Historical Geology/Ice ages

In this article we shall look at what an ice age is, what causes ice ages, and how we can identify the evidence for ice ages in the geological record.

The reader will probably find it useful to look back at the main article on glaciers before reading further.

Definition of an ice age
An ice age, or to use the more technical term, a glaciation, may be defined as a time when continental glaciers are present. This, of course, means that we are currently experiencing an ice age, since there are continental glaciers on Antarctica and Greenland. To be more precise, we are living in an interglacial: a warmer interval within an ice age when the glaciers have retreated towards the poles.

Glaciations in the geological record
Geologists have identified and dated a number of ice ages in the geological record:


 * The Huronian glaciation (or Makganyene glaciation) extended from 2400 to 2100 million years ago


 * The Cryogenian glaciation lasted from 850 to 635 million years ago.


 * The Andean-Saharan glaciation was from 460 to 430 million years ago.


 * The Karoo Ice Age lasted from 360–260 million years ago.


 * The Pliocene-Quaternary ice age is the one now in progress, and started about 2.6 million years ago.

The methods of dating events have already been explained in other articles; later in this article we shall discuss how we identify ice ages.

Causes of ice ages
We have quite a good idea of the causes of the present ice age, which can be explained by continental drift affecting the oceanic circulation. The causes of previous ice ages are more obscure and debatable.

This is because causality as such is not preserved in the geological record. We may be able to see that event A happened and was followed by event B, but the fact (if it is a fact) that event A caused event B must be established on theoretical grounds, by understanding the relationship between the events of the type in question. For example, it is not hard to deduce a causal relationship between the feet of dinosaurs and dinosaur footprints, because we have a good general understanding of the relationship between feet and footprints. But our understanding of climatic effects is hardly as secure.

A further point to consider is this: ice ages are rare and irregular in the geological record. This suggests that there may well not be a single cause for ice ages. If there was a constant reason why ice ages happen, then they would be a permanent feature of the geological record, or at least occur as cyclic events; the fact that they are sporadic suggests that they had singular causes. The best we can do in investigating the causes of an ice age is to look at the events leading up to it and see which of these events might, in principle, be the cause of that particular ice age.

Consider, for example, the Huronian ice age. Two explanations have been proposed for this.

(1) There is evidence (as was discussed in the article on banded iron formations) that oxygen levels rose dramatically in the time leading up to the Huronian. Abundant free oxygen would have combined with the methane (CH4) in the atmosphere, converting it to carbon dioxide and water. Methane is a potent greenhouse gas; that is, it helps to keep the climate warm. Carbon dioxide is also a greenhouse gas, but much less potent than methane. The result of the rise of oxygen and the decline of methane should, therefore, have been a decline in global temperature.

(2) There was remarkably little volcanic activity between 2.45 and 2.2 billion years ago. This can be shown by analysis of zircons, which being resistant to weathering and erosion outlast the igneous rocks in which they are formed, and which can be accurately dated as explained in the article on uranium-based dating methods. The scarcity of zircons formed between 2.45 and 2.2 billion years ago as compared to their relative abundance both before and after strongly suggests a time of low volcanic activity.

Now volcanoes are prone to emit the greenhouse gas carbon dioxide; if volcanic activity virtually ceased, while the process of chemical weathering continued to remove carbon dioxide from the atmosphere, this would lead to a reduction of the greenhouse effect and a decline in global temperatures.

There is, of course, absolutely no reason why both these mechanisms shouldn't have worked together to produce the Huronian glaciation; but while this is possible, it is also possible that one or the other was of such a greater degree of significance that it would be reasonable to call it the cause of the Huronian glaciation.

Whichever of these mechanisms was the main cause, it would be an example of a one-off event. The transition from an atmosphere with little or no free oxygen to one with abundant free oxygen happened, and indeed could happen, only once in the history of the Earth. But a quarter-billion years of low volcanic activity is equally rare in the history of the Earth. It follows that studying the possible causes of this one ice age tells us nothing about the causes of the others, nor can studying the causes of the ice age that we're currently in tell us anything about the causes of the Huronian ice age.

While it can be difficult to say why a particular ice age occurred, it is relatively easy to determine that it occurred. We shall now turn our attention to the nature of the evidence.

Ice ages: how do we know?
As explained in the main article on glaciers, we can observe sedimentary and erosional features associated with glaciers existing today, or which have melted within the time that people have been keeping records of their location: features including striations, moraines of glacial till, deposits of outwash, dropstones in lakes, etc.

When we look at the evidence for recent glacial advances, we see the same sedimentary and erosional signs, but further south than the southern limit of the ice sheets in the present interglacial. By looking at the location of the lobe-shaped terminal moraines, we can find the southern extent of the ice sheets.

These observations would really be sufficient on their own; but in addition to this we can use proxies to investigate the advance and retreat of the glaciers. Biogeographical proxies such as pollen are particularly useful; as the events in question are so geologically recent, we can recognize the pollen of modern species and know exactly under which conditions they flourish.

Another source of information is cosmogenic surface dating. Glaciers are a powerful erosive force, and, as the striations show, scrape sediment off right down to the bedrock. When the glacier retreats in an interglacial, the bedrock is exposed to cosmic rays, and so we can use cosmogenic surface dating to date the retreat of the glaciers.

One interesting relic of the recent glacial retreat is the existence of so-called "sky islands". These are mountains found in the southern United States and Mexico which are home to a distinctive flora and fauna, and which are surrounded by dry grassland or scrubland which the sky island fauna can't cross: hence the name sky islands. But if the species can't migrate from mountain to mountain, how did these species achieve their current range? In the light of paleoclimatology, the solution becomes obvious. When the climate of the southern U.S. was cooler than it is today, the climate of the plains would have been congenial to the species now found in sky islands. But when temperatures increased in the interglacial, they had nowhere to go but the cold moist refuge of the mountains.

Another indication of the recent glacial advance and retreat is isostatic rebound. Recall from our discussion of the structure of the Earth that the lithosphere is essentially floating on top of the denser athenosphere. When a region of the Earth is burdened by the weight of continental glaciers, this ought to push the lithosphere down into the athenosphere. When the glaciers retreat, the weight is lifted, and the lithosphere should very slowly bob back up. Geologists can measure just this happening, at a rate of about 1 cm/year, in regions which (according to sedimentary and other indications) were covered by sheet glaciers before the present interglacial.

As an example of what the burden of ice sheets can do to topography, consider the map to the right: it shows the present elevation of Greenland as it would be if the ice sheets weren't present. As you can see, Greenland is depressed below sea-level at the center where the ice sheets are thickest.

When we look further into the past, we should expect many of these indications to be absent: isostatic rebound, for example, wouldn't take a quarter of a billion years, and so would not remain as evidence of the Karoo ice age. Similarly, as explained in the article on biogeographical proxies, these become less and less useful the further back into the past we wish to look.

However, we still have the moraines, the striations, the dropstones, etc. Are these sufficient to diagnose an ice age? It is reasonable to say yes: for we find these associated with present glaciers, and we also find them associated with the other indications listed above (sky islands, isostatic rebound, proxies, cosmogenic surface dating) when the deposition of the till is recent enough for these additional indications of the former presence of glaciers to have survived to the present. It would be really astonishing if in the cases where we lack this supplementary information, the sedimentary and erosional indications of glaciation were produced by some other process; it is reasonable to conclude that they are signs of an ice age, both because of their empirical association with the present ice age, and because (with the arguable exception of drumlins) we have a clear understanding of how and why glaciers produce these effects.