and structure of materials.

Despite these difficulties several important observations concerning the behaviour of basal ice and glacier substrates have been made. Most experiments have been driven by two different concerns: understanding the contribution of subglacial sediment deformation to glacier behaviour, which has been reviewed by Murray (1997), and understanding the rheological behaviour of ice-debris mixtures.

There has been considerable research on the rheology of ice-sediment mixtures, both in the laboratory and in the field. The results of this research have provided evidence to suggest that the behaviour of basal ice is very sensitive to debris and solutes entrained within the ice. However, laboratory experiments conducted on the behaviour of basal ice have produced contradictory results. For example, Goughnour & Andersland (1968) concluded that enhanced deformation was associated with low concentrations of debris, whereas Hooke et al. (1992) and Nickling & Bennett (1984) concluded that the creep rate of ice decreased with the addition of fine sand and rock debris respectively.

Empirical studies of the mechanical behaviour of basal ice suggest a strong sensitivity to debris content. Holdsworth (1974) measured deformation of basal ice in a tunnel excavated in Meserve Glacier (—17°C) in the McMurdo dry valleys and found that an amber ice layer experienced enhanced deformation, which he attributed to high solute and debris concentrations. He suggested that the exponent in Glen's flow law for the amber ice was between 5 and 6. More recently a reassessment of the Meserve Glacier basal ice suggests slow but detectable sliding of the amber ice over the substrate (Cuffey et al., 1999) and that the enhanced deformation of the amber layer could be attributed to the small size of ice crystals in the ice (Cuffey et al., 2000c). Brugman (1983) reported preferential deformation along debris-rich ice layers and Lawson (1996) suggested that the presence of debris can enhance creep rates. In situ measurements made by Echelmeyer & Wang (1987) in a tunnel in Urumqui No. 1 Glacier in China demonstrated that 60-85% of surface velocity was due to enhanced deformation in ice-laden till, motion in discrete shear planes (10-25% of glacier motion) and sliding at the ice-sediment interface. Similarly Waller & Hart (1999) measured deformation at the margin of an outlet glacier in Greenland and found that most motion could be attributed to sliding or subglacial sediment deformation and relatively little motion could be explained by creep within the basal ice.

In addition to the work described above, several studies have made direct measurements of movement close to glacier beds using tilt meters, plough meters and drag spools deployed in holes drilled from the surface (e.g. Blake et al., 1992, 1994; Fischer & Clarke, 1994). Engelhardt & Kamb (1998) used a tethered stake to measure deformation at the base of Ice Stream B and concluded that the majority of surface velocity could be accounted for by sliding or deformation within a thin (30 mm thick) layer of subglacial till. More recently Porter & Murray (2001) used tilt sensors hammered into the glacier substrate of Bakaninbreen, Svalbard to monitor near-bed deformation. They interpreted data from the drag spool as evidence of movement predominantly in the form of internal deformation and calculated viscosities that suggested the subglacial sediment may have been partly frozen. They concluded that where the bed is cold the sediment-rich basal layer grades into ice-rich sediment without a clear boundary.

Recent excavations of tunnels in the basal zones of several glaciers in the McMurdo dry valleys have provided opportunities for detailed deformation measurements to be made close to the glacier bed. The basal velocity profile, together with tunnel deformation measurements in Suess Glacier, demonstrates that deformation in the lower 4.5 m of the glacier is characterized by progressive simple shear and that strain is heterogeneous. Four distinctive strain domains occur in the basal ice: a high-strain domain associated with the presence of the amber ice where the velocity profile approximates a power law (Fig. 65.1); a low strain domain associated with the presence of the solid facies (2.9-2.4m); a moderate strain domain associated with the stratified facies; and sliding at the interface between the amber ice and the solid facies (Fig. 65.1). At the boundary between the rapidly deforming amber ice and the underlying frozen sand and gravel numerous air-filled cavities are present, both where the frozen layer protruded into the amber ice and where the contact was relatively flat. At this boundary dial gauges recorded movement from 0.93 to 5.65 mmyr-1. The measurements can be interpreted as sliding velocities or zones of high shear concentrated in a very thin layer. However, the coincidence of the high strain rates with the presence of slickensides on the cavity roofs clearly indicates that sliding has occurred at the interface. The slickensides have formed as ice sliding over the frozen sediment layer has moulded itself around the roughness elements on top of the layer and produce an imprint of the form roughness in the cavity.

The underlying control of the deformation processes and the consequent deformation structures appears to be the strength of the ice and sediment mixture, the averages of which are 0.9 MPa for the amber ice, 1.28 MPa for the stratified ice and 2.53MPa for the solid facies (Fitzsimons et al., 2001). Material with a low viscosity (low shear strength) supports the highest strain rates and the material with the highest viscosity (high shear strength) supports the lowest strain rates (Table 65.1). Motion within the strat-

Table 65.1 Peak shear strength values from basal ice samples from Suess Glacier

Ice facies

Peak strength

Debris content


(% volume)













Figure 65.7 The basal ice solid facies with well-preserved planar bedding in Wright Lower Glacier.

ified basal ice is discrete (slickenslides and offsets in velocity profiles measured using strain markers) (see figs 4, 5 & 6 in Fitzsimons et al., 1999).

At the margin of Wright Lower Glacier (basal ice —17°C) plumblines and displacement transducers were used to monitor deformation through layers of frozen sand entrained into the basal zone (Figs 65.6 & 65.7). These measurements show that there was no detectable creep within the frozen sediment layers. However, displacement transducers mounted across cracks in the frozen sediment (Fig. 65.8) demonstrated the cracks were opening at a rate of 22mmyr—1 measured over an 11 month period. The structure of ice adjacent to the cracks suggests that the ice is creeping into the cracks as they open (Fig. 65.9).

Measurements made at the base of Taylor Glacier, which has a basal temperature of—18°C at the margin, have also demonstrated the presence of sliding interfaces within the basal zone. These interfaces are characterized by the development of cavities on the stoss and lee sides of clasts that protruded through the sliding boundary. Sliding velocities up to 167mmyr—1 have been measured using a combination of plumblines, engineering dial gauges and linear variable displacement transducers. Close to the glacier margin the sliding constitutes approximately 10% of glacier motion. Deformation structures that developed in the basal zone over the monitoring period demonstrate that debris-rich ice deformed at considerably greater rates than adjacent clean ice (Figs 65.10 & 65.11).

Taken together, the observations and measurements made in the tunnels in glaciers in the McMurdo dry valleys demonstrate a considerably more complex pattern of deformation than is suggested by the literature on cold-based glaciers. In the cases of the Suess, Wright Lower and Taylor glaciers if observations were made from drill holes from the glacier surface it is very likely that a very different picture of glacier behaviour would have emerged because a drill would not have penetrated the solid facies. Under these circumstances measurements of the first contact between ice and sediment would have been interpreted as the glacier bed. Even

Figure 65.8 Crack in the solid facies in Wright Lower Glacier. The glacier flow direction is from right to left.

Above And Below Glacier
Figure 65.9 Ice from above and below intruding into a crack in the solid facies in Wright Lower Glacier. The glacier flow direction is from right to left.

in the case of measurements and observations made in the tunnels which provide excellent access, our experience in measuring motion at the base of a glacier is that the closer we examine the glacier bed the more problematic the concept of a distinct bed

Figure 65.10 Photograph of a tunnel in the left margin of Taylor Glacier showing the 'key-hole' structure produced from more rapid deformation associated with debris-rich ice. The tunnel walls were vertical initially. Photograph taken 11 months after the tunnel was excavated. The glacier flow direction is from the ladder toward the reader.

Figure 65.10 Photograph of a tunnel in the left margin of Taylor Glacier showing the 'key-hole' structure produced from more rapid deformation associated with debris-rich ice. The tunnel walls were vertical initially. Photograph taken 11 months after the tunnel was excavated. The glacier flow direction is from the ladder toward the reader.

Figure 65.11 Circular strain markers deformed into ellipses after 10 days in debris-rich basal ice in Taylor Glacier. The glacier flow direction is from left to right.

becomes—both in terms of physical characteristics and in terms of deformation patterns (Fig. 65.1).

Was this article helpful?

0 0

Post a comment