The principal landform formed by meltwater flow beneath a glacier is the esker. Eskers are the deposits of former subglacial, englacial or supraglacial channels. They are usually slightly sinuous ridges of glaciofluvial sediment that undulate in height along their length. Their orientation is controlled by glacier slope and the pattern of water pressure potential within the glacier; they may therefore show little respect for subglacial topography and need not trend downslope.
Two broad types of esker exist: (i) single-ridge eskers; and (ii) braided (anastomosing) eskers. Braided eskers consist of a network of ridges that merge and bifurcate. They are typically short in length, with braided reaches less than 1 km long, and they are often associated with areas of kame and kettle topography. Single-ridge eskers vary from less than 1 km to hundreds of kilometres in length. Long eskers are typically 400-700 m wide and 40-50 m high, whereas smaller eskers (< 300 m long) are usually 40-50 m wide and only 10-20 m high.
In general eskers may vary in cross-profile along their length and may occasionally have a beaded form. This beaded form may simply consist of wider sections at regular intervals along the length of the esker or alternatively consist of a chain of short lengths of esker ridge between which the ridge is barely visible.
In a few cases eskers have been recorded on top of, and therefore infilling, subglacial N-channels. For example, quarry excavations within the Blakeney Esker in Norfolk, England, reveal a series of small meltwater channels cut into the till surface on which the esker rests. This indicates that the esker formed subgla-cially and that deposition followed a period of subglacial meltwater erosion. Esker-like bodies of sediment have also been recorded from within units of lodgement till (see Section 8.1.1: Figure 8.12).
Eskers are typically composed of a core of poorly sorted sand and gravels. Above this core, sorted sands and gravels may occur. Sometimes these have arched bedding that dips outwards from the centre of the ridge. Sands and gravels are usually well rounded and have palaeocurrent orientations parallel to the trend of the esker. The esker may be capped by a thin veneer of till. In general, however, the sedimen-tology of eskers is highly variable and generalisations are difficult. This perhaps reflects the variety of depositional environments in which they form and the high flow regimes or energy levels present.
Single-ridge eskers form when a supraglacial, englacial or subglacial channel or tunnel becomes blocked. Sedimentation in supraglacial channels is easily understood, but sedimentation within englacial or subglacial tunnels is more problematic. By applying theories developed to explain the flow of solids within pipes it has been suggested that sediment of all sizes may be transported as a single mass within englacial or subglacial tunnels (see Section 8.2). The high concentration of debris has a buoyant effect allowing larger particle to be transported more easily. Deposition of these particles is impossible because any constriction of the tunnel caused by deposition would simply increase the water velocity and thereby re-entrain the sediment. However, deposition of all the material in transit takes place if flow within the pipe is suddenly blocked. Obstruction of the tunnel by an ice fall or similar blockage will cause the deposition of large volumes of poorly sorted sand and gravel. Finer grained arched gravel and sands are predicted by pipe flow theory when discharge falls, a situation that would occur at the end of the ablation season. It has been suggested that eskers can form very rapidly due to the high sediment loads within meltwater streams. In this way an esker could form rapidly by the deposition of gravel masses associated with permanent or temporary blockage of a tunnel and by the deposition of arched gravel units during declining discharge at the end of the melt season. Subsequently the esker is uncovered by ice retreat, or lowered to the ground surface by ablation if the tunnel is located in an englacial position. Modification of the esker during lowering is likely, although observations in Iceland suggest that englacial tunnel fills can be lowered quickly with little remobilisation of the esker sediment as the ice retreats. Similarly, supra-glacial channel fills can be infilled with sediment and lowered during ablation.
The formation of braided eskers is a more complex problem because it is difficult to envisage the formation of a braided network of subglacial or englacial tunnels. Examples of contemporary eskers from Iceland, however, formed by the infill of englacial and supraglacial tunnels show simple bifurcation. It has also been suggested that subglacial braided channels and associated eskers may form as a result of catastrophic floods. This model suggests that when a single channel cannot accommodate the high discharges of water and sediment during a flood event new channels are produced causing a multichannelled subglacial system to develop, although the exact mechanisms of this process remains unclear. An alternative hypothesis is that braided eskers develop supraglacially either by the development through time of a cross-cutting channel pattern or by the lowering of a supraglacial outwash fan that has deeply incised sediment-filled channels, which are inverted during meltout to form a braided esker.
Observations from within braided eskers in the Southern Uplands of Scotland support this last idea. Here there are two large braided esker systems; one centred at Carstairs and the other at Newbiggin (Box 9.8). The Carstairs system is over 7.5 km long and several hundred metres wide. The major ridges are generally 10-15 m high, but may locally increase in height to over 30 m. The ridges are sharp crested and steep sided. The principal ridges lie within a broad belt of kame and kettle topography, in which low sinuous ridges can occasionally be identified. Internally the main ridges are composed of a core of boulder-rich gravel in which 40-50% of the material is over 250 mm in size and boulders up to 2 m in diameter are also common. The surrounding kames contain much finer grained sands and gravels more typical of that found on braided outwash surfaces. Faulting within these mounds suggests that the sediment has undergone subsidence associated with the meltout of buried ice. It has been suggested that the whole sequence formed as a supraglacial outwash fan subject to catastrophic or high-magnitude low-frequency flood events. During periods of normal flow a system of channels and bars developed on the ice-cored fan surface in which fine gravel and sands were deposited. This was followed by a period of catastrophic flow, which cut a series of deep channels into the buried ice beneath the fan. These channels were then filled with coarse boulder gravel as flow magnitude fell. With the return of normal flow conditions channels were infilled by further low-magnitude deposits. When the outwash surface was abandoned, perhaps as a consequence of the catastrophic flows, meltout of the buried ice inverted the topography. The large boulder-filled flood channels were inverted to form the main esker ridges, while the small channel fills produced the kame and kettle topography around them.
In summary eskers may form in a variety of different settings. These can be summarised as: (i) deposition in subglacial tunnels; (ii) deposition in englacial tunnels and subsequent lowering; (iii) deposition in supraglacial channels and subsequently lowering; and (iv) deposition in ice-walled re-entrants at the ice margin.
Was this article helpful?