Glaciofluvial Icemarginal Landforms

There are two main sources of glacial meltwater at a glacier snout; water emerging from subglacial meltwater portals or from supraglacial channels fed by surface melt and the emergence of englacial conduits. As glacial meltwater emerges at the ice margin its velocity typically falls due to changes in confining pressure, gradient and increases in bed or channel friction. As a consequence, meltwater streams deposit sediment rapidly to form a variety of outwash-related landforms. The morphology and sedimentary composition of these landforms will depend on: (i) the topographic setting of the ice margin and its geometry; (ii) the presence or absence of buried ice; and (iii) the total amount of sediment in transport within the meltwater which is a function of discharge and sediment availability. It is worth noting that some outwash systems are heavily affected by outburst floods, or jolkullhaups, generated by subglacial volcanism in volcanically active regions, or by the breaching of ice- and moraine-dammed lakes. These high-magnitude but low-frequency events have a profound impact on the evolution of outwash sequences.

The presence or absence of buried ice and the rate at which it melts is an important control on the morphology of ice-marginal glaciofluvial landforms. As a glacier retreats the drainage pattern will change and on the outwash surface river channels and bars are abandoned. The final morphology of these abandoned out-wash surfaces depends upon the amount of buried ice beneath them and whether this ice has melted out before they were abandoned. The conceptual model in Figure 9.27 develops this idea further and shows two glacier margins at which an outwash surface is being deposited. As the two glaciers retreat, the drainage system evolves and this outwash surface is abandoned. At the glacier margin with no buried ice, the outwash morphology reflects the fluvial depositional processes that deposited it, displaying bars, channels and river terraces. This contrasts with the ice margin where buried ice is present. Here, outwash morphology may follow two different evolutionary paths depending upon whether the buried ice has melted out before or after the outwash surface is abandoned. If the ice melts out after the fan is abandoned then the outwash surface will be deformed by subsidence. This may be confined to the occasional kettle hole - an enclosed hollow formed by the meltout of buried ice - if the proportion of buried ice is small, but if it is large then the whole surface will be deformed into an area of kame and kettle topography. If the meltout of the buried ice is complete prior to the outwash surface being abandoned then there will be little evidence of subsidence on the surface, although it will be evident in the sedimentary structures of the landform (Figure 8.19). As kettle holes form, the meltwater streams will tend to be diverted into them. Deposition will proceed rapidly within these kettle holes because standing water within them causes a reduction in the velocity of stream flow. In this way areas of subsidence are infilled as they form. The two cases illustrated in Figure 9.27

1: No buried ice 2: Buried ice

1: No buried ice 2: Buried ice

3: Undulating outwash surfaces 4: Kame and kettle topography

Figure 9.27 Conceptual model showing the evolution of two outwash surfaces, one of which is underlain by buried ice the other is not. The key control on the outwash morphology that results is the timing between the meltout of the buried ice and the abandonment of the outwash surface. If the surface is abandoned before all the ice has melted out then a kame and kettle topography results. Alternatively if the buried ice melts out before the surface is abandoned its presence may not be visible in the outwash surface because meltwater streams tend to infill the kettle holes as they form.

3: Undulating outwash surfaces 4: Kame and kettle topography

Figure 9.27 Conceptual model showing the evolution of two outwash surfaces, one of which is underlain by buried ice the other is not. The key control on the outwash morphology that results is the timing between the meltout of the buried ice and the abandonment of the outwash surface. If the surface is abandoned before all the ice has melted out then a kame and kettle topography results. Alternatively if the buried ice melts out before the surface is abandoned its presence may not be visible in the outwash surface because meltwater streams tend to infill the kettle holes as they form.

illustrate end-members of a continuum. The character of the resulting landsurface depends on: (i) the rate at which the buried ice melts; (ii) the rate of glacier retreat and the rate at which the outwash surfaces are abandoned; and (iii) the rate of fluvial deposition and its distribution across the outwash surface. Two groups of outwash landforms therefore can be recognised: (i) outwash plains and fans; and (ii) kames and kame terraces.

Outwash fans build up in front of a stationary ice margin (Figures 9.28, 9.29 and 9.30). The apex of each fan is focused on the point at which the meltwater emerges. Fans develop because coarse material is deposited relatively close to the meltwater portal, whereas the finer fraction is transported further. Outwash fans may merge away from the glacier and grade into large braided river sequences forming an

Outwash fan Ice-contact face

Käme & kettle topography

Figure 9.28 The formation and morphology of a simple outwash fan.

Figure 9.29 Two examples of hochsander or outwash fans in front of outlet glaciers of the Myrdalsjokull Ice Cap in Iceland. [Modified from: Kj^r et al. (2004) Sedimentary Geology, 172, figure 13, p. 159]

outwash plain or sandur (Figure 9.30). The term sandur is Icelandic in origin and means 'a sandy-gravelly area formed by proglacial streams'. A distinction can be made between ice-marginal outwash fans, sometimes referred to as hochsander fans, and outwash plains or sandurs. A hochsander is typically supraglacially fed and architecturally similar to an alluvial fan, with a semi-conical shape, a restricted radial length of 0.5-1.5 km, a plano-convex cross-profile and a slope gradient between 1° and 5°. The ice-contact face is frequently underlain by buried ice, which melts when the fan is abandoned to give an ice-contact face composed of kame and kettle topography (Figure 9.28). The surface of the fan has a shallow relief formed by abandoned channels and bars, and internally the fan is dominated by planar or low-angle cross-bedded sands with gravel interbeds coupled with varying amounts of diamicts. Figure 9.29 provides a schematic plan of an Icelandic outwash fan and shows a strong proximal to distal transition of fluvial processes and associated facies. This proximal to distal transition is seen within the facies assemblage, which tends to show an upward fining as lower-energy distal facies succeed more proximal ones, although there are many exceptions to this rule. Good examples of proximal fans have been described from the margins of the Myrdalsjokull ice cap in Iceland (Figure 9.29). Here the development of these fans is favoured by an environment dominated by surpraglacial rather than subglacial drainage and the contrast in fan morphology between Kotlujokull and Slettjokull reflects the fact that the latter ice margin is stationary or actively receding. It is possible also to recognise a continuum of fan types on the basis of discharge. At one end of the spectrum are 'dry' fans with a steep gradient and limited distal extent (Figure 9.30). These are composed predominantly of diamicts and poorly sorted sands and gravels produced by sediment gravity flows transporting supraglacial debris off-ice. At the other end of the spectrum are lower gradient, laterally extensive fans that are composed of sorted gravel and sand units. These show a proximal to distal fining. These latter fans may merge along an ice margin to form a 'moraine-like' landform with an asymmetric cross-profile that marks a stationary ice margin (Figure 9.30).

Outwash fans contrast with broader outwash plains, which are topographically less constrained and frequently grade from proximal fans. There is also some evidence to suggest that the development of outwash plains is favoured where melt-water emerges from large well-defined subglacial meltwater portals rather than from supraglacial sources. It has been argued recently that outwash plains are more common in the geological record than at present, reflecting different glacial dynamics and also the availability of accommodation space in which sediment may accumulate (Figure 9.30). Many modern glaciers, for example in Iceland or Alaska, where much of the research on outwash fans has taken place to date, have laterally restricted proglacial forelands, with relative steep longitudinal gradients between glacier and adjacent coastal areas. Here the development of outwash fans, with facies characteristics similar to alluvial fans, is common place. However, during the last glacial cycle large mid-latitude ice sheets terminated on land in both North America and Europe, where topography was less confined and the distance to the sea much greater. Here glaciofluvial deposition away from ice-marginal outwash fans was dominated by large unconstrained braided river systems, producing broad outwash plains. This point has been demonstrated with respect to outwash

Figure 9.30 The formation and morphology of outwash fans and sandurs. [Modified from: Krzyszkowski and Zielinski (2002) Sedimentary Geology, 149, figures 4 and 5, pp. 80 and 83. Zielinski and van Loon (2003) Boreas, 32, figures 9 and 10, pp. 605 and 606]

deposits in the Polish Lowlands, which have very different facies architecture to modern outwash fans. Here vertical facies successions are chaotic, without the distinct proximal to distal transitions common to some modern outwash fans. The deposits were built up by extensive (> 50 km) braided river systems flowing from proximal fans. Lithological and facies changes along these rivers are very different from modern outwash fans and represent a depositional environment in which the longitudinal energy regime was more constant; that is they do not have the rapid decrease in longitudinal gradient common to modern fans. In these outwash plains the hydrodynamic gradient tends to be a transverse rather than a longitudinal one; the main, high-energy gravel-bed channels alternate laterally with secondary low-energy sand-bed channels with extensive interchannel areas subject to aeolian processes. The location of these channels changed through time, with channel avulsion associated with flood events and changes in base level or discharge regime causing regular lateral shifts in facies rather than simple proximal to distal transitions. This raises some interesting issues with respect to the appropriateness of modern analogues for the interpretation of the Pleistocene record in the context of glaciofluvial deposits.

So far in this discussion we have assumed that deposition of outwash occurs on a glacial foreland without topographic restraint. However, ice frequently retreats away from reverse bedrock slopes or from earlier glacial landforms and consequently outwash may accumulate in more restricted locations, for example between an ice margin and a lateral hillside. In these situations, sediment deposited by a receding glacier may be left as a kame terrace. The morphology of a kame terrace depends largely upon: (i) the size of the meltwater stream; (ii) the steepness of ice margin; and (iii) the angle of the slope or hill side against which the glacier rests. If the meltwater stream is large, the ice margin is steep and the valley side is gently sloping then a large broad kame terrace will result. Buried ice will be present only close to the ice margin. In this case, a broad terrace surface will be produced when the ice retreats. It will have an outer edge that consists of a narrow belt of kame and kettle topography formed by subsidence of the buried ice at the terrace edge. The sedimentary architecture of these kame terraces is dominated on the ice-contact side of the landform by subsidence structures consisting of extensional faults and folds (see Figure 8.19). Alternatively, if fluvial deposition occurs over a large area of the ice margin then a narrow terrace bordered by a very broad belt of kame and kettle topography will form. Where kame terraces are formed against steep valley sides by meltwater streams, then small irregular fragments of terrace often form. These are often associated with linear kames formed in ice-walled channels that parallel the ice margin. Internally these kame terraces and kames contain sediments typical of ice-proximal meltwater, and also sediment derived from supraglacial debris and flow-till deposits (see Section 8.1.2).

One of the finest examples of a large assemblage of kame terraces is found along the shores of Loch Etive in Scotland. During the Younger Dryas a valley glacier terminated at the mouth of Loch Etive. Large kame terraces mark the lateral margins of this former glacier. The outside edges of these terraces are heavily kettled and are marked by narrow belts of hummocky topography. This hummocky topography also contains ridges with a sinuous planform. These ridges formed when englacial tunnels or supraglacial channels supplying meltwater to the terrace became blocked with sediment. As the glacier retreated and the buried ice melted, these sediment-filled tunnels or channels were lowered to the ground to form ridges of sediment that follow the former course of the channel or tunnel. The terraces at Loch Etive merge down-valley into a large ice-contact outwash fan that partially blocks the entrance to the loch.

Kame terraces need not only form at lateral ice margins. They also form at a glacier snout wherethey rest against a reverse slope such as the ice-contact face of an outwash fan. Kame terraces can be distinguished from outwash terraces formed by the dissection of outwash fans and plains, because: (i) kame terraces are formed by deposition not erosion; (ii) kame terraces usually contain kettle holes along their front margins and are higher here than at their valley-side margin, whereas out-wash terraces may be kettled but not preferentially at their front edges and also tend to be highest at their valley-side margins; (iii) outwash terraces are frequently attitudinally matched across the valley, whereas kame terraces are not; and (iv) kame terraces contain subsidence structures at their margins when viewed in section, whereas outwash terraces do not.

Kame terraces have only one ice-contact margin in contrast to kames, which have two. Kames form whenever glaciofluvial deposition occurs within an ice-walled channel or depression. They may form at a lateral ice margin, but may also form large areas of kame and kettle topography in front of a glacier. A topography of ice-cored ridges may develop where an ice margin is experiencing large amounts of icefrontal compression and thrusting (Figure 8.15). Meltwater streams will weave between these ridges depositing sediment between them. As the ice-cored ridges melt, the sediment-filled depressions or channels may either: (i) be inverted to form kames; or (ii) be spread over the surface by slumping and flow to form an undulating hummocky surface. A continuum exists between these two extremes.

Where well-defined kames form they may reveal the position of former river channels. The strongest meltwater flows will have the least buried ice beneath them due to the rapid ablation caused by flowing water. Consequently they will often form the highest kames. It is therefore sometimes possible to reconstruct the former patterns of drainage from the elevation of kames.

The pattern of kames, when viewed from above, may either be controlled or uncontrolled. On the one hand, if the pattern of ice-cored ridges, usually controlled by the pattern of debris bands and shearing within the glacier, is regular then the pattern of kames that will result will also have a regular pattern and will therefore be controlled. If on the other hand the pattern of ice-cored ridges is irregular or random, perhaps because the distribution of debris on the surface of the glacier is complex, then the topography of kames that will result will also be irregular or uncontrolled. The margin of Elisbreen in Oscar II Land, Svalbard contains a series of ice-cored ridges parallel to the ice margin. The pattern of meltwater streams and lakes at this ice margin is controlled by the pattern of ice-cored ridges. Glaciofluvial sediment is accumulating between these ridges and ice-cored moraines are progressively melting out with distance from the current ice margin. As this occurs the topography is inverted and the former fluvial channels are being left as upstanding mounds or kames, which trace the former pattern of drainage and are orientated parallel to the former ice margin and its ice-cored moraines. The sedimentary architecture of kames also reflects their history of subsidence and the removal of lateral ice support. Typical subsidence structures found within kames are illustrated in Figure 9.31.

Till

^ Faults

Figure 9.31 Typical sedimentary structures found within kames. [Modified from: Boulton (1972) Journal of the Geological Society of London, 128, figure 4, p. 370]

So far we have restricted this discussion of glaciofluvial sedimentation to glaciers with simple discharge regimes. However, some glaciers experience more complex discharge regimes, with periodic episodes of catastrophic or jokulhlaup flow (Chapter 4). These high-magnitude, low-frequency discharge events typically introduce large volumes of sediment into a glacier foreland and may have a profound effect on the evolution of glaciofluvial landforms.

0 0

Post a comment