An unfrozen rock bed

This is the classic domain for studies of glacier erosion because much early glaciological study was concerned with temperate, rock-floored, valley glaciers. The classic conception is of erosion processes dominated by 'plucking', where the ice prises away rock tools from the glacier bed, and abrasion in which these tools, embedded in the glacier sole, scratch and abrade the substratum to produce striated, smoothed and streamlined landforms. The debris generated by plucking and abrasion itself breaks down to produce a characteristic 'crushing distribution' of grain sizes (Hal-dorsen, 1981), and is transported away by inclusion in a basal ice layer by processes of relegation (Sharp et al., 1989). In temperate glaciers, this debris-rich basal ice layer is typically of the order of centimetres in thickness, with concentrations of the order of 10-40% by volume (Table 2.1).

Implicit in the various approaches to abrasion of the bed by sliding ice is that there is a strong contrast between a state in which the glacier sole consists of a continuous debris carpet, and one where clasts are well separated in relatively clean ice. In the former case, a 'sandpaper' model is appropriate, where friction and abrasion rate are a function of effective pressure at the ice-bed interface, the driving stress and ice velocity (Drewry, 1986). Occasional large clasts that penetrate through the basal debris carpet because of the way in which they concentrate stress (Boulton et al., 1979) will create larger deeper grooves in bedrock. In the latter case, the flow of ice around sparse, frictionally retarded clasts in traction over the bed will determine the abrasion rate (Boulton, 1974; Hallet, 1979). In the former mode, the tangential friction generated by a continuous basal debris carpet is a significant contribution to the resistance to glacier movement (section 2.2). The sandpaper mode is likely to produce greater rates of erosion than the sparse mode, generating even more debris, potentially clogging the basal transport system and thereby depositing subglacial till.

There has not, as yet, been an effective test for an abrasion law. Bedforms and glaciated valley profiles seem to provide only a weak test, as a wide variety of laws is able to generate typical ero-sional forms (Harbor, 1992). One of the difficulties in modelling valley glaciers and the overdeepening they produce has been the absence of negative feedback that would prevent the abrasion process from continuing to overdeepen the bed locally. Alley et al. (2003) have suggested that negative feedback can be provided by the triggering of a supercooling mechanism as overdeepened slopes increase, a mechanism that is suggested to inhibit erosion and promote deposition.

0 0

Post a comment