Historical Geology/Glaciers

Satellite view of the Nimrod Glacier, Antarctica.

A glacier is a mass of ice moving on the surface of the Earth. In this article, we shall discuss how glaciers form and move, the geological features associated with glaciers, and how recognizing these features can allow us to tell where glaciers have been in past ages of the Earth's history.

Formation and motion of glaciers

A glacier forms at an accumulation point, that is, a place where more snow accumulates than melts. This snow then piles up and compacts under its own weight to form ice.

Even if this happened on a perfectly level surface, as the ice mounted up it would eventually start to squidge outwards under the pressure of its own weight; and often glaciers will form on mountaintops, where gravity is also a factor. Under the effects of pressure and/or gravity, the ice will flow. A glacier flows in two ways: by sliding along its base, and by "plastic flow" of the molecules of ice within the glacier.

You may recall from science class that ice tends to melt under pressure; this means that the base of a glacier is often lubricated by water.

The overall speed of a glacier can be measured by simple methods: hammer a stake into a glacier, wait a while, come back, and see how far it's moved. The speeds so measured range from centimeters to meters per day, depending on the glacier.

One significant different between the flow of ice and the flow of water is this: a river is pulled downwards by gravity. This happens to glaciers too, when flowing downhill; but glaciers are also pushed by the pressure behind them: as a result, glaciers can and do flow uphill.

Once in motion, the ice in the glacier will keep flowing until it reaches a point where the ice ablates: either it reaches the sea, breaking up into bergs, or it reaches a zone where the climate is warm enough to melt the advancing glacier. In the latter case, the end of such a glacier represents an equilibrium state at which the rate of melting is just sufficient to balance the rate of flow of the glacier.

Now, while this equilibrium is maintained the glacier as a whole will stay still. The ice in the glacier will move, starting off at the accretion point and ending up at the ablation point, but the glacier as a whole stays in one place: it is like a conveyor belt of ice moving from accumulation to ablation.

The length of the glacier will change with the climate: for example, if the climate gets warmer around the ablation end of the glacier, then the glacier won't be able to progress as far before reaching a zone in which the rate of melting equals the rate of flow, so the glacier will retreat (note that the ice in the glacier will still be moving forward while this is going on). Conversely, of course, a drop in temperature will let the glacier get further from the accumulation point. Changes at the accumulation site of the glacier will also affect its length: the more it snows at the accumulation point, the greater the volume of flowing ice, and the further it will get before it melts. It follows that global cooling will cause glaciers to extend further from accumulation points, and global warming will see them extend less far, or vanish entirely if the temperature rises so much that the snow melts at the former point of accumulation.

Types of glacier

A glacier originating where the snow accumulates on a mountaintop and flows down the mountain is known as a valley glacier or an alpine glacier: the terms are synonymous.

The larger glaciers, such as can be found today covering the surface of Greenland and Antarctica, are known as continental glaciers, or sheet glaciers or ice sheets. Again, these are synonymous and the diversity of terms does not indicate that some sort of distinction is being drawn. Whereas valley glaciers flow downwards from the point of accumulation, the sheet glaciers of Greenland and Antarctica flow outwards in all directions from the point of accumulation.

Erosion associated with glaciers

Striation caused by a glacier.

A glacier moving over a landscape will sweep up topsoil and loose rocks as it goes, transporting them towards the ablation end of the glacier and revealing the bedrock beneath. This bedrock will be polished by the passage of the ice over it; it will also be grooved and scored in the direction of travel by the rocks contained in the glacier: such grooves are known as striations or striae, and the rocks are said to be striated. The picture to the right shows an example of striation:

The passage of a glacier will produce debris of all sizes, from enormous chunks of rock plucked from the bedrock to very fine rock flour produced by the grinding action of the glacier. A valley glacier will also carry along any rock fragments that fall from the valley walls. Some glaciers, known reasonably enough as "rock glaciers", consist mostly of rocks cemented together by ice.

One erosional feature commonly associated with glaciers is the roche moutonée, caused when a glacier slides over a hummock of rock. As the glacier slides up the hill, it polishes and striates it; flowing down the other side, it plucks fragments from the rock, leaving a steeper and more ragged face in its wake.

A valley glacier will create a bowl known as a cirque at the accumulation point, with about a quarter of the rim of the bowl missing in the direction in which the glacier leaves the cirque. A mountain eroded by glaciers will have a rugged, jagged topography, with knife-edge ridges where two cirques or glacial valleys adjoin. The valleys carved out by a valley glacier will have a characteristic U-shaped cross-section quite different from that produced by a river, which produces V-shaped valleys.

Deposition associated with glaciers

The rock fragments transported and deposited by glaciers are known as till. This is unrounded and unsorted by size. This may not sound very remarkable, until we reflect that the action of wind or water cannot produce sediments like this: such an unsorted jumble of shards is characteristic of glacial action.

Moraines, Glacier Bay National Park, Alaska.

In a valley glacier, there will be concentrations of till along the border of the glacier, where it has been plucked or ground from the valley sides. Such an accumulation of till is known as a lateral moraine. Where two valley glaciers meet, lateral moraines will merge into medial moraines in the middle of the larger glacier so formed, as shown in the photograph to the right.

Ground moraine is till deposited over a wide area either when the till on the underside of the glacier lodges against something, or when a glacier retreats in response to climatic changes. Ground moraine is often found in small hills, shaped somewhat like the back of a spoon, known as drumlins. No-one is really sure how these form, but the fact that they are composed of till and found in conjunction with other signs of glacial action confirm that they are glacial in origin; also, they are invariably found oriented with their long axis in the direction of the flow of the glacier (as determined by study of striations, roches moutonées, etc). At the ablation end of a glacier, the sediments transported by the glacier will be dumped to form an end moraine, resulting in a ridge of till in the same convex shape as the lobes typically found at the end of a glacier.

Beyond the ablation zone, where the glacier melts, the water from it will be carried away, typically in a braided stream (a term which will be explained in more depth in the article on rivers). This will carry with it the lighter sediments, known as outwash, which will be deposited in front of the glacier as an outwash plain.

One interesting sedimentary feature can be seen in lakes fed by glacial waters (proglacial lakes). In summer, relatively coarse outwash of sand and gravel will be deposited in these lakes; in winter, when the lake freezes over, the calm conditions below the ice allow fine particles of clay and of organic material to settle. The result is the formation of couplets of sedimentary material, one fine, one relatively coarse, repeated over and over, each couplet being known as a varve. Because the deposition of varves is an annual event, the study of varves is of interest in dating, as will be discussed in a later article.

Walden Pond, Massachusetts, a kettle lake made famous by the writings of Henry David Thoreau.

Kettles are another feature we associate with glaciers. When a glacier retreats, we often observe that it leaves a large block of ice behind it. The outwash sediment from the retreating glacier will then build up around the orphaned block of ice. When it has melted, which may take many years, the result is a depression in the outwash plain: this is a kettle. If it lies below the water table, it will fill up to produce a small lake with outwash banks.

Former glaciers: how do we know?

It is not difficult to detect the passage of a glacier, even if it is no longer present: for if you take away the glacier, you are still left with the patterns of deposition and erosion that glaciers produce; and these are highly distinctive and cannot be produced by other mechanisms. We may note that we can see glaciers disappearing today: for example, Glacier National Park in Montana has at the time of writing only 26 named glaciers, down from 150 in the year 1850; so our statements about what evidence former glaciers leave behind are by no means hypothetical, but rather are based on direct observation.

With all former glaciers, whether valley glaciers or continental glaciers, we see characteristic patterns of erosion: we see such things as bedrock polishing, striation, roches moutonées, and so forth. We also see till. As we have remarked, till is a very distinctive sort of sediment, which cannot be produced by the action of wind and water, as shown by its unsorted, unrounded nature. The arrangement of till can also be quite distinctive; a pile of till in a crescent-shaped end moraine admits of no other explanation except that a glacier deposited it there.

Erratic boulders, when we find them, present another blatant clue. An erratic boulder is one which, in terms of the rock of which it is composed, has nothing in common with the geology of its surroundings, and which must have been transported to its present location over some distance; in some cases, hundreds of kilometers. Ice, as we can observe, can transport such enormous rocks; water and wind do not.

We are therefore left with some unambiguous signs of former glaciers. When we see smoothed bedrock marked with striations leading to a semi-circular moraine of unrounded and unsorted rocks, many of them way out of their original geological context, beyond which is what looks suspiciously like an outwash plain, there is really no other conclusion that we can draw, except that we are looking at where a glacier once flowed and terminated.

Tarn, North Cascades National Park, Washington.

Besides the erosional and depositional features already mentioned, glaciers leave behind some highly distinctive landforms. Consider for example the photograph to the right. There is no longer a glacier present, and, indeed, a lake has formed at the former accumulation point. (Such a lake is known as a tarn.)

Despite the absence of any actual ice, the reader should have no trouble in recognizing the landforms associated with a valley glacier, which are as clear and distinctive as an elephant's footprint. Here, surely, is the great bowl of a cirque, and just where we would expect to find it, near the peak where the temperatures are lowest; in the foreground, where the rim of the cirque is open, we see a valley with the distinctive U-shaped cross-section of a glacial valley, where the glacier once exited the cirque.

Sheet glaciers do not carve out the same forms, but they do leave some large-scale clues behind. They often sweep away soil and other sediment, leaving large expanses of bare rock. In doing so, they also erase the drainage systems that were present in the landscape before their arrival, so that after they retreat, the landscape is poorly drained: such features are a sign to geologists to look for other indications of glaciation.

We should mention a couple more signs of glaciation. The first of these is isostatic rebound. Ice is heavy, and for reasons that we shall go into more thoroughly in later articles, the weight of a continental glacier should press the crust of the Earth down into the mantle, and, when the glacier is gone, the Earth's crust should slowly "bounce" back up. This happens rapidly enough as to leave its marks over mere centuries: so, in parts of Scandinavia, we can see former harbors now standing uselessly distant from the sea. Today, the rate of rebound in post-glacial areas is measured directly by a GPS monitoring system called BIFROST: the maximum rate of rebound is about 1cm/year. This on its own would not prove the former presence of glaciers, but in combination with the less ambiguous signs of erosion and deposition, the phenomenon of rebound does confirm the hypothesis of glaciation.

There is one more prediction that we can make and confirm. We can use geological dating methods, of which more will be said in later articles, to establish the times at which sheet glaciers covered northern North America and Eurasia. Now, if we are correct in attributing these striations, moraines, and so forth to glacial action, then we ought to find that at the same time, we have other evidence of a colder climate, such as flora and fauna adapted to colder climates; lower sea levels caused by water being locked up in continental glaciers; temperature-dependent changes in the composition of shells; and so forth. And this is exactly what we do find, providing an independent confirmation of an ice age.

We shall deal more fully with these topics in later articles on geological dating methods, on paleoclimatology, and on ice ages; for now we shall simply note that these techniques can be used, and that they confirm what we can learn from studying landforms, sediments, and erosional features.